Follicle-stimulating Hormone Interacts with Exoloop 3 of the Receptor*

The human follicle-stimulating hormone (FSH) receptor consists of two distinct domains of ∼330 amino acids, the N-terminal extracellular exodomain and membrane-associated endodomain including three exoloops and seven transmembrane helices. The exodomain binds the hormone with high affinity, and the resulting hormone/exodomain complex modulates the endodomain where receptor activation occurs. It has been an enigma whether the hormone interacts with the endodomain. In a step to address the question, exoloop 3 of580KVPLITVSKAK590 was examined by Ala scan, multiple substitution, assays for hormone binding, cAMP and inositol phosphate (IP) induction, and photoaffinity labeling. We present the evidence for the interaction of FSH and exoloop 3. A peptide mimic of exoloop 3 specifically and saturably photoaffinity-labels FSH α but not FSH β. This is in contrast to photoaffinity labeling of FSH β by the peptide mimic of the N-terminal region of the receptor. Leu583 and Ile584 are crucial for the interaction of FSH and exoloop 3. Substitutions of these two residues enhanced the hormone binding affinity. This is due to the loss of the original side chains but not the introduction of new side chains. The Leu583 and Ile584 side chains appear to project in opposite directions. Ile584 appears to be so specific and to require flexibility and stereo specificity so that no other amino acids can fit into its place. Leu583 is less specific. The improvement in hormone binding by substitutions was offset by the severe impairment of signal generation of cAMP and/or inositol phosphate. For example, the Phe or Tyr substitution of Leu583 improved the hormone binding and cAMP induction but impaired IP induction. On the other hand, the substitutions for Ile584 and Lys590abolished the cAMP and IP induction. Our results open a logical question whether Leu583, Ile584, and Lys590 interact with the exodomain and/or the hormone. The answers will provide new insights into the mechanisms of hormone binding and signal generation.

The FSH1 receptor (FSHR) 1 and other glycoprotein hormone (luteinizing hormone/chorionic gonadotropin and thyroid-stim-ulating hormone) receptors belong to a structurally unique subfamily of G protein-coupled receptors. Unlike other receptor subfamilies, they comprise two equal halves, an N-terminal extracellular half (exodomain) and a C-terminal membraneassociated half (endodomain) (1)(2)(3)(4). The exodomain is ϳ350 amino acids long and alone is capable of high affinity hormone binding (5)(6)(7)(8) with hormone selectivity (9 -11) but without hormone action (7,12). Receptor activation occurs in the endodomain (13), which is structurally equivalent to the entire molecule of many other G protein-coupled receptors (14). Glycoprotein hormones initially bind to the exodomain, and then the resulting hormone/exodomain complex modulates the endodomain (13), which activates adenylyl cyclase (AC) to generate cAMP and phospholipase C␤ (PLC␤) to produce inositol phosphate and diacylglycerol. Therefore, the ternary interactions among the hormone, exodomain, and endodomain are crucial for successful signal generation. However, there is little information on the subject, particularly concerning the FSH receptor. Since the exodomain lacking the endodomain is capable of high affinity hormone binding (5)(6)(7)(8)15), the high affinity hormone binding appears to be independent of the endodomain. Contrary to this view, it has been reported that FSH and human chorionic gonadotropin binding to their cognate receptors is regulated by certain residues of exoloops 2 and 3 of the endodomain (15,16). Furthermore, the hinge region of the exodomain interacts with exoloop 2 and modulates cAMP induction (17)(18)(19). These results suggest that the exodomain interacts with the exoloops and modulates them for signal generation. Yet it is unclear whether the hormone complexed with the exodomain also contacts the exoloops.
In this study, we set out to investigate whether exoloops interact with the hormone at all. In a step toward this goal, we examined exoloop 3 of FSHR for its involvement in the activation of the two effectors and interaction with FSH. It is the shortest of the three exoloops, consisting of 11 amino acids, and has been implicated in the cAMP signal generation (15,20,21). Our observations show, for the first time, the interaction of exoloop 3 with FSH, in particular the FSH ␣ subunit, the mode of this interaction, and its role in the AC and PLC␤ activation.

