Hormone Interactions to Leu-rich Repeats in the Gonadotropin Receptors

I. ANALYSIS OF LEU-RICH REPEATS OF HUMAN LUTEINIZING HORMONE/CHORIONIC GONADOTROPIN RECEPTOR AND FOLLICLE-STIMULATING HORMONE RECEPTOR^*

Yong Sang Song^§, Inhae Ji, Jeremy Beauchamp^¶, Neil W. Isaacs^¶, and Tae H. Ji

From the  Department of Chemistry, University of Kentucky, Lexington, Kentucky 40506-0055, the ^§ Cancer Research Center, Seoul National University College of Medicine, Seoul 110-744, Korea, and the ^¶ Department of Chemistry, University of Glasgow, Glasgow G128QQ, United Kingdom

Received for publication, May 3, 2000, and in revised form, June 12, 2000

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The luteinizing hormone receptor (LHR) and follicle-stimulating hormone receptor (FSHR) have an ~350-amino acid-long, N-terminal extracellular exodomain. This exodomain binds hormone with high affinity and specificity and contains eight to nine putative Leu-rich repeat (LRR) sequences. LRRs are known to assume the horseshoe structure in ribonuclease inhibitors, and the inner lining of the horseshoe consists of the -stranded Leu/Ile-X-Leu/Ile motif. In the case of ribonuclease inhibitors, these  strands interact with ribonuclease. However, it is unclear whether the putative LRRs of LHR and FSHR play any role in the structure and function. In this work, the -stranded Leu/Ile residues in all LRRs of the human LHR and FSHR were Ala-scanned and characterized. In addition, the 23 residues around LRR2 of LHR were Ala-scanned. The results show that -stranded Leu and Ile residues in all LRRs are important but not equally. These Leu/Ile-X-Leu/Ile motifs appear to form the hydrophobic core of the LRR loop, crucial for the LRR structure. Interestingly, the hot spots are primarily in the upstream and downstream LRRs of the LHR exodomain, whereas important LRRs spread throughout the FSHR exodomain. This may explain the distinct hormone specificity despite the structural similarity of the two receptors.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The follicle-stimulating hormone receptor (FSHR)¹ and the luteinizing hormone/chorionic gonadotropin receptor (LHR) play crucial roles in reproduction of mammals, including humans. Human reproduction is a major concern as we have been trying to both control and promote human fertility. For example, humans are experiencing a dramatic population increase, yet ~16% of families suffer from infertility problems. FSHR is present in granulosa cells of the ovary (1) and Sertoli cells of the testis (2). In contrast, LHR is present in luteal and granulosa cells in the ovary and thecal cells in the testis. In the ovary, the expression of FSHR and LHR is dependent on the developmental stage of primordial, primary, secondary (preantral), small antral, large antral, and preovulatory follicles during the ovulation cycle (3). Therefore, the appearance, action, and disappearance of these receptors are crucially related with the granulosa cell differentiation, follicular development, and ovulation.

LHR and FSHR belong to the structurally unique glycoprotein hormone receptor subfamily of the G protein-coupled receptor family (4). Unlike other receptor subfamilies, they comprise two equal halves, an extracellular N-terminal half (exodomain) and a membrane associated C-terminal half (endodomain) (5-9) as shown in Fig. 1A. The exodomain is ~350 amino acids long and, alone, is capable of high affinity hormone binding (10-13) with hormone selectivity (14-16) but without hormone action (12, 17, 18). Hormone signal is generated in the endodomain (19), which is structurally equivalent to the entire molecule of many other G protein-coupled receptors such as rhodopsin and adrenergic receptors (4). Growing evidence suggests that glycoprotein hormones initially bind to the exodomain (4) and that the resulting hormone-exodomain complex undergoes a conformational change (20) and interacts with the endodomain. This secondary interaction is thought to generate signal in the endodomain (4, 19, 21). Interestingly, the high affinity hormone binding at the exodomain is modulated by the endodomain (22, 23). This suggests that the exodomain, in particular the hormone contact points, and endodomain are intimately related to each other during the initial and secondary interactions. To understand the underlying mechanism, it is essential to identify the exact hormone contact points in the exodomain.

