| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Departments of Pharmacology (M.A.) and Physiology and Biophysics (D.L.S.), The University of Iowa, Iowa City, Iowa 52242-1109; and Dipartimento di Chimica (F.F.), Universitá di Modena e Reggio Emilia, Via Campi 183 41100 Modena, Italy
Correspondence: Address all correspondence and requests for reprints to: Dr. Mario Ascoli, Department of Pharmacology, The University of Iowa, 2-319B BSB, 51 Newton Road, Iowa City, Iowa 52242-1109. E-mail: mario-ascoli{at}uiowa.edu
| Abstract |
|---|
|
|
|---|
I. Introduction
II. Structure and Biogenesis of the Lutropin/Choriogonadotropin Receptor (LHR)
A. The LHR protein
B. The LHR gene and the LHR mRNA
III. Expression of the LHR
IV. Signaling Pathways Activated by the LHR
V. Binding of LH/CG to the LHR
VI. Activation of the LHR
VII. Regulation of the LHR
A. Transcriptional regulation
B. Posttranscriptional regulation
VIII. Summary and Conclusions
| I. Introduction |
|---|
|
|
|---|
Research on the LHR has proceeded at a fast pace since our last review was published, and a few additional reviews on different aspects of the biology of the LHR have been published by other investigators (9, 10, 11, 12). In this review we summarize the large body of literature now available on the LHR and we use these data to generate models for the structure, functions, and regulation of this important receptor.
| II. Structure and Biogenesis of the LHR |
|---|
|
|
|---|
|
By convention, residue 1 of the rLHR is taken as the N-terminal residue (R1) of the mature rLHR. The identity of this residue was determined by sequencing the receptor protein purified from the rat ovary (2, 15). The cloning of the cDNA for the rLHR (2) revealed that the precursor form of the rLHR contains a 26-residue signal peptide, and computer algorithms that predict the most likely site of cleavage of signal peptides (http://www.cbs.dtu.dk/services/SignalP/) correctly predict the above-mentioned arginine residue to be the N terminus of the mature rLHR. In contrast, the identity of the N terminus of the mature hLHR is not known with certainty because the hLHR has never been purified and subjected to protein sequencing. Thus, by convention, the amino acid residues of the hLHR have been numbered from the initiator methionine of the hLHR precursor sequence obtained by virtual translation of the open reading frame of the cognate cDNA (6, 7). Algorithms that predict the most likely site of cleavage of signal peptides predict the N terminus of the mature hLHR to be L25. For optimal alignment, L25 of the hLHR corresponds to L3 of the rLHR (Fig. 1
). Because of these shifts, one must move the numbering of amino acid residues up or down by 22 positions to find equivalent hLHR and rLHR residues (for example, L457 in the hLHR is equivalent to L435 in the rLHR).
The polypeptide chains of the mature rLHR and hLHR are predicted to be composed of 674 and 675 residues, respectively. Although the predicted molecular mass of these polypeptide chains is approximately 75 kDa, the molecular mass of the mature hLHR and rLHR is expected to be higher because the LHR is a glycoprotein. Mammalian cells transfected with the cDNAs for the rLHR or the hLHR display three distinct LHR species with molecular masses (estimated from SDS gels) of 6575 kDa, 8595 kDa, and 165200 kDa (Fig. 2
and Refs. 16 and 17, 18, 19, 20, 21, 22, 23, 24, 25). The 85- to 95-kDa band is the mature LHR present at the cell surface. This form of the LHR can be readily labeled by surface biotinylation of intact cells (19) and it can be degraded by proteases under conditions that proteolyze only cell-surface proteins (16, 21). The 85- to 95-kDa band of the LHR is also susceptible to digestion with neuraminidase and peptide-N-glycosidase-F, two glycosidases that are known to act on the mature-type of carbohydrate side chains generally associated with cell-surface glycoproteins. In contrast, the 85- to 95-kDa band is not susceptible to endoglycosidase-H (EndoH), a glycosidase that removes the high-mannose type of carbohydrate side chains associated with immature glycoproteins (16, 17, 24, 25). Biosynthetic labeling of transfected cells with radioactive amino acids revealed that the 65- to 75-kDa form of the LHR is a precursor of the cell-surface receptor (16, 25). This form of the LHR is located intracellularly because it is insensitive to surface proteolysis, it cannot be detected by surface biotinylation of intact cells and is insensitive to neuraminidase digestion (16, 19, 21, 24, 25). The finding that the 65- to 75-kDa band is readily susceptible to EndoH digestion (16, 17, 24) suggests that this is an immature glycoprotein localized in the endoplasmic reticulum. The 165- to 200-kDa LHR species has been identified as an oligomer/aggregate of the 65- to 75-kDa LHR. This identification is based on three findings. First, biosynthetic labeling experiments reveal that the time course of labeling of the 65- to 75-kDa and 165- to 200-kDa forms of the rLHR are basically identical (16, 26). Second, like the 65- to 75-kDa LHR, the 165- to 200-kDa LHR is susceptible to degradation by EndoH, is resistant to degradation by proteolysis of intact cells, and is undetectable by biotinylation of intact cells (16, 19, 21, 26). Third, the 165- to 200-kDa LHR is still detectable in cells transfected with mutants of the LHR that prevent the maturation of the 65- to 75-kDa precursor into the 85- to 95-kDa cell-surface receptor (26). It should be stressed, however, that whereas studies on the properties of the 65- to 75-kDa and 85- to 95-kDa bands of the LHR have been accomplished using cells that are transiently or stably transfected with the rLHR or hLHR, the identification of the 165- to 200-kDa LHR as an oligomer/aggregate of the 65- to 75-kDa immature LHR has been accomplished only using cells expressing the recombinant rLHR. Thus, until the 165- to 200-kDa form is more fully characterized in cells expressing the hLHR, one must consider the possibility that this form of the hLHR is not necessarily an oligomer/aggregate of the 65- to 75-kDa immature receptor. This possibility is in fact likely, given the finding that the 65- to 75-kDa immature receptor is less prevalent in cells transfected with the hLHR than in cells transfected with the rLHR (Fig. 2
).
|
In summary then, the mature form of the LHR present at the cell surface is a glycoprotein with an apparent molecular mass of 8595 kDa that arises from the maturation and transport of a 68- to 75-kDa precursor glycoprotein that is localized in the endoplasmic reticulum. At steady state, the relative abundance of the mature 85- to 95-kDa LHR is much lower than that of the two other forms (i.e., 6575 kDa and 165200 kDa) in cells transfected with the rLHR, but the mature form of the LHR is more abundant than the two other forms in cells transfected with the hLHR (Fig. 2
). This observation is consistent with the results of 125I-hCG binding experiments performed in intact 293 cells transfected with identical amounts of the same expression vector encoding for the rLHR or hLHR (Fig. 2
). It is important to note that the 68- to 75-kDa precursor of the rLHR can bind hCG with the same affinity as the mature 85- to 95-kDa rLHR (21), but the binding affinity of ovine (o) LH for the mature rLHR is higher than its binding affinity for the rLHR precursor (23). Clearly then, although the precursor of the rLHR has attained a conformation that permits hormone binding, this conformation is not the same as that of the mature form of the rLHR.
The rate of maturation of the rLHR and the turnover of the mature and immature forms of the rLHR have been studied in some detail in transfected 293 cells (16, 26, 27, 28). These studies have revealed that the conversion of the immature to the mature form of the rLHR is a slow and inefficient process (half-time of
120 min) that seems to require the association of the immature receptor with at least one chaperone protein (calnexin, see Ref. 29). The immature rLHR is degraded with a half-life of approximately 60 min, and a large proportion of it is never converted to the mature receptor. The 85- to 95-kDa cell-surface rLHR is a fairly stable molecule that is degraded with a half-life of approximately 400 min. The degradation of the immature rLHR (but not that of the mature rLHR) was recently shown to be enhanced by a protein called p38JAB1 (28). The maturation and turnover of the hLHR have not been studied in detail, but the half-life of degradation of the cell-surface hLHR appears to be much slower than that of the rLHR (19). Lastly, it is important to note that the 68- to 75-kDa LHR precursor is not an artifact of overexpression in transfected cells because this form of the LHR has also been detected in Western blots of porcine testes (17, 30) and rat ovaries (16).
The extracellular domains of the rLHR and hLHR each have six consensus sites for N-linked glycosylation. These are fully conserved between the two receptors (as shown by the shaded regions in Fig. 1
). Although it is known that the hLHR contains N-linked carbohydrates, studies determining whether all six potential sites for carbohydrate attachment are used have not been performed on this receptor species. In contrast, this question has been addressed for both the rLHR and pLHR. Using the pLHR (which also contains the same six conserved consensus sequences for N-linked carbohydrates shown in Fig. 1
) purified from stably transfected L cells, Milgrom and colleagues (31) showed, by mass spectrometric analyses, that five of the six sites on the mature pLHR contain carbohydrate. Of these, three are monoantennary and three are multiantennary (31). The pLHR site lacking carbohydrate was found to be N299 (which corresponds to N277 and N299 in the rLHR and hLHR, respectively). Studies by one of our groups (24) using site-directed mutagenesis and endoglycosidase digestion of the rLHR expressed in 293 cells showed that all six sites of the rLHR are glycosylated. However, a study by another group (32) examining the recombinant rLHR expressed in insect cells reported that one of the six sites (corresponding to N77) was not glycosylated. Although it is possible that insect and mammalian cells do not utilize the same sites for N-linked glycosylation, it is also possible that the different results obtained arise from differences in the methodology used to detect glycosylation. In the study of Zhang et al. (32), the presence/absence of carbohydrate on a given site was ascertained by differences in the overall molecular weight of the wt-rLHR and mutants of the rLHR in which the putative glycosylation sites were individually mutated. Because the contribution of a single oligosaccharide chain to the overall molecular weight of a large glycoprotein such as the rLHR is rather small (
2 kb), and insect cells do not glycosylate the rLHR to the same extent as mammalian cells (32), it would be difficult to detect the molecular weight contribution of a given oligosaccharide chain by the methods used. Indeed, the detection of glycosylation at each of the rLHR sites in mammalian cells was only readily detected after treatment of the cells with N-chlorosuccinimide to release a fragment of the extracellular domain containing the glycosylated sites (24). This increased the ratio of carbohydrate to protein and made small changes in molecular weight due to carbohydrate content more readily discernable. The presence of carbohydrate at N77 of the rLHR expressed in 293 cells could also be detected by comparing the molecular weight of a mutant form of the rLHR in which all potential sites for glycosylation other than N77 were mutated with that of a mutant in which all potential glycosylation sites were mutated (24). Moreover, the molecular weight of a mutant of the rLHR in which all potential sites for glycosylation other than N77 were mutated was still sensitive to endoglycosidase treatment (24).
With regards to the potential functional role of LHR glycosylation, studies with the hLHR have shown that the individual mutation of any one of the six predicted sites of N-linked carbohydrate attachment have no effect on hCG binding affinity or hCG-stimulated cAMP production in 293 cells expressing the mutants (33). Similar results were observed with the rLHR (24). Treatment of membranes prepared from human corpus luteum with peptide-N-glycosidase-F (an enzyme that removes all forms of N-linked carbohydrates) does not alter the binding affinity, specificity, or signal transduction properties of the membranes (33), suggesting that, once the hLHR has folded correctly and transported to the plasma membrane, the carbohydrates are dispensable for receptor function. Using the rLHR expressed in mammalian 293 cells, Davis et al. (24) examined the potential contributions of glycosylation on receptor folding and expression, in addition to hormone binding and signal transduction. It was observed that detergent solubilized extracts of cells expressing mutants of the rLHR in which all six glycosylation sites were simultaneously mutated were devoid of any hCG binding activity. However, the lack of binding activity was most likely due to the cumulative deleterious effects of the multiple mutations on the structure of the polypeptide backbone of the receptor rather than to the lack of carbohydrates. Indeed, cells treated with tunicamycin under conditions where only nonglycosylated rLHR is synthesized were found to bind hCG with high affinity (24). Notably, the tunicamycin-treated cells also responded to hCG with a stimulation of cAMP comparable to that of cells expressing a similar density of cell-surface wt rLHR. Therefore, the N-linked carbohydrates of the rLHR do not appear to be essential for the folding of the receptor or for its transport to the plasma. These results further suggest that, similar to the hLHR, once at the plasma membrane, the rLHR does not require N-linked carbohydrates for hormone binding or for hormone-stimulated second messenger production. In contrast to these results, Ji et al. (34) reported that tunicamycin treatment of mouse Leydig tumor cells resulted in a loss of cell-surface receptors. The cells used in the studies by this group of investigators (34) express much lower levels of LHR than those used by Davis et al. (24). Because culturing cells with tunicamycin not only prevents glycosylation but also results in a reduction in receptor expression (because of a general inhibition of protein synthesis), it is possible that the tunicamycin treatment of the cells in the study of Ji et al. decreased cell-surface expression of the rLHR to levels below detection. It should also be noted that Zhang et al. (32) reported that the extracellular domain of the rLHR expressed in insect cells treated with tunicamycin could not bind hCG. These observations cannot be attributed to the use of the extracellular domain (as opposed to the full-length rLHR) because Davis et al. (24) showed that, like the full-length rLHR, detergent extracts of 293 cells expressing the extracellular domain of the rLHR and treated with tunicamycin bound hCG with high affinity. Rather, the apparent discrepancies between the two reports can most likely be attributed to the different assays used to determine hCG binding activity of the receptor. Zhang et al. (32) examined hormone binding by ligand blots, which required the receptor to be refolded after electrophoresis was done under denaturing conditions. This renaturation step may require the presence of carbohydrates. When considered together, these results indicate that the LHR does not have a single carbohydrate chain that is essential for binding hormone or for signal transduction. These results stand in contrast to those obtained with hCG, where a single N-linked carbohydrate on the
-subunit is essential for the ability of hCG to stimulate Gs coupling by the LHR (35, 36). Taken altogether, therefore, it is reasonable to conclude that the nonglycosylated rLHR can be properly folded and expressed at the cell surface (albeit at reduced levels) and that the cell-surface nonglycosylated receptor can bind hormone and transduce signals. It should be noted that whereas these observations rule out an essential role for N-linked carbohydrates for the folding and trafficking of the rLHR, they do not rule out the possibility that the carbohydrates facilitate these processes. Indeed, the reduced expression of the rLHR at the cell surface observed in tunicamycin-treated cells could not only be due to the inhibition of protein synthesis but may also be due to less efficient folding and/or trafficking of the nonglycosylated rLHR. Other studies have since shown that calnexin, a chaperone protein that resides in the endoplasmic reticulum, associates with immature forms of the rLHR (29) and hLHR (D. L. Segaloff, unpublished observations). Because the effects of calnexin on the folding and trafficking of glycoproteins is thought to be mediated (at least in part) by the interaction of calnexin with glucose residues of N-linked carbohydrates (37), it is reasonable to speculate that the N-linked carbohydrates of the rLHR may play a facilitative role in these processes.