Mutagenesis and Functional Expression of Human FSH Receptor-
Each mutant human FSHR cDNA was prepared in a pSELECT vector using the nonPCR-based Altered Sites mutagenesis system (Promega), sequenced on a Beckman CEQ 2000XL capillary sequencer, and subcloned into pcDNA3 (Invitrogen) as described (22). 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 g/ml G-418 and then used for hormone binding and cAMP assay. All assays were carried out in duplicate and repeated 4 -6 times. Means and standard variations were calculated. scribed previously for radioiodination of human chorionic gonadotropin (23). Denatured FSH was prepared by boiling in 8 M urea for 30 min. Stable cells were assayed for 125 I-FSH binding in the presence of increasing concentrations of nonradioactive FSH. The K d values were determined by Scatchard plots. Truncated exodomain was solubilized in Nonidet P-40 and assayed for hormone binding as described previously (16). 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 FSH 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 ϫ g for 10 min at 4°C, the supernatant was collected, dried under vacuum, and resuspended in 10 l of the cAMP assay buffer, which was provided by the manufacturer. cAMP concentrations were determined with an 125 I-cAMP assay kit (Amersham Biosciences) following the manufacturer's instructions and validated for use in our laboratory. Exoloop 3 of FSHR was modeled based on the crystal structure of rhodopsin as a template (24).
Inositol Phosphate Assay-Stable cells were plated in 12-well plates and grown in inositol-free Dulbecco's modified Eagle's medium (Atlanta Biologicals) supplemented with 8% heat-inactivated horse serum and 2 Ci/ml [ 3 H]inositol (PerkinElmer Life Sciences) for 48 h to 40 -50% confluency. After removing the medium, the cells were incubated in 1 ml of fresh wash buffer consisting of Dulbecco's modified Eagle's medium without inositol and 15 mM HEPES (pH 7.3) for 1 h at 37°C. This medium was removed, and 0.3 ml of wash buffer containing 20 mM LiCl was added and incubated for 15 min at 37°C. After the cells were stimulated with increasing concentrations of hormone for 30 min at 37°C, the incubation was terminated by the removal of medium and the addition of 0.25 ml of 0.6 N HCl to each well. The cells were scraped and transferred into microcentrifuge tubes, and the wells were again washed with 0.25 ml of 0.6 N HCl. The combined washes were treated with 0.9 ml of a mixture of chloroform:methanol (2:1), vortexed, and centrifuged at 1000 ϫ g for 5 min at room temperature. The top aqueous layer, which was free of phospholipids, was removed, and the remaining chloroform layer was treated with 0.2 ml of methanol:water (1:1), vortexed, and centrifuged, as above. This aqueous layer was added to the previous aqueous layer, and the samples were dried in a vacuum concentrator. The dried samples were redissolved in 0.5 ml of 50 mM Effects of Ala substitutions on IP production, K d , and cAMP induction. As shown in A, the exoloop 3 amino acids, 580 KV-PLITVSKAK 590 , were individually substituted with Ala except the A588G substitution, and the mutant receptors were assayed for IP t , IP 1 , IP 2 , and IP 3 as described under "Experimental Procedures." As shown in B, the ratios of K d wt/mut (blank bar), maximum cAMP mut/wt (gray bar), and IPt mut/wt (black bar) of the mutants were presented in bars. The ratios above 1.0 indicate that the binding affinity of the mutants is better than the wild type affinity, and the maximum cAMP and IP levels of the mutants are higher than those of the wild type.  Procedures." For hormone binding, counts of empty tubes (background) were ϳ50 CPM, and nonspecific binding was ϳ75 CPM including background. Maximum specific binding CPM are normally in the range of 1,400 -500 CPM. Nontransfected cells did not show specific binding of FSH. Each experiment was performed in duplicate, and values were determined for K d , receptors/cell, EC 50 for cAMP synthesis, and maximum cAMP level. After experiments were repeated 6 -10 times, the means and standard deviations were calculated. NS indicates not significant.