Sequence comparison (Fig. 1, A and B) and modeling suggest that the exodomain contains imperfect Leu/Ile-rich repeats (LRRs) (5, 24-28), which are flanked by short N- and C-terminal regions. LRRs are a structural motif in a few crystal structures of proteins, namely ribonuclease inhibitors (29). However, the imperfect LRR motif is present in a large family of proteins and therefore could be a common structural motif. LRRs comprise divergent 22-29 amino acids with a short  strand consisting of two Leu and/or Ile residues in the alternate sequence (Leu/Ile-X-Leu/Ile), which is linked by linkers to  helices in parallel to the  strand. In addition, there are four hydrophobic residues at imperfectly matching positions. Therefore, the primary homology among LRRs is the Leu/Ile-X-Leu/Ile motif in  strands. The LRRs found in the crystal structure of ribonuclease inhibitors assume the three-quarter donut structure with the inner lining of  strands and the outer surface of  helices (29, 30). The inner lining of ribonuclease inhibitor interacts with ribonuclease at multiple contact points (31, 30). Based on the structure of ribonuclease inhibitors, the possibility has been raised that the inner lining of LRRs in LHR and FSHR may contact LH/CG and FSH (26-28, 32), in particular, with the concave side of the hormones (33).

Unfortunately, there is no experimental evidence for the existence of the LRR structure in LHR and FSHR and the LRRs' interaction with hormone. Mutational and photoaffinity analyses of LHR show that a number of regions and amino acids in the exodomain, particularly in the N-terminal half of the exodomain, either interact with hormone (34, 35) or are important for hormone binding (36-39). Interestingly, some of these hormone contact sites are outside of the LRRs (34, 35), whereas certain regions and ionic residues in the LRR region are directly or indirectly important for hormone binding (37, 39). Furthermore, the presence of Leu versus Ile in the eight Leu/Ile-X-Leu/Ile motifs is conserved except one residue between LHR and FSHR. This underscores the potential importance of LRRs in these receptors. These results, however, do not address whether putative LRRs and  strand Leu/Ile-X-Leu/Ile motifs are indeed involved in hormone binding. It is also enigmatic how the two receptors produce totally distinct specificities with the 94% homologous Leu//Ile-X-Leu/Ile motifs.

In this study, we examined the conserved  strand Leu/Ile residues of LRRs in the human LHR and FSHR. The results show that  strand Leu and Ile residues in all LRRs are important, suggesting multiple contact sites and crucial roles of LRRs in the receptors. Furthermore, some LRRs are more crucial than others and the crucial LRRs are primarily present in the upstream half LRRs in the LHR exodomain, in contrast to their presence in the C-terminal half LRRs of the FSHR exodomain. This suggests the distinct hormone binding specificity of the two receptors resides at spatially separate sites. In addition, the Ala scan of the sequence around LRR2 in the LHR shows that not only the Ile-X-Ile sequence in the  strand but also other conserved hydrophobic residues are indispensable for hormone binding. These results are compatible with the LRR model structure proposed by Bhowmick et al. (28) and show for the first time that LRRs in the gonadotropin receptors, particularly the  Leu/Ile-X-Leu/Ile motifs, play a crucial role.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Mutagenesis and Functional Expression of Receptors-- Each mutant human LHR and FSHR cDNA was prepared in a pSELECT vector using the non-polymerase chain reaction-based Altered Sites Mutagenesis System (Promega), sequenced, and subcloned into pcDNA3 (Invitrogen) as described previously (40). After subcloning pcDNA3, the mutant cDNAs were sequenced again. Plasmids were transfected into human embryonic kidney (HEK) 293 cells by the calcium phosphate method. Stable cell lines were established in minimum essential medium containing 10% horse serum and 500 µg/ml G-418 and then used for hormone binding and cAMP assays. In addition to stable cell lines expressing receptors, some receptors were transiently expressed using varying amounts of plasmid to express different concentrations of mutant receptors. All assays were carried out in duplicate and repeated four to six times. Means and standard variations were calculated.

Hormone Binding and Intracellular cAMP Assay-- hCG and human FSH were provided by the National Hormone and Pituitary Program and radioiodinated as described previously (41). Cells were assayed for ¹²⁵I-hormone binding in the presence of increasing concentrations of nonradioactive hormone. Scatchard plots determined the K_d values. For intracellular cAMP assay, cells were washed twice with Dulbecco's modified Eagle's medium and incubated in the medium containing isobutylmethylxanthine (0.1 mg/ml) for 15 min. Increasing concentrations of hCG were then added, and the incubation was continued for 45 min at 37 °C. After removing the medium, the cells were rinsed once with fresh medium without isobutylmethylxanthine, lysed in 70% ethanol, freeze-thawed in liquid nitrogen, and scraped. After pelleting cell debris at 16,000 × g for 10 min at 4 °C, the supernatant was collected, dried under vacuum, and resuspended in 10 µl of the cAMP assay buffer that was provided by the manufacturer. cAMP concentrations were determined with an ¹²⁵I-cAMP assay kit (Amersham Pharmacia Biotech) following the manufacturer's instruction and validated for use in our laboratory.

