This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Copyright Permission Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Ascoli, M. Articles by Segaloff, D. L. Search for Related Content PubMed PubMed Citation Articles by Ascoli, M. Articles by Segaloff, D. L.

The Lutropin/Choriogonadotropin Receptor, A 2002 Perspective

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

Top Abstract I. Introduction II. Structure and Biogenesis... III. Expression of the... IV. Signaling Pathways Activated... V. Binding of LH/CG... VI. Activation of the... VII. Regulation of the... VIII. Summary and Conclusions References

 
Reproduction cannot take place without the proper functioning of the lutropin/choriogonadotropin receptor (LHR). When the LHR does not work properly, ovulation does not occur in females and Leydig cells do not develop normally in the male. Also, because the LHR is essential for sustaining the elevated levels of progesterone needed to maintain pregnancy during the first trimester, disruptions in the functions of the LHR during pregnancy have catastrophic consequences. As such, a full understanding of the biology of the LHR is essential to the survival of our species. In this review we summarize our current knowledge of the structure, functions, and regulation of this important receptor.

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

Top Abstract I. Introduction II. Structure and Biogenesis... III. Expression of the... IV. Signaling Pathways Activated... V. Binding of LH/CG... VI. Activation of the... VII. Regulation of the... VIII. Summary and Conclusions References

 
THIS IS THE third time that we have written a review on the lutropin/choriogonadotropin receptor (LHR) in this journal. In the first review (1), which was published in February of 1989, we contrasted and compared the conflicting data that existed on the structure of the mammalian LH/CG receptor (LHR). The controversy on the structure of the LHR that was obvious in that review was laid to rest later in 1989 with the simultaneous publication of two papers reporting the cloning, sequencing, and expression of cDNAs for the rat [r (2)] and porcine [p (3)] LHR. These two papers conclusively established that the LHR is a single polypeptide chain with an overall structure and topology that made it a member of the rhodopsin/ß₂-adrenergic receptor subfamily (4, 5) of G protein-coupled receptors (GPCRs). The cloning of the rLHR and pLHR cDNAs was quickly followed by the cloning of cDNAs for the human (h) LHR (6, 7). The novel experimental tools so generated were also rapidly used to try to understand the molecular basis of the functions and regulation of the LHR, and our second review on this subject (8), which was published in June of 1993, summarized the rather large amount of data that was generated shortly after the cloning of the first two cDNAs for the LHR.

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

Top Abstract I. Introduction II. Structure and Biogenesis... III. Expression of the... IV. Signaling Pathways Activated... V. Binding of LH/CG... VI. Activation of the... VII. Regulation of the... VIII. Summary and Conclusions References

 
A. The LHR protein
A search of the National Center for Biotechnology Information nucleotide database revealed 63 entries for "LH receptor." Most of these represent partial or complete sequences for LHR cDNAs for a variety of domestic animals (i.e., cow, pig, sheep, chicken, and turkey), animals that are widely used as experimental models in modern biological research (i.e., rat, mouse, and Xenopus), and other interesting animals such as the catfish, mink, marmoset, and black bear. Because most of the experiments to date have been performed with the rLHR and hLHR, an amino acid sequence alignment of these two receptors is presented in Fig. 1.