Tris-HCl, pH 8, and applied to Dowex AG 1-X8 formate (Bio-Rad) columns. The microcentrifuge tubes were washed twice with 0.5 ml of the same buffer, and the washes were applied to the columns for a total of 1.5 ml. The columns were sequentially washed with 4.5 ml of H 2 O and 4.5 ml of 60 mM ammonium formate and 5 mM sodium tetraborate to elute the free inositol and the glycerol phosphoinositol. IP 1 , IP 2 , and IP 3 were sequentially eluted with 4 ml of 0.1 N formic acid in 0.2, 0.75, and 1.1 M ammonium formate, respectively, and collected in 1-ml fractions. Aliquots of 200 l were counted for radioactivity in 1.5 ml of Ultima AF scintillation fluid (Packard). Peak radioactivities were used for the data analysis.
Derivatization and Radioiodination of Peptide-A peptide mimic corresponding to the exoloop 3 sequence of 580 KVPLITVSKAK 590 (FSHR exo3 ) was synthesized, to which a Tyr residue was attached to the C terminus for radioiodination. The N terminus of the peptide was acetylated, and the C terminus was amidated. NHS-ABG was synthesized as described previously (23) and freshly dissolved in dimethyl sulfoxide to a concentration of 50 mM and NHS-ABG was freshly dissolved in 0.1 M sodium phosphate (pH 7.5) to a concentration of 20 mM. These reagent solutions were immediately used to derivatize receptor peptides. In the dark, 10 l of NHS-ABG was added to 30 g of receptor peptides in 40 l of 0.1 M sodium phosphate (pH 7.5). The mixture was incubated with NHS-ABG for 30 min at 25°C. The following were added to the derivatization mixture: 1 mCi of Na[ 125 I]iodine in 10 l of 0.1 M NaOH and 7 l of chloramine-T (1 mg/ml) in 10 mM Na 2 HPO 4 , pH 7.4. After 20 s, 7 l of sodium metabisulfite (2.5 mg/ml) in 10 mM Na 2 HPO 4 , pH 7.4, was introduced to terminate radioiodination. Derivatized and radioiodinated ABG-125 I-FSHR exo3 solution was mixed with 60 l of 16% sucrose solution in PBS and fractionated on a Sephadex Superfine G-10 column (0.6 ϫ 15 cm) using PBS.
Photoaffinity Labeling of FSH-The following solutions were sequentially introduced to siliconized glass tubes: 20 l of 0.9% NaCl and 10 mM Na 2 HPO 4 , pH 7.4 (PBS), 10 l of FSH (10 ng/l) in PBS, and 10 l of ABG-125 I-FSHR exo3 (10 ng/l) in PBS. Competitive inhibition experiments were carried out as described for the photoaffinity labeling experiments except that 10 l instead of 20 l of PBS was introduced to each tube, and the mixture was incubated with 10 l of increasing concentrations of nonradioactive receptor peptides. The mixtures were incubated at 37°C for 90 min in the dark, irradiated with Mineralight R-52 UV lamp for 3 min as described previously (25), and solubilized in 2% SDS, 100 mM dithiothreitol, and 8 M urea. The samples were electrophoresed on 8 -12% polyacrylamide gradient gels. Gels were dried on filter paper and exposed to an imaging screen overnight, which was scanned on a phosphorimaging device.
Deglycosylation-The FSH ␣ and ␤ subunits co-migrate on SDS-PAGE. To separate them on the gel, FSH was deglycosylated with PNGase F before and after it was photoaffinity-labeled. Enzymatic cleavage was done by incubation of the labeled FSH complex with 20 or 50 units of PNGase F (New England Biolabs, Inc., Beverly, MA) in 40 l for 18 h at 37°C. The samples were solubilized in SDS under the reducing condition and electrophoresed on 15% gel containing 9 M urea.