Radioimmunoassay for FLAG-G Receptor-- FLAG-LHR was prepared by inserting the FLAG epitope (42), Asp-Tyr-Lys-Asp-Asp-Asp-Asp-Lys (5'-GAC TAC AAG GAC GAT GAC GAT AAG-3'), between the C terminus (Ser²⁶) of the signal sequence and the N terminus (Arg²⁷) of mature receptors. We have shown that the FLAG tag did not significantly impair the activity of wild type and mutant LH/CG-Rs (34). Also, the FLAG epitope (43) has successfully been used as a marker to identify, trace, and purify recombinant proteins carrying the tag without significantly impairing their biological activities of other molecules (44, 45). Mouse anti-FLAG monoclonal antibody M2 (Eastman Kodak Co.) was iodinated with ¹²⁵I (41), and ¹²⁵I-anti-FLAG antibodies were purified on a Sephadex G-150 column. Binding of ¹²⁵I-anti-FLAG to HEK 293 cells expressing FLAG-LH receptors was carried out as described previously (34).

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The eight LRRs of LHR and FSHR contain the Leu/Ile-X-Leu/Ile motif, which constitutes the core of the putative  strands present in all LRRs of the glycoprotein hormone receptors and other LRR proteins (29). In fact, the presence of Leu versus Ile in the eight Leu/Ile-X-Leu/Ile motifs is conserved except one residue between LHR and FSHR. The exception is that the LRR1 motif is Leu-X-Leu in LHR and Leu-X-Phe in FSHR (Fig. 1, A and B). This 94% homology is striking, particularly in light of the less than 50% homology in the exodomain sequences (46) and the distinct hormone binding specificities of LHR and FSHR. In addition to the eight LRRs, there is putative LRR9 in the LHR (28), but it lacks the Leu/Ile-X-Leu/Ile motif and the Ala-Phe sequence. Therefore, we focused on LRRs 1-8 in this study.

View larger version (40K): [in this window] [in a new window]   Fig. 1.   Structure and sequence of the gonadotropin receptors. The LHR and FSHR consist of a ~350-amino acid-long extracellular domain (exodomain) and a membrane-associated domain (endodomain) of equal size. The exodomain is thought to include eight to nine LRRs, each consisting of a  strand and an  helix. The endodomain consists of three exoloops, three cytoloops, seven transmembrane domains, and the C-terminal tail. A, the exodomain sequence of human LHR is aligned based on the LRR motif with conserved residues in bold letters. B, likewise, the LRR sequences of human FSHR are aligned. C, a schematic model of the LRRs in the exodomain and the seven transmembrane domains in the endodomain.

