Modulation of High Affinity Hormone Binding

The lutropin/choriogonadotropin receptor is a seven-transmembrane receptor and consists of two major domains of similar size, an extracellular exodomain and a membrane-associated endodomain which includes 3 exoloops. The uniquely large exodomain is responsible for high affinity hormone binding whereas receptor activation occurs at the endodomain. However, little is known about the relationship between the exodomain and endodomain. It was reported that hormone binding to the exodomain was improved when the endodomain was truncated. This result suggests that hormone binding to the exodomain was influenced by the endodomain. To test this hypothesis, amino acids of exoloop 2 were examined by Ala substitutions. The binding affinity was enhanced by some Ala substitutions but attenuated by others. These results indicate that exoloop 2 influences the hormone binding to the exodomain. Particularly, the high affinity hormone binding at the exodomain is constrained by a group of amino acids, Ser484, Asn485, Lys488, Ser490, and Ser499. Computer modeling suggests these residues may be positioned on one side of exoloop 2. It also influences the affinity for cAMP induction and the maximal cAMP production in distinct ways, in addition to its influence on the hormone binding affinity. The distinct ways of influencing these functions are sometimes in conflict and compromised to attain the maximal affinity for cAMP induction. As a result, the exodomain attains the maximal affinity for hormone binding when the endodomain is truncated and cAMP induction is disengaged.

The LH/CG 1 receptor belongs to a subfamily of glycoprotein hormone receptors within the seven-transmembrane (TM) receptor family. Unlike other seven-transmembrane receptors, glycoprotein hormone receptors consist of two equal halves, an extracellular N-terminal half (exodomain) and a membrane associated C-terminal half (endodomain) (1,2). The exodomain alone is capable of high affinity hormone binding (3)(4)(5) with no hormone action (5,6). On the other hand, the endodomain is the site for receptor activation (7). However, little is known about the relationship between the exodomain and endodomain.
In this study, the hormone binding affinity of the truncated exodomain which lacks the endodomain is shown to be slightly and consistently higher than that of the wild type receptor. This result suggests that the endodomain influences the hormone binding at the exodomain. To identify the responsible amino acid residues the 20 amino acids of exoloop 2 in the wild type receptor, from Ser 484 to Gln 503 , were Ala scanned.

Mutagenesis and Functional Expression of LH/CG Receptors-Mu-
tant LH/CG-R cDNAs were prepared in pSELECT vector using the Altered Sites Mutagenesis System (Promega), sequenced, subcloned into pcDNA3 (Invitrogen) as described (8), and sequenced again to verify mutation sequences. In one of mutants, a stop codon was introduced immediately after the Gly 336 codon to produce the truncated exodomain, Arg 1 -Gly 336 . Mutant LH/CG receptor constructs were transfected into human embryonic kidney 293 cells by the calcium phosphate method. Stable cell lines were established in minimum essential medium containing 10% horse serum and 500 g/ml Geneticin (G-418).
125 I-hCG Binding and Intracellular cAMP Assay-Stably transfected cells were assayed for 125 I-hCG binding in the presence of 150,000 cpm of 125 I-hCG (9) and increasing concentrations of cold hCG. hCG, batch CR 127, was supplied by the National Hormone and Pituitary Program. Nontransfected cells did not show specific binding of hCG. For intracellular cAMP assay, ϳ50,000 cells were washed twice with Dulbecco's modified Eagle's media and incubated in the media containing isobutylmethylxanthine (0.1 g/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 media, the cells were rinsed once with fresh media without isobutylmethylxanthine, lysed in 70% ethanol, freezethawed 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 (Amersham). cAMP concentrations were determined with an 125 I-cAMP assay kit (Amersham) following the manufacturer's instructions and validated for use in our laboratory. All assays were carried out in duplicate and repeated 4 -6 times. Means and standard deviations were calculated and analyzed by Student's t test to determine the statistical significance (p 1 ) of repeats of the same samples. In addition, values for mutants and the exodomain were compared with the corresponding values of the wild type receptor using ANOVA with 95% confidence to determine the statistical significance (p 2 ) of the differences. 125 I-hCG Binding to Solubilized LH/CG Receptor-Transfected cells were washed twice with ice-cold 150 mM NaCl, 20 mM HEPES, pH 7.4 (buffer A). Cells were scraped on ice and collected in buffer A containing protease inhibitors (1 mM phenylmethylsulfonyl fluoride, 5 mM N-ethylmaleimide, and 10 mM EDTA) and pelleted by centrifugation at 1,300 ϫ g for 10 min. Cells from a 10-cm plate were resuspended in 0.6 ml of buffer A containing 1% Nonidet P-40, 20% glycerol, and the above protease inhibitors (buffer B), incubated on ice for 15 min, and diluted with 5.4 ml of buffer A containing 20% glycerol plus the protease inhibitors (buffer C). The mixture was centrifuged at 100,000 ϫ g for 60 min. The supernatant (500 l) was mixed with 150,000 cpm of 125 I-hCG and 6.5 l of 0.9% NaCl and 10 mM Na 2 HPO 4 at pH 7.4 containing increasing concentrations of unlabeled hCG. After incubation at 4°C for 12 h, the solution was thoroughly mixed with 250 l of buffer A containing bovine ␥-globulin (5 g/ml) and 750 l of buffer A containing 20% polyethylene glycol 8,000. After incubation at 4°C for 10 min, samples were pelleted at 1,300 ϫ g for 30 min and supernatants removed. Pellets were resuspended in 1.5 ml of buffer A containing 20% polyethylene glycol 8,000, centrifuged, and counted for radioactivity.

RESULTS
hCG Binds Better to the Truncated Exodomain Than to the Wild Type Receptor-The truncated exodomain, Arg 1 -Gly 336 , was expressed in 293 cells. Since it remains within the cells and is not secreted (4,5,10), the cells expressing it were solubilized in Nonidet P-40 and assayed for 125 I-hCG binding (Fig. 1). The mean K d value, 571 pM (p 1 Ͻ 0.01), of the truncated exodomain in solution was determined from six repeat experiments, each in duplicate. This is different from the K d value of the wild type receptor solubilized in Nonidet P-40, 943 pM (p 1 Ͻ 0.01). To determine whether the difference in the two K d values was statistically significant, they were examined using ANOVA with 95% confidence. The result indicates that the two K d values and the difference between them are statistically significant (p 2 Ͻ 0.01). Furthermore, the difference could not be attributed to experimental variations due to varying receptor concentrations, since the concentrations of the truncated exodomain and wild type receptor were within the acceptable range (Fig. 1). The results are also consistent with the previous reports by several laboratories (4, 5, 10), although the difference was not emphasized and explored at the time. These results, taken together, demonstrate that the hormone binding affinity of the truncated exodomain is slightly and consistently better than that of the wild type receptor.
Why Does the Truncated Exodomain Bind Better?-There are two notable differences in the truncated exodomain and the extracellular exodomain of the wild type receptor: the location and structure. The former is present within the cell and lacks the endodomain, whereas the latter is present on the cell sur-face and covalently linked to the endodomain. One might raise an unlikely possibility that the location of the exodomain prior to the solubilization in Nonidet P-40 solution affected the binding affinity. To test this hypothesis, the K d values of several LH/CG receptors with various Pro to Phe substitutions (11) were examined. These mutants are expressed both at the cell surface (extracellular) and in cells (intracellular) but the ratios of the extracellular/intracellular concentrations vary depending on which Pro is mutated ( Table I). The K d ratio of receptors on intact cells and those in Nonidet P-40 solution increases as does the intracellular receptor concentration. This result indicates that the K d values of intracellular receptors are higher than those of extracellular receptors. Therefore, the binding affinity of intracellular receptors is worse than that of the corresponding extracellular receptors. Furthermore, it indicates that the extracellular or intracellular location of the exodomain alone is not responsible for the different K d values.
Alternatively, the different K d values of the truncated intracellular exodomain and the extracellular exodomain of the wild type receptor may come from the interaction with the endodomain. For example, the exodomain of the wild type receptor may interact with the endodomain which in turn influences the hormone binding affinity whereas the truncated intracellular exodomain does not have any exodomain to interact with. Such an influence by the endodomain on the exodomain is likely to occur at the extracellular portion of the endodomain, including the three exoloops. To determine such a region(s) and amino acid residues, the 20 amino acids of exoloop 2 were individually substituted with Ala to produce 20 substitution mutants. Effects of Ala Substitution for Upstream Amino Acids of Exoloop 2-Among 11 mutant receptors with Ala substitution for the upstream amino acids, 6 were surface-expressed in reasonable receptor concentrations, 12,100 -64,800 receptors/ cell (Fig. 2). K d values of these surface-expressed receptors were diverse in the range of 250 to 840 pM. LH/CG-R N485A , LH/CG-R K488A , and LH/CG-R S490A , displayed lower K d values, indicating that their hormone binding affinities are better than that of the wild type receptor. On the other hand, the K d value of LH/CG-R V489A was 2-fold higher, an indication for a low binding affinity. In contrast to these surface-expressed mutants, hCG binding to the cells transfected with the other 5 mutants was marginal or not detected. For example, hCG bound to LH/CG-R I491A on intact cells with a normal affinity but the receptor concentration was Ͻ10% of the concentration of the wild type receptor. 125 I-hCG binding to intact cells was hardly detectable for LH/CG-R S484A , LH/CG-R Y486A , LH/CG-R C492A , and LH/CG-R P494A .
To determine whether any of these mutant receptors were trapped inside the cells or defective in hCG binding, stably transfected cells were solubilized in Nonidet P-40 and assayed for hormone binding. As shown in Fig. 2, C and D, all five mutants in the detergent solution bound hCG (p 1 Ͻ 0.05) and the receptor concentrations were not significantly different from the concentrations of the wild type receptor (p 1 Ͻ 0.05). In addition, the statistical significance of the each mutant data was analyzed twice for different purposes. First to determine the statistical significance (p 1 values) of repeat data for each, mutants were analyzed by Student' t test. In addition, the values for the exodomain were compared with the corresponding values of the wild type receptor using ANOVA with 95% confidence to determine the statistical significance of the difference (p 2 values).
These results indicate that the five mutant receptors were expressed but not efficiently transported to the cell surface.
The K d value of LH/CG-R S484A was 2-fold lower than that of the wild type receptor, an indication of a 2-fold improved affinity for hCG binding. In contrast to the improved binding affinity, LH/CG-R P494A showed a 2-fold lower affinity as the K d value was 2-fold higher. The other three mutants in solution showed slightly higher K d values. These results indicate that the Ala substitutions somewhat impacted the hormone binding affinity of the mutant receptors, in addition to impairing their expression on the cell surface.
In contrast to the diverse K d values, EC 50 values for cAMP induction by most of surface-expressed mutant receptors except LH/CG-R V489A are 3-5-fold higher than the EC 50 value of the wild type receptor (Fig. 2E). This is a significant reduction in the affinity for cAMP induction and indicates that these Ala substitutions impaired, but did not improve, the affinity for cAMP induction. In contrast to this normal binding with poor cAMP induction, LH/CG-R V489A bound hCG with a 2-fold lower affinity yet induced cAMP normally (p 1 Ͻ 0.001).
Ala Substitution for Downstream Amino Acids of Exoloop 2-All 9 mutant receptors with Ala substitution for the downstream amino acids were surface-expressed and the receptor concentrations were reasonable, being in the range of 12,200 -37,000/cell (Fig. 3). Their affinities for hormone binding were generally similar to that of the wild type receptor. However, the K d value of LH/CG-R S499A was lower (280 pM) than that of the wild type receptor. This result indicates that the S499A substitution improved the binding affinity as did the S484A, K488A, and S490A substitutions. In contrast to the improved binding affinities, the K d values of LH/CG-R M495A and LH/CG-R S502A were higher, 840 and 630 pM, respectively. All of the mutants were capable of inducing cAMP and the EC 50 values for cAMP induction were either similar to or slightly higher, up to 3-fold, than that of the wild type receptor. The maximum cAMP induction levels were similar to or lower than that produced by the wild type receptor.
Effect of Surface Receptor Concentrations-Surface concen-  (11). The extracellular concentration of receptors present on the plasma membrane was determined from the receptors which 125 I-hCG bound to intact cells. The total receptor concentration was determined with 125 I-hCG binding to receptors present in Nonidet P-40 solution of whole cells. The intracellular concentration of receptors present in cells was estimated by substracting the extracellular receptor concentration from the total receptor concentration. R stands for receptor.  To determine whether levels of surface receptor concentrations had any effect on hormone binding, cells were transiently transfected with varying concentrations of receptor plasmids and selected for those expressing ϳ40,000 receptors/ cell (Table II). Cells were assayed for hormone binding to intact cells. The K d values (Table II) for the wild type receptor LH/ CG-R N485A and LH/CG-R M487A , LH/CG-R K489A , and LH/CG-R D496A were similar to the K d values of the same receptors which were expressed at diverse levels ( Figs. 2 and 3). These results show that the surface receptor concentration did not impact the hormone binding affinity. To test whether receptor transport to the surface membrane might have had any effect on the hormone binding affinity, cells expressing the receptors were solubilized and assayed for hormone binding to solubilized receptors (Table II). The results show that the % K d values of receptors were similar to those of receptors expressed on the cell surface, an indication for no significant effect of receptor transport on hormone binding.
Double Ala Substitutions for Nearby Asp and Glu-Ionic amino acids of the LH/CG receptor have been implicated for important roles (8,(12)(13)(14). However, the D496A substitution and E498A substitution individually had marginal effects on surface expression, hCG binding, and cAMP induction. Therefore, both Asp 496 and Glu 498 were substituted with Ala to produce a double substitution mutant, LH/CG-R D496A/E498A . These double substitutions substantially attenuated the surface expression of the mutant and reduced the affinity for cAMP induction. Remarkably, the hormone binding affinity improved by more than 2-fold, underscoring the importance of their potential role to attenuate the high affinity hormone binding of the natural receptor. It is not clear whether these two nearby anionic residues play an important and mutually substitutable role in the hormone binding affinity.
Verification of Mutagenesis-Our site-directed mutagenesis requires a synthetic oligonucleotides with 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, mutant cDNAs were reverted to the wild type cDNA which was in turn used to transfect cells. All of the revertants behaved the same as the wild type receptor in surface expression, hormone binding, and cAMP induction (data not included), indicating that there were no mutations other than the intended Ala substitutions.

Influence of Exoloop 2 on Hormone Binding-Some
Ala substitutions for exoloop 2 amino acids impacted the hormone binding affinity, noticeably enhancing or reducing the binding affinity whereas others did not have a significant effect (Fig. 4). To facilitate the comparison of these diverse effects, percent K d values for mutants were calculated by dividing their K d values with that of the wild type receptor (Table III and Fig. 4). Most noticeable were the S484A, K488A, S490A, S499A, and D496A/ E498A substitutions. They improved the binding affinity up to 2-fold and these improvements were statistically significant (p 2 FIG. 3. Ala scan of downstream residues of exoloop 2. The downstream 9 amino acids of exoloop 2, from Met 495 to Gln 503 , were individually substituted with Ala and the resulting mutant receptors were stably expressed on human 293 cells. They were assayed and the data analyzed as described in the legend to Fig. 2. In addition, a double substitution mutant was generated in which both Asp 496 and Glu 498 were replaced with Ala. Ͻ 0.05 to p 2 Ͻ 0.001) according to ANOVA analysis. This is specific since the same Ala substitutions mostly attenuated, but never improved, the affinity for cAMP induction and other substitutions did not improve the binding affinity. It is interesting and could be significant that the extent of the improvement in the binding affinity of some of the mutant receptors is in the range of the enhanced binding affinity of the truncated exodomain as compared with the binding affinity of the extracellular exodomain of the wild type receptor. This correlation of the improvement and the extent of the improvement in the binding affinity after truncation of the endodomain and some Ala substitutions suggest that the high affinity hormone binding to the exodomain is influenced and furthermore, attenuated by the endodomain, including exoloop 2, of the wild type receptor. Exoloop 2 may interact with the exodomain or the exodomain-hormone complex to modulate the structure. If so, the exodomain may assume a structure more favorable for hormone binding in some mutants, including LH/CG-R S484A and LH/CG-R D496A/E498A . This is consistent with the recent suggestion that exoloop 2 contacts the hormone (15). Structure of Exoloop 2-It is interesting to see that all of the three Ser to Ala substitutions in exoloop 2, S484A, S490A, and S499A, improved the binding affinity. It is not clear whether there is a correlation between the three Ser to Ala substitutions and the improved binding affinities and whether there is a structural relationship of the three residues. If there are, the three Ser residues and possibly some other residues including Lys 488 (Fig. 4) might be arranged in the exterior of a structure to influence hormone binding and the affinity.
To help envision such a structure and to test whether the structure is possible, exoloop 2 was modeled based on the observation that the K d values of the mutant receptors vary according to their positions (Fig. 4) in putative secondary structures. Exoloop 2 was anchored to the 4th and 5th TMs of bacteriorhodopsin (16). We focused on Ser 484 , Asn 485 , Lys 488 , Ser 490 , and Ser 499 for which Ala substitution improved the binding affinity. One model was developed using the program "O" (17) with ideal 3 10 helix and ␤-sheet bond angles and lengths, with the most common side chain rotamers chosen for each residue (the upper models in Fig. 5). No energy minimization calculations were performed, as ideal geometry was given to the model from the outset and interactions between secondary structure elements are minimal. Here, Ser 484 , Lys 488 , Ser 490 , and Ser 499 are present in the exterior of exoloop 2 and their side chains facing toward the same side. Asn 485 , however, is anomalous and orients toward the interior of the loop. Particular attention was paid to the fact that Leu 493 , Asp 496 , and Ser 499 have a periodicity of 3 and the Ala substitution for them retains or slightly decreases the wild type K d . These residues may mark one side of an ␣ or 3 10 helix with periodicities of 3.4 and 3.0, respectively. Similarly, S484A, K488A, and S490A substitutions resulted in lower than wildtype K d values and the residues might mark one side of a ␤-pleated sheet. However, in this case, there is an anomalous increased K d by the Y486A substitution and a decrease by the N485A substitution. With these observations, the structure of exoloop 2 has been modeled to give a ␤-strand over the first seven residues of the loop followed by a short coil region and two turns of 3 10 helix. The loop has been constructed so that residues Ser 484 , Lys 488 , Ser 490 , Leu 493 , Asp 496 , and Ser 499 have side chains close to each other. Therefore, all may have similar effects on the interactions of the loop with the local protein environment, but the exact nature of these interactions cannot be predicted with the available data. This simple model ignores the anomalies at Asn 485 and Tyr 486 .
Another method was developed for modeling nucleic acid loop structures (18) and incorporated the AMBER energy minimization (19). Crystal structure of the seven-helix bundle from bacteriorhodopsin (16) was used for anchoring the loop. Using a reduced-coordinates approach, random loop structures with the targeted pair of fixed ends were generated. Those loop structures with energies below a threshold were selected and subjected to a short run of energy equilibration using Metropolis Monte Carlo simulation (18) and then, 1,000 cycles of energy minimization using AMBER (19). It produced a structure with Ser 484 , Asn 485 , Lys 488 , Ser 490 , and Ser 499 on one side of the loop surface (the lower models in Fig. 5). Therefore, it is possible that Ser 484 , Asn 485 , Lys 488 , Ser 490 , and Ser 499 are coordinated to position themselves in a specific loop structure and to influence the hormone binding affinity. In contrast to the improved binding affinities of some of the substitutions, the V489A, P494A, and M495A substitutions attenuated the binding affinity by ϳ2-fold. Interestingly, these residues are clustered together in the middle of exoloop 2 in both models, suggesting their intriguing role in the binding affinity.
Effects of Ala Substitutions on cAMP Induction-The significant attenuation by Ala substitution of the affinity for but not maximal level of cAMP induction indicate that once the mutant receptors were activated, even if poorly, they effectively produced cAMP. Overall, exoloop 2 is more important for the affinity for cAMP induction than for the maximal cAMP induction. This further suggests distinct mechanisms for the affinity and maximal level of cAMP production. These distinct mechanisms are more obvious when percent values for K d , EC 50 , and maximal cAMP production as well as their ratios are compared (Table III and Fig. 4). The ratios of percent EC 50 /percent K d for most of the mutants were Ͻ1, indicating more severe reductions in the affinities for cAMP induction than the hormone binding affinities.
Integral Roles of Exoloop 2 on the Hormone Binding Affinity, cAMP Induction Affinity, and Maximal cAMP Production-Our data indicate the importance of exoloop 2 for the hormone binding affinity, the affinity for cAMP induction, and the maximal cAMP production. This is probably accomplished by interaction of the exodomain and endodomain, particularly exoloop 2. The roles of exoloop 2 on each of these functions are distinct as are the mechanisms of exoloop 2 to influence them. Some of them appear to be in conflict. As a result, the hormone binding affinity, the affinity for cAMP induction, and the maximal cAMP production seem to be compromised. Ala substitutions both positively and negatively impact the hormone binding affinity and the maximal level of cAMP. However, the affinity for cAMP induction is never augmented by the substitutions. Therefore, the hormone binding affinity and the maximal level of cAMP were compromised to reach the best affinity for cAMP induction. For example, the hormone binding affinity for the exodomain reaches the maximum only when the endodomain is removed and cAMP induction is completely disengaged from the hormone binding.
Disulfide Bridge between Exoloops 1 and 2-The C492A substitution resulted in the loss of surface expression but not the high affinity hormone binding as found in other substitutions for Cys (20). It indicates the importance of Cys 492 in surface expression but not in hormone binding. Cys residues corre-  (17) with ideal 3 10 helix and ␤-sheet bond angles and lengths, with the most common side chain rotamers chosen for each residue. The lower model was built using a method similar to that developed for modeling nucleic acid loop structures (18). sponding to Cys 492 of exoloop 2 and Cys 416 of exoloop 1 of the LH/CG-R have been implicated to form a disulfide bridge in the other seven TM receptors (21). If this is true for the LH/CG receptor, the disulfide may not be essential for hormone binding. Another possibility is that Cys 492 of exoloop 2 and Cys 416 of exoloop 1 may not form a disulfide or may form disulfides with other Cys residues of the exodomain. It will be of interest to see whether the putative disulfide plays a role in surface expression of the LH/CG receptor, as do disulfides of hCG in hCG processing (22).
Effects of Ala Substitutions on the Surface Expression-Some of the substitutions attenuated or blocked surface expression of the resulting mutants. As a result, most of LH/CG-R S484A , LH/CG-R Y486A , LH/CG-R I491A , LH/CG-R C492A , LH/CG-R P494A , and LH/CG-R D496A /LH/CG-R E498A were trapped in the cells. Therefore, these amino acids are crucial for surface expression of the receptor.
Conclusion-Our data demonstrate that exoloop 2 of the LH/CG-R distinctly influences the hormone binding affinity, the affinity for cAMP induction, and the maximal cAMP production, probably by the interaction of exoloop 2 with the exodomain and other parts of the endodomain such as TM 6 and TM 7. These distinct ways of influencing the functions are sometimes in conflict and compromised to attain the maximal affinity for cAMP induction. The high affinity hormone binding at the exodomain is constrained by some of the exoloop 2 residues, particularly, Ser 484 , Asn 485 , Lys 488 , Ser 490 , and Ser 499 . As a result, the exodomain attains the maximal affinity for hormone binding when the endodomain is truncated and cAMP induction is disengaged.