Perhaps the most salient feature of the extracellular domain of the LHR is the presence of several repeats of a structural motif of about 25 residues in length that is rich in hydrophobic amino acids and is called the leucine-rich repeat (LRR). Although it was initially thought that the extracellular domain of the rLHR had 14 or 15 of these repeats (2, 38), more recent knowledge suggests the presence of only 8 or 9 repeats (39, 40, 41). It is now generally agreed that the extracellular domain of the mature LHR can be divided into three regions: an N-terminal cysteine-rich region (labeled as such in Fig. 1
), followed by 8 or 9 LRRs and a C-terminal cysteine-rich region (labeled as such in Fig. 1
) that is also known as the hinge region. The most commonly accepted alignment for the LRRs of the extracellular domain of the rLHR and hLHR is shown in Fig. 3
.
|
-helical segment. These units in turn form a superhelix where all the ß-strands and
-helical segments are oriented in a parallel fashion leading to the formation of a nonglobular, horseshoe-shaped molecule. In this molecule, all the ß-strands line the inside or convex surface, whereas all the
-helical segments line the outside or concave surface (44). Although the LRRs of the extracellular domain of the LHR are somewhat shorter than those found in the porcine ribonuclease inhibitor, the crystal structure of this protein has provided the basis for several attempts to model the individual LRRs of the LHR and the overall three-dimensional structure of its extracellular domain (38, 40, 45, 46). Commonly accepted models for these two structures are shown in Fig. 4
|
|
|
-helical segments (63). The first seven of these
-helical segments encompass what has been traditionally viewed as the seven TM segments of the GPCR superfamily. The eighth helix is entirely cytoplasmic and lies between the cytoplasmic end of TM7 and the conserved intracellular cysteine residue that is palmitoylated (see below and Fig. 1
-helical segments predicted for the hLHR is also shown in Fig. 1
No serious attempts have been made to model the three-dimensional structure of the C-terminal tail of the LHR. This region of the LHR is the most divergent between the rLHR and the hLHR (see Fig. 1
) and is the site of two known posttranslational modifications, palmitoylation and phosphorylation. An intracellular cysteine residue present in the juxtamembrane region of the C-terminal tail of the rodhopsin/ß2-adrenergic receptor-like subfamily of GPCRs is among the most highly conserved residues of this subfamily of GPCRs, and all members of this subfamily examined to date have been shown to be palmitoylated at this site. The palmitate present at this highly conserved position is thought to be imbedded in the membrane, and thus, the amino acid residues present between the cytoplasmic end of TM helix 7 and these conserved cysteines are thought to form a fourth IL for this subfamily of GPCRs (reviewed in Ref. 4). The LHR is unusual in having two adjacent cysteines in this position (Fig. 1
). Although the palmitoylation of the hLHR has not been studied, the rLHR expressed in 293 cells has been shown to be palmitoylated at both of these residues (22, 26). The finding that palmitoylation can be detected only on the mature 85- to 95-kDa form of the rLHR (22, 25, 26) suggests that this posttranslational modification occurs during the maturation and transport of the immature receptor to the plasma membrane or once the LHR has reached the plasma membrane. Mutation of the palmitoylation sites of the rLHR had no effect on hCG binding or hCG-stimulated signal transduction (22, 26), but it was reported to enhance the rate of internalization of hCG (22).
Like many other GPCRs studied to date (reviewed in Refs. 65, 66, 67), the rLHR (68) and the hLHR (19) have been shown to be phosphorylated when expressed in transfected cells. Attempts to detect phosphorylation of the LHR in porcine follicular membranes, however, have failed (69). In transfected cells, basal phosphorylation of the mature, cell-surface (8595 kDa) rLHR or hLHR is readily detectable, but phosphorylation of the immature (6875 kDa) receptor is not detectable (19, 68). The phosphorylation of the cell-surface rLHR or hLHR can be enhanced by stimulation of the cells with hCG or phorbol esters (19, 68). Stimulation of transfected cells with cAMP analogs also induces a small and rather variable increase in phosphorylation of the rLHR (68). These analogs have not been tested on cells expressing the hLHR. The identity of the kinase(s) that phosphorylate the LHR in LH/CG-stimulated cells are not known, but they are presumed to be members of the GPCR kinase (GRK) family (67). Overexpression of one of the members of this family [GPCR kinase 2 (GRK2)] has been shown to enhance the LH/CG-induced phosphorylation of the hLHR (19). The LH/CG-induced phosphorylation of the rLHR occurs only on serine residues (70), and the residues of the rLHR that become phosphorylated in LH/CG-stimulated cells have been identified as Ser635, Ser639, Ser649, and Ser652 (see blue residues in Fig. 1
and Refs. 70, 71, 72, 73). The phosphorylation sites of the hLHR have not been studied in as much detail, but the mutation of the five serines that are equivalent to those phosphorylated in the rLHR (see blue residues in Fig. 1
) causes a drastic reduction in basal and LH/CG-stimulated phosphorylation of the hLHR (19). The functional consequences of phosphorylation are discussed in Section VII.B.
B. The LHR gene and the LHR mRNA
The human LHR is encoded by a single gene located in the short arm of chromosome 2 (2p21) (74). In the map of the human genome the ID for the LHR is ENSG00000095001, and its detailed location can be viewed in sequence AC073082 on chromosome 2.
The general organization of the hLHR (75) and rLHR genes (76, 77, 78) are very similar. These genes are about 80 kb in size and each consists of 10 introns and 11 exons (75, 76, 77, 78). The entire serpentine and C-terminal domains of the LHR are encoded in exon 11. This exon also codes for the C-terminal end of the hinge region of the extracellular domain. The N-terminal cysteine-rich region, all of the LRRs, and the N-terminal end of the hinge region of the extracellular domain arise from the splicing of exons 110 (Fig. 1
).
Unlike the coding sequences, the sequences of the 5'-flanking regions of the genes for the rLHR and the hLHR are only approximately 60% homologous (75, 77, 79). The 5'-flanking region of the rLHR resembles those seen in housekeeping genes in that it lacks TATA and CAAT box consensus sequences, displays multiple transcriptional start sites, and has a promoter region rich in G/C nucleotides (nt) and containing several Sp1 sites. The 5'-flanking region of the hLHR gene isolated from a human lymphocyte library by Milgrom and colleagues (75) was shown to be CG rich and to contain two putative TATA boxes and a CAAT box and one site of transcription initiation when analyzed using mRNA from testicular cells. Subsequent studies by Dufau and colleagues (80) on the 5'-flanking region of the hLHR gene that they had cloned from a human placental library suggested the presence of multiple transcriptional start sites when using mRNA from human choriocarcinoma JAR cells, human ovary, or human testes. Taken altogether, these data suggest that the 5'-flanking region of the hLHR gene, like the rLHR gene, is most likely also similar to housekeeping genes.
Transcription of the rLHR and pLHR genes gives rise to multiple mRNA species in gonadal tissues, and the relative abundance of each of the six transcripts differs between ovarian and testicular tissues (2, 81). These different transcripts are thought to occur as a result of the use of different transcriptional start sites, alternate splicing of the gene, and/or differences in polyadenylation (2, 3, 17, 79, 82, 83, 84, 85). Interestingly, variants of the rLHR and pLHR were identified that encode for only the extracellular domain (17, 79). In the pig, 40% of the mRNA transcripts correspond to such variants (17). In the rat ovaries and testis, a 1.2-kb species encoding the extracellular domain is present but at low abundance. However, in MA-10 Leydig tumor cells, this 1.2-kb species is the predominant mRNA (79). Results obtained using heterologous cells transfected with cDNAs encoding for these truncated receptors results have varied between observations reporting total intracellular retention of the cognate protein (86), secretion into the media (87), or both secretion and intracellular retention (17). Three alternatively spliced variants of the hLHR have also been reported in gonadal cells, but these have not been fully characterized (88, 89).
Although the presence of transcripts encoding alternately spliced variants of the LHR gene, especially ones encoding only the extracellular domain, raise intriguing questions regarding their possible physiological roles, it is important to point out that at this writing there are no data yet that demonstrate the presence of proteins arising from such transcripts.
The marmoset monkey provides what may be the only documented example of a splice variant of the LHR that is translated into a stable protein. Like the genes for the human and rat LHR, the LHR gene of the marmoset monkey is made up of 11 exons, but exon 10 is spliced out of the mature mRNA (90). Despite this deletion, the mature marmoset LHR can bind hCG with a high affinity, comparable to that of the hLHR, and the binding of hCG can be translated into increases in cAMP and inositol phosphate production (90). When transfected into COS cells, however, the expression of the marmoset LHR at the cell surface is much lower than that of the hLHR (90), and the deletion of exon 10 from the hLHR results in a decrease in the cell-surface expression of the hLHR (91). Thus, although the deletion of exon 10 seems to impair the trafficking of the receptor to the cell surface, it does not affect the ability of the receptor to bind agonist or to promote cAMP accumulation. Thus, the finding that a naturally occurring homozygous deletion of exon 10 of the hLHR gene is associated with male hypogonadism is likely to be explained by the corresponding reduction in the density of cell-surface receptors rather than by changes in hormone binding and/or signal transduction (92). It should be stressed, however, that the mutant of the hLHR that lacks exon 10 is due to a mutation in the hLHR gene and is not a splice variant of the wt gene.
Several polymorphisms of the hLHR gene have been reported, and those that result in a change in amino acid sequence are summarized in Fig. 6
. A 6-nt in-frame insertion/deletion between codons 18 and 19 of exon 1 results in the expression of two hLHR variants that differ by the presence (LQ) or absence (
LQ) of a Leu-Gln pair near the C terminus of the signal peptide (75, 80, 93). This polymorphism was originally thought to represent an additional gene for the hLHR (80), but the sequence of the human genome has conclusively excluded this possibility. The genotype frequency of this polymorphism in a Caucasian population is 0.088 (LQ/LQ), 0.558 (
LQ/
LQ), and 0.354 (LQ/
LQ) (93). However, the LQ allele has, thus far, not been found in a population of Japanese subjects (93). The presence/absence of the LQ allele does not appear to affect the expression or functional properties of the hLHR as assessed by hCG binding and cAMP responses of cells expressing either variant (93). This is an interesting finding because a longer (11-amino acid) in-frame insertion in the same position results in the intracellular retention of the hLHR and leads to Leydig cell hypoplasia (Fig. 6
and Ref. 94).
|
A number of naturally occurring mutations of the hLHR gene associated with human reproductive disorders have also been reported. They will be briefly summarized here to illustrate what these mutants tell us about the processing of the hLHR (this section), as well as the mechanisms of hormone binding (Section V) and receptor activation (Section VI). A detailed discussion on the functional properties of these mutants and their role in the pathophysiology of the pituitary-gonadal axis can be found in two recent reviews (52, 96).
A tabulation of the location of the naturally occurring loss-of-function mutations of the hLHR reported to date shows that these mutations are not restricted to a particular area of the hLHR (92, 94, 97, 98, 99, 100, 101, 102, 103, 104, 105). They are located throughout the polypeptide chain, as shown in Fig. 6
. Although some of these mutations prevent hCG binding and/or hCGinduced signal transduction, a salient feature that is common to all of them is that they impair the maturation and/or transport of the hLHR precursor so that the expression of the hLHR at the cell surface is always reduced. A comparison of the cell-surface expression of some of these loss-of-function mutants (101) reveals that this parameter can vary from about 50% of the hLHR-wt (i.e., the I625K mutant) to less than 1% of the hLHR-wt (i.e., the A593P mutant).
In contrast to the heterogeneous location of the naturally occurring loss-of-function mutations of the hLHR (Fig. 6
), all naturally occurring gain-of-function mutations reported to date (106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121) are localized to exon 11 in the serpentine region of the hLHR (Fig. 6C
). These mutants display different degrees of constitutive activity and, together with models of the three-dimensional structure of the serpentine region of the hLHR, have provided important information about the mechanisms involved in the activation of the hLHR (see Section VI). Although the restricted location of these mutations is in agreement with the perceived importance of the serpentine region of the hLHR in signal transduction (see Section VI), their restricted location may also be a function of the methods used to search for mutations. Because earlier efforts to find activating mutations of the hLHR were restricted to this region of the gene, it is possible that further studies will reveal the presence of activating mutations of the hLHR elsewhere. Such a finding can indeed be forecasted by the recent demonstration that certain laboratory-designed mutations of the hinge region of the hLHR can also induce constitutive activation (122, 123). In addition, naturally occurring activating mutations have been recently identified in the hinge region of the extracellular domain of the structurally related TSHR (124, 125, 126).
| III. Expression of the LHR |
|---|
|
|
|---|
There is increasing evidence that the LHR may be present in extragonadal tissues as well, both in the reproductive tract and elsewhere. The suggestion of LHR expression has been based in many cases upon the detection of LHR mRNA. In a few cases, the expression of the LHR protein has also been examined using antibodies or by radiolabeled hCG binding assays. Within the female reproductive tract, the LHR has been reported to be present in the bovine, porcine, rat, mouse, rabbit, and human uterus by a number of different laboratories (129, 130, 131, 132, 133, 134, 135, 136, 137). Whereas the LHR in human uterus has been observed in the endometrial endothelium (134), in mice the LHR has been observed in uterine stroma and subepithelial cells (129). Whether these differences reflect differences among species or result from different experimental techniques remains to be determined. It should also be noted that controversy surrounds the issue of LHR expression in the human uterus. Whereas Rao and colleagues (134) have reported the detection of LHR in human uterus, Stewart et al. (138) were not able to detect the LHR in the human uterine samples they examined. Two different groups have also reported the presence of the LHR in other human or mouse female reproductive tract tissues (129, 139). Rao and colleagues (139) reported the presence of the LHR in the mucosa of the human fallopian tube. In the mouse, the LHR was not detected in the oviductal mucosa, but rather, was found in the serosa and in subepithelial cells of the oviduct (129). It should also be pointed out that Rao and colleagues (reviewed in Refs. 128, 140 , and 141) have reported that the LHR is present in a variety of other tissues including human sperm, human seminal vesicles, rat and human prostate, human prostate carcinomas, human skin, human breast cell lines, lactating rat mammary gland, human adrenals, neural retina, neuroendocrine cells, and rat brain. The presence of LHR mRNA and protein has also been documented in breast cancer cell lines, individual human breast cancer biopsies, and benign breast lesions by other investigators (142). In addition, a study by Frazier et al. (143) suggested the presence of the LHR mRNA in the human thyroid.
Certainly, the purported presence of the LHR in these various tissues raises intriguing questions about the physiological role(s) of the gonadotropins and the LHR. It is interesting to note, though, that there have been no reported clinical observations of abnormalities in these different systems in individuals with either loss-of-function or gain-of-function mutations of the LHR gene (52, 96). This finding suggests that if the LHR does indeed play a physiological role in these systems, it may be rather subtle. Determining the potential physiological roles of the LHR in nongonadal tissues in LHR-knockout mice or in humans harboring loss-of-function mutations of the LHR has been problematic because the affected individuals are infertile and they suffer from a secondary deficiency in gonadal hormones that are normally stimulated by LH. The potential development of mice with tissue-specific knockouts of the LHR gene may provide further insights into this problem, however.