RESULTS
Effects of Ala Substitutions on Production of IP 1 , IP 2 , and IP 3 -FSHR exoloop 3 consists of 11 amino acids, 580 KV-PLITVSKAK 590 , which are conserved among species except Ala 589 (Fig. 1). The sequence is also conserved among the glycoprotein hormone receptor family except 587 SKA 589 near the C terminus. The previous Ala scan has demonstrated that exoloop 3 constrains the hormone binding at the exodomain and plays a crucial role in cAMP induction (15). However, little is known about the mechanism or its role in IP induction. To address these questions, the Ala substituents of individual residues, except for the A589G substitution, were stably expressed on HEK 293 cells. The cells were assayed for inositol phosphates, IP 1 , IP 2 , IP 3 , and IP t . Most of the mutant receptors, except the V581A,P582A substitutions, were incapable of inducing IP production in response to FSH ( Fig. 2A). In contrast, V581A was capable of producing noticeable levels of IP 1 and IP 2 and an insignificant level of IP 3 . P582A produced a detectable level of IP 1 but not IP 2 and IP 3 . These results raise the question of whether the non-responding mutant receptors were expressed on the cell surface in this study, although they were in the previous study (15). Therefore, the cells stably transfected with the mutants were assayed for 125 I-FSH binding as well as FSH-dependent cAMP production.
Distinct Effects of Ala Substitutions on Hormone Binding, IP Induction, and cAMP Induction-For easy comparison of the data, the ratios of K d wild type/mutant (K d wt/mut ), maximum IPt mut/wt , and maximum cAMP mut/wt were calculated (Fig. 2B).
The results show that all of the cells bound the hormone, indicating the surface expression of the mutant receptors. Of interest is that the L583A and I584A mutations improved the hormone binding affinity by 2-3-fold. This is in striking contrast to the loss of IP induction by most of the mutants except the V581A and P582A mutants. On the other hand, the mutational effect is less severe on the activation of adenylyl cyclase to produce cAMP. For example, most of the mutants were capable of producing some cAMP, although less than the wild type. The three mutants, L583A, I584A, and K590A, did not produce cAMP. Therefore, the activation of PLC␤ is more sensitive to Ala substitution than is the activation of AC and hormone binding. The results also show different mechanisms, in particular the sites, of the PLC␤ activation, AC activation, and hormone binding. We cannot, however, unequivocally dismiss the possibility that the lack of the IP induction was due to the limitation of the detectable IPs. To visualize the difference, exoloop 3 was computer-modeled (Fig. 3A). The results showed the contrasting topography of the sensitive residues for the signal generation and hormone binding. The residues crucial for the PLC␤ signal cover most of exoloop 3 except the Nterminal region (Fig. 3B). On the other hand, the residues sensitive to the AC signal are confined in the middle and C terminus of the exoloop (Fig. 3C). Leu 583 and Ile 584 are most sensitive to hormone binding and are located near the middle of the exoloop (Fig. 3D). Their side chains protrude in opposite directions. The sensitive residues appear to be accessible from one side of the exoloop, suggesting the intriguing possibility that they might be modulated from the side of the exoloop by the exodomain and/or the hormone. Particularly, Leu 583 and Ile 584 are sensitive to all of the three functions: hormone binding, PLC␤ activation, and AC activation. In addition to Leu 583 and Ile 584 , Lys 590 is important to the activation of PLC␤ and AC.