View larger version (36K): [in this window] [in a new window]   Figure 1. Amino acid sequence alignment of the rLHR and hLHR. Sequences for the rLHR (accession no. P16235) and the hLHR (accession no. P22888) were obtained from the SWISS-PROT databank. The dashes indicate identical residues. The boundaries of the three distinct regions of the extracellular domain discussed in the text (the N-terminal cysteine-rich region, leucine-rich motif region, and hinge region) are marked by the green, red, and green arrows, respectively, and the cysteine residues present in the N-terminal cysteine rich and hinge regions are highlighted in green. The sequences coded for by the different exons are delineated by the vertical bars and are labeled 1–11. The seven TM regions are delineated by black boxes and labeled TM-1–TM-7. The three ELs and four ILs that connect the TM regions are labeled as EL-1–EL-3 and IL-1–IL-4, respectively. The blue boxes delineate predicted helical segments based on the rhodopsin homology model discussed in Section VI. The consensus sequences for N-linked glycosylation are highlighted by the gray boxes. Polar residues that are highly conserved among the rhodopsin/ß2-adrenergic receptor subfamily of GPCRs and may be involved in receptor activation are highlighted in pink and correspond to the pink residues shown in Fig. 7. Other residues that are highly conserved among the rhodopsin/ß2-adrenergic receptor subfamily are shown in purple. The serine residues that are phosphorylated in response to agonist stimulation are highlighted in blue. The residues highlighted in yellow are those that determine the slow rate of internalization of the rLHR and the fast rate of internalization of the hLHR. The underlined (GTALL) sequence present in the C-terminal tail of the hLHR highlights a short linear sequence that, when grafted into the rLHR, reroutes the internalized hCG-rLHR complex from a lysosomal degradation pathway to a recycling pathway.

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 65–75 kDa, 85–95 kDa, and 165–200 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).

View larger version (29K): [in this window] [in a new window]   Figure 2. Expression of the rLHR and hLHR in transfected 293 cells. The 293 cells were transiently transfected with epitope(myc)-tagged forms of the rLHR (21 ) or hLHR (19 ). The expression of the receptors was documented using Western blots of whole-cell lysates developed with a monoclonal antibody (9E10) to the myc epitope or by saturation binding analysis of 125I-hCG to the intact, transfected cells.

In summary then, the mature form of the LHR present at the cell surface is a glycoprotein with an apparent molecular mass of 85–95 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., 65–75 kDa and 165–200 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 p38^JAB1 (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.

View larger version (57K): [in this window] [in a new window]   Figure 3. Alignment of the LRRs in the extracellular domain of the rLHR and hLHR. Amino acid sequence alignment of the LRRs of the extracellular domains of the rLHR and hLHR. Dashes indicate identical residues and dots indicate gaps introduced for optimal alignment. The gray boxes highlight the location of consensus sequences for N-linked glycosylation. The model depicting nine LRRs is presented (40 41 53 ). The alignment for the eight-LRR model is basically the same as that shown here, except for the absence of LRR-9 (39 ). The boxes at the top labeled ß-sheet, turn, and -helix correspond to the proposed three-dimensional structure of the individual LRRs and extracellular domain shown in Fig. 4. The residues enclosed in boxes represent the conserved hydrophobic residues present in LRRs, and the numbers at the top correspond to the numbers associated with these residues in the structure shown on panel b of Fig. 4A. Also note that there are two cysteine residues present in this region (marked with an asterisk in LRR-4 and LRR-5). Residues in green have no effect on binding when mutated. Residues in pink abolish or decrease binding without affecting expression. Residues in orange abolish binding, but their effect on receptor expression was not tested (39 40 186 187 290 ).

View larger version (24K): [in this window] [in a new window]   Figure 4. Structural representation of a single LRR (A) and the leucine-rich motif region (B) of the LHR. A, The structural representation of the LRR depicted on the left (a) is that derived from the three-dimensional structure of the ribonuclease inhibitor (44 ). The structure of the LRR depicted on the right (b) is a proposal for the structure of the LRRs of the extracellular domain of the LHR and is based on the structure of the LRR depicted in panel a. The conserved hydrophobic residues are numbered based on the convention of Kobe and Deisenhofer (44 ), and these numbers also correspond to those shown at the top of the conserved residues on the amino acid sequence alignment shown in Fig. 3. B, A ribbon diagram of the LRR region of the extracellular domain of the LHR is shown. Note that the ß-sheets (arrows) and the -helices (spirals) align on opposite faces of the horseshoe-shaped structure. The location of the three consensus sequences for N-linked glycosylation and the two cysteine residues present in this region are also shown. [Reproduced with permission from N. Bhowmick et al.: Mol Endocrinol 10:1147–1159, 1996 (40 ). © The Endocrine Society.]