Leu/Ile-X-Leu/Ile Motifs in LHR LRRs-- To examine the importance of the Leu and Ile residues in the putative  strands of LHR, they were individually substituted with Ala. Each of the Ala substitution mutants was stably expressed in HEK 293 cells and assayed for ¹²⁵I-hCG binding on intact cells or after solubilization in Triton X-100 and for hCG-dependent cAMP induction. For example, the cells transfected with the vector carrying the Leu²⁹  Ala mutant cDNA were incubated with a constant amount of ¹²⁵I-hCG in the presence of increasing concentrations of unlabeled hCG. The resulting displacement data (Fig. 2A) did not show hCG binding (Fig. 2B). To test whether the Leu²⁹  Ala mutant was capable of binding hCG but trapped in cells, the cells transfected with the Leu²⁹  Ala mutant vector were solubilized in Triton X-100 and assayed for ¹²⁵I-hCG binding (Fig. 2, C and D). Again, ¹²⁵I-hCG binding was not observed. Since the cAMP assay is significantly more sensitive than the ¹²⁵I binding assay, we tested if the cells produce cAMP in response to hCG. cAMP production was insignificant up to 10 nM hCG treatment (Fig. 2E). Although the cells produced cAMP, 2-5% of the maximum cAMP level of the wild type receptor, when treated with 100 nM, we considered it insignificant. This result, however, suggests the surface expression of the mutant receptor. We and others (28, 34, 47) have shown that mutant receptors that do not bind ¹²⁵I-hCG on intact cells and in solution are more likely to be defective in hormone binding than unexpressed. Therefore, mutant receptors that do not bind hCG on intact cells and in solution are generally considered to be incapable of binding hormone. However, to demonstrate their expression on the cell surface, antibodies were utilized. Because most of anti-LHR antibodies lack high affinity and specificity, the FLAG tag (42) was attached to the mutant receptor. The FLAG tag can be detected by monoclonal anti-FLAG antibody band, and it does not interfere with the expression and activity of receptors (34, 48). The FLAG-Leu²⁹  Ala mutant was expressed and assayed for binding monoclonal ¹²⁵I-anti-FLAG antibody (34). The antibody bound to the intact cells expressing the FLAG-Leu²⁹  Ala mutant as well as the FLAG-wild type receptor, indicating the surface expression of the FLAG-Leu²⁹  Ala mutant (Table I). On the other hand, the antibody did not bind to the cells expressing the wild type LH/CG-R lacking the FLAG tag nor the FLAG-Asp¹⁷  Ala that is known to be trapped in cells but not expressed on the cell surface (34). These results show that the Leu²⁹  Ala mutant was expressed on the surface but was defective in binding hCG. The fact that anti-FLAG antibody recognized the mutant indicates that the mutant was folded to a conformation similar to that of the wild type FLAG receptor, at least in the FLAG region.

View larger version (50K): [in this window] [in a new window]   Fig. 2.   Analysis of Leu and Ile in the  strands of LRR1-LRR4. Leu and Ile of the putative  strands in LRR1 and LRR2 of LHR were individually substituted with Ala, and the resulting mutant receptors were stably expressed on HEK 293 cells. Cells were assayed for 125I-hCG binding in the presence of increasing concentrations of nonradioactive hCG (A), and Scatchard analysis (B) was plotted against specific binding. Cells were also solubilized in Triton X-100 and assayed for 125I-hCG binding in the presence of unlabeled hCG (C), and the results were converted to Scatchard plots (D). In addition, intact cells were treated with increasing concentrations of unlabeled hCG, and intracellular cAMP was measured (E) as described under "Experimental Procedures." Experiments were repeated four to six times in duplicate. NS stands for not significant. F-J, the same as A-E for Leu/Ile residues in LRR3 and LRR4.

Next, the Leu³¹  Ala mutant was examined. ¹²⁵I-hCG bound to intact cells expressing the Leu³¹  Ala mutant with a K_d value of 690 pM as compared with the wild type value of 600 pM (Fig. 2, A and B). The mutant solubilized in Triton X-100 was also capable of binding the hormone with the affinity similar to the wild type affinity. It also produced cAMP in response to hCG, although the affinity and maximum cAMP level were significantly lower than the wild type receptor. These results show that these two Leu residues of LRR1 responded dissimilarly to Ala substitution. In contrast to the dissimilar sensitivity to Ala substitution for the two Leu residues of LRR1, Ala substitution for the two Ile residues in LRR2 abolished hCG binding on intact cells and in solution as well as cAMP induction, although the two mutants were detected on the cell surface by anti-FLAG antibody (Table I).

In LRR3, the Leu/Ile pair comprises Ile⁷⁸ and Ile⁸⁰. The Ile⁷⁸  Ala mutant was capable of binding ¹²⁵I-hCG and inducing cAMP (Fig. 2, F-J), but the affinities were lower than the wild type affinities. On the other hand, the Ile⁸⁰  Ala mutant was incapable of hormone binding and cAMP induction, despite its expression on the cell surface (Table I). In LRR4, both of the Leu¹⁰³  Ala and Ile¹⁰⁵  Ala mutants lost the receptor activities as did the Ile⁵³  Ala and Ile⁵⁵  Ala mutants of LRR2.