The clearest evidence for the extragonadal expression of a functional LHR comes from studies done on a postmenopausal woman with Cushings syndrome and ACTH-independent macronodular adrenal hyperplasia (144). Her clinical history suggested that she also developed Cushings syndrome during each of four full-term pregnancies, but chronic hypercortisolism became obvious only several years after menopause. These findings led to the hypothesis that the ectopic expression of the LHR in her adrenal cortex, coupled with the elevated levels of hCG (during pregnancy) or LH (during menopause), were responsible for the hypercortisolism and adrenal hyperplasia (144). Although the expression of the LHR in the adrenal cortex of this patient was not directly demonstrated, the following functional evidence clearly supports the hypothesis proposed. The administration of GnRH, hCG, or recombinant hLH, but not hFSH, resulted in an increase in serum testosterone, estradiol, and cortisol levels. Administration of a long-acting GnRH analog (leuprolide) initially increased LH, FSH, and cortisol levels but eventually suppressed the levels of LH and FSH and normalized cortisol. Lastly, GnRH did not stimulate cortisol secretion when LHR levels were suppressed by chronic administration of leuprolide, thus ruling out a direct effect of GnRH on the adrenal cortex. These data clearly show that cortisol production was controlled by LH/CG and strongly suggest the presence of a functional LHR in the adrenal cortex of this individual. In fact, long-term treatment of this patient with leuprolide controlled the hypercortisolism and obviated the need for a bilateral adrenalectomy (144). A number of other cases in which clinical findings may be explained by the inappropriate expression of the LHR in the adrenal cortex or in adrenocortical tumors have been recently reviewed (145). Lastly, ectopic expression of functional LHR in the adrenal cortex also appears to occur in transgenic or knockout mouse models with elevated levels of gonadotropins (146, 147, 148, 149).
| IV. Signaling Pathways Activated by the LHR |
|---|
|
|
|---|
The LHR was one of the first GPCRs shown to independently activate two G protein-dependent signaling pathways, adenylyl cyclase and PLC. The first conclusive demonstration of this phenomenon was documented by Gudermann et al. using L cells (153) or Xenopus oocytes (154) expressing the recombinant mouse LHR. This observation has now been extensively reproduced by a number of investigators using a variety of cell lines transfected with either mLHR (153, 155), rLHR (70, 72, 156), or the hLHR (64, 115, 157, 158, 159). Although all investigators agree that the LHR can activate Gs (160, 161, 162), and thus induce the activation of adenylyl cyclase, the identity of the other G protein(s) that are activated by the LHR and which of these mediate the activation of the inositol phosphate pathway is still somewhat controversial. Using a well-established photoaffinity labeling/immunoprecipitation technique, Gudermanns group (160) reported that the mouse LHR expressed in L cells or the endogenous LHR present in bovine luteal membranes activated Gs and Gi2 but not Gq/11, G12, or G13. Using the same methodology, Hunzicker-Dunn and colleagues (161, 162) reported that the endogenous LHR present in porcine follicular membranes activates Gs, Gi, G13, and Gq/11. The time course of the LHR-mediated activation of Gq/11, however, was exceptionally slow compared with that of Gs and Gi and to that of other receptors that activated Gq/11. Also, the subtype of Gi activated by the LHR was not examined in these experiments.
The identity of the G protein and G protein subunits that mediate the effects of the LHR on PLC have been carefully examined in only two cell types expressing the recombinant mLHR, L cells (153, 160) and Sf9 cells (155). Although the LHR-mediated activation of PLC was initially reported to be pertussis toxin insensitive in transfected L cells (153) and in Xenopus oocytes expressing the recombinant mLHR (154), this finding was later corrected (160). The LHR-mediated accumulation of inositol phosphates and Ca2+ mobilization in L cells transfected with the mLHR were later shown to be inhibited by pertussis toxin (160). The mLHR-mediated activation of PLC in transfected mouse L cells was also shown to be inhibited by scavengers of Gß/
(159), and L cells were shown to express two of the different isoforms of PLC (ß2 and ß3) that can be activated by Gß/
subunits. Lastly, overexpression of PLC-ß2 was shown to potentiate the LHR-mediated increase in inositol phosphate accumulation (160). These findings led to the proposal that the LHR-mediated activation of PLC is mediated by the Gß/
subunits liberated from the LHR-induced activation of Gs and Gi (160). Additional experiments performed in Sf9 cells coinfected with the LHR and different G protein subunits revealed that coexpression of G
i2, but not G
s, G
11, G
i1, G
i3, and G
q, resulted in the potentiation of the LHR-stimulated PLC activity (155). Curiously, however, coexpression of two different combinations of Gß/
subunits did not further increase the ability of G
i2 to enhance the LHR-stimulated PLC activity (155). This latter finding is not quite consistent with the proposal that the LHR-induced activation of PLC is mediated by Gß/
(159). Thus, as of now, it seems possible that the LHR-induced activation of PLC may be mediated by the ß/
subunits liberated by the activation of Gi2 (and possibly Gs) or by
q liberated during the activation of Gq.
It is also important to note that although the LHR-mediated activation of adenylyl cyclase is detectable in all cell types examined, the LHR-mediated activation of PLC is not always detectable. For example, two different groups (163, 164) failed to detect an effect of hCG on inositol phosphate accumulation in mouse Leydig tumor cells (MA-10) expressing the endogenous LHR even at high concentrations of hCG. It is not known whether this response is due to the lack of expression of the appropriate G proteins or PLC. These cells do express an isoform of PLC that can be activated by
q, however, as documented by the finding that their inositol phosphate response can be readily activated by the activation of an endogenous GPCR that couples to Gq [i.e., one of the arginine vasopressin type 1 receptors (163)]. It is also possible that the inability of the LHR to activate PLC in MA-10 cells is due to their low density of endogenous LHR (165) because in transfected cells an LHR-induced inositol phosphate response is highly dependent on the density of LHR at the cell surface and is detectable only when cells expressing a high receptor density are exposed to high concentrations of hCG (70, 71, 72, 166). In fact, very recent studies from one of our labs (167) have shown that MA-10 cells expressing high densities of the recombinant hLHR-wt can indeed respond to hCG with an increase in inositol phosphate accumulation. Another problem with the interpretation of results regarding activation of the inositol phosphate vs. the cAMP response relates to the sensitivity of these two assays. Measurements of cAMP levels are usually done using rather sensitive techniques such as a RIA or enzyme-linked immunoassays, whereas measurements of inositol phosphates, which depend on radiolabeling the inositol-containing cellular lipids with [3H]inositol, are much less sensitive. Lastly, in contrast to the well-documented role of cAMP as a mediator of the actions of LH/CG on steroidogenesis, there is little information available about actions of LH/CG that may be mediated by the activation of PLC. Because the activation of PLC requires high concentrations of LH/CG, it has been proposed that this signaling pathway is activated only in females during the preovulatory LH surge or during pregnancy (153). Because maternal hCG is also important in the development of the normal male phenotype, it is possible that exposure of the male fetus to the high levels of maternal hCG may also result in the stimulation of the PLC cascade in fetal Leydig cells.
The classification of naturally occurring mutants of the LHR as gain- or loss-of-function mutants (see above) and the impact of laboratory-designed mutations on the activation of the LHR have been traditionally based on measurements of cAMP accumulation as an index of receptor activation. Only a few experiments have compared the effects of a given LHR mutation (naturally occurring or laboratory-designed) on the cAMP and inositol phosphate responses, and the results obtained indicate that some mutations have a similar effect on the hCG-induced activation of these two pathways, whereas others have divergent effects. For example, a truncation of the C-terminal tail of the rLHR at residue 653 has little or no effect on the hCG-induced cAMP response, but it severely blunts the hCG-induced inositol phosphate response, whereas a truncation at residue 628 enhances both the hCG-induced cAMP and inositol phosphate response (71). Similarly, some single point mutations of residues present in the EL3 of the rLHR (156), IL3 of the hLHR (168), or TM3 of the hLHR (64) have been reported to have differential effects on the hCG-induced cAMP and inositol phosphate responses. These findings raise the possibility that structural features of the LHR that mediate the cAMP response are different from those that mediate the inositol phosphate response. This issue is particularly important in view of a recent proposal (115) stating that constitutive activity toward the inositol phosphate pathway may be restricted only to naturally occurring somatic gain-of-function mutations that are associated with Leydig cell adenomas. This proposal arose from the finding that a gain-of-function mutation of the hLHR associated with Leydig cell adenomas (D578H) displayed constitutive activity toward the cAMP and inositol phosphate responses, whereas a similar gain-of-function mutation of the hLHR associated with Leydig cell hyperplasia (D578Y) displayed constitutive activity only toward the cAMP response (115). It is important to note, however, that other publications from the same group of investigators (157, 159) have reported a small degree of constitutive activation of the D578Y mutant on the inositol phosphate response. In addition, a recent study utilizing another gain-of-function mutation of the hLHR associated with Leydig cell hyperplasia (L457R) has shown that constitutive activation of the inositol phosphate response could not be detected when measuring inositol phosphate accumulation by the traditional assays involving the metabolic labeling of cells with [3H]inositol, but could be detected using a more sensitive reporter gene assay indicative of PKC activation (64), whereas recent studies using MA-10 cells transfected with the L457R, D578Y, and D578H mutants have also shown that they all display constitutive activation of the inositol phosphate response (167). Lastly, the only other naturally occurring gain-of-function mutation of the hLHR that has been examined for constitutive activation of the inositol phosphate response (i.e., the D564G mutant) has been reported to have no constitutive activity toward this pathway by two different groups of investigators (159, 168).
Clearly, more experiments need to be done to further understand the molecular basis of the LHR-mediated activation of PLC and the physiological consequences of this activation. Also, as the number of pathways activated by GPCRs in a G protein-dependent and -independent fashion grows (reviewed in Refs. 151, 152 , and 169, 170, 171, 172), the effects of the LHR on these pathways will need to be evaluated. So far, the only other signaling pathway that has been shown to be activated by the LHR in cells expressing the endogenous or transfected hLHR is the MAPK cascade (167, 173). In fact, when expressed in MA-10 cells, two of the naturally occurring hLHR mutants associated with Leydig cell hyperplasia (D578Y and L457R) and the naturally occurring hLHR mutant associated with Leydig cell adenomas (D578H) were also shown to display constitutive activation of this pathway (167). The molecular basis of the activation of this pathway and the consequences of such activation have not, however, been explored.
| V. Binding of LH/CG to the LHR |
|---|
|
|
|---|
The functional analysis of mutants of the LHR (and other glycoprotein hormone receptors) has often been confounded by the finding that many mutations impair the maturation and/or transport of the receptor to the cell surface, and thus, the mutant receptors are often trapped in the endoplasmic reticulum as the immature 68-kDa precursor (see above and Refs. 8, 23, 25, 26 , and 179). Thus, a decrease in hCG binding in intact cells transfected with a given receptor mutant may or may not be due to an effect of that mutation on hCG binding. Because the binding of hCG can be readily measured in extracts of cells prepared with a nonionic detergent and glycerol (180), and because the 68-kDa precursor of the wt rLHR can bind hCG with the same affinity as the 85-kDa cell-surface receptor (21), the most accurate way to determine whether a given mutation affects hCG binding is to measure this parameter in extracts of transfected cells prepared with a nonionic detergent and glycerol rather than on the intact transfected cells (reviewed in Ref. 8). This complexity in data interpretation can be easily illustrated using studies on two naturally occurring loss-of-function mutations of the hLHR. The deletion of exon 10 of the hLHR is associated with Leydig cell hypoplasia (92), and heterologous cells transfected with a mutant of the hLHR that lacks exon 10 display a reduction in maximal hCG binding when measured in intact cells, but not when measured in cell extracts (91). Measurements of the binding affinity of hCG to extracts of cells transfected with the mutant hLHR lacking exon 10 also show that these cells bind hCG with the same affinity as those transfected with the full-length hLHR-wt (91). The deletion of exon 8 of the hLHR is also associated with Leydig cell hypoplasia (100), and heterologous cells transfected with a mutant of the hLHR that lacks exon 8 display a marked reduction in hCG binding in intact cells and in detergent extracts that precludes measurements of binding affinities (100). Because the expression of the mutant lacking exon 8 was not independently confirmed, we cannot make any conclusions about the involvement of this exon on hCG binding. We can, however, readily conclude that exon 10 is not necessary for hCG binding.
Truncations of the extracellular domain of the LHR as well as deletions and chimeras of the full-length LHR have contributed to a better definition of the regions of the extracellular domain of the LHR that participate in the binding of LH/CG (Table 1
). Ji and co-workers (123, 181) examined several C-terminal truncations of the extracellular domain of the rLHR that systematically deleted each of the 10 exons that encode for this domain. They reported that the binding affinity of these constructs for hCG decreased in a stepwise fashion. Constructs containing exons 17, 18, and 19 (resulting in truncations of the extracellular domain at residues 267, 205, or 180) lowered hCG binding affinity 10- to 20-fold. Constructs containing exons 16 and 15 (resulting in truncations of the extracellular domain at residues 157 and 131, respectively) lowered hCG binding affinity 100- to 200-fold. Constructs containing exons 14, 13, and 12 (resulting in truncations of the extracellular domain at residues 106, 81, and 56, respectively) lowered hCG binding affinity 300- to 600-fold, whereas a construct consisting of only exon 1 (residues 132) did not bind hCG at all. A similar type of analysis was conducted by Thomas et al. (182), who systematically deleted portions of the extracellular domain based on the initial assignment of 14 LRRs in this region (see Section II.A). Their experiments were done using the full-length receptor, and their results are also summarized in Table 1
but, for consistency, the regions deleted are tabulated based on the current assignment of nine LRRs (see Section II.A). Other groups of investigators have also used progressive deletions of the extracellular domain of the rLHR (174), chimeras of the rFSHR and rLHR (174, 183), and deletions of the full-length hLHR (91, 123) to help define the LH/CG binding site, and their data are also summarized in Table 1
.
|
An additional region that is roughly flanked by residues 160190 of the rLHR and encoded mostly by exon 8 has also been examined by two different groups of investigators using C-terminal truncations of the extracellular domain (181) or by deleting it from the full-length receptor (182). Both groups concluded that the removal of this region decreases the hCG binding affinity by about one order of magnitude (Table 1
).
The remaining portion of the extracellular domain of the rLHR is encoded for by exons 9 and 10. It encompasses a portion of LRR8, all of LRR9, and all of the hinge region (see Fig. 1
). Although Jis group (123, 181) reported that the deletion of exons 9 or 10 reduced the binding affinity of the extracellular domain of the rLHR by about one order of magnitude, Braun et al. (174) reported that the removal of this region had little or no effect on hCG binding affinity (see Table 1
). Thomas et al. (182) also reported that the deletion of exons 9 and 10 from the full-length rLHR had no effect on hCG binding affinity, and a chimera of the rLHR and rFSHR that encompasses only exons 18 of the rLHR was reported to bind hCG with an affinity that is similar to that of the rLHR-wt (183). Lastly, the deletion of exon 10 from the hLHR was also reported to have no effect on hCG binding affinity (91). Thus, the bulk of the evidence indicates that the region encoded by exons 9 and 10 (i.e., most of LRR8 all of LRR9 and the hinge region) is not necessary for the high affinity binding of hCG to the LHR.
All the binding data summarized in Table 1
were obtained using detergent/glycerol extracts of the transfected cells and their interpretation is, therefore, not confounded by the possibility that these mutants may not be expressed at the cell surface (see above). Also, because binding affinities were measured in all cases, these data are also not affected by potential differences in the levels of expression of the construct analyzed. The only instance in which conclusions can be ambiguous is when there is no detectable binding in extracts of transfected cells. Therefore, we cannot conclude that rLHR (1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32) or an rLHR/rFSHR chimera containing only residues 146674 of the rLHR are unable to bind hCG (see Table 1
) because the expression of these constructs was not independently verified (174, 181). In contrast, the lack of detectable hCG binding to many of the deletion mutants of the full-length rLHR (see Table 1
) is not due to lack of expression of these mutants because their expression was independently verified using Western blots (182).