View larger version (13K): [in this window] [in a new window]   Figure 5. Phylogenetic relatedness of LGRs from different species. This dendogram was constructed based on full-length amino sequence comparisons of the different LGRs shown. The LGRs can be basically divided into three distinct families. One family is composed of the mammalian LHR, TSHR, FSHR, the fly LGR1, the nematode LGR (nLGR), and the sea anemone LGR (AeLGR). The second family is composed of mammalian LGR4–6 and fly LGR2. The third family is composed of mammalian LGR7 and snail LGR. [Reproduced with permission S. Nishi et al.: Endocrinology 141:4081–4090, 2000 (53 ). © The Endocrine Society.]

View larger version (43K): [in this window] [in a new window]   Figure 7. Rhodopsin-based homology model of the TM region of the hLHR-wt and mutants thereof. A, The TM regions of the hLHR as viewed from the cytosol (left) or the plane of the membrane (right) are shown. The figure on the right is presented with the cytoplasm at the top and extracellular space at the bottom, similar to structural representations of rhodopsin. The helical segments shown by the cylinders and labeled 1–7 correspond to the helices delineated by the blue boxes in Fig. 1. The pink residues are polar amino acids that are highly conserved in the rhodopsin/ß2-adrenergic receptor subfamily of GPCRs and correspond to the pink residues shown in Fig. 1. The green residues are those that result in constitutive activation when mutated and correspond to the green residues shown in Fig. 6. B, The TM regions of the hLHR as viewed from the cytosol. The helical segments shown by the cylinders and labeled 1–7 correspond to the helices delineated by the blue boxes in Fig. 1. In the models for the loss-of-function (I625K) and gain-of-function mutants (M398T, L457R, D564G, and D578G), the position of the mutated residue is indicated by a blue sphere. The pink shading depicts the solvent-accessible surface computed over amino acids R464, T467, and I468 in TM3 and K563 in TM6, as discussed in the text and in Refs. 60 and 64 .

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 (85–95 kDa) rLHR or hLHR is readily detectable, but phosphorylation of the immature (68–75 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 1–10 (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).

View larger version (36K): [in this window] [in a new window]   Figure 6. Location of the naturally occurring polymorphisms and loss-of-function and gain-of-function mutations of the hLHR. The location and identity of naturally occurring mutations of the hLHR are shown. Polymorphisms are depicted in yellow, loss-of-function mutations are depicted in orange, and gain-of-function mutations are depicted in green. Other annotations are described in the legend to Fig. 1.

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

Top Abstract I. Introduction II. Structure and Biogenesis... III. Expression of the... IV. Signaling Pathways Activated... V. Binding of LH/CG... VI. Activation of the... VII. Regulation of the... VIII. Summary and Conclusions References

 
For many years the LHR was thought to be localized strictly to gonadal cells. In the testes, the LHR is thought to be restricted to Leydig cells. In the ovary, expression of the LHR occurs in theca cells, interstitial cells, differentiated granulosa cells, and luteal cells. Certainly, the main physiological roles of the LHR can be attributed to its actions in the ovaries and the testes. Thus, women who are homozygous for loss-of-function mutations of the LHR are infertile. 46,XY individuals homozygous for severe inactivating mutations of the LHR are also infertile and present as pseudohermaphrodites, whereas less severe inactivating mutations result in micropenis and/or hypospadias. These phenotypes can be explained by the inability of the fetal Leydig cells to respond to maternal hCG with increased testosterone production, a steroid that is required for the differentiation of external genitalia to the male phenotype (52, 96). Transgenic mice in which the LHR gene has been knocked out are also infertile (127, 128). The primary difference observed between the LHR loss-of-function phenotype between mice and humans is the lack of pseudohermaphroditism in the male LHR-knockout mice.