In LRR5 and LRR6, individual Ala substitutions for Leu¹²⁸, Ile¹³⁰, Leu¹⁵⁴, and Leu¹⁵⁶ partially impaired, but did not abolish, the hormone binding and cAMP induction (Fig. 3, A-E). Interestingly, the affinities of hormone binding to intact cells of these mutants were similar, approximately 2-fold lower than the wild type affinity. This result indicates similar impacts of the Ala substitutions. In addition, the binding affinity does not appear to be dependent on the receptor concentration on the cell surface. To test the relationship between the receptor concentration and the binding affinity, cells were transfected with varying concentrations of plasmids carrying the wild type receptor and Leu¹²⁸  Ala mutant. Cells expressing varying surface concentrations of the receptors showed similar K_d values (Table II). Ala substitutions for the LRR7 and LRR8 residues showed split results (Fig. 3). In LRR7, Leu¹⁷⁷  Ala partially impaired hormone binding and cAMP induction, whereas Leu¹⁷⁹  Ala entirely abolished them. The Leu²⁰²  Ala mutant of LRR8 completely lost the ability to bind the hormone and induce cAMP, but the effect was only partial for the Ile²⁰⁴  Ala substitution.

View larger version (53K): [in this window] [in a new window]   Fig. 3.   Analysis of Leu and Ile in the  strands of LRR5-LRR8. Leu and Ile of the  strands of LRR5-LRR8 were individually Ala-scanned and analyzed as described in the legend to Fig. 2.

Overall, the results show three distinct LRR groups. In LRR2 and LRR4, both residues of the Leu/Ile-X-Leu/Ile motif are sensitive to Ala substitution; only one residue is sensitive in LRR1, LRR3, LRR7, and LRR8; and neither is sensitive in LRR5 and LRR6. To readily compare various binding affinities of the mutant receptors, the wild type K_d value was divided by each mutant's K_d value (Fig. 4). The lower the K_d^wt/K_d^mut ratio implies the lower hormone binding affinity of the mutant. All of the mutants have a K_d ratio less than the unity, consistent with their lower binding affinities than the wild type affinity. Most of the K_d ratios are in the range of 0.2-0.5 except those incapable of hormone binding. This clearly indicates the significant impact of Ala substitution for the conserved Leu/Ile-X-Leu/Ile motif of the putative  strand of LRRs in LHR. It also suggests the importance of the motif in most of LRRs, in particular those in the region of LRR2, LRR3, and LRR4 where all Ala substitutions except Ile⁷⁸  Ala completely abolished hormone binding. The results also indicate that the partial and complete losses of cAMP induction by the Ala mutants were due to impaired hormone binding rather than defective cAMP induction, per se. There was no mutant receptor that was capable of binding hormone but incapable of inducing cAMP. This suggests that the Leu/Ile-X-Leu/Ile motifs enable hormone binding, not inducing cAMP.

View larger version (49K): [in this window] [in a new window]   Fig. 4.   Comparative binding affinities of LRR mutants: comparative binding affinities of LRR mutants. The wild type Kd value was divided with the Kd value of each mutant. Solid bars represent Kd values of receptors on intact cells and open bars for solubilized receptors.

Fidelity of Mutagenesis and Other Controls-- Our site-directed mutagenesis utilizes synthetic oligonucleotides containing a mutant sequence for each mutation and, furthermore, does not involve polymerase chain reaction. After mutagenesis, the mutant and flanking sequences are verified by sequencing. In addition, the same sequence is confirmed once more after a mutant cDNA is subcloned into the expression vector. Therefore, it is highly unlikely that a mutant cDNA might have undergone an unintended mutation(s) during the mutagenesis and subcloning. To confirm this mutation, cDNAs were reverted to the wild type cDNA, which, in turn, was used to transfect cells. The revertants behaved the same as the wild type receptor in surface expression, hormone binding, and cAMP induction (data not included), indicating that there was no mutation other than the intended substitutions. In addition, intact or solubilized HEK 293 cells that were transfected with the vector containing no receptor cDNA or that were not transfected did not bind hCG and FSH (data not included).