Multiple Mutational Analysis and Specificity of Leu 583 , Ile 584 , Lys 590 , and Pro 582 -The Ala substitution for Leu 583 or Ile 584 enhanced the hormone binding affinity by 2-3-fold but impaired signal generation for IP and cAMP. These two residues have large hydrophobic side chains, which are exposed on the surface according to the computer model (Fig. 3). Hydrophobic side chains are generally incompatible with surface exposure, especially to water molecules. However, surface exposure provides a hydrophobic contact site. To examine the TABLE I IPt production of Leu 583 mutants Leu 583 was substituted with Tyr, Phe, Ala, Glu, Arg, or deleted and the mutants were expressed on HEK 293 cells. All of the mutants were expressed on the cell surface and bound 125 I-FSH as described in Fig. 4. The assay for total IP production in response to increasing concentrations of FSH showed that only the L 583 Y mutant was capable of producing IP t . NS, not significant. roles of the side chains, Leu 583 and Ile 584 were substituted with a panel of amino acids with various side chains: negative or positive, hydroxyl, neutral, ring or aliphatic groups. In addition, the residues were deleted in deletion mutants, which is helpful in assessing the effect of removing the original side chain without introducing a new side chain. The first question raised was whether the improved binding affinity upon the Ala substitution was due to the introduction of the Ala side chain or loss of the original side chains. As shown in Fig. 4, A, B, D, and E, all of the substitutions of Leu 583 decreased the K d values, thus improving the binding affinity. Even when Leu 583 was deleted, the affinity improved by more than 4-fold, suggesting that the loss of the Leu 583 side chain contributes to the improved binding affinity. In contrast to the improvement in hormone binding, most of the mutants did not induce noticeable amounts of cAMP (Fig. 4C). Exceptions were the L583F and L583Y mutants that induced significant amounts of cAMP with reasonable EC 50 values. Besides the L583F and L583Y mutants, the L583Q mutant induced a marginal level of cAMP (Fig. 4F). These results suggest the need for a specific group such as the Leu side chain or a ring group for the AC activation. When the Tyr, Phe, Ala, Glu, Arg, and deletion mutants were assayed for IP t production, only the L583Y mutant produced a small amount of IP t (Table I). This result suggests a similarity in the interactions to activate AC and PLC␤. All substitutions for and deletion of Ile 584 decreased the K d values by up to 5-fold (Fig. 5). In contrast to this improved hormone binding, none of the mutants induced cAMP. Therefore, the deletion of Ile 584 , not the introduction of new side chains, was likely responsible for the improved binding affinity. Furthermore, Ile 584 appears to be crucial for cAMP induction (Fig. 5C). To further test these hypotheses, the adjacent Pro 582 was substituted with various amino acids (Fig. 6). The muta-tional effect on hormone binding was diverse and less dramatic. Some mutations decreased the K d value, whereas others increased it. Several mutants were capable of inducing cAMP, whereas several others failed to induce the second messenger. Clearly, Pro 582 appears to play a role different from those of Leu 583 and Ile 584 . In addition to Pro 582 , Lys 590 was examined with multiple substitutions. Lys 590 is located at the far end of exoloop 3, at the boundary with the transmembrane 7. The K d values of the substituents varied widely from 1.6 to 50 nM as compared with the wild type value of 4 nM, whereas none of the mutants induced significant amounts of cAMP (Fig. 7). Therefore, Lys 590 is also essential for activation of AC. In addition, the mutants with Cys, Phe, Leu, Tyr, Ala, and Arg substitutions for Lys 590 were assayed for the production of IP 1 , IP 2 , IP 3 , and IPt. None of the mutants induced significant amounts of any IP species (data not included).
Photoaffinity Labeling of FSH with Exoloop 3 Peptide-To test the possible interaction of exoloop 3 with the hormone, 125 I-ABG-FSHR exo3 was incubated with FSH and irradiated with UV for increasing time periods. Samples were solubilized in SDS under the reducing condition and electrophoresed. The autoradiographic phosphorimage of the gel (Fig. 8A) revealed the labeling of the FSH band. The two subunits of the human FSH preparation comigrate on SDS-PAGE. The band was not labeled when the sample was not irradiated with UV, suggesting the requirement for UV irradiation. The extent of the labeling was dependent on the irradiation time, reaching the maximum after 60 s of irradiation. The results show that the labeling is saturable.
To determine the nature of the labeling, increasing concentrations of the hormone were labeled with a constant amount of 125 I-ABG-FSHR exo3 (Fig. 8B). Conversely, increasing concentrations of 125 I-ABG-FSHR exo3 were used to label a constant amount of FSH (Fig. 8C). The labeling plateaued under both conditions, indicating saturable labeling. To examine the relationship of the labeling with other exoloops and receptor peptide, FSH was incubated with 125 I-ABG-FSHR exo3 in the presence of increasing concentrations of unlabeled FSHR peptides corresponding to exoloops 1, 2, and 3 as well as the N-terminal sequence Ser 9 -Lys 40 , FSHR 9 -40 , which is known to interact with FSH. 2 Increasing concentrations of the peptides inhibited the photoaffinity labeling in a dose-dependent manner and eventually blocked the labeling with varying affinity (Fig. 9), suggesting a specificity. FSHR exo2 is the most potent inhibitor, suggesting the possibility of its strong interaction with the hormone. Furthermore, 125 I-ABG-FSHR exo3 failed to label denatured FSH that does not bind to the receptor, despite high concentrations of the peptide (Fig. 10A), suggesting the speci-ficity of the affinity labeling for biologically active FSH. FSH was denatured by boiling in 8 M urea for 30 min. To test whether the denatured FSH remained in solution, the mixture of radioactively labeled FSH and unlabeled FSH was denatured, varying volumes of the mixture were transferred to other tubes, and the radioactivity was counted. The transfer was quantitative with a 99 -100 efficiency, indicating that denatured FSH was present in the photoaffinity labeling tube. 125 I-ABG-FSHR exo3 did not label urokinase, nor did it label phospholipases A, C, and D (Fig. 10B). In addition, it failed to noticeably label human growth hormone (Fig. 10B). The exoloop 3 peptide inhibited 125 I-FSH binding to the receptor on intact cells in a dose-dependent manner. These results show that the binding to and labeling of FSH by the peptide were specific to bioactive FSH.
Since the ␣ and ␤ subunits of human FSH comigrate on SDS-PAGE, it is unclear which of the subunits was labeled. To determine the identity of the labeled subunit(s), FSH was labeled with 125 I-ABG-FSHR exo3 , deglycosylated with PNGase F, and electrophoresed. The labeled band corresponded to the ␣ subunit (Fig. 10B). Deglycosylated human FSH separates into two bands on SDS-PAGE, the higher molecular weight ␤ subunit in the upper band and the smaller ␣ subunit in the lower band, which was verified by monoclonal antisubunit antibodies. 2

DISCUSSION
Our results show that the exoloop 3 is crucially, yet differently, involved in hormone binding and induction of cAMP and IP. They show that FSHR exoloop 3 contacts the ␣ subunit of FSH as part of the ternary complex consisting of FSH, the exodomain, and endodomain. Particularly, Leu 583 and Ile 584 are more important than other amino acids and project from one side of exoloop 3 in opposite directions. Interestingly, the substitutions of the two residues significantly improved the hormone binding, which is due to the loss of the original side chains rather than the introduction of the side chains from the substitutions. For example, all substitutions of Ile 584 with Tyr, Pro, Ala, Cys, Ser, Gln, Asp, Glu, and Arg enhanced the hormone binding affinity, as did the deletion of Ile 584 as shown by the K d wt/mut ratios in Fig. 11A. To analyze the nature of the effect, the side chain hydrophobicity (27, 28) of substituting amino acids was plotted against the K d values. The plot showed  (Table II). The first group had a negative hydrophobic effect on the binding affinity, whereas the second had a positive hydrophobic effect, suggesting a complex, specific microenvironment and interaction of the Ile 584 side chain. The interaction appears to be partly hydrophobic and may involve other specificity such as stereospecificity.
Substitutions of Pro 582 show diverse results, independent of the hydrophobicity of the side chain. On the other hand, substitutions of Lys 590 showed two distinct groups (Fig. 11B): the first group of FYQEDR with the most severe hydrophobicity/K d coefficient of Ϫ8.1 and the second group of LCAPSEDR with a coefficient of Ϫ0.87 (Table II). In the first group, ring groups such as the phenyl and phenolic side chains of Phe and Tyr, respectively, severely impaired the binding affinity, considerably more than any substitutions of Leu 583 and Ile 584 did. The second group also negatively impacted the binding affinity, but the effects were mild. A striking difference of the two groups is the substitutions with Phe and Leu. The K d value of FSHR K590F was 50 nM in contrast to 7.2 nM for FSHR K590L , raising a question of whether the side chain flexibility and geometry play a role. The deletion of Lys 590 , thus, likely relieves the constraint and improves the K d value as shown by the deletion mutant.
All of the substitutions, except L583Y, diversely impaired the cAMP induction (Fig. 11C). This adverse impact was seen the least on the substitutions of Pro 582 . Among the various Leu 583 mutants, only L583F and L583Y were capable of inducing cAMP production. This suggests that a hydrophobic side chain larger than a methyl group is necessary at the position, regardless of whether the side chain is aliphatic or aromatic. This is in contrast to the adverse effect of a ring group at Lys 590 on the hormone binding, clearly indicating the requirements for distinct groups at Leu 583 and Lys 590 . All other substitutions of Leu 583 lead to insignificant cAMP induction. In addition, every substitution of Ile 584 and Lys 590 abolished cAMP, showing the irreplaceable nature of Ile 584 and Lys 590 . In particular, the fact that the K590R substitution failed to induce cAMP production raises the question of whether the positive charge of Lys 590 is necessary for cAMP induction. The substitutions of Lys 590 with Ala, Cys, Phe, Leu, Tyr, and Arg abrogated IP induction, suggesting the irreplaceable role of Lys 590 for both IP and cAMP induction. This is in contrast to the differential effects on cAMP and IP induction by Ala substitutions for the exoloop 3 amino acids, Lys 580 -Lys 590 .
In conclusion, our observations in this study show the interaction of the FSHR exoloop 3 with FSH, specifically the ␣ subunit. This is consistent with the decade-long view that the common ␣ subunit of the glycoprotein hormones is likely to induce the common hormone action, including cAMP induction (26). Leu 583 and Ile 584 are crucial for this interaction, in different ways. For example, Ile 584 appears to be so specific, requiring some flexibility and stereospecificity so that no other amino acids fit into its place. On the other hand, Leu 583 is less specific. Substitutions of these residues often enhanced the hormone binding affinity. However, these improvements in the hormone binding were offset by severe impairment of signal generations for cAMP and/or IP. For example, the Phe or Tyr substitution of Leu 583 improved the hormone binding and cAMP induction but impaired IP induction. On the other hand, substitutions for Ile 584 and Lys 590 abolished the cAMP and IP induction. These conflicting effects on the hormone binding and signal generation likely have constrained any further improvements in receptor functions. Our results open the next logical question of FIG. 11. Hydrophobicity analysis of substitutions of Pro 582 , Leu 583 , Ile 584 , and Lys 590 . As shown in A, the K d wt/mut ratios of mutant receptors with a panel of amino acids at Pro 582 , Leu 583 , Ile 584 , and Lys 590 were compared in a bar graph. As shown in B, the maximum cAMP mut/wt ratios of mutant receptors were compared in a bar graph. del, deleted.  10. Identification of the labeled FSH subunit and futile labeling of denatured FSH. As shown in A, denatured FSH that is not capable of binding and activating FSHR was labeled with increasing concentrations of 125 I-AB-FSHR exo3 as described in the legend for whether Leu 583 , Ile 584 , and Lys 590 interact with the exodomain and/or the hormone. The answers will provide new insights into the mechanisms of hormone binding and signal generation.