As already summarized above, the involvement of the N-terminal cysteine-rich region in hCG binding can be deduced from the finding that deletion of residues 111 of the rLHR abolishes hCG binding (Table 1
and Ref. 182). This conclusion is also supported by the finding that a synthetic peptide corresponding to residues 2138 inhibits hCG binding to rat luteal membranes (184) and the finding that a synthetic peptide corresponding to residues 1836 of the rLHR can be cross-linked to hCG (185). With the exception of mutation of individual cysteine residues (186), the effects of other point mutations of this region on hCG binding have not been examined. The individual mutation of each of the four cysteines present in this region of the rLHR were found to abolish hCG binding in detergent/glycerol extracts of transfected cells (186). Because the expression of the mutants was not independently confirmed, it is not known whether these residues participate in hCG binding or whether receptor expression was abolished (186).
The effects of discrete point mutations of the LRR region on hCG binding affinity have been examined by several investigators (39, 40, 186, 187), and a summary of their data is presented in Fig. 3
. Individual mutation of the residues shown in green has some (2- to 4-fold reduction) or no effect on hCG binding affinity. Those shown in red abolish binding with little or no effect on receptor expression. Those shown in orange also abolish binding, but because the expression of the mutants was not independently confirmed it is not known whether the mutated residues participate in hCG binding or whether receptor expression was abolished. The two most striking features of these results are 1) mutation of the two cysteines present in this region (in LRR-3 and LRR-4) have no effect on hCG binding (186); and 2) most mutations that abolish hCG binding are located in the region of the individual LRRs that form the ß-strand (39, 40, 186, 187). The location of these residues is interesting because the ß-strands form the inner lining of the putative horseshoe-like structure formed by the extracellular domain of the LHR (these are represented by the green arrows in Fig. 4B
).
Mutations of a number of residues present in the hinge region of the hLHR or rLHR, including the six cysteine residues (cf. Fig. 1
), have been found to provoke a small reduction (i.e., no more than 2-fold) or to have no adverse effect on hCG binding affinity (122, 123, 186). Because the hinge region is encoded for by the 3' end of exon 9, all of exon 10 and the 5' end of exon 11 (cf. Fig. 1
), these results are in agreement with the finding that removal of exons 911 has no effect on hCG binding (Table 1
and Refs. 91, 174 , and 182).
In summary, all investigators agree that the N-terminal extracellular domain of the LHR (i.e., exons 110) represents the main binding site for hCG, and most investigators agree that exons 9 and 10 are not necessary for binding. This reduces the hormone binding site to roughly the N-terminal 210 residues (LRRs 18) of the extracellular domain (cf. Figs. 1
and 3
). As mentioned above, the extracellular domain of the LHR is thought to form a horseshoe-like structure with the ß-strands and the
-helices of the individual LRRs forming the inner and outer linings of the horseshoe, respectively (see arrows and helices in Fig. 4B
). Modeling of the extracellular domain of the LHR with the known threedimensional structure of hCG (47) has led to the proposal that the binding of hCG to the LHR occurs at the inner lining of the horseshoe and the primary binding site involves intermolecular interactions between hCG and the residues present in the ß-strands of the individual LRRs (38, 40, 45, 46). The results of mutations of individual LRR residues (summarized in Fig. 3
) clearly support this model.
Although it is clear that the N-terminal extracellular domain of the LHR represents the primary hormone binding site, there is some evidence that suggests that other portions of the receptor participate in, or otherwise influence, hormone binding to the extracellular domain. One of the first reports to raise this possibility came from Jis group (188), which reported that intact cells transfected with rLHR constructs in which residues 6297 or 6336 were deleted could bind hCG with very low affinity (
1 µM). These data were interpreted to mean that the serpentine domain of the rLHR contains a secondary low affinity binding site (188), but the data are not conclusive for two reasons. First, there was no independent assessment of either the expression or the surface localization of the transfected constructs. Second, these constructs also encoded for 532 residues of the N-terminal extracellular domain, and it is thus possible that these residues (as opposed to those present in the serpentine domain) were responsible for the low affinity binding reported. Other experiments from Jis group (189) have suggested that detergent/glycerol extracts of cells expressing a truncated form of the rLHR encoding for only the extracellular domain of the rLHR (residues 1338) bind hCG with a somewhat higher affinity (
2-fold) than detergent/glycerol extracts of cells expressing the full-length rLHR. This finding led to the suggestion that the serpentine domain has a slight inhibitory effect on the binding of hCG to the extracellular domain of the LHR. Alanine scanning mutagenesis of the second EL of the rLHR did in fact show that, in the context of the full-length receptor, the individual mutation of many of the residues present in this loop either enhances or decreases the binding affinity of hCG. The magnitude of these effects was rather small (
2-fold), however. In another study involving alanine scanning mutagenesis of the second EL of the rLHR and the hLHR, one of our laboratories (20) examined the binding affinity of hCG to constructs of the hLHR in which four selected residues of this loop were individually mutated to alanine. One of the four mutants induced a 3- to 4-fold reduction in hCG binding affinity. Other groups (176), including one of our own (177), have failed to detect differences in the binding affinity of hCG or hLH to the full-length rLHR or its extracellular domain expressed alone (177) or as a fusion protein with the TM domain of CD8 (176). One of our groups has also concluded that the deletion of portions of the three ELs have no effect on the binding affinity of the rLHR for hCG (23). In a study from one of our laboratories (177), we did notice, however, that the binding of oLH to detergent/glycerol extracts of cells transfected with the extracellular domain of the rLHR (i.e., rLHR-t338, see Table 1
) is different from the binding of oLH to the full-length rLHR. Thus, whereas the full-length receptor-bound oLH with two apparent affinities, rLHR-t338 bound oLH with a single affinity, which was intermediate between the high and low affinities of the full-length rLHR. These data were originally interpreted as being consistent with the notion that the serpentine region of the rLHR contributes to a second low affinity hormone binding site (177). More recent studies, however, show that the lack of high affinity binding of oLH to rLHR-t338 is not unique to this mutant but is common to several other mutants that, like rLHR-t338, are retained intracellularly and are not processed into the mature cellsurface receptor (23). Because there is little or no difference in the binding affinities of hCG for the cell surface and the intracellular rLHR precursors (see Section II.A and Ref. 21), the intracellular retention of the extracellular domain of the rLHR could not account for the reported differences in the binding affinity of hCG for this construct and the full-length rLHR. Lastly, there are two studies (190, 191) that have examined the effects of mutating a highly conserved aspartate present in the TM2 of the rLHR (D383, shaded pink in Fig. 1
) on hCG binding affinity. Jis group (190) reported that the mutation of this residue (to asparagine) resulted in a 200-fold decrease in the binding affinity of the rLHR for hCG, whereas one of our groups (191) reported that the same mutation had no effect on hCG binding affinity. We did report, however, that a previously documented (192) Na+-induced increase in the binding affinity of oLH for the rLHR was completely abolished by the D383N mutation (191).
In summary then, although there is some evidence that suggests that either the ELs and/or the TM helices of the LHR affect the binding of LH or hCG to its extracellular domain, there is no agreement on how this effect may be brought about and on which regions of the receptor are involved.
| VI. Activation of the LHR |
|---|
|
|
|---|
As with other GPCRs, the regions of the LHR that are thought to interact with and activate G proteins are presumably located in the four ILs (see Fig. 1
) and the juxtacytoplasmic regions of the TM helices. Although the precise areas of the LHR that activate Gs (or other G proteins, see Section IV) have not been mapped, there are data to suggest that the carboxyl-terminal region of IL3 (see Fig. 1
) is involved in G protein activation. Thus, it has been shown that some (but not all) synthetic peptides derived from the TM helices and/or ILs of the LHR can directly activate Gs in isolated membranes (194). For example, synthetic peptides corresponding to TM1 or TM3 cannot activate Gs, but synthetic peptides corresponding to the juxtacytoplasmic regions of TM4, TM5, TM6, and TM7 were active in this assay. More importantly, a peptide corresponding to the carboxyl portion of IL3 plus the juxtacytoplasmic portion of TM6 is more effective in activating Gs than a peptide corresponding to only the juxtacytoplasmic region of TM6 (194). The importance of IL3, as well as TM5/TM6 interactions in LHR-mediated G protein activation, has also been documented by other laboratories using site-directed mutagenesis of these regions, the context of the full-length LHR (168), or in the context of chimeras of the LHR and the FSHR (195). All of these findings are in general agreement with the view that IL3 is involved in the activation of G proteins by other GPCRs (reviewed in Refs. 4, 5 , and 196). For the ß2-adrenergic receptor, it had been suggested that the Gs-activating properties of the carboxyl-terminal region of IL3 may be dependent upon its ability to form an amphiphilic helix (197, 198, 199, 200). This does not appear to be the case for the LHR, however, because mutations that would abolish the ability of this portion of the rLHR to assume an amphiphilic structure do not diminish its signaling properties (201). Lastly, it should also be stressed that IL3 is not necessarily the only region of GPCRs that is involved in G protein activation. For example, rhodopsin peptides corresponding to IL2 (the IL that connects TM3 with TM4) and IL3 (the IL that connects TM5 with TM6), but not IL1 (the IL that connects TM1 with TM2) can bind and/or activate transducin (202). Likewise, docking simulations between constitutively active mutants of the
1b-adrenergic receptor and several G proteins suggest that IL2 and IL3, but not IL1, play a role in G protein recognition (203).
Our understanding of the mechanisms involved in the activation of the LHR has been greatly facilitated by the discovery of mutations that cause constitutive activation of this receptor (see Fig. 6
) and by recent advances in molecular modeling. Until recently, no high-resolution structure of a GPCR was available, and thus, ab initio approaches were employed for building static (59) and dynamic (60) models of the hLHR TM helices and interconnecting loops. In a study done by one of our groups (60), comparisons were made between the unliganded inactive hLHR, two naturally occurring inactivating hLHR mutations, and several naturally occurring constitutively active hLHR mutants. These comparisons allowed for the determination of common structural rearrangements that accompanied the constitutive activation of the hLHR by different mutations (60). Since then, Palczewski et al. (63) have published the crystal structure of visual rhodopsin determined at 2.8-Å resolution. Taking advantage of this major advance in the field, another model for the hLHR (henceforth referred to as the rhodopsin-based hLHR model) has been recently built using rhodopsin as a template (Ref. 64 and F. Fanelli, submitted manuscript). The average arrangements of the seven TM helices of the unliganded inactive (R0) state of the hLHR, as predicted by the rhodopsin-based hLHR model, are shown in Fig. 7A
. The figure depicts the predicted overall structure as observed from the intracellular side (left panel) as well as that observed in the plane of the membrane (right panel). In the right panel, the cytoplasmic face of the membrane is shown at the top to keep the orientation constant relative to that of rhodopsin.
Differences between the rhodopsin-based model and the earlier ab initio model of the hLHR (60) are essentially restricted to the environment of the receptor sites susceptible to activating mutations, as well as the environment of a highly conserved arginine residue present in the IL2 of many GPCRs (R464in the hLHR, highlighted in pink in Figs. 1
and 7A
) and other polar conserved amino acids (also highlighted in pink in Figs. 1
and 7A
). Despite these differences, both models support the hypothesis that the hLHR sites that are susceptible to activating mutations (green circles in Fig. 7A
) are in interhelical positions close to amino acids that are highly conserved among the rhodopsin/ß2-adrenergic receptor subfamily and appear to be involved in signaling (pink circles in Fig. 7A
and pink highlighting in Fig. 1
). The importance of some of these highly conserved polar residues on LHR-mediated signaling has already been documented by mutagenesis. For example, mutation of hLHR residue D405 (or the corresponding rLHR residue D383) in TM2 to Asn has been shown to impair signaling (19, 191, 204). Mutation of rLHR E441 (equivalent to hLHR E463) to Gln reduced cell-surface expression, and its impact on signaling remains difficult to interpret because this property could not be examined in cells expressing an equivalent density of rLHR-wt (205). Mutation of rLHR R442 (equivalent to hLHR R464) to His also reduced cell-surface expression and almost completely abolished signaling when compared with cells expressing an equivalent density of rLHR-wt (204), and mutation of Y601 to Ala in TM7 of the rLHR (equivalent to Y623 in the hLHR) has also been shown to impair signaling (206).
Like the ab initio model, the homology-based model also suggests that interactions between TM3 and TM6, TM3 and TM7, and TM6 and TM7 play an essential role in stabilizing the structure of the inactive hLHR because the majority of the residues that are susceptible to activating mutations are in or close to these interfaces (green circles in Fig. 7A
). The importance of these interhelical interactions in stabilizing the structure of the inactive LHR has also been suggested by mutagenesis studies of selected residues present in TM6 and TM7 of the rLHR (206, 207) or the hLHR (195).
Figure 7B
shows the average minimized structures (viewed from the cytoplasmic side) of several constitutively active mutants of the hLHR (D564G, M398T, L457R, and D578G), one inactivating mutation (I625K), and the unliganded wt hLHR. A comparison of these predicted structures suggests that activating mutations are able to induce structural modifications at the interfaces between the cytosolic extensions of TM3 and TM6 (60, 64). One of the structural modifications that occurs with constitutive activation is a change in the interaction patterns involving R464, the highly conserved residue present in the E/D-R-Y/W motif present in the TM3/IL2 interface (highlighted in pink in Figs. 1
and 7A
). In the inactive structures, R464 is predicted to be involved in a double salt bridge with both an adjacent E463 as well as with D564 in the cytosolic extension of TM6, a residue that results in constitutive activation when mutated (107, 118, 159, 168). A similar interaction pattern is found in the high resolution crystal structure of rhodopsin (63) between R135 (homologous to R464 in the hLHR) and both the adjacent E134 (homologous to E463 in the hLHR) and E247 (homologous to D564 in the hLHR). In the constitutively active hLHR mutants, R464 is predicted to lose its interactions with one or both of these anionic amino acids, thus becoming more exposed to the solvent. Thus, the homology model suggests that the R464-D564 salt bridge contributes to the stabilization of the inactive state of the hLHR, linking the cytosolic extension of TM3 and TM6. This hypothesis is consistent with the finding that substitution of D564 with Gly, Ala, Val, Leu, Phe, Lys, orAsn, mutations that would break or weaken the interaction with R464, result in constitutive activation, whereas substitution of D564 with Glu, which is not predicted to weaken this interaction, does not lead to constitutive activation (159, 168).
The destabilization or breakage of the interactions involving R464 that accompanies activation of the hLHR is predicted to result in an increase in the solvent accessibility of the cytosolic extensions of TM3 and TM6 (indicated by the pink dots in Fig. 7B
). This can be best demonstrated by determining a composite solvent-accessible surface value, computed over selected residues in the cytosolic extensions of TM6 and TM3 facing the core of the helix bundle (60, 64). Results of such computations reveal that the solvent-accessible surface value for constitutively active mutants is higher than the corresponding value for the hLHR-wt or for inactivating mutations (pink dots in Fig. 7B
, and Refs. 60 and 64). Thus, when considered together, these results show that structural perturbations caused by different constitutively active mutants of the hLHR (resulting from the loss and/or gain of interactions between the mutated amino acid and its environment) induce a common structural change, including a weakening of the interactions of R464 with other residues (see above) and an increase in the solvent accessibility of the cytosolic extensions of TM3 and TM6, as shown by the pink dots in Fig. 7B
. These results suggest that a structural modification at the interface between TM3 and TM6 is important in receptor activation and/or G protein recognition. This hypothesis is consistent with a number of studies demonstrating that a rearrangement in the relative position of TM3 and TM6 is a fundamental step in GPCR activation (4, 208, 209) and agrees well with the data (see above) implicating the cytosolic ends of TM5 and TM6, as well as IL3, as being involved in G protein activation by the LHR (168, 194, 195, 210).