Sequence around Ile⁵³-X-Ile⁵⁵ in LRR2-- As a first step to examine crucial LRR2 and LRR4, amino acid residues (Val³⁸-Arg⁶³) flanking Ile⁵³-Glu⁵⁴-Ile⁵⁵ of LRR2 were Ala-scanned. Ala substitution for each of Val³⁸, Pro⁴⁰, Ser⁴¹, and Gln⁴² did not abolish hormone binding as did Ala substitution for Ile³⁹ (Fig. 5, A-D), although the affinities for hormone binding to Ala substituents on intact cells were similarly ~3-fold lower than the wild type affinity. In addition, the mutants were capable of inducing cAMP in response to hCG binding, indicating that the Ala substitutions were tolerable. In contrast, the Ile³⁹  Ala mutant failed to bind hCG both on intact cells and even after solubilization in Triton X-100, showing the mutant's inability of hCG binding. The next five Ala substituents behaved similarly to the first five Ala substituents. Arg⁴⁵  Ala, Gly⁴⁶  Ala, Leu⁴⁷  Ala, and Asn⁴⁸  Ala were tolerable for hCG binding with 2-4-fold lower affinities and cAMP induction, whereas Phe⁴⁴  Ala abrogated hCG binding and cAMP induction (Fig. 5, F-J). Among the next five Ala substituents, Glu⁴⁹  Ala, Val⁵⁰  Ala, Ile⁵¹  Ala, Lys⁵²  Ala, and Ile⁵³  Ala (Fig. 6, A-E), Ile⁵³  Ala was intolerable as Ile⁵³ was part of the Leu/Ile-X-Leu/Ile motif. On the other hand, the others were all tolerable. The results were the same for the next five Ala substitutions for Glu⁵⁴, Ile⁵⁵, Ser⁵⁶, Gln⁵⁷, and Ile⁵⁸ (Fig. 6, F-J). Ile⁵⁵  Ala was intolerable, whereas the other Ala substitutions reduced the affinities for hormone binding and cAMP induction. Ala substitutions for the last five residues produced different results. Asp⁵⁹  Ala and Arg⁶³  Ala were tolerable, whereas the Ala substitutions for the three tandem residues, Ser⁶⁰, Leu⁶¹, and Glu⁶², abolished hCG binding and cAMP induction. The Ala substitution mutants were again reverted to the wild type, and the revertants behaved the same as the wild type receptor.

View larger version (54K): [in this window] [in a new window]   Fig. 5.   Ala scan and characterization of residues around LRR2 of LHR. Amino acids from Val38 to Asn48 around LRR2 were individually substituted with Ala, and the resulting mutants were assayed as described in the legend to Fig. 2.

View larger version (54K): [in this window] [in a new window]   Fig. 6.   Ala scan and characterization of residues around LRR2 of LHR. Amino acids from Glu49 to Arg63 around LRR2 were individually substituted with Ala, and the resulting mutants were assayed as described in the legend to Fig. 2.

The comparative K_d^wt/K_d^mut ratios (Fig. 7) show that all Ala substitutions resulted in the reduction of hormone binding affinity by 2-fold or more. In particular, Ala substitution for Ile³⁹, Phe⁴⁴, Ile^53, Ile⁵⁵, Ser⁶⁰, Leu⁶¹, and Glu⁶² abolished hormone binding, although they were expressed on the cell surface according to the anti-FLAG antibody binding assay (data not shown). Interestingly, Ile³⁹, Phe⁴⁴, Ile⁵³, Ile⁵⁵, and Leu⁶¹ are conserved among all LRRs (Fig. 1B), whereas Ser⁶⁰ and Glu⁶² are not, an indication of the importance of the conserved hydrophobic residues. The result is consistent with the importance of the conserved residues and, therefore, the LRRs in the LHR. The fact that the conserved residues are located in the putative  strand and  helix signifies their structural role.

View larger version (28K): [in this window] [in a new window]   Fig. 7.   Comparative binding affinities of LRR2 mutants. The wild type Kd value was divided with the Kd value of each mutant described in the legends to Figs. 5 and 6. Solid bars represent Kd values of receptors on intact cells and open bars for solubilized receptors.

Leu/Ile-X-Leu/Ile Motifs in LRRs of FSHR-- The Leu/Ile-X-Leu/Ile sequences of FSHR are exactly the same as those of LHR except LRR1. In contrast to this near perfect sequence, the sequence homology of the exodomains of the two receptors is less than 50% (46), the selectivity for LH/CG and FSH is nearly 100%, and the affinity is extremely high with the K_d value of ~500 pM. To understand these seemingly contrasting results and determine whether the nearly identical Leu/Ile-X-Leu/Ile sequences of the two receptors possibly contribute distinguish the two hormones, we examined the LRRs of FSHR. In LRR1, the Leu³²  Ala mutant on intact cells and after solubilization was capable of binding ¹²⁵I-FSH (Fig. 8, A-D). Interestingly, the K_d value was 1,010 pM as compared with the wild type K_d value of 2,420 pM, indicating a 2-fold improvement in the hormone binding affinity. However, cAMP induction was marginal (Fig. 8E). The Ala substitution for Phe³⁴ abolished FSH binding and cAMP induction (Fig. 8). The cells transfected with the mutant cDNA plasmid did not bind ¹²⁵I-FSH nor those solubilized in Triton X-100, although the FLAG-Phe³⁴  Ala mutant was detected by ¹²⁵I-anti-FLAG antibody on the cell surface (data not shown).