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 Cushing’s syndrome and ACTH-independent macronodular adrenal hyperplasia (144). Her clinical history suggested that she also developed Cushing’s 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

Top Abstract I. Introduction II. Structure and Biogenesis... III. Expression of the... IV. Signaling Pathways Activated... V. Binding of LH/CG... VI. Activation of the... VII. Regulation of the... VIII. Summary and Conclusions References

 
Although most investigators agree that the LHR-mediated effects on the differentiated function of Leydig and granulosa cells are mediated mostly (if not entirely) by the activation of the Gs/adenylyl cyclase/cAMP/PKA pathway, it is abundantly clear now that this is not the only pathway activated by the LHR and that additional pathways may be involved in other LHR-dependent events such as the proliferation and/or differentiation of target cells. This issue is specially meaningful at a time when there is a growing body of evidence suggesting that GPCRs can affect the proliferation and differentiation of endocrine cells and that they may do so by using signaling mechanisms that are much more complex than previously recognized. For example, it is now generally accepted that a given GPCR can independently activate more than one subfamily of heterotrimeric G proteins (reviewed in Ref. 150), and that heterotrimeric G proteins may in fact not be the only mediators of GPCR signaling (reviewed in Refs. 151 and 152).

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, Gudermann’s 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 Gi₂, but not Gs, G 11, Gi1, Gi3, and Gq, 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 Gi₂ 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

Top Abstract I. Introduction II. Structure and Biogenesis... III. Expression of the... IV. Signaling Pathways Activated... V. Binding of LH/CG... VI. Activation of the... VII. Regulation of the... VIII. Summary and Conclusions References

 
The amino acid sequence and predicted topology of the LHR deduced from the cloning efforts quickly led to the hypothesis that the large extracellular domain of this receptor was responsible for the recognition and high affinity binding of LH and CG. This hypothesis was initially tested in one of our laboratories (86) by measuring ¹²⁵I-hCG binding in cells transfected with a construct (designated rLHR-t338) that encoded for only residues 1–338 (i.e., the extracellular domain) of the rLHR. Because the construct used was devoid of coding sequences of the serpentine and C-terminal tail, but contained the coding sequence for the signal peptide, we expected that the truncated receptor would be secreted into the culture medium. This, however, was not found to be the case. Instead, the extracellular domain of the rLHR was found to be retained inside of the cells. Thus, binding assays performed on the conditioned medium and intact cells failed to detect any binding. The only way to detect ¹²⁵I-hCG binding to cells transfected with rLHR-t338 was to solubilize the cells with a nonionic detergent and glycerol (86). Binding assays performed on solubilized cells transfected with either rLHR-wt or rLHR-t338 revealed that rLHR-t338 binds hCG with the same affinity as the full-length rLHR (K_d = 0.2–0.5 nM). The ability of the extracellular domain of the rLHR, pLHR, and hLHR to bind hCG with very similar (or identical) affinities to those of their full-length counterparts has now been confirmed by many investigators using a variety of constructs encoding for the extracellular domain (17, 78, 87), chimeras containing the extracellular domain of the LHR with the serpentine and C-terminal tails of the FSHR (174) or the ß₂-adrenergic receptor (175), and fusion proteins containing the extracellular domain of the LHR with the single TM domain of CD8 (176). As expected, the extracellular domain of the LHR is also responsible for dictating hormonal specificity. Thus, whereas the extracellular domain of the rLHR binds LH and CG with high affinity, it does not bind FSH (177). Likewise, hormone binding and hormone-stimulated cAMP accumulation in cells transfected with LHR/FSHR chimeras is dictated by the identity of the extracellular domain of the chimeras (174, 178). Lastly, a soluble form of the extracellular domain of the hLHR can inhibit the binding of hCG to cells expressing the full-length hLHR, but it cannot inhibit the binding of hFSH to cells expressing the full-length hFSHR (176). Conversely, a soluble form of the extracellular domain of the hFSHR can inhibit the binding of hFSH to cells expressing the full-length hFSHR, but it cannot inhibit the binding of hCG to cells expressing the full-length hLHR (176).

The functional analysis of mutants of the LHR (and other glycoprotein hormone receptors