It is important to note that the first report of a mutation causing constitutive activation of a GPCR showed that substitution of an alanine residue present in the third IL of the
1b-adrenergic receptor with any of the other 19 amino acids resulted in constitutive activation (211). However, many subsequent studies performed with several other GPCRs (212, 213, 214, 215), including the LHR (64, 157, 159), have failed to show such lack of selectivity in residues that can confer constitutive activity. For example, substitution of D564 (in IL3 hLHR) with Gly, Ala, Val, Leu, Phe, Lys, or Asn results in constitutive activation, whereas substitution with Glu does not (159, 168). Likewise, substitution of D578 (in TM6 of the hLHR) with Glu, Ser, Leu, Tyr, or Phe results in constitutive activation, whereas substitution with Asn does not (157). Lastly, recent studies from one of our groups show that a positively charged residue is required for the substitution of L457 (in TM3) for the hLHR to be constitutively active (64). Consistent with the experimental data, computer simulations suggest that a salt bridge formed between the cationic amino acid in position 457 in TM3, and the anionic D578 in TM6 is primarily responsible for stabilizing the active states of hLHR mutants in which L457 has been mutated to Lys, Arg, or His (64). These observations suggest that constitutive activity of the hLHR in particular (and GPCRs in general) can arise not only by the disruption of interhelical bonds stabilizing the inactive state but also by the formation of new bonds that can stabilize an active state of the receptor.
It should be clear from the foregoing discussion that the study of constitutively active mutants of the LHR has provided us with a fertile ground for the identification of regions of the LHR that may be involved in G protein activation. Although the conclusions made so far must still be considered tentative, they will continue to be tested and will eventually be excluded or shown to be correct. An important aspect of LHR activation that is still poorly understood, however, is how the binding of LH/CG to the extracellular domain of the LHR leads to the presumed rearrangement of the serpentine domain that culminates in G protein activation. Recent mutagenesis studies from several laboratories have implicated specific residues of the N- and C-terminal ends of the hinge region of the extracellular domain of the LHR (122, 123, 216) or in the third EL (217, 218) as being dispensable for hCG binding but critical for receptor activation. Studies done with the rLHR showed that individual mutations of a Pro, Glu, or Asp flanking the last Cys residue present in the C-terminal end of the hinge region (cf. Fig. 1
) impair hCG-induced activation without affecting receptor expression or hCG binding (216), whereas studies done with the hLHR show that mutation of a Ser located two residues upstream of the first Cys residue present at the N-terminal end of the hinge region (cf. Fig. 1
) can result in constitutive activation of this receptor without affecting receptor expression or hCG binding (122, 123). These results are particularly interesting in view of the fact that these residues are fully conserved among the LHR from different species and among the three glycoprotein hormone receptors (122, 123, 216). Studies done by two different laboratories involving mutagenesis of selected residues present in the three ELs of the rLHR have also implicated a Lys residue present in the EL3/TM6 interface (cf. Fig. 1
) as being involved in hCG-induced activation without affecting receptor expression or hCG binding (217, 218). Alanine scanning mutagenesis of EL2 in the rLHR (189) and the hLHR (20) have also implicated several residues present in this region as being important for signal transduction. Lastly, a Glu present in the TM2/EL1 interface (cf. Fig. 1
) was reported to be involved in receptor activation by Jis group (219). This, however, could not be reproduced by one of our laboratories (179).
Based on these studies and those summarized in Section V, one may consider at least three models for the agonistinduced activation of the LHR.
Model 1.
LH/CG binds to the extracellular domain of the LHR, and the bound hormone interacts with and activates the serpentine region. This model is appealing in that it takes into consideration the high degree of homology of the three glycoprotein hormone receptors and the four glycoprotein hormones. One can envision a scenario in which the specificity of binding is dictated by the amino acid sequence of the extracellular domain of a given glycoprotein hormone receptor and the amino acid sequence of the ß-subunit of the hormone, whereas the activation step is common to all glycoprotein hormone receptors and mediated by specific interactions of conserved amino acids present in the common
-subunit of the four glycoprotein hormones and conserved amino acids present in the serpentine region of the three glycoprotein hormone receptors (45, 46, 174, 220). This model is supported by the finding that certain mutations of the hormones (LH or CG) can impair receptor activation without affecting hormone binding (36, 220, 221, 2). In agreement with the view that receptor activation may be mediated by the
-subunit of LH/CG, most of the hormone mutations that prevent activation without affecting binding are located in the
-subunit (36, 220, 221, 222). This model is also supported by the finding that intact cells transfected with constructs of the rLHR in which residues 6297 or 6336 were deleted could respond to extremely high concentrations of hCG with an increase in cAMP accumulation (188). As already mentioned, however (see Section V), it is not known whether these constructs were expressed or properly localized at the cell surface. Moreover, because these constructs also encoded for 532 residues of the N-terminal extracellular domain, it is possible that these extracellular residues (as opposed to those present in the serpentine domain) were responsible for the small increases in cAMP accumulation detected at high hormone concentrations.
Model 2.
LH/CG binds to the extracellular domain of the LHR, thus allowing the extracellular domain of the LHR to activate the serpentine region. This model is similar to Model 1 except that the extracellular domain of the LHR, instead of the bound hormone, is the activator of the serpentine region. Model 2 is supported by the finding (see above) that discrete mutations of the extracellular domain of the LHR can impair signal transduction without impairing hormone binding (216). This finding implies that some residues of the extracellular domain of the LHR can interact with and activate the serpentine domain. It would also suggest that, in the inactive state, the extracellular domain of the LHR is not as tightly associated with the serpentine region as in the activated state. This hypothesis is supported by the observation that the hLHR-wt loses binding activity more rapidly and to a greater extent than constitutively active mutants when intact cells expressing these constructs are treated with protease (Ref. 27 and D. L. Segaloff, unpublished observations).
Model 3.
The extracellular domain of the LHR may hold the serpentine domain in an inactive state, and the binding of LH/CG to the extracellular domain activates the serpentine domain by relaxing this interaction (38, 122, 223). Model 3 is supported by the finding (see above) that discrete mutations of the extracellular domain of the LHR can induce constitutive activation (122, 123). A prediction from this model is that expression of the serpentine and C-terminal regions of the LHR would result in constitutive signaling. Such a result has not been observed by two different groups of investigators that expressed constructs containing the entire C-terminal and serpentine domains plus a few residues of the extracellular domain of the rLHR (188), or a construct encoding for only the serpentine and C-terminal domains of the hLHR (224). As already mentioned above, a caveat with the interpretation of these experiments is that there was no independent assessment of either the expression or the surface localization of the transfected constructs. In considering this model, it is worth noting that the expression of a construct of the hTSHR lacking the extracellular domain results in a level of constitutive signaling that is much higher than that detected in cells expressing the full-length hTSHR (225). In contrast to the results discussed above with the lack of constitutive activity of truncated forms of the LHR, the data obtained with the TSHR are readily interpretable because the cell-surface expression of the two constructs in question was independently confirmed (225). The data obtained with the TSHR may not be extrapolated to the LHR, however, because there appear to be a number of differences in the mode of activation of the LHR when compared with the TSHR or the FSHR (122, 168, 226).
| VII. Regulation of the LHR |
|---|
|
|
|---|
Transcriptional regulation of the LHR is thought to be crucial to the appropriate cell-specific expression of the LHR. The cis- and trans-acting elements governing the basal transcription of the rLHR gene have been examined by Dufau and colleagues (82, 239, 240) in murine Leydig tumor MLTC cells and nonexpressing Chinese hamster ovary (CHO) cells and by one of our groups in primary cultures of rat granulosa cells (241). Both groups have reported the promoter activity of the rLHR gene to be within the first 155176 nt of the 5'-flanking region of the gene (where the numbering of the 5'-flanking region is relative to the start of translation due to the presence of multiple transcriptional start sites). Interestingly, there are regions upstream of the promoter that repress basal transcription of the rLHR gene. In rat granulosa cells, a repressor region was broadly mapped to within the large area between nt -186 and -1375 (241). In MLTC cells, repressor regions were mapped to between nt -173 and -626 and -626 and -990 (82). An additional repressor region located between nt -2056 and -1237 was found in nonexpressing CHO cells but not in the MLTC cells (82). Within the promoter region of the rLHR gene, there are three Sp1 consensus sequences (designated Sp1a, -b, and -c) for Sp1 proteins, two consensus sequences for steroidogenic factor 1, binding and one for AP-2. EMSAs using extracts from rat granulosa cells show that complexes consisting of Sp1 and Sp3 transcription factors are formed at each of the three Sp1 sites (241). Mutations of any of the three Sp1 sites causes a marked reduction in reporter gene activity in the rat granulosa cells, and the combined mutagenesis of all three sites reduced basal transcription of the reporter gene to less than 10% of control (241). These results suggest that the concerted actions of the three Sp1 sites play a critical role in the basal transcription of the rLHR gene in rat granulosa cells. Although Sp1 sites were also shown to play a key role in the basal transcription of the rLHR gene in MLTC cells (239, 240), some differences between the rat granulosa cells and MLTC cells were observed. Thus, in the MLTC cells, a functional role for the Sp1b site was not observed. Furthermore, in MLTC cells it was found that Sp1 bound not only to the canonical Sp1 binding site at nt -143 to -128, but also to the upstream sequence at nt -151 to -143, and that mutation of this upstream sequence rather than the canonical Sp1 sequence reduced basal promoter activity of a reporter gene. Studies in MLTC and CHO cells further suggested that tissue-specific basal rLHR promoter activity is further regulated by three additional domains, which have been called the R domain (nt -1266 to -1377), C-box (nt -42 to -73), and the M1 domain (nt -24 to -42) (11, 82, 239, 240).
The basal transcriptional activity of the hLHR gene has been examined by Dufau and colleagues in JAR cells and simian virus 40-transformed placental cells (80, 242, 243). Similar to the rLHR, the promoter region of the hLHR appears to be within the first 176 bp of the 5'-flanking region (80). This region of the hLHR gene contains consensus sequences for two Sp1 sites, three AP-2 sites, and one estrogen receptor response element half-site (EREhs). Similar to the rLHR gene, the Sp1 sites of the hLHR gene are also involved in regulating the basal transcription of the gene, and EMSAs suggest the binding of both Sp1 and Sp3 to each of the two sites (243). Reporter gene constructs in which one or both sites were disrupted show that they each contribute to basal hLHR gene transcription (243). Upstream of these two Sp1 sites is the EREhs at nt -171. Using a yeast one-hybrid screen, Zhang and Dufau (242) identified three nuclear orphan receptors (EAR2, EAR3/COUP-TF1, and TR4) that bind to this EREhs. EMSAs were used to show that endogenous EAR2 and EAR3/COUP-TF1 from JAR cells and from human testis and TR4 from human testis bind to the hLHR EREhs (242). Functional analyses suggest that the binding of EAR2 and EAR3/COUP-TF1 to the hLHR EREhs repress, whereas the binding of TR4 to this site stimulates, hLHR gene transcription (242, 244). These studies suggest that the relative abundance of these three nuclear orphan receptors may be significant in determining the basal transcription of the hLHR gene in different cell types.
Ultimately, the identity of promoter regions of the LHR gene required for appropriate gender and tissue distribution must be confirmed in vivo by the expression of heterologous genes in transgenic mice. Using this approach, Huhtaniemi and colleagues (245) have constructed three different ßgalactosidase fusion genes containing 7.4 kb, 1.2 kb, or 173 bp of the 5'-flanking sequence of the mLHR gene and studied their expression in vivo in transgenic mice and in vitro in MLTC Leydig tumor cells and KK-1 granulosa tumor cells (245). Interestingly, in the transgenic mice, all three constructs directed ß-galactosidase expression to adult Leydig cells, but not to fetal Leydig cells or to ovarian cells. The expression in adult Leydig cells decreased with increasing lengths of the promoter. Similarly, the constructs were robustly expressed in MLTC cells (where the highest expression was with the shortest 173 bp of the 5'-flanking sequence) but minimally expressed in KK-1 cells. These data suggest that basal transcriptional activity of the mLHR promoter lies within the first 173 nt upstream of the coding sequence of the mLHR, a result that is consistent with analysis of the mLHR gene in transfected mouse Leydig tumor cells (246) and the rLHR and hLHR genes in several cell types (see above). The results observed in the transgenic mice, though, further suggest that regulatory regions outside of the 7.4-kb 5'-flanking region of the mLHR are required for directing mLHR expression to fetal Leydig cells and to the ovary.
In addition to directing the basal expression of the LHR gene, transcriptional regulation of the LHR gene is involved in the FSH-dependent induction of LHR in differentiating granulosa cells. During the growth and differentiation of granulosa cells in the developing follicles, there is an estrogen- and FSH-dependent acquisition of LHR. These actions of FSH can be mimicked, at least in part, by agents that increase intracellular levels of cAMP. The increase in LHR binding activity during this process is accompanied by an increase in LHR mRNA (227, 247). Using granulosa cells cultured from estrogen-primed immature rats, Shi and Segaloff (247) demonstrated by nuclear run-on assays that the FSH or cAMP treatment of these cells causes an approximately 10-fold increase in transcription of the endogenous LHR gene. Although these observations do not rule out a potential role for LHR induction also being mediated by increased LHR mRNA stability, they demonstrate that a cAMP-mediated induction of the LHR gene is clearly important in this process. The increased transcription of the LHR gene in response to FSH or cAMP is not observed until 24 or more hours after treatment, and increases in LHR mRNA and hCG binding activity display a similar lag time (247). Because primary cultures of rat granulosa cells transfected with a reporter gene construct containing 2.1 kb of 5'-flanking sequence of the rLHR gene respond to cAMP treatment, this system is amenable for use in identifying cis- and trans-acting elements mediating the cAMP-dependent induction of the rLHR gene. Cells thusly transfected show incremental increases in the cAMP-mediated fold induction of reporter gene activity as the length of the 5'-flanking region is extended from -40 bp to -2056 bp, suggesting the presence of multiple cAMP-responsive cis elements (241). Similarly, EMSAs using probes corresponding to overlapping portions of the 2.1-kb 5'-flanking sequence and extracts from control vs. cAMP-treated granulosa cells show the presence of multiple complexes whose intensities are increased upon cAMP treatment (241). Some of the elements in the rLHR gene mediating cAMP responsiveness have been identified. These include the three Sp1 sites mentioned above (Sp1a at nt -83, Sp1b at nt -103, and Sp1c at nt -174). These three sites have also been shown to bind both Sp1 and Sp3 and to be involved in the basal transcription of the rLHR gene in primary cultures of rat granulosa cells (241). Individual mutation of any one of the Sp1 sites causes a marked reduction in the 8-Br-cAMP-provoked induction of reporter gene activity, and the simultaneous mutation of all three sites reduces the induction in an additive fashion (241). Two other novel cAMP responsive elements identified thus far are located at nt -142 to -146 and at nt -924 to -933. The site in the proximal promoter region partially overlaps with the Sp1c site and, therefore, was termed an Sp1c adjacent site [SAS (248)]. Although it partially overlaps with the Sp1c site, neither Sp1 nor Sp3 are required for the cAMP-inducible complex that forms at the SAS site. Mutations that disrupt the nuclear protein/DNA complex formation at the SAS site without disturbing those at the Sp1c site reduce the cAMP-inducible reporter gene activity approximately 2-fold. The cAMP-inducible element at the distal site was found to have a somewhat different consensus sequence than the SAS site, but to bind the same nuclear protein(s). Therefore, it was termed an SAS-like element (248). The importance of this site in mediating cAMP responsiveness is underscored by the finding that its mutation abolishes cAMP responsiveness when a 2-kb fragment of 5'-flanking region of the rLHR is used to drive the expression of a reporter gene (248). The identity of the factor(s) binding to the SAS and SAS-like sites has not yet been determined. However, preliminary evidence suggests that it may be related to a family of G-string-related transcription factors (248).