View larger version (56K): [in this window] [in a new window]   Fig. 8.   Ala scan and characterization of the  strand residues of FSHR. Leu, Ile, and their equivalent residues of the  strands of LRR1-LRR8 of FSHR were individually Ala-scanned and analyzed as described for LRR residues of LHR in the legend to Fig. 2.

In LRR2, the Ile⁵⁸  Ala mutant, but not the Ile⁶⁰  Ala mutant, was capable of binding hormone and inducing cAMP (Fig. 8, A-E). The binding affinity of the Ile⁵⁸  Ala mutant was nearly 3-fold better than the wild type receptor, but the cAMP induction was marginal, as was the Leu³²  Ala mutant of LRR1. The Ala substitution of the second residue of LRR2, Ile⁶⁰  Ala, abolished hormone binding and cAMP induction, as did the Ala substitution for the second residue of LRR1, Phe³⁴  Ala.

In LRR3 the Ile⁸³  Ala and Ile⁸⁵  Ala mutants failed to bind the hormone and induce cAMP. In LRR4 the result of the Ala substitutions was the same as the result of Ala substitutions for LRR1 andLRR2 residues. For example, the Leu¹⁰⁸  Ala mutant bound FSH and induced cAMP, whereas the Ile¹¹⁰  Ala did not bind the hormone nor induced cAMP. In LRR5 the two Ala substitution mutants were capable of binding the hormone and cAMP induction (Fig. 8, F-J). However, their binding affinities and cAMP induction were quite different. The Ala substitutions for Leu¹⁵⁸ and Leu¹⁶⁰ of LRR6 had split results. The Leu¹⁵⁸  Ala mutant was incapable of binding the hormone and inducing cAMP, whereas the Leu⁶⁰  Ala mutant bound FSH and induced cAMP. In LRR7 and LRR8 none of the Ala mutants were capable of binding FSH and inducing cAMP.

Taken together, Ala substitutions for the Leu/Ile-X-Leu/Ile motifs in LRRs significantly impact the hormone binding and/or cAMP induction of FSHR. Interestingly, some of the Ala substitutions noticeably improved the hormone binding affinity. However, the ability to induce cAMP by the mutants was consistently impaired. The K_d ratios of the wild type receptor and individual mutants (Fig. 4B) show that LRR5, LRR6, LRR7, and LRR8 in the C-terminal region of the exodomain were more severely impaired than those LRRs in the N-terminal regions of the exodomain were.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

In this study, we set out to determine whether putative LRRs and  strand Leu/Ile-X-Leu/Ile motifs are indeed required for hormone binding. In addition, we hoped to learn how the 94% homologous Leu//Ile-X-Leu/Ile motifs of LHR and FSHR might contribute to their distinct hormone specificities.

Ala substitutions for all Leu and Ile residues, except Leu³¹, in the LRR  strands impaired the hormone binding activity of LHR, although the Ala mutants were expressed on the cell surface. Some Ala substitutions abolished hormone binding, whereas others had less significant effects as shown in the computer models (Fig. 9, A and B). These results suggest that the Leu and Ile residues are important, although in varying degree. Since the conserved and important residues are spread around an LRR (Fig. 1A), some of them are probably necessary for the structural integrity of LRRs. The examination of the residues around the Ile-X-Ile sequence of LRR2 reveals the importance of the two alternate Ile residues, since the Ala substitutions for the intervening and flanking residues had less dramatic or marginal effects (Fig. 9C). In addition to the Ile-X-Ile sequence, Ala substitutions for Ile³⁹, Phe⁴⁴, Ile⁵³, Ile⁵⁵, and Leu⁶¹, which are conserved among LRRs, resulted in the complete loss of the hormone binding without exception. These residues form the hydrophobic core of the LRR2. In contrast, the substitution effects of other residues were less significant. These results support the LRR structure in LHR and its significant role in hormone binding.