In summary, although considerable progress has been made in our understanding of the mechanisms underlying both basal and hormonal regulation of the transcription of the LHR gene in recent years, much more needs to be done to piece together the role of the many elements and factors that govern the cell-specific expression and hormonal regulation of the LHR mRNA.
B. Posttranscriptional regulation
Desensitization is a word that is used to describe the ability of target cells to turn off an agonist response in the face of continuous agonist exposure. Desensitization is an important component of the regulation of hormone actions and it can occur at multiple levels. For example, experiments done 2025 yr ago demonstrated the existence of two steps in the steroidogenic pathway that contribute to the desensitization of steroidogenic responses observed in male rats injected with LH/CG or in freshly isolated rat or mouse Leydig cells or cultured Leydig tumor cells exposed to LH/CG. These include a reduction in the activity and/or levels of 17
-hydroxylase/17,20-lyase and a reduction in the amount of cholesterol available for steroidogenesis (reviewed in Refs. 249, 250, 251). Additional studies performed using a variety of gonadal cells in culture or subcellular fractions thereof have defined two additional regulatory events that occur at the level of the LHR that may contribute to desensitization. One of them, henceforth referred to as uncoupling, is relatively fast and is caused by the impairment in the ability of the LHR to interact with and activate its cognate G proteins. The other, henceforth referred to as down-regulation, is relatively slow and is due to a reduction in the density of cell-surface receptors. As already discussed above (Section VII.A), changes in the levels of cell-surface receptors can be brought about by changes in transcription, but as discussed below, posttranscriptional regulation of the density of cell-surface receptors also occurs. It now appears that the LH/CG-induced uncoupling and posttranscriptional down-regulation of the LHR are mediated by a family of GPCR-binding proteins known as the nonvisual or ß-arrestins. There are two members of this family, referred as ß-arrestin-1 (also known as arrestin-2) and ß-arrestin-2 (also known as arrestin-3).
The large number of studies conducted with the adrenergic receptors and other GPCRs have delineated the following pathway leading to the formation a GPCR/ß-arrestin complex (reviewed in Refs 65, 66, 67, 172, 252 , and 253). Shortly after G protein activation, GPCRs are phosphorylated in serine and/or threonine residues that are usually, but not always, located in the C-terminal tail. This phosphorylation event, which is catalyzed by a family of serine/threonine kinases known as GRKs, enhances the affinity of the GPCRs for the ß-arrestins, thus leading to the recruitment of the ß-arrestins from the cytosol to the plasma membrane and the formation of a stoichiometric GPCR/ß-arrestin complex. The GPCR/ß-arrestin complex is a common molecular intermediate in at least three events that modulate GPCR functions. First, it uncouples the GPCRs from their cognate G proteins by preventing G protein binding (254). Second, it may serve as a scaffold for the attraction and activation of other signaling components [such as Src and members of the MAPK cascade (255, 256)]. Third, it targets the GPCRs to clathrin-coated pits for subsequent internalization (172, 253). Two of these events, the ß-arrestin-dependent uncoupling and internalization, seem to be involved in the posttranscriptional regulation of the LHR.
Although all investigators agree that the LH/CG-induced uncoupling of the LHR is mediated by the ß-arrestins, there is a fair amount of controversy on whether the LHR undergoes phosphorylation in response to agonist stimulation and on the involvement of LHR phosphorylation in ß-arrestin binding and uncoupling. In the model proposed by Hunzicker-Dunn and colleagues (257, 258), the LH/CG-induced uncoupling of the endogenous LHR present in porcine follicular cells is GTP- and ß-arrestin-dependent, but is independent of the phosphorylation of the LHR. A lack of involvement of LHR phosphorylation and a need for GTP is supported by their inability to detect LH/CG-induced phosphorylation of the endogenous LHR in porcine follicular membranes (69), the lack of effect of protein kinase inhibitors on LH/CG-induced uncoupling (259), the failure of a nonhydrolyzable ATP analog (AMP-PNP) to block LH/CGinduced uncoupling (69, 260, 261), and the need for GTP analogs to induce optimal desensitization (260, 261, 262). In more recent experiments, they have been able to show that the LH/CG-induced uncoupling of the porcine LHR can be enhanced by addition of purified ß-arrestins or inhibited by addition of an antibody to the ß-arrestins (263). They were also able to show that addition of a fragment of the LHR that can bind the ß-arrestins (i.e., the third IL) to porcine follicular membranes inhibits LH/CG-induced uncoupling (264). Further studies from Hunzicker-Dunns laboratory (257, 258, 265) that examined the molecular basis of the GTP dependence of the LH/CG-induced uncoupling implicated ADP ribosylation factor 6 (Arf6, a small GTPase) and Arf-nucleotide binding site opener (ARNO, its guanine nucleotide exchange factor) in LH/CG-induced uncoupling. In their most current model, the LHR is proposed to directly activate ARNO, which in turn activates Arf6 by promoting GTP/GDP exchange. The activated Arf6 then promotes the release of ß-arrestin from a membrane docking site, thus allowing it to bind to the unphosphorylated LHR and to promote uncoupling (257, 258, 265). In a recent publication, however, Lefkowitz and colleagues (266) reported that ARNO can be found in a complex with ß-arrestin, thus suggesting the presence of an alternative pathway whereby ARNO binds to the LHR using the LHR-bound ß-arrestin as an adaptor.
Studies from one of our laboratories have shown that, when expressed in 293 cells, the rLHR (68) and the hLHR (19) become quickly phosphorylated upon addition of hCG. The phosphorylation sites have been mapped to the four (rLHR) or five (hLHR) serine residues present in the C-terminal tail (these residues are depicted in blue in Fig. 1
). The importance of the phosphorylation of these residues in the agonistinduced uncoupling of the rLHR expressed in 293 cells is supported by a number of mutagenesis studies. Thus, the removal of the four phosphorylated serines (by truncation of the C-terminal tail, see Refs. 70, 71 , and 267), their simultaneous mutation to alanine residues (72), or the individual mutation of two of them (S635 or S639, see Ref. 73) retards the time course of LH/CG-induced uncoupling but does not affect the magnitude of this effect detected at long time points. A role for rLHR phosphorylation in the process of uncoupling is also supported by the finding that ATP (not GTP) is required for optimal uncoupling in membranes prepared from 293 cells transfected with the rLHR (268).
Because both groups of investigators agree that the uncoupling of the LHR is mediated by the formation of a ßarrestin/LHR complex, the conflicting data summarized above can be basically reduced to the question of whether phosphorylation of the LHR is necessary for ß-arrestin binding. Unfortunately, this issue has not been directly examined with either of the two species of LHR used in the studies summarized above (i.e., the rLHR or the pLHR). With the rLHR, it is known, however, that another ß-arrestin-dependent event (i.e., internalization) is also impaired by the simultaneous mutation of all four phosphorylation sites (72) or by the individual mutation of three of them (S635, S639, or S649, see Ref. 73). With the pLHR, it is known that ß-arrestin can bind to an unphosphorylated synthetic peptide corresponding to the third IL (264), but this finding does not preclude the likely possibility that other receptor regions could also participate in ß-arrestin binding. Although the low levels of expression of the rLHR in transfected cells (cf. Fig. 2
) make it difficult to examine its interaction with the ß-arrestins, the high levels of expression of the hLHR in transfected cells (cf. Fig. 2
) facilitate measurements of this interaction (269). The recent analysis of a number of hLHR mutants that impair hLHR activation, phosphorylation, or both have in fact shown that the formation of a complex between the hLHR and ß-arrestin-2 is dependent mostly on receptor activation rather than on receptor phosphorylation (269). Although it would be tempting to extrapolate this finding to conclude that phosphorylation of the rLHR or the pLHR is not needed for their interaction with the ß-arrestins, this extrapolation does not seem warranted in view of recent evidence showing that some of the functional properties of the LHR show substantial interspecies variation. For example, the Kd for hCG binding is about one order of magnitude lower for the rLHR than for the hLHR (cf. Fig. 2
), the half-time of internalization of hCG mediated by the rLHR is approximately 10 times longer than that mediated by the hLHR (Fig. 8
and Ref. 270); and the internalized hCG-hLHR complex is routed to a recycling pathway, whereas the internalized hCG-rLHR complex is routed to a degradation pathway (Fig. 9
and Ref. 271). Because the two models presented above have been derived from studies conducted using the LHR from two different species (i.e., porcine or rat), it is entirely possible that they are both correct. Thus, the binding of ß-arrestins to the pLHR (and subsequent desensitization) may be fairly independent of receptor phosphorylation, whereas the binding of ß-arrestins to the rLHR (and subsequent desensitization) may be more dependent on receptor phosphorylation. It is also possible that the differences observed are due to the use of isolated membranes by Hunzicker-Dunn and colleagues (257, 258, 265) and intact cells by one of our groups (19, 68). These issues will not be resolved until the requirements for ß-arrestin binding to the rLHR and pLHR are fully understood and/or all experiments are done using the same experimental paradigm. Lastly, studies on the involvement of ARNO and Arf6 in the desensitization of the LHR have been done only using porcine follicular membranes, and the generality of this model needs to be tested in other systems. As already mentioned above, however, a recent paper (266) suggests that ARNO and the ß-arrestins form a complex, thus providing the molecular basis by which all GPCRs can interact with and activate ARNO.
|
|
In contrast to many other internalized agonist-receptor complexes that dissociate in the acidic environment (pH 5-6) that prevails in endosomes (280), the agonist-LHR complex is rather insensitive to dissociation in this environment (274), and the complex formed by hCG and the mouse, rat, or porcine LHR that is internalized into endosomes is delivered to the lysosomes in the intact form (i.e., without dissociation of the agonist and receptor, see Refs. 272, 273, 274, 275). The more acidic environment that prevails in lysosomes, as well as proteolysis, promote the dissociation of the agonist-receptor complex, and both subunits of the agonist are eventually degraded to single amino acids (281). Although the lysosomal degradation of the internalized LHR has not been formally documented, the net result of this pathway is to target the cell-surface LHR to a compartment where it can be degraded (i.e., the lysosomes), and as such, the endocytosis of the agonist-LHR complex is involved in down-regulation of the cell-surface LHR that occurs when gonadal cells are exposed to agonist (249). The relative importance of internalization to the process of down-regulation has been particularly well studied in MA-10 cells (235). Although it is clear that hCG, acting through cAMP as a second messenger, decreases the transcription of the endogenous LHR gene and the levels of LHR mRNA in MA-10 cells (see Section VII.A and Refs. 233, 234, 235), this effect is quantitatively unimportant to the hCG-induced down-regulation of the LHR detected in this cell type (235).
The importance of LHR internalization to the process of down-regulation was underscored recently by a series of studies comparing these two processes in mouse (MA-10) or rat (R2C) Leydig tumor cells that express the endogenous LHR and in 293 cells expressing the recombinant rLHR (282). Using standardized assay conditions, one of our groups showed that the extent of agonist-induced down-regulation of the LHR is less in 293 cells expressing the recombinant rLHR-wt than in rat or Leydig tumor cells expressing the endogenous LHR. The difference in the extent of agonist-induced down-regulation of the LHR observed in these different cell lines was shown to be due mostly to differences in the rate at which they internalize the agonist-receptor complex (282). Thus, there is a positive correlation between the rate of internalization of the agonist-LHR complex and the extent of down-regulation in three different cell lines (MA-10, R2C, and transfected 293 cells) that express the mouse or rat LHR-wt. Second, all mutations of the rLHR that enhance the rate of internalization of the agonist-rLHR complex in transfected 293 cells also enhance the extent of agonist-induced down-regulation of the rLHR in these cells. Third, three manipulations that enhance the rate of internalization of the agonist-rLHR complex in transfected 293 cells (i.e., cotransfections with GRK2 or ß-arrestin-1 or -2) also enhance the extent of down-regulation of the rLHR-wt in these cells. Lastly, cotransfection with a related construct that has little or no effect on internalization (visual arrestin) does not enhance down-regulation. If the rate of internalization is indeed important to the extent of down-regulation, one would predict that manipulations that slow down internalization should also impair down-regulation. This has been found to be the case, however, in only some instances. For example, the slow rate of internalization of a weak partial agonist in MA-10 cells (278) is accompanied by a reduction in the extent of down-regulation (235). In contrast, when using transfected 293 cells, where the rate of internalization of the agonist-receptor complex is already slow compared with that of target cells, mutations of the rLHR that decrease the rate of internalization by less than approximately 2-fold are not accompanied by a reduction in the extent of down-regulation. However, one mutation that decreased the rate of internalization by 3- to 4-fold did abolish down-regulation.
The fate of the internalized LHR is another important determinant of the extent of hCG-induced down-regulation. As already mentioned above, the complex formed by hCG and the rat, mouse, or porcine LHR accumulates in the lysosomes, which is ultimately responsible not only for the degradation of hCG (272, 273, 275, 281, 282) but also for the net loss of cell-surface LHR that ensues after exposure of rodent or porcine target cells or cells expressing the recombinant rodent or porcine LHR to agonists (235, 277, 282). The targeting of these hCG-receptor complexes to the lysosomes is rather unique because, once internalized, most GPCRs are recycled back to the plasma membrane rather than being routed to the lysosomes (65, 172, 283, 284). Only a few other GPCRs (such as the thrombin receptor and the endothelin type A receptor) are routed to the lysosomes and targeted for degradation (285, 286, 287, 288, 289). Surprisingly, recent studies from one of our laboratories have shown that the fate of the highly related hLHR is different from that of the rLHR in that a substantial portion of the hCG-hLHR complex is routed to a recycling pathway rather than to a degradation pathway (Fig. 9
and Ref. 271). Although more work needs to be done to fully understand the cell and molecular basis of the differential fate of the internalized hCG-hLHR complex and the hCG-rodent/porcine LHR complex, it is already known that, when grafted onto the C-terminal tail of the rLHR, a short linear sequence present in the C-terminal tail of the hLHR (GTALL, underlined in Fig. 1
) can reroute the internalized hCG-rLHR complex from the lysosomal degradation pathway to a recycling pathway (271). The removal of the GTALL sequence from the hLHR failed to affect its routing, however (271).
The last decade has led to the deduction of the amino acid sequences of many proteins, and the data obtained clearly support the maxim that structure dictates function. Because the LHR is a member of the large family of GPCRs, it is not surprising that some of the posttranscriptional events that regulate the LHR are common to other GPCRs. As summarized above, however, the LHR has some unique properties that set it apart from other GPCRs and justify the study of its posttranscriptional regulation. Two unique features of the LHR that should be considered in this respect are 1) the size and complexity of its two cognate ligands (LH and CG), and 2) the high affinity (cf. Fig. 2
) of ligand binding to the LHR. Thus, whereas agonist dissociation to other GPCRs is an integral part of the process of desensitization, the high affinity binding of LH and CG to the LHR results in what can be considered to be an irreversible activation of this receptor during the short time frame needed for phosphorylation, ß-arrestin binding, desensitization, phosphorylation, and internalization. As such, it is likely that the mechanisms involved in the posttranslational regulation of the LHR have evolved to take this difference into account. Lastly, the remarkable differences in the rates of internalization (Fig. 8
and Ref. 270) and the fates of the internalized rLHR and hLHR (Fig. 9
and Ref. 271) should serve as a warning in extrapolating results obtained with studies done using the pLHR, mLHR, or rLHR to the hLHR.
| VIII. Summary and Conclusions |
|---|
|
|
|---|
| Acknowledgments |
|---|
| Footnotes |
|---|
Abbreviations: Arf6, ADP ribosylation factor 6; ARNO, Arf-nucleotide binding site opener; CHO, Chinese hamster ovary; EL, extracellular loop; EndoH, endoglycosidase H; EREhs, estrogen receptor response element half-site; FSHR, FSH receptor; GPCR, G protein-coupled receptor; GRK, GPCR kinase; h, human; IL, intracellular loop; LHR, lutropin/choriogonadotropin receptor; LGR, LRR-containing GPCR; LRR, leucine-rich repeat; nt, nucleotide; o, ovine; p, porcine; r, rat; SAS, Sp1c adjacent site; SNP, single nucleotide polymorphism; TM, transmembrane; TSHR, TSH receptor; wt, wild type.
| References |
|---|
|
|
|---|
-helices. Nature 389:203206[CrossRef][Medline]
-carbon template for the transmembrane helices in the rhodopsin family of Gprotein-coupled receptors. J Mol Biol 272:144164[CrossRef][Medline]
Tyr mutation of the lutrophin/choriogonadotrophin receptor gene. J Med Genet 35:340341
isoleucine) in the extracellular domain of the thyrotropin receptor. J Clin Invest 100:16341649[Medline]
subunit promoter/simian viruse 40 T-antigen fusion gene. Endocrinology 138:35213531
-subunit promoter/simian virus 40 T-antigen fusion gene. Mol Cell Endocrinol 12:801809
and pertussis toxin-catalyzed ADP-ribosylation of Gi
*. Mol Endocrinol 11:538549
subunits of hetertrimeric G proteins stimulated the mitogen-activated protein kinase pathway in COS-7 cells. J Biol Chem 269:78517854
1B-adrenergic receptor by all amino acid substitutions at a single site. J Biol Chem 267:14301433This article has been cited by other articles:
![]() |
C. Wang, B. Xu, B. Zhou, C. Zhang, J. Yang, H. Ouyang, G. Ning, M. Zhang, J. Shen, and G. Xia Reducing CYP51 inhibits follicle-stimulating hormone induced resumption of mouse oocyte meiosis in vitro J. Lipid Res., November 1, 2009; 50(11): 2164 - 2172. [Abstract] [Full Text] [PDF] |
||||
![]() |
N. A. Rahman and C.V. Rao Recent progress in luteinizing hormone/human chorionic gonadotrophin hormone research Mol. Hum. Reprod., November 1, 2009; 15(11): 703 - 711. [Abstract] [Full Text] [PDF] |
||||
![]() |
B. Piotrkowski, C. M Monzon, R. M Pagotto, C. G Reche, M. Besio, C. B Cymeryng, and O. P Pignataro Effects of heme oxygenase isozymes on Leydig cells steroidogenesis J. Endocrinol., October 1, 2009; 203(1): 155 - 165. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. Bachelot, J. Beaufaron, N. Servel, C. Kedzia, P. Monget, P. A. Kelly, G. Gibori, and N. Binart Prolactin independent rescue of mouse corpus luteum life span: identification of prolactin and luteinizing hormone target genes Am J Physiol Endocrinol Metab, September 1, 2009; 297(3): E676 - E684. [Abstract] [Full Text] [PDF] |
||||
![]() |
P. R. Manna, I. T. Huhtaniemi, and D. M. Stocco Mechanisms of Protein Kinase C Signaling in the Modulation of 3',5'-Cyclic Adenosine Monophosphate-Mediated Steroidogenesis in Mouse Gonadal Cells Endocrinology, July 1, 2009; 150(7): 3308 - 3317. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. E. Dickinson, A. J. Stewart, M. Myers, R. P. Millar, and W. C. Duncan Differential Expression and Functional Characterization of Luteinizing Hormone Receptor Splice Variants in Human Luteal Cells: Implications for Luteolysis Endocrinology, June 1, 2009; 150(6): 2873 - 2881. [Abstract] [Full Text] [PDF] |
||||
![]() |
P. R. Manna, M. T. Dyson, and D. M. Stocco Regulation of the steroidogenic acute regulatory protein gene expression: present and future perspectives Mol. Hum. Reprod., June 1, 2009; 15(6): 321 - 333. [Abstract] [Full Text] [PDF] |
||||
![]() |
G. Kleinau and G. Krause Thyrotropin and Homologous Glycoprotein Hormone Receptors: Structural and Functional Aspects of Extracellular Signaling Mechanisms Endocr. Rev., April 1, 2009; 30(2): 133 - 151. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. Guan, X. Feng, X. Wu, M. Zhang, X. Zhang, T. E. Hebert, and D. L. Segaloff Bioluminescence Resonance Energy Transfer Studies Reveal Constitutive Dimerization of the Human Lutropin Receptor and a Lack of Correlation between Receptor Activation and the Propensity for Dimerization J. Biol. Chem., March 20, 2009; 284(12): 7483 - 7494. [Abstract] [Full Text] [PDF] |
||||
![]() |
T. B. Salisbury, A. K. Binder, J. C. Grammer, and J. H. Nilson GnRH-Regulated Expression of Jun and JUN Target Genes in Gonadotropes Requires a Functional Interaction between TCF/LEF Family Members and {beta}-Catenin Mol. Endocrinol., March 1, 2009; 23(3): 402 - 411. [Abstract] [Full Text] [PDF] |
||||
![]() |
C. V. Bishop, J. D. Hennebold, and R. L. Stouffer The effects of luteinizing hormone ablation/replacement versus steroid ablation/replacement on gene expression in the primate corpus luteum Mol. Hum. Reprod., March 1, 2009; 15(3): 181 - 193. [Abstract] [Full Text] [PDF] |
||||
![]() |
L. J. Martin, N. Boucher, B. El-Asmar, and J. J. Tremblay cAMP-Induced Expression of the Orphan Nuclear Receptor Nur77 in MA-10 Leydig Cells Involves a CaMKI Pathway J Androl, March 1, 2009; 30(2): 134 - 145. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. H. S. Costa, A. C. Latronico, R. M. Martin, A. S Barbosa, M. Q Almeida, C. F. P. Lotfi, H. P Lima Valassi, M. Y. Nishi, A. M. Lucon, S. A. Siqueira, et al. Expression profiles of the glucose-dependent insulinotropic peptide receptor and LHCGR in sporadic adrenocortical tumors J. Endocrinol., February 1, 2009; 200(2): 167 - 175. [Abstract] [Full Text] [PDF] |
||||
![]() |
G. Agrawal and R. R. Dighe Critical Involvement of the Hinge Region of the Follicle-stimulating Hormone Receptor in the Activation of the Receptor J. Biol. Chem., January 30, 2009; 284(5): 2636 - 2647. [Abstract] [Full Text] [PDF] |
||||
![]() |
H. Lin, M.-C. Chen, and C.-T. Ku Cyclin-Dependent Kinase 5 Regulates Steroidogenic Acute Regulatory Protein and Androgen Production in Mouse Leydig Cells Endocrinology, January 1, 2009; 150(1): 396 - 403. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. Neumann, G. Kleinau, S. Costanzi, S. Moore, J.-k. Jiang, B. M. Raaka, C. J. Thomas, G. Krause, and M. C. Gershengorn A Low-Molecular-Weight Antagonist for the Human Thyrotropin Receptor with Therapeutic Potential for Hyperthyroidism Endocrinology, December 1, 2008; 149(12): 5945 - 5950. [Abstract] [Full Text] [PDF] |
||||
![]() |
K. Sayasith, N. Bouchard, M. Dore, and J. Sirois Regulation of Bovine Tumor Necrosis Factor-{alpha}-Induced Protein 6 in Ovarian Follicles during the Ovulatory Process and Promoter Activation in Granulosa Cells Endocrinology, December 1, 2008; 149(12): 6213 - 6225. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. W. Warrenfeltz, S. A. Lott, T. M. Palmer, J. C. Gray, and D. Puett Luteinizing Hormone-Induced Up-Regulation of ErbB-2 Is Insufficient Stimulant of Growth and Invasion in Ovarian Cancer Cells Mol. Cancer Res., November 1, 2008; 6(11): 1775 - 1785. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. Royer, A. Lefevre-Minisini, G. Caltabiano, T. Lacombe, Y. Malthiery, F. Savagner, L. Pardo, and P. Rodien The Cloned Equine Thyrotropin Receptor Is Hypersensitive to Human Chorionic Gonadotropin; Identification of Three Residues in the Extracellular Domain Involved in Ligand Specificity Endocrinology, October 1, 2008; 149(10): 5088 - 5096. [Abstract] [Full Text] [PDF] |
||||
![]() |
L. J Martin and J. J Tremblay Glucocorticoids antagonize cAMP-induced Star transcription in Leydig cells through the orphan nuclear receptor NR4A1 J. Mol. Endocrinol., September 1, 2008; 41(3): 165 - 175. [Abstract] [Full Text] [PDF] |
||||
![]() |
L. J. Martin, N. Boucher, C. Brousseau, and J. J. Tremblay The Orphan Nuclear Receptor NUR77 Regulates Hormone-Induced StAR Transcription in Leydig Cells through Cooperation with Ca2+/Calmodulin-Dependent Protein Kinase I Mol. Endocrinol., September 1, 2008; 22(9): 2021 - 2037. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. Hamel, I. Dufort, C. Robert, C. Gravel, M.-C. Leveille, A. Leader, and M.-A. Sirard Identification of differentially expressed markers in human follicular cells associated with competent oocytes Hum. Reprod., May 1, 2008; 23(5): 1118 - 1127. [Abstract] [Full Text] [PDF] |
||||
![]() |
L.-M. Chen, R.-S. Wang, Y.-F. Lee, N.-C. Liu, Y.-J. Chang, C.-C. Wu, S. Xie, Y.-C. Hung, and C. Chang Subfertility with Defective Folliculogenesis in Female Mice Lacking Testicular Orphan Nuclear Receptor 4 Mol. Endocrinol., April 1, 2008; 22(4): 858 - 867. [Abstract] [Full Text] [PDF] |
||||
![]() |
X. Feng, T. Muller, D. Mizrachi, F. Fanelli, and D. L. Segaloff An Intracellular Loop (IL2) Residue Confers Different Basal Constitutive Activities to the Human Lutropin Receptor and Human Thyrotropin Receptor through Structural Communication between IL2 and Helix 6, via Helix 3 Endocrinology, April 1, 2008; 149(4): 1705 - 1717. [Abstract] [Full Text] [PDF] |
||||
![]() |
L. H. Heitman, J. Oosterom, K. M. Bonger, C. M. Timmers, P. H. G. Wiegerinck, and A. P. IJzerman [3H]Org 43553, the First Low-Molecular-Weight Agonistic and Allosteric Radioligand for the Human Luteinizing Hormone Receptor Mol. Pharmacol., February 1, 2008; 73(2): 518 - 524. [Abstract] [Full Text] [PDF] |
||||
![]() |
H. L Henderson, D. J Hodson, S. J Gregory, J. Townsend, and D. J Tortonese Gonadotropin-Releasing Hormone Stimulates Prolactin Release from Lactotrophs in Photoperiodic Species Through a Gonadotropin-Independent Mechanism Biol Reprod, February 1, 2008; 78(2): 370 - 377. [Abstract] [Full Text] [PDF] |
||||
![]() |
K. Angelova, F. Fanelli, and D. Puett Contributions of Intracellular Loops 2 and 3 of the Lutropin Receptor in Gs Coupling Mol. Endocrinol., January 1, 2008; 22(1): 126 - 138. [Abstract] [Full Text] [PDF] |
||||
![]() |
G. Pidoux, P. Gerbaud, O. Marpeau, J. Guibourdenche, F. Ferreira, J. Badet, D. Evain-Brion, and J.-L. Frendo Human Placental Development Is Impaired by Abnormal Human Chorionic Gonadotropin Signaling in Trisomy 21 Pregnancies Endocrinology, November 1, 2007; 148(11): 5403 - 5413. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. A. Andric, M. M. Janjic, N. J. Stojkov, and T. S. Kostic Protein kinase G-mediated stimulation of basal Leydig cell steroidogenesis Am J Physiol Endocrinol Metab, November 1, 2007; 293(5): E1399 - E1408. [Abstract] [Full Text] [PDF] |
||||
![]() |
K. Handschuh, J. Guibourdenche, V. Tsatsaris, M. Guesnon, I. Laurendeau, D. Evain-Brion, and T. Fournier Human Chorionic Gonadotropin Produced by the Invasive Trophoblast But Not the Villous Trophoblast Promotes Cell Invasion and Is Down-Regulated by Peroxisome Proliferator-Activated Receptor-{gamma} Endocrinology, October 1, 2007; 148(10): 5011 - 5019. [Abstract] [Full Text] [PDF] |
||||
![]() |
L. Wang, A. K. Nair, and K. M. J. Menon Ribonucleic Acid Binding Protein-Mediated Regulation of Luteinizing Hormone Receptor Expression in Granulosa Cells: Relationship to Sterol Metabolism Mol. Endocrinol., September 1, 2007; 21(9): 2233 - 2241. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. Zhang, Y.-X. Tao, G. L. Ryan, X. Feng, F. Fanelli, and D. L. Segaloff Intrinsic Differences in the Response of the Human Lutropin Receptor Versus the Human Follitropin Receptor to Activating Mutations J. Biol. Chem., August 31, 2007; 282(35): 25527 - 25539. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. Gromoll, L. Lahrmann, M. Godmann, T. Muller, C. Michel, S. Stamm, and M. Simoni Genomic Checkpoints for Exon 10 Usage in the Luteinizing Hormone Receptor Type 1 and Type 2 Mol. Endocrinol., August 1, 2007; 21(8): 1984 - 1996. [Abstract] [Full Text] [PDF] |
||||
![]() |
K. Shiraishi and M. Ascoli Lutropin/Choriogonadotropin Stimulate the Proliferation of Primary Cultures of Rat Leydig Cells through a Pathway that Involves Activation of the Extracellularly Regulated Kinase 1/2 Cascade Endocrinology, July 1, 2007; 148(7): 3214 - 3225. [Abstract] [Full Text] [PDF] |
||||
![]() |
P. R Manna, Y. Jo, and D. M Stocco Regulation of Leydig cell steroidogenesis by extracellular signal-regulated kinase 1/2: role of protein kinase A and protein kinase C signaling J. Endocrinol., April 1, 2007; 193(1): 53 - 63. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. Kern, A. I. Agoulnik, and G. D. Bryant-Greenwood The Low-Density Lipoprotein Class A Module of the Relaxin Receptor (Leucine-Rich Repeat Containing G-Protein Coupled Receptor 7): Its Role in Signaling and Trafficking to the Cell Membrane Endocrinology, March 1, 2007; 148(3): 1181 - 1194. [Abstract] [Full Text] [PDF] |
||||
![]() |
E.-J. Yang, B. T. Nasipak, and D. B. Kelley Direct action of gonadotropin in brain integrates behavioral and reproductive functions PNAS, February 13, 2007; 104(7): 2477 - 2482. [Abstract] [Full Text] [PDF] |
||||
![]() |
P. Nurwakagari, A. Breit, C. Hess, H. Salman-Livny, D. Ben-Menahem, and T. Gudermann A conformational contribution of the luteinizing hormone-receptor ectodomain to receptor activation J. Mol. Endocrinol., February 1, 2007; 38(2): 259 - 275. [Abstract] [Full Text] [PDF] |
||||
![]() |
C. Stocco, C. Telleria, and G. Gibori The Molecular Control of Corpus Luteum Formation, Function, and Regression Endocr. Rev., February 1, 2007; 28(1): 117 - 149. [Abstract] [Full Text] [PDF] |
||||
![]() |
G. Kleinau, M. Brehm, U. Wiedemann, D. Labudde, U. Leser, and G. Krause Implications for Molecular Mechanisms of Glycoprotein Hormone Receptors Using a New Sequence-Structure-Function Analysis Resource Mol. Endocrinol., February 1, 2007; 21(2): 574 - 580. [Abstract] [Full Text] [PDF] |
||||
![]() |
G. Kleinau, M. Claus, H. Jaeschke, S. Mueller, S. Neumann, R. Paschke, and G. Krause Contacts between Extracellular Loop Two and Transmembrane Helix Six Determine Basal Activity of the Thyroid-stimulating Hormone Receptor J. Biol. Chem., January 5, 2007; 282(1): 518 - 525. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. Bonomi, M. Busnelli, L. Persani, G. Vassart, and S. Costagliola Structural Differences in the Hinge Region of the Glycoprotein Hormone Receptors: Evidence from the Sulfated Tyrosine Residues Mol. Endocrinol., December 1, 2006; 20(12): 3351 - 3363. [Abstract] [Full Text] [PDF] |
||||
![]() |
N. Andric and M. Ascoli A Delayed Gonadotropin-Dependent and Growth Factor-Mediated Activation of the Extracellular Signal-Regulated Kinase 1/2 Cascade Negatively Regulates Aromatase Expression in Granulosa Cells Mol. Endocrinol., December 1, 2006; 20(12): 3308 - 3320. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. Berndt, S. P. d'Hauterive, S. Blacher, C. Pequeux, S. Lorquet, C. Munaut, M. Applanat, M. A. Herve, N. Lamande, P. Corvol, et al. Angiogenic activity of human chorionic gonadotropin through LH receptor activation on endothelial and epithelial cells of the endometrium FASEB J, December 1, 2006; 20(14): 2630 - 2632. [Abstract] [Full Text] [PDF] |
||||
![]() |
C. Galet and M. Ascoli A Constitutively Active Mutant of the Human Lutropin Receptor (hLHR-L457R) Escapes Lysosomal Targeting and Degradation Mol. Endocrinol., November 1, 2006; 20(11): 2931 - 2945. [Abstract] [Full Text] [PDF] |
||||
![]() |
P. Cameo, M. Szmidt, Z. Strakova, P. Mavrogianis, K. L. Sharpe-Timms, and A. T. Fazleabas Decidualization Regulates the Expression of the Endometrial Chorionic Gonadotropin Receptor in the Primate Biol Reprod, November 1, 2006; 75(5): 681 - 689. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. Van Durme, F. Horn, S. Costagliola, G. Vriend, and G. Vassart GRIS: Glycoprotein-Hormone Receptor Information System Mol. Endocrinol., September 1, 2006; 20(9): 2247 - 2255. [Abstract] [Full Text] [PDF] |
||||
![]() |
P. R Manna, S. P Chandrala, Y. Jo, and D. M Stocco cAMP-independent signaling regulates steroidogenesis in mouse Leydig cells in the absence of StAR phosphorylation. J. Mol. Endocrinol., August 1, 2006; 37(1): 81 - 95. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. Freamat, H. Kawauchi, M. Nozaki, and S. A Sower Identification and cloning of a glycoprotein hormone receptor from sea lamprey, Petromyzon marinus. J. Mol. Endocrinol., August 1, 2006; 37(1): 135 - 146. [Abstract] [Full Text] [PDF] |
||||
![]() |
C. Y. Andersen, P. Humaidan, H. B. Ejdrup, L. Bungum, M.L. Grondahl, and L.G. Westergaard Hormonal characteristics of follicular fluid from women receiving either GnRH agonist or hCG for ovulation induction Hum. Reprod., August 1, 2006; 21(8): 2126 - 2130. [Abstract] [Full Text] [PDF] |
||||
![]() |
K. Shiraishi and M. Ascoli Activation of the Lutropin/Choriogonadotropin Receptor in MA-10 Cells Stimulates Tyrosine Kinase Cascades that Activate Ras and the Extracellular Signal Regulated Kinases (ERK1/2) Endocrinology, July 1, 2006; 147(7): 3419 - 3427. [Abstract] [Full Text] [PDF] |
||||
![]() |
P. M. Apaja, J. T. Tuusa, E. M. Pietila, H. J. Rajaniemi, and U. E. Petaja-Repo Luteinizing Hormone Receptor Ectodomain Splice Variant Misroutes the Full-Length Receptor into a Subcompartment of the Endoplasmic Reticulum Mol. Biol. Cell, May 1, 2006; 17(5): 2243 - 2255. [Abstract] [Full Text] [PDF] |
||||
![]() |
H. Jaschke, S. Neumann, S. Moore, C. J. Thomas, A.-O. Colson, S. Costanzi, G. Kleinau, J.-K. Jiang, R. Paschke, B. M. Raaka, et al. A Low Molecular Weight Agonist Signals by Binding to the Transmembrane Domain of Thyroid-stimulating Hormone Receptor (TSHR) and Luteinizing Hormone/Chorionic Gonadotropin Receptor (LHCGR) J. Biol. Chem., April 14, 2006; 281(15): 9841 - 9844. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. Conti, M. Hsieh, J.-Y. Park, and Y.-Q. Su Role of the Epidermal Growth Factor Network in Ovarian Follicles Mol. Endocrinol., April 1, 2006; 20(4): 715 - 723. [Abstract] [Full Text] [PDF] |
||||
![]() |
H. Jaeschke, S. Neumann, G. Kleinau, S. Mueller, M. Claus, G. Krause, and R. Paschke An Aromatic Environment in the Vicinity of Serine 281 Is a Structural Requirement for Thyrotropin Receptor Function Endocrinology, April 1, 2006; 147(4): 1753 - 1760. [Abstract] [Full Text] [PDF] |
||||
![]() |
T. Mizutani, K. Shiraishi, T. Welsh, and M. Ascoli Activation of the Lutropin/Choriogonadotropin Receptor in MA-10 Cells Leads to the Tyrosine Phosphorylation of the Focal Adhesion Kinase by a Pathway that Involves Src Family Kinases Mol. Endocrinol., March 1, 2006; 20(3): 619 - 630. [Abstract] [Full Text] [PDF] |
||||
![]() |
K. Saner-Amigh, B. A. Mayhew, F. Mantero, F. Schiavi, P. C. White, C. V. Rao, and W. E. Rainey Elevated Expression of Luteinizing Hormone Receptor in Aldosterone-Producing Adenomas J. Clin. Endocrinol. Metab., March 1, 2006; 91(3): 1136 - 1142. [Abstract] [Full Text] [PDF] |
||||
![]() |
P. R. Manna, S. P. Chandrala, S. R. King, Y. Jo, R. Counis, I. T. Huhtaniemi, and D. M. Stocco Molecular Mechanisms of Insulin-like Growth Factor-I Mediated Regulation of the Steroidogenic Acute Regulatory Protein in Mouse Leydig Cells Mol. Endocrinol., February 1, 2006; 20(2): 362 - 378. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. De Leener, L. Montanelli, J. Van Durme, H. Chae, G. Smits, G. Vassart, and S. Costagliola Presence and Absence of Follicle-Stimulating Hormone Receptor Mutations Provide Some Insights into Spontaneous Ovarian Hyperstimulation Syndrome Physiopathology J. Clin. Endocrinol. Metab., February 1, 2006; 91(2): 555 - 562. [Abstract] [Full Text] [PDF] |
||||
![]() |
N. M. Robert, L. J. Martin, and J. J. Tremblay The Orphan Nuclear Receptor NR4A1 Regulates Insulin-Like 3 Gene Transcription in Leydig Cells Biol Reprod, February 1, 2006; 74(2): 322 - 330. [Abstract] [Full Text] [PDF] |
||||
![]() |
T. L. Mazzuco, O. Chabre, J.-J. Feige, and M. Thomas Aberrant Expression of Human Luteinizing Hormone Receptor by Adrenocortical Cells Is Sufficient to Provoke Both Hyperplasia and Cushing's Syndrome Features J. Clin. Endocrinol. Metab., January 1, 2006; 91(1): 196 - 203. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. K. Nair and K. M. J. Menon Regulation of Luteinizing Hormone Receptor Expression: EVIDENCE OF TRANSLATIONAL SUPPRESSION IN VITRO BY A HORMONALLY REGULATED mRNA-BINDING PROTEIN AND ITS ENDOGENOUS ASSOCIATION WITH LUTEINIZING HORMONE RECEPTOR mRNA IN THE OVARY J. Biol. Chem., December 30, 2005; 280(52): 42809 - 42816. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. P N Themmen An update of the pathophysiology of human gonadotrophin subunit and receptor gene mutations and polymorphisms Reproduction, September 1, 2005; 130(3): 263 - 274. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. Costagliola, E. Urizar, F. Mendive, and G. Vassart Specificity and promiscuity of gonadotropin receptors Reproduction, September 1, 2005; 130(3): 275 - 281. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. B Rulli and I. Huhtaniemi What have gonadotrophin overexpressing transgenic mice taught us about gonadal function? Reproduction, September 1, 2005; 130(3): 283 - 291. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. Bielinska, E. Genova, I. Boime, H. Parviainen, S. Kiiveri, J. Leppaluoto, N. Rahman, M. Heikinheimo, and D. B. Wilson Gonadotropin-Induced Adrenocortical Neoplasia in NU/J Nude Mice Endocrinology, September 1, 2005; 146(9): 3975 - 3984. [Abstract] [Full Text] [PDF] |
||||
![]() |
F. X. Donadeu and M. Ascoli The Differential Effects of the Gonadotropin Receptors on Aromatase Expression in Primary Cultures of Immature Rat Granulosa Cells Are Highly Dependent on the Density of Receptors Expressed and the Activation of the Inositol Phosphate Cascade Endocrinology, September 1, 2005; 146(9): 3907 - 3916. [Abstract] [Full Text] [PDF] |
||||
![]() |
K. Sayasith, J. G. Lussier, and J. Sirois Role of Upstream Stimulatory Factor Phosphorylation in the Regulation of the Prostaglandin G/H Synthase-2 Promoter in Granulosa Cells J. Biol. Chem., August 12, 2005; 280(32): 28885 - 28893. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. Ascoli Learning New Tricks from an Old Dog: The Processing of the Intracellular Precursor of the Luteinizing Hormone Receptor (LHR) into the Mature Cell-Surface LHR Is a Regulated Process Endocrinology, August 1, 2005; 146(8): 3221 - 3223. [Full Text] [PDF] |
||||
![]() |
S. J. Marx and W. F. Simonds Hereditary Hormone Excess: Genes, Molecular Pathways, and Syndromes Endocr. Rev., August 1, 2005; 26(5): 615 - 661. [Abstract] [Full Text] [PDF] |
||||
![]() |
B. Karges, S. Gidenne, C. Aumas, F. Haddad, P. A. Kelly, E. Milgrom, and N. de Roux Zero-Length Cross-Linking Reveals that Tight Interactions between the Extracellular and Transmembrane Domains of the Luteinizing Hormone Receptor Persist during Receptor Activation Mol. Endocrinol., August 1, 2005; 19(8): 2086 - 2098. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. Yamashita, K. Nakamura, Y. Omori, K. Tsunekawa, M. Murakami, and T. Minegishi Association of Human Follitropin (FSH) Receptor with Splicing Variant of Human Lutropin/Choriogonadotropin Receptor Negatively Controls the Expression of Human FSH Receptor Mol. Endocrinol., August 1, 2005; 19(8): 2099 - 2111. [Abstract] [Full Text] [PDF] |
||||
![]() |
P. M. Apaja, J. T. Aatsinki, H. J. Rajaniemi, and U. E. Petaja-Repo Expression of the Mature Luteinizing Hormone Receptor in Rodent Urogenital and Adrenal Tissues Is Developmentally Regulated at a Posttranslational Level Endocrinology, August 1, 2005; 146(8): 3224 - 3232. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. Zhang, D. Mizrachi, F. Fanelli, and D. L. Segaloff The Formation of a Salt Bridge Between Helices 3 and 6 Is Responsible for the Constitutive Activity and Lack of Hormone Responsiveness of the Naturally Occurring L457R Mutation of the Human Lutropin Receptor J. Biol. Chem., July 15, 2005; 280(28): 26169 - 26176. [Abstract] [Full Text] [PDF] |
||||
![]() |
E. M. Pietila, J. T. Tuusa, P. M. Apaja, J. T. Aatsinki, A. E. Hakalahti, H. J. Rajaniemi, and U. E. Petaja-Repo Inefficient Maturation of the Rat Luteinizing Hormone Receptor: A PUTATIVE WAY TO REGULATE RECEPTOR NUMBERS AT THE CELL SURFACE J. Biol. Chem., July 15, 2005; 280(28): 26622 - 26629. [Abstract] [Full Text] [PDF] |
||||
![]() |
Y. Kazeto, R. Goto-Kazeto, P. Thomas, and J. M Trant Molecular characterization of three forms of putative membrane-bound progestin receptors and their tissue-distribution in channel catfish, Ictalurus punctatus J. Mol. Endocrinol., June 1, 2005; 34(3): 781 - 791. [Abstract] [Full Text] [PDF] |
||||
![]() |
H.-F. Kwok, W.-K. So, Y. Wang, and W. Ge Zebrafish Gonadotropins and Their Receptors: I. Cloning and Characterization of Zebrafish Follicle-Stimulating Hormone and Luteinizing Hormone Receptors-- Evidence for Their Distinct Functions in Follicle Development Biol Reprod, June 1, 2005; 72(6): 1370 - 1381. [Abstract] [Full Text] [PDF] |
||||
![]() |
C. Galet and M. Ascoli The Differential Binding Affinities of the Luteinizing Hormone (LH)/Choriogonadotropin Receptor for LH and Choriogonadotropin Are Dictated by Different Extracellular Domain Residues Mol. Endocrinol., May 1, 2005; 19(5): 1263 - 1276. [Abstract] [Full Text] [PDF] |
||||
![]() |
H. Zhou, Y. Jiang, W. K W Ko, W. Li, and A. O L Wong Paracrine regulation of growth hormone gene expression by gonadotrophin release in grass carp pituitary cells: functional implications, molecular mechanisms and signal transduction J. Mol. Endocrinol., April 1, 2005; 34(2): 415 - 432. [Abstract] [Full Text] [PDF] |
||||
![]() |
|