View larger version (27K): [in this window] [in a new window]   Fig. 9.   Exodomain models. For models A for LHR, B for FSHR, and C for LRR2 of LHR, the receptor exodomain was homology-modeled, as described previously (28), based on the coordinates of ribonuclease inhibitor from the Brookhaven data base as the template with the O and Frodo software packages (51) using an Indigo 2-Extreme (Silicon Graphics) and easv10 (Evans and Sutherland). They were modified based on the experimental data presented in this study. In each case, yellow indicates an natural residue, red indicates Ala substitutions with a maximum effect on hormone binding (Kd of hormone binding to cells) and/or expression, blue indicates a minimum effect on hormone binding, and colors in between (red > crimson > magenta > blue) indicate the effects between these extremes. Note that in this scheme the wild type FSH receptor would be a crimson color. FSH apparently binds better to the mutant receptors, so these are going to be bluer than the wild type. LRR 2 of LHR is shown with the green  strand and red  helix in C.

The next question is whether or not the crucial LRRs interact with the hormone. The results of this study could not provide a direct answer. The LRR2 study, however, shed some light on the issue. The residues that are most sensitive to Ala substitution are hydrophobic and appear to form the hydrophobic core of LRR2. Since an LRR assumes an oblong loop (Fig. 9), it is unlikely that all sides of an LRR contact with the hormone. In particular, the hydrophobic core residues, including the conserved Ile⁵³ and Ile⁵⁵ in the  strand, are unlikely to interact with the hormone. If an LRR interacts with hormone at all, the residues involved will be present in the inner lining of the LRR horseshoe or on the linkers. In fact, our model predicts that Lys⁵², Glu⁵⁴, Ser⁵⁶, and Gln⁵⁷ are to face the hormone. However, the Ala substitutions for these residues did not completely impair the hormone binding capacity. Therefore, they do not appear to provide the major force to the hormone-receptor interaction.

One may raise an issue whether the Ala substitutional effects were due to changes in the global structure of LHR rather than local structure around the substituted residue. This is a general and contentious issue concerning the interpretation of the result from a substitutional study. Although it is difficult to test for an unequivocal answer, there is evidence supporting local conformational changes induced by substitution of a single amino acid. For example, crystal structures of other proteins show that single amino acid substitutions generally cause subtle and local conformational alterations (49, 50). In our hand, mutations did not impact the recognition of the FLAG epitope at the exodomain N terminus by anti-FlAG antibodies (34, 48). The previous studies also suggest that the targeting machinery of LHR is capable of detecting some conformational changes, which hormone binding could not. This indicates that the machinery is more sensitive to such conformational changes of LHR than the hormone binding could recognize. However, in the absence of foolproof evidence and conformational antibodies, it is premature to dismiss the potential interference of misfolding on the local and global conformation of the mutants.

Despite the 94% homology in the Leu/Ile-X-Leu/Ile sequences between the two receptors, LHR and FSHR are capable of recognizing their cognate ligands, LH/CG and FSH, respectively, with a high affinity and no cross-reactivity. Therefore, it is necessary to know how the homologous Leu/Ile-X-Leu/Ile sequences in the putative hormone contact sites contribute to the discrete hormone specificity. The K_d ratios in Fig. 4 provide some answers, at least in part. For example, the location of the important LRRs is distinct. The upstream LRRs of LHR are more sensitive to Ala substitution than the downstream LRRs, whereas the downstream LRRs of FSHR are more sensitive. This suggests different contact points in LRRs of the two receptors. The role of LRRs may also differ as Ala substitutions improved the binding affinity of some FSHR mutants, whereas Ala substitutions attenuated the binding affinity of LHR. This result suggests a clear difference in the structural role of LRRs in the two receptors, which could explain their dramatically distinct hormone binding specificities. Despite the higher binding affinities of the FSHR mutants and the lower binding affinities of the LHR mutants, the EC₅₀ values for cAMP induction increased for both receptors. Therefore, both receptors appear to share some common mechanism to generate the cAMP signal.

	FOOTNOTES

^* This work was supported by Grants HD-18702 and DK-51469 from the National Institutes of Health.The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

To whom all correspondence should be addressed: Dept. of Chemistry, University of Kentucky, Lexington, KY 40506-0055. Tel.: 859-257-3163; Fax: 859-527-3229; E-mail tji@pop.uky.edu.

Published, JBC Papers in Press, July 3, 2000, DOI 10.1074/jbc.M003772200

	ABBREVIATIONS

The abbreviations used are: FSH, follicle-stimulating hormone; FSHR, FSH receptor; h, human; CG, chorionic gonadotropin; LH, luteinizing hormone; LHR, LH receptor; LRR, Leu-rich repeat; HEK, human embryonic kidney.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES