Changes to Gonadotropin-Releasing Hormone (GnRH) Receptor ...

0013-7227/08/$15.00/0 Printed in U.S.A.

Endocrinology 149(6):3118 –3129 Copyright © 2008 by The Endocrine Society doi: 10.1210/en.2008-0002

Changes to Gonadotropin-Releasing Hormone (GnRH) Receptor Extracellular Loops Differentially Affect GnRH Analog Binding and Activation: Evidence for Distinct Ligand-Stabilized Receptor Conformations Kevin D. G. Pfleger, Adam J. Pawson, and Robert P. Millar Medical Research Council Human Reproductive Sciences Unit (K.D.G.P., A.J.P., R.P.M.), Centre for Reproductive Biology, The Queen’s Medical Research Institute, Edinburgh EH16 4TJ, United Kingdom; and Western Australian Institute for Medical Research/Centre for Medical Research (K.D.G.P.), University of Western Australia, QEII Medical Centre, Nedlands, Western Australia 6009, Australia GnRH and its structural variants bind to GnRH receptors from different species with different affinities and specificities. By investigating chimeric receptors that combine regions of mammalian and nonmammalian GnRH receptors, a greater understanding of how different domains influence ligand binding and receptor activation can be achieved. Using human-catfish and human-chicken chimeric receptors, we demonstrate the importance of extracellular loop conformation for ligand binding and agonist potency, providing further evidence for GnRH and GnRH II stabilization of distinct active receptor conformations. We demonstrate examples of GnRH receptor gain-of-function mutations that apparently improve agonist potency independently of affinity, implicating a role for extracellular loops in stabilizing the inactive

G

NRH CONTROLS THE reproductive endocrine system and elucidation of how it binds and activates its cognate receptor is of prime importance for the treatment of conditions such as precocious puberty, endometriosis, uterine fibroids, and prostate cancer (1). GnRH II [(His5,Trp7,Tyr8)GnRH] is a second form of GnRH endogenously expressed in humans; however, its function has yet to be clearly defined (2– 4). GnRH receptors have been cloned from a variety of species (5). Mammalian and nonmammalian GnRH receptors have different affinities for different structural variants of GnRH (6), including GnRH and GnRH II. These differences in binding affinity are likely to be due to differences in ligand binding residues and/or conformation of the ligand binding site. The latter can influence affinity by altering the configuration of receptor contact sites or affecting the ease with which a ligand can access these sites. Consequently, chimeric receptors that combine regions of mammalian and nonmammalian GnRH receptors can be used to shed light on the mechanism of ligand binding and receptor activation. In a previous study (6), we identified that [D-Trp6]GnRH and [D-Lys6]GnRH II are particularly high affinity superagonist analogs of GnRH and GnRH II respectively at the human reFirst Published Online March 20, 2008 Abbreviations: ECL, Extracellular loop; TM, transmembrane domain. Endocrinology is published monthly by The Endocrine Society (http:// www.endo-society.org), the foremost professional society serving the endocrine community.

receptor conformation. We also show that entire extracellular loop substitution can overcome the detrimental effects of localized mutations, thereby demonstrating the importance of considering the conformation of entire domains when drawing conclusions from point-mutation studies. Finally, we present evidence implicating the configuration of extracellular loops 2 and 3 in combination differentiating GnRH analog binding modes. Because there are two endogenous forms of GnRH ligand but only one functional form of full-length GnRH receptor in humans, understanding how GnRH and GnRH II can elicit distinct functional effects through the same receptor is likely to provide important insights into how these ligands can have differential effects in both physiological and pathological situations. (Endocrinology 149: 3118 –3129, 2008)

ceptor. This study also showed that the aforementioned four ligands had significantly different binding affinities at the catfish receptor I, compared with the mammalian receptors examined. Therefore, regions of catfish receptor I were substituted into the human receptor in the present study. Peptide and nonpeptide antagonists interact with receptors in different orientations compared with agonists (7–9). Antagonist 135-18 [(Ac-D-Nal(2)1,D-4-Cl-Phe2,D-Pal(3)3,Ile5, D-Lys(iPr)6,Lys(iPr)8,D-AlaNH210)GnRH] is of interest because it acts as an antagonist at the human receptor but an agonist at the Xenopus type I, chicken, and marmoset type II GnRH receptors (10 –12). Thus, this antagonist is able to stabilize active conformations of the Xenopus I, chicken, and marmoset type II receptors but not the human receptor. How the interaction of this ligand differs at mammalian, compared with nonmammalian receptors, is likely to provide insights into the differences between the active and inactive receptor conformations. Two residues conserved between most G protein-coupled receptors are Asp2.50 and Asn7.49 (13); however, mammalian type I GnRH receptors have the reciprocal arrangement of Asn2.50 and Asp7.49 (see Materials and Methods for explanation of numbering scheme). Mutation of Asn2.50 to Asp resulted in very low receptor expression and no detectable ligand binding. Mutation of Asp7.49 to Asn in combination with Asn2.50Asp restored ligand binding, but the mutant receptor was uncoupled from G␣q activation of phospholipase C (14 –16). This indicated

3118

Pfleger et al. • Effect of GnRH Receptor ECL Substitution

that these residues are in close proximity in the functional receptor (14, 16, 17). Type II mammalian and nonmammalian GnRH receptors differ from the mammalian type I GnRH receptors, having Asp2.50 and Asp7.49. In catfish receptor I, mutation of Asp2.50 to Asn abolished ligand binding, probably as a result of incorrect receptor folding. The mutation of Asp7.49 to Asn did not appear to affect receptor expression; however, it did reduce second-messenger generation (18). These differences between mammalian and nonmammalian GnRH receptors at this crucial locus are important to consider in the context of the present study because they suggest that the conformations of these receptors differ (17, 18). Many studies have now used point mutations to elucidate important ligand binding sites for GnRH analogs. Asp2.61 (19), Asn2.65 (20, 21), and Lys3.32 (22, 23), which are all conserved between mammalian and nonmammalian GnRH receptors, are key examples of residues implicated in binding both GnRH and GnRH II (Fig. 1). However, less consideration has been made of the important role played by receptor domain conformation, particularly with respect to the extracellular loops (ECLs). A prime example of where ECL conformation has been shown to be a critical determinant of receptor function is for the interaction between Asp/Glu7.32 in extracellular loop 3 (ECL3) of mammalian receptors and Arg8 in GnRH (19, 20). Indeed, it was speculated that the chicken receptor would lack this residue due to its apparent inability to select between GnRH and [Gln8]GnRH. However, its subsequent cloning revealed the presence of Glu7.32 (10) and an alternative explanation regarding differences in

Endocrinology, June 2008, 149(6):3118 –3129

3119

receptor conformation was put forward. In mammalian type I receptors, Asp/Glu7.32 is followed by proline [Ser-(Asp/ Glu)-Pro], whereas in nonmammalian type I receptors (see Ref. 5 for definitions of subtypes), it is preceded by proline [Pro-(Asp/Glu)-Tyr]. This difference is likely to alter the orientation of the Asp/Glu side chain and thereby affect the interaction with Arg8 in GnRH (21–23), even though this interaction still appears to occur with catfish receptor I (24). These findings illustrate the importance of investigating the role of ECL configuration and not just focusing on changes to single residues in isolation. Furthermore, changes to ECLs that dissociate agonist affinity from signaling capacity have been observed with different G protein-coupled receptors. For example, a double mutation in GnRH receptor ECL3 did not significantly reduce the affinity for GnRH II, but coupling to inositol phosphate production was abolished (22). Similarly, a lutropin/ choriogonadotropin receptor with a single point mutation in ECL3 was found to bind human choriogonadotropin with the same affinity as wild-type receptor, but coupling to adenylate cyclase was greatly reduced (25). In contrast, substitution of ␤2-adrenergic receptor ECL3 with that of ␣1aadrenergic receptor (26), mutations in ECL2 of thrombin receptor (27) and complement factor 5a receptor (28), and mutations in ECL1 or -2 in TSH receptor (29) resulted in constitutive activity. Such observations indicate that ECL conformational changes can influence receptor signaling independently of ligand binding, presumably by altering the conformation of the adjacent transmembrane domains (TMs)

FIG. 1. Schematic representation of the human GnRH receptor amino acid sequence. The consensus amino acid numbering scheme described by Ballesteros and Weinstein (30) used the residues shown in black squares as reference points (see Materials and Methods). Residues shown in black circles constitute the human ECL sequences that are substituted by either catfish receptor I ECLs (shown on the outside of the loops) or chicken receptor ECL2 (shown on the inside of ECL2). A dash represents the absence of a residue at that position. Residues shown in gray circles are examples of positions identified as potentially influencing GnRH analog binding, either directly or by affecting configuration of the binding pocket.

3120


and hence intracellular domains attached to those TMs. Furthermore, the emergence of constitutive activity with certain ECL mutations (26 –29) implicates a role for ECLs in stabilizing the inactive receptor conformation (26, 28). A number of residues in the ECLs and TMs have been shown to influence the binding of GnRH analogs, either directly or by configuring contact sites (5). Furthermore, there are precedents for changes to ECLs of G proteincoupled receptors differentiating between agonist binding and receptor activation (22, 25–29). The present study used human-catfish and human-chicken chimeric receptors to gain new insights into how ECL configurations differentially influence the affinity and potency of GnRH, GnRH II, and conformationally constrained GnRH and GnRH II superagonists as well as the affinity of antagonist 135-18. In particular, ECL substitution appears to result in a gain of function for GnRH II analogs; ECL conformation as a whole has been shown to overcome the effect of localized point mutations, and the configuration of ECL2 and -3 in combination has been implicated in differentiating GnRH analog binding. Materials and Methods Consensus amino acid numbering scheme This scheme facilitates comparisons between G protein-coupled receptors and has been described previously (30). The most conserved residue in each TM is assigned the TM number followed by the index 50. Therefore, in the human GnRH receptor Asn53, Asn87, Arg139, Trp164, Pro223, Pro282, and Pro320 are designated Asn1.50, Asn2.50, Arg3.50, Trp4.50, Pro5.50, Pro6.50, and Pro7.50 (Fig. 1). Other residues are then numbered relative to these positions, e.g. Asp2.61(98) (where appropriate, the amino acid number for the particular receptor has been included in parentheses).

GnRH analogs GnRH (pGlu-His-Trp-Ser-Tyr-Gly-Leu-Arg-Pro-GlyNH2), GnRH II [(His5,Trp7,Tyr8)GnRH], and (D-Trp6)GnRH were supplied by Bachem (Saffron Walden, Essex, UK). (D-Lys6)GnRH II was a gift from the University of Cape Town (Cape Town, South Africa). Antagonist 135-18 [(Ac-D-Nal(2)1,D-4-Cl-Phe2,D-Pal(3)3,Ile5,D-Lys(iPr)6,Lys(iPr)8, D-AlaNH210)GnRH] was a gift from R. Roeske (Indiana University, Indianapolis, IN).

GnRH receptor cDNA The human (31) and chicken (10) GnRH receptor cDNA constructs were gifts from the University of Cape Town (Cape Town, South Africa). The catfish GnRH receptor I cDNA (32) was a gift from Jan Bogerd (Utrecht University, Utrecht, The Netherlands). The human receptor cDNA with Gln5.35(208) mutated to Glu (12) and the human receptor/ chicken receptor ECL2 cDNA (10) were provided by Thomas Ott (Medical Research Council Human Reproductive Sciences Unit, Edinburgh, UK). The human receptor/chicken receptor ECL3 cDNA was modified from a construct provided by Bernhard Fromme (University of Cape Town, Cape Town, South Africa).

Production of chimeric GnRH receptor cDNA The human GnRH receptor cDNA construct had previously been engineered to include a number of silent mutations, thereby introducing restriction sites at the TM/ECL boundaries (12). Oligonucleotide primers spanning these sites were designed and synthesized (Genosys, Cambridge, UK) with each coding, partly for human GnRH receptor TM and partly for catfish GnRH receptor ECL: TM2/ECL1 (sense), 5⬘-CATGCCACTGGATGGGGTGTGGAATGTGAC-3⬘; ECL1/TM3 (antisense), 5⬘-ACTGAGTACTTTGCACATGGCGTCTC-3⬘; TM4/ECL2


(sense), 5⬘-CAGTTGTACATCTTCAGGATGATTAAGGCCAAAGG3⬘; ECL2/TM5 (antisense), 5⬘-TAAAAGGCCTCCTGCCAGTGCTGT3⬘; TM6/ECL3 (sense), 5⬘-CTACTACGTACTAGGCATTTGGTATTGGTTTGATCCACAGATGCTGCA-3⬘; ECL3/TM7 (antisense), 5⬘AGTGGTTAACATAATCAGGGATCA-3⬘. DNA coding for the catfish GnRH receptor ECLs was generated by PCR using these primers, Deep Vent DNA polymerase (New England Biolabs, Hitchin, Hertfordshire, UK), and the catfish GnRH receptor I cDNA as template. The products were successively ligated into the engineered human GnRH receptor cDNA construct in place of the DNA coding for the analogous human ECLs. To overcome the problem of additional restriction sites in the plasmid, the pBluescript vector (Stratagene, Cambridge, UK) was used, with subcloning into pZErO-2 (Invitrogen, Groningen, The Netherlands) for substitution of ECL1. The constructs were then subcloned into pcDNA-1/Amp (Invitrogen) for transfection into COS-7 cells. For production of the human receptor containing the His5.34(207)Glu and Gln5.35(208)Glu mutations, an antisense oligonucleotide primer was designed and synthesized that spanned a StuI restriction site and the sequence coding for the two mutations: 5⬘-TAAAAGGCCTCCTCCCACCATTGTG-3⬘. A PCR fragment was generated using this primer, the T3 primer, Deep Vent DNA polymerase, and the engineered human GnRH receptor cDNA as template. The PCR fragment was digested with StuI and EcoRI, and ligated into the engineered human GnRH receptor cDNA construct. The human receptor/chicken receptor ECL2 and -3 cDNA was generated by digesting the human receptor/chicken receptor ECL2 cDNA with StuI and EcoRI, followed by ligation of the isolated receptor cDNA fragment into the human receptor/chicken receptor ECL3 cDNA. All receptor cDNA constructs were confirmed by automated sequencing using a Prism 310 genetic analyzer (Applied Biosystems, Warrington, UK).

Cell culture and transfection COS-7 cells were seeded in 100-mm2 dishes at a density of 1.2 ⫻ 106 cells/dish. Cells were maintained at 37 C, 5% CO2 in DMEM containing 10% fetal calf serum, 0.3 mg/ml glutamine, 100 IU/ml penicillin, and 100 ␮g/ml streptomycin (Sigma-Aldrich Corp., Poole, Dorset, UK). After 24 h, the cells were transiently transfected with GnRH receptor cDNA (10 ␮g of DNA per 100-mm2 dish) using Superfect (QIAGEN, Crawley, West Sussex, UK) according to the manufacturer’s instructions (30 ␮l Superfect per 100 mm2 dish for 8 h). After a further 48 h, cells were scraped in PBS, pelleted, and stored at ⫺70 C.

Receptor binding assays Competition binding assays were carried out as described previously (6). Briefly, crude membrane suspension was incubated overnight at 4 C with 125I-(His5,D-Tyr6)GnRH (⬃120,000 cpm/tube) and varying concentrations of unlabeled GnRH analogs in triplicate. The suspensions were then filtered through a membrane harvester (Brandel, St. Albans, Hertfordshire, UK) and bound radioactivity counted using a multi-␥counter [PerkinElmer Corp. (Wallac, Inc.), Cambridge, UK]. 125I-(His5,DTyr6)GnRH has a Kd of 0.19 nm and was used at a concentration of 22 pm (33). Maximum specific binding ranged between approximately 5000 and 10,000 cpm/tube with nonspecific binding ranging between approximately 2000 and 4000 cpm/tube. No specific binding was detected with COS-7 cells transfected with vector only. Membrane concentration was varied to control for expression levels such that similar maximal specific radioligand binding was observed at all receptors.

Total inositol phosphate signaling assays Agonist stimulation of total inositol phosphate production was carried out as described previously (34, 35). Briefly, transiently transfected COS-7 cells were incubated with inositol-free DMEM containing 1% dialyzed heat-inactivated fetal calf serum and 0.5 ␮Ci/well myo-[3H]inositol (Amersham Pharmacia Biotech, Piscataway, NJ) for 48 h. Medium was removed and the cells washed with 1 ml buffer (140 mm NaCl, 20 mm HEPES, 4 mm KCl, 8 mm glucose, 1 mm MgCl2, 1 mm CaCl2, and 1 mg/ml BSA) containing 10 mm LiCl and incubated for 1 h at 37 C in 0.5 ml buffer containing 10 mm LiCl and agonist at the indicated concentration. Reactions were terminated by the removal of agonist and the



addition of 1 ml ice-cold 10 mm formic acid, which was incubated for 30 min at 4 C. Total [3H]inositol phosphates were separated from the formic acid cell extracts on AG 1-X8 anion exchange resin (Bio-Rad Laboratories, Hercules, CA) and eluted with a 1 m ammonium formate/ 0.1 m formic acid solution. The associated radioactivity was determined by liquid scintillation counting. Although receptor expression levels at the plasma membrane were not measured directly, binding studies indicate that the mutant receptors assayed for total inositol phosphate production were expressed at similar levels to the wild-type human receptor.

Data reduction and statistical analysis One-site competition binding curves and sigmoidal dose-response curves were generated by Prism graphing software (GraphPad Software, Inc., San Diego, CA) using nonlinear regression. Statistical analysis was carried out using two-tailed, unpaired Student’s t tests with Welch’s correction (does not assume equal variances).

Results ECL substitution differentially affects GnRH analog binding

Chimeric GnRH receptors consisting of the human receptor containing catfish receptor I ECLs in all single, double, and triple combinations were produced and expression in COS-7 cells allowed functional characterization. The binding affinity of GnRH and (D-Trp6)GnRH at the wild-type catfish GnRH receptor I was 7- and 17-fold lower than that at the wild-type human GnRH receptor, respectively (Table 1). Incorporation of catfish receptor ECL3 into the human receptor with any combination of other ECLs resulted in a significant decrease in the binding affinity of GnRH, compared with that at the wild-type human receptor. In contrast, a significant decrease in the binding affinity of (D-Trp6)GnRH, compared with that at the wild-type human receptor, was observed only after substitution with all three catfish receptor ECLs. For both GnRH and (D-Trp6)GnRH, triple ECL substitution reduced the binding affinity at the chimeric receptor to a level not significantly different from that observed at the wildtype catfish receptor (Fig. 2). The binding affinity of GnRH II and (D-Lys6)GnRH II at the wild-type catfish receptor was 104- and 27-fold higher than that at the wild-type human receptor respectively (Table 1). For both GnRH II and (D-Lys6)GnRH II, in contrast to that observed for GnRH and (D-Trp6)GnRH, the binding affinity at the chimeric receptor containing all three catfish receptor

3121

ECLs was significantly different from that observed at the wild-type catfish receptor (P ⬍ 0.05; Fig. 2). The binding affinity of GnRH II at the triple ECL-substituted chimeric receptor was 4-fold higher than at the wild-type human receptor (Fig. 2C). The binding affinity of (D-Lys6)GnRH II at the triple ECL-substituted chimeric receptor was not significantly different from that at the wild-type human receptor (Fig. 2D). Thus, whereas introduction of the catfish receptor ECLs into the human receptor produces a phenotype for GnRH and (D-Trp6)GnRH indistinguishable from the wild-type catfish receptor, these changes do not completely convey catfish receptor binding characteristics with regard to GnRH II and (D-Lys6)GnRH II. A significant increase in the binding affinity of GnRH II, compared with that at the wild-type human receptor, was observed only at chimeric receptors containing both catfish receptor ECL2 and -3 (Table 1). Surprisingly, a significant decrease in the binding affinity of (D-Lys6)GnRH II, compared with that at the wild-type human receptor, was observed at chimeric receptors containing the catfish receptor ECL3 in combination with the human receptor ECL2 (human ⫹ catfish ECL3 and human ⫹ catfish ECL1 ⫹ 3 in Table 1). This particular combination of ECLs appears to be detrimental for the binding of (D-Lys6)GnRH II because binding at chimeric receptors containing both catfish receptor ECL2 and -3 was not significantly different from that observed at the wild-type human receptor (Table 1). The binding affinity of antagonist 135-18 at the wild-type catfish GnRH receptor was 47-fold lower than that at the wild-type human GnRH receptor (Table 1). As for GnRH II and (D-Lys6)GnRH II, and in contrast to that observed for GnRH and (D-Trp6)GnRH, the binding affinity of antagonist 135-18 at the chimeric receptor containing all three catfish ECLs was significantly different from that observed at the wild-type catfish receptor (P ⬍ 0.05). The affinity of antagonist 135-18 for this triple ECL-substituted chimeric receptor was just 2-fold lower than at the wild-type human receptor (Fig. 2E). A significant decrease in the binding affinity of antagonist 135-18, compared with that at the wild-type human receptor, was observed only at chimeric receptors containing catfish receptor ECL1 and -2 (Table 1). Surprisingly, a significant increase in the binding affinity of antagonist 135-18, compared with that

TABLE 1. Binding of agonists, superagonists and antagonist 135–18 at human-catfish chimeric GnRH receptors GnRH IC50 (nM)

GnRH receptor

Wild-type human Human ⫹ catfish Human ⫹ catfish Human ⫹ catfish Human ⫹ catfish Human ⫹ catfish Human ⫹ catfish Human ⫹ catfish Wild-type catfish

ECL1 ECL2 ECL3 ECL1 ECL1 ECL2 ECL1

⫹ ⫹ ⫹ ⫹

93.8 ⫾ 24.1 (1.00) 110.2 ⫾ 10.2 (0.85) 120.5 ⫾ 31.8 (0.78) 273.1 ⫾ 14.4 (0.34)a 2 174.7 ⫾ 10.5 (0.54) 3 342.5 ⫾ 18.7 (0.27)a 3 313.6 ⫾ 50.2 (0.30)a 2⫹3 395.1 ⫾ 61.6 (0.24)a 633.8 ⫾ 90.6 (0.15)a

(D-Trp6)GnRH IC50 (nM)

GnRH II IC50 (nM)

(D-Lys6)GnRH II IC50 (nM)

Antagonist 135–18 IC50 (nM)

2.6 ⫾ 1.1 (1.00) 11.8 ⫾ 6.9 (0.22) 1.8 ⫾ 0.2 (1.44) 5.3 ⫾ 0.5 (0.49) 3.6 ⫾ 0.6 (0.72) 4.8 ⫾ 1.6 (0.54) 14.2 ⫾ 4.8 (0.18) 35.3 ⫾ 9.3 (0.07)a 45.3 ⫾ 12.9 (0.06)a

135.6 ⫾ 15.8 (1.00) 145.1 ⫾ 37.5 (0.93) 138.4 ⫾ 23.3 (0.98) 157.0 ⫾ 20.2 (0.86) 133.1 ⫾ 37.9 (1.02) 113.2 ⫾ 21.5 (1.20) 54.2 ⫾ 11.1 (2.50)a 35.7 ⫾ 7.8 (3.80)a 1.3 ⫾ 0.2 (104)a

19.0 ⫾ 4.8 (1.00) 40.0 ⫾ 10.4 (0.48) 18.1 ⫾ 3.8 (1.05) 45.3 ⫾ 6.6 (0.42)a 6.8 ⫾ 1.2 (2.80) 49.8 ⫾ 5.5 (0.38)a 19.9 ⫾ 1.9 (0.95) 10.7 ⫾ 2.2 (1.78) 0.7 ⫾ 0.3 (27.1)a

8.1 ⫾ 1.0 (1.00) 18.7 ⫾ 9.3 (0.43) 11.0 ⫾ 0.1 (0.74) 2.7 ⫾ 0.2 (3.00)a 43.1 ⫾ 2.6 (0.19)a 2.3 ⫾ 0.4 (3.52)a 9.1 ⫾ 2.3 (0.89) 18.1 ⫾ 1.2 (0.45)a 377.5 ⫾ 65.3 (0.02)a

Radioligand binding assays were performed on homogenized membranes of COS-7 cells transiently transfected with GnRH receptor cDNA. IC50 values are mean ⫾ SEM of between three and six experiments carried out in triplicate and were calculated as described in Materials and Methods. The IC50 is essentially the same as Ki because the concentration of radioligand is very low (22 pM) and the Kd is 0.19 nM. Fold increases in binding affinity relative to that at the wild-type human receptor are shown in parentheses. a Significantly different from affinity at wild-type human receptor (P ⬍ 0.05).

3122



FIG. 2. Binding curves for agonists, superagonists, and antagonist 135-18 at wild-type and human-catfish chimeric GnRH receptors. Binding of GnRH (A), (D-Trp6)GnRH (B), GnRH II (C), (D-Lys6)GnRH II (D), and antagonist 13518 (E) at the wild-type human (white squares), wild-type catfish I (black circles), and triple ECL-substituted chimeric (white circles) GnRH receptors. Radioligand binding assays were performed on homogenized membranes of COS-7 cells transiently transfected with GnRH receptor cDNA. Curves are representative of between three and six experiments, each carried out in triplicate (error bars are SEM of triplicates).

at the wild-type human receptor, was observed at chimeric receptors containing the catfish receptor ECL3 in combination with the human receptor ECL2 (human ⫹ catfish ECL3 and human ⫹ catfish ECL1 ⫹ 3 in Table 1). This particular combination of ECLs appears to improve the binding of antagonist 135-18 as binding at chimeric receptors containing both catfish receptor ECL2 and -3 was not significantly different from, or was significantly lower than, that observed at the wild-type human receptor. ECL substitution differentially affects activation by GnRH analogs

Replacement of ECL1 in single and double chimeric receptors did not appear to contribute to any changes in the binding affinity of any of the agonists tested (affinities at ECL1 ⫹ 2 and ECL1 ⫹ 3 substituted receptors were similar to those at ECL2 and ECL3 substituted receptors, respectively, and the affinities at the ECL1 substituted receptor were not significantly different from those at the wild-type human receptor). Therefore, signaling assays were carried out only to assess the effects of ECL2, ECL3, ECL2 ⫹ 3, and triple ECL substitution. No increase in inositol phosphate production was observed after treatment of the wild-type catfish receptor with antagonist 135-18 (data not shown).

The potency of GnRH (EC50 ⫽ 168.6 ⫾ 30.4 nm) and (D-Trp6)GnRH (EC50 ⫽ 29.3 ⫾ 6.7 nm) activating the wildtype catfish receptor was 130- and 17-fold lower than when activating the wild-type human receptor respectively (Table 2 and Fig. 3). Incorporation of catfish receptor ECL2 or ECL3 into the human receptor resulted in significant decreases in the potency of GnRH of 5- or 27-fold, respectively, whereas incorporation of ECL2 and ECL3 in combination resulted in a 102-fold decrease, similar to that observed for the wild-type catfish receptor (Table 2). In contrast, none of the combinations of ECL substitution significantly altered the potency of (D-Trp6)GnRH (Table 2 and Fig. 3). The potency of GnRH II (EC50 ⫽ 2.1 ⫾ 0.6 nm) and (D-Lys6)GnRH II (EC50 ⫽ 0.7 ⫾ 0.2 nm) activating the wild-type catfish receptor was 3.5- and 7-fold higher than when activating the wild-type human receptor, respectively (Table 2 and Fig. 3). Only substitution with all three catfish receptor ECLs produced significantly increased potencies for these agonists, resulting in EC50 values not significantly different from those observed when activating the wild-type catfish receptor (Table 2 and Fig. 3). Incorporation of catfish receptor ECL2 into the human receptor resulted in a marginal decrease in GnRH II potency and no decrease in (D-Lys6)GnRH II potency.



3123

TABLE 2. Potency of inositol phosphate production resulting from activation of human-catfish chimeric GnRH receptors by agonists and superagonists GnRH EC50 (nM)

(D-Trp6)GnRH EC50 (nM)

GnRH II EC50 (nM)

(D-Lys6)GnRH II EC50 (nM)

1.3 ⫾ 0.2 (1.00) 6.1 ⫾ 1.1 (0.21)a 34.6 ⫾ 1.5 (0.04)a 132.9 ⫾ 24.9 (0.01)a 45.9 ⫾ 6.9 (0.03)a 168.6 ⫾ 30.4 (0.01)a

1.7 ⫾ 0.6 (1.00) 0.6 ⫾ 0.2 (2.80) 2.4 ⫾ 0.4 (0.71) 12.7 ⫾ 4.6 (0.14) 3.2 ⫾ 0.5 (0.55) 29.3 ⫾ 6.7 (0.06)a

7.3 ⫾ 0.9 (1.00) 16.9 ⫾ 1.4 (0.43)a 7.3 ⫾ 0.8 (1.00) 16.9 ⫾ 5.7 (0.43) 2.2 ⫾ 0.7 (3.25)a 2.1 ⫾ 0.6 (3.50)a

4.7 ⫾ 0.7 (1.00) 2.3 ⫾ 0.5 (2.08) 6.4 ⫾ 1.7 (0.74) 4.5 ⫾ 0.8 (1.05) 0.6 ⫾ 0.2 (7.43)a 0.7 ⫾ 0.2 (6.80)a

GnRH receptor

Wild-type human Human ⫹ catfish Human ⫹ catfish Human ⫹ catfish Human ⫹ catfish Wild-type catfish

ECL2 ECL3 ECL2 ⫹ 3 ECL1 ⫹ 2⫹3

Total inositol phosphate signaling assays were performed on COS-7 cells transiently transfected with GnRH receptor cDNA. EC50 values are mean ⫾ SEM of between three and four experiments carried out in duplicate and were calculated as described in Materials and Methods. Fold increases in potency relative to the wild-type human receptor are shown in parentheses. a Significantly different from potency at wild-type human receptor (P ⬍ 0.05).

Complete substitution of ECL2 overcomes or reduces the detrimental impact of localized mutations

As we have previously described (6), the binding affinity of GnRH II at the human receptor containing Gln5.35(208)Glu was significantly lower than that observed at the wild-type human receptor (P ⬍ 0.05; Table 3 and Fig. 4A). However, the binding affinity of GnRH II at the human receptor containing the entire catfish receptor ECL2 [which includes Gln5.35(208)Glu] was not significantly different from that observed at the wild-type human receptor (Table 3 and Fig. 4A). In contrast, the IC50 for (D-Lys6)GnRH II was not significantly affected by either mutation (Fig. 4B) nor indeed was the IC50 for GnRH [wild-type human, 93.8 ⫾ 24.1 nm; Gln5.35(208)Glu mutant, 115.1 ⫾ 12.8 nm; catfish ECL2 chimera, 120.5 ⫾ 31.8 nm]. Mutation of His5.34(207)Glu in combination with Gln5.35(208)Glu reduced the binding of 125I-(His5,D-Tyr6)GnRH

FIG. 3. Inositol phosphate dose-response curves for agonists and superagonists activating wild type and human-catfish chimeric GnRH receptors. Dose-response curves for GnRH (A), (D-Trp6)GnRH (B), GnRH II (C), and (D-Lys6)GnRH II (D) activating the wildtype human (white squares), wild-type catfish I (black circles), and triple ECL-substituted chimeric (white circles) GnRH receptors. Total inositol phosphate signaling assays were performed on COS-7 cells transiently transfected with GnRH receptor cDNA. Curves are representative of between three and four experiments, each carried out in duplicate (error bars are SEM of duplicates).

to levels below the limits of detection. Alternative radioligands were not tested, but inositol phosphate production after treatment with GnRH II or (D-Lys6)GnRH II was also found to be below the limits of detection (data not shown). In contrast, specific binding was observed at the human receptor containing the entire chicken receptor ECL2 [which includes His5.34(207)Glu and Gln5.35(208)Glu]. The binding affinity of GnRH II at this chimeric receptor was 5-fold lower than that at the wild-type human receptor [compared with 3-fold lower for the Gln5.35(208)Glu mutation alone], but again the binding affinity of (D-lys6)GnRH II was not significantly lower than that at the wild-type human receptor (Table 3 and Fig. 4). Both GnRH II and (D-Lys6)GnRH II have considerably higher binding affinities at the wild-type chicken receptor, compared with the wild-type human receptor (Table 3 and Fig. 4).

3124



TABLE 3. Binding of GnRH II and (D-Lys6)GnRH II at ECL2 mutant GnRH receptors GnRH receptor

GnRH II IC50 (nM)

(D-Lys6)GnRH II IC50 (nM)

Wild-type human Human Gln5.35(208)Glu Human ⫹ catfish ECL2 (contains Gln5.35(208)Glu) Human His5.34(207)Glu/Gln5.35(208)Glu Human ⫹ chicken ECL2 (contains His5.34(207)Glu/Gln5.35(208)Glu) Wild-type catfish Wild-type chicken

135.6 ⫾ 15.8 (1.00) 350.1 ⫾ 15.5 (0.39)a 138.4 ⫾ 23.3 (0.98) ND 693.4 ⫾ 74.7 (0.20)a 1.3 ⫾ 0.2 (104)a 0.75 ⫾ 0.21 (181)a

19.0 ⫾ 4.8 (1.00) 20.0 ⫾ 2.0 (0.95) 18.1 ⫾ 3.8 (1.05) ND 44.8 ⫾ 10.7 (0.42) 0.7 ⫾ 0.3 (27.1)a 0.19 ⫾ 0.03 (100)a

Radioligand binding assays were performed on homogenized membranes of COS-7 cells transiently transfected with GnRH receptor cDNA. IC50 values are mean ⫾ SEM of between three and seven experiments carried out in triplicate and were calculated as described in Materials and Methods. The IC50 is essentially the same as Ki because the concentration of radioligand is very low (22 pM) and the Kd is 0.19 nM. Fold increases in binding affinity relative to that at the wild-type human receptor are shown in parentheses. Data for human Gln5.35(208)Glu GnRH receptor have been published previously (6) and are included for comparison. a Significantly different from affinity at wild-type human receptor (P ⬍ 0.05). ND, Not determined: specific radioligand binding below limits of detection (n ⫽ 3).

Discussion ECL substitution differentially affects GnRH analog binding and potency

The GnRH binding affinity and potency of inositol phosphate stimulation were significantly lower at the chimeric

FIG. 4. Binding curves for GnRH II and (D-Lys6)GnRH II at wild type and ECL2 mutant GnRH receptors. Binding of GnRH II (A) and (DLys6)GnRH II (B) at the wild-type human (white squares), wild-type catfish I (black circles), wild-type chicken (black triangles), human Gln5.35(208)Glu (black squares), human ⫹ catfish ECL2 (white circles), and human ⫹ chicken ECL2 (white triangles) GnRH receptors. Radioligand binding assays were performed on homogenized membranes of COS-7 cells transiently transfected with GnRH receptor cDNA. Curves are representative of between three and seven experiments, each carried out in triplicate (error bars are SEM of triplicates).

receptor containing the catfish receptor ECL3, compared with that at the wild-type human receptor. This agrees with previous suggestions that ECL3 is important for configuring the GnRH receptor for high affinity binding of GnRH (19, 20, 22, 23). Asp/Glu7.32 in mammalian type I receptors is believed to interact with Arg8 in GnRH (19, 20). In the human receptor Asp7.32 is followed by proline (Ser-Asp-Pro), whereas in the catfish receptor, it is preceded by proline (Pro-Asp-Tyr). This difference is likely to alter the orientation of the aspartate side chain, thereby affecting the interaction with Arg8 in GnRH (21–23). In contrast to that observed for binding affinity, the inositol phosphate signaling potency of GnRH was decreased by ECL2 substitution. This effect appears to be in addition to the influence of ECL3 substitution as combined ECL2 and ECL3 incorporation resulted in a larger decrease than observed with either single substitution. In contrast to that observed with GnRH, the binding affinity and potency of (D-Trp6)GnRH was not changed significantly as a result of substituting ECL3. This also agrees with previous observations that conformational constraint of GnRH overcomes the requirement for Arg8 to interact with Asp7.32 for high-affinity binding (20) and questions the appropriateness of including this interaction in ligand-receptor binding models involving conformationally constrained GnRH analogs (5). A significant decrease in the binding affinity of (DTrp6)GnRH, compared with that at the wild-type human receptor, was observed only after substitution with all three catfish receptor ECLs. As with GnRH, (D-Trp6)GnRH had an affinity at this triple ECL-substituted chimeric receptor that was not significantly different from that at the wild-type catfish receptor. The distinct configuration of all three catfish receptor ECLs presumably influences binding of these ligands by altering the spatial arrangement of ECL contact sites, indirectly altering the configuration of contact sites within the TMs and/or influencing the ease with which the ligands can access contact sites within the TMs. However, this effect on binding affinity was not reflected in inositol phosphate signaling potency, consistent with the notion that conformational requirements for ligand binding can be distinguished from requirements for receptor activation (22, 25–29).


ECL substitution appears to result in a gain of function for GnRH II analogs

For GnRH II, (D-Lys6)GnRH II and antagonist 135-18, in contrast to that observed for GnRH and (D-Trp6)GnRH, triple ECL substitution resulted in binding affinities that remained significantly different from those observed at the wild-type catfish receptor. However, the increased potency of GnRH II and (D-Lys6)GnRH II with triple ECL substitution resulted in a gain of function to a level not significantly different from that observed at the wild-type catfish receptor. Interpretation of the GnRH II data is complicated by the observation that there is a 104-fold difference in affinity between the wild-type human and catfish receptors, compared with a 3.5-fold difference in potency. Consequently, the 3.8-fold increase in binding affinity of GnRH II at the triple ECL substituted receptor is not a substantial effect, compared with the 104-fold increase at the wild-type catfish receptor. In contrast, a similar increase in potency of GnRH II (3.3-fold) at the triple ECL substituted receptor is sufficient to result in a similar potency to that observed at the wild-type catfish receptor. Evidence for a differential effect on affinity and potency is more compelling for (D-Lys6)GnRH II as an insignificant increase in affinity of 1.8-fold was observed at the triple ECL substituted receptor (compared with 27-fold for wild-type catfish receptor), whereas the increase in potency was 7.4-fold (compared with 6.8-fold for wild-type catfish receptor). Consequently, these observations appear to be further examples of mutagenesis enabling agonist binding affinity to be dissociated from receptor activation (22, 25–29). They are also consistent with the concept of constitutive receptor activity (36), an extreme example of mutagenesis improving potency to a greater extent than affinity to a point at which ligand binding is no longer required for receptor activation. Such constitutive activity has yet to be demonstrated for the GnRH receptor. Indeed, mutagenesis of a conserved residue that results in this phenomenon in other G protein-coupled receptors was found to uncouple signaling to the inositol phosphate pathway (37). ECLs and TMs are likely to influence each other’s configuration

The ability of ECL substitutions to improve agonist potency to such an extent implies that the ECLs play an important role in configuring the TMs and consequently the intracellular sites of G protein coupling. This is consistent with various studies of other G protein-coupled receptor ECL mutations (26 –29), including data generated from chimeras of ␤2- and ␣1a-adrenergic receptors (26), and supports the notion that ECLs contribute to stabilization of the inactive receptor conformation (26, 28). Such a concept also has implications for large G protein-coupled receptor agonists that are incapable of inserting into the receptor core to any degree and are likely to exert their effects via interaction with the ECLs, directly or via binding to the N terminus that interacts with the ECLs. The mechanism of activation for these receptors may well involve disruption of the ECL configuration and consequently intramolecular interactions stabilizing the inactive receptor conformation. In addition to the ECLs influencing the configuration of the


3125

TMs, the reciprocal relationship is also likely to occur. The spatial arrangement of the ECLs will, in part, be dictated by the relative positioning of the TMs to which they are anchored. Evidence suggests that mammalian type I and nonmammalian receptor TMs have different configurations (10, 18). Therefore, the three catfish receptor ECLs substituted into the human receptor may not be in the same spatial arrangement as in the wild-type catfish receptor and such a difference in ECL configuration despite conservation of primary structure may explain the lack of improvement in GnRH II, (D-Lys6)GnRH II or antagonist 135-18 affinity. The same ECL contact sites may be used by the different agonists in both human and catfish receptors, but it may be the positioning of these sites that predominantly determines their differential contribution to binding affinity. Alternatively, if contact sites are introduced, they may be incorrectly configured for the same reasons. Furthermore, additional GnRH II and (D-Lys6)GnRH II contact sites may exist within the TMs or N terminus. Our observations support the concept that receptor conformations that bind ligands with high affinity are not necessarily the same as those that bind G proteins or other signaling molecules with high affinity. Indeed, the general correlation between high agonist affinity and potency is likely to be an example of coordinated evolution because there is little requirement for a high-affinity receptor that does not signal and constitutive receptor activity is undesirable in the vast majority of situations (38). The capacity for dissociation of binding and signaling also supports the concept that one receptor can couple to different signaling pathways after activation by different ligands (39), a particularly important concept for understanding the potential physiological roles of GnRH and GnRH II in a system that appears to include only one cognate receptor in humans (40). ECL substitution has little effect on antagonist 135-18 binding

It is surprising that the binding affinity of antagonist 135-18 at the triple ECL-substituted chimeric receptor is only 2-fold lower than at the wild-type human receptor, implying that this antagonist binds wholly to sites within the TMs and N terminus or that it interacts with ECL contact sites that are conserved between human and catfish receptors but configured differently due to the different positioning of the ECLs by the TMs. The latter scenario may explain how antagonist 135-18 acts as an antagonist at the human receptor, a partial agonist at the Xenopus type I receptor and a full agonist at the chicken and marmoset type II receptors (10 –12). The binding of the ligand may result in stabilization of distinct receptor conformations due to the different configuration of the conserved contact sites in the different receptors. Lys3.32 is believed to be crucial for agonist binding but not for antagonist binding (41). Perhaps the different configuration of the conserved ECL contact sites alters the ability of antagonist 135-18 to interact with such TM contact sites that differentiate between agonists and antagonists.

3126


Differences between GnRH and GnRH II binding modes observed despite conservation of binding sites

This study has highlighted differences between GnRH analogs [including (D-Trp6)GnRH], GnRH II analogs [including (D-Lys6)GnRH II] and antagonist 135-18. The different affinities of GnRH analogs for the human receptor, compared with catfish receptor I, can be explained by the different ECLs. This is not the case for GnRH II analogs or antagonist 135-18, implying that these ligands form different interactions with GnRH receptors, compared with those formed by GnRH analogs. This is unsurprising for antagonist 135-18 because antagonists have been shown to form different receptor interactions, compared with agonists (7, 8). However, the residues that appear to be particularly important for GnRH binding, namely Asp2.61, Trp2.64, Asn2.65, Lys3.32, Asn5.39, Trp6.48, Tyr6.58, and Asp7.32, are all conserved between the human and catfish I receptors (6), and evidence has been presented for GnRH II interacting with at least Asp2.61 (42), Asn2.65 (43), and Lys3.32 (44). Therefore, GnRH and GnRH II would appear to have different ligand conformations, form different additional ligand-receptor interactions, and/or interact with conserved contact sites that are in different spatial arrangements in human and catfish I receptors. In a previous study, we suggested that residues five to eight of GnRH II are preconfigured for interaction with nonmammalian receptors but that D-Lys6 substitution may alter this conformation to improve binding at mammalian type I receptors (6). The N and C termini may be configured differently in GnRH, compared with GnRH II. Nuclear magnetic resonance studies of GnRH and chicken GnRH I [(Gln8)GnRH] show these two ligands having similar turn conformations around Gly6 but very different conformations of the N and C-termini (5, 45). Chicken GnRH I differs from GnRH by a single residue. Therefore, it is conceivable that GnRH II, which has three residues different from GnRH, also has different N- and C-terminal conformations. As discussed above, mammalian and nonmammalian type I receptors are likely to have different configurations. Therefore, we suggest that the conformation of GnRH selects for (and/or is induced/selected by) the human receptor and that of GnRH II selects for (and/or is induced/selected by) the differently configured catfish receptor I. Configuration of ECL2 and -3 in combination differentiates GnRH analog binding modes

Although ECL substitution did not account for the difference in binding affinities of GnRH II, (D-Lys6)GnRH II and antagonist 135-18 at the human receptor, compared with the catfish receptor, there were significant changes in the binding affinity of these ligands at chimeric receptors, compared with that at the wild-type human receptor. A significant increase in the binding affinity of GnRH II, compared with that at the wild-type human receptor, was observed at chimeric receptors containing both catfish ECL2 and -3. This implies that changing the conformation of these two ECLs in combination alters the spatial arrangement of ECL contact sites, indirectly alters the configuration of contact sites within the TMs and/or influences the ease with which the ligand can access


contact sites within the TMs. As shown in Fig. 1, a number of residues in TMs 5, 6, and 7 (TMs that are likely to be influenced by changes in the configuration of the adjacent ECL2 and -3) affect ligand binding (46 –50). Therefore, the present study supports models showing this region forming part of a ligand binding pocket (14, 24, 44, 46 –51) and implies that the human receptor ECL2/3 configuration is less conducive to GnRH II binding than the catfish receptor ECL2/3 configuration. Further evidence for the concept that GnRH and GnRH II select for (and/or are selected by) distinct receptor conformations (40, 52, 53) has been provided recently, after the discovery of TM residues that appear to differentially influence the binding of GnRH and GnRH II (49, 50). Of particular note in the context of the present study is Phe6.40(272) in TM6. Mutation of this residue to Ala resulted in a considerable increase in affinity for GnRH II and (Tyr8)GnRH but not GnRH, (His6)GnRH or (Trp7)GnRH (50). Of the receptors cloned to date, Phe6.40 is completely conserved in mammalian type I GnRH receptors, whereas Ile or Val always occupies this position in mammalian type II and nonmammalian GnRH receptors, including catfish receptor I (5). Therefore, ECL configuration appears to have a significant effect on the conformation of a region that is clearly important for differentiating GnRH and GnRH II binding. The binding affinity of (D-Lys6)GnRH II at the triple ECLsubstituted chimeric receptor was not significantly different from that at the wild-type human receptor. D-Lys6 appears to alter the conformation of GnRH II or form an additional interaction with the human receptor so that triple catfish ECL substitution does not significantly improve binding. If ECL2/3 configuration in the human receptor is less conducive to the binding of GnRH II, D-Lys6 appears to overcome this. It would then be a further example of conformational constraint around position six of the ligand overcoming changes in ECL conformation, as observed with conformationally constrained GnRH analogs overcoming the requirement for Arg8 to interact with ECL3 (20). Catfish ECL substitution was expected to increase the binding affinity of (D-Lys6)GnRH II. Therefore, a surprising observation is that the artificial configuration of human ECL2 and catfish ECL3 (termed configuration C in Table 4) appears to reduce the binding affinity of (D-Lys6)GnRH II at chimeric receptors, compared with that at the wild-type human receptor. This also implies that (D-Lys6)GnRH II has a different conformation to GnRH II or that the D-Lys6 forms an additional interaction with the receptor, which is disrupted by this particular ECL configuration. The significant decrease in the binding affinity of antagonist 135-18 at chimeric receptors containing catfish ECL1 and -2, compared with that at the wild-type human receptor, implies that the configuration of these two ECLs alters the spatial arrangement of ECL contact sites, indirectly alters the conformation of contact sites within the TMs and/or influences the ease with which the ligand can access contact sites within the TMs. This configuration does not appear to affect agonist binding, again illustrating that antagonist 135-18 has a different binding mode, compared with the agonists. The artificial configuration between human ECL2 and catfish ECL3 (termed configuration C in Table 4) results in a sur-



3127

TABLE 4. (D-Lys6)GnRH II and antagonist 135–18 discriminate between three apparently different configurations of GnRH receptor Relative affinitya

Configuration A: wild-type catfish GnRH receptor

Configuration B: wild-type human GnRH receptor

Configuration C: chimeras containing human ECL2 and catfish ECL3b

(D-Lys6)GnRH II Antagonist 135–18

⫹⫹⫹⫹⫹ ⫹

⫹⫹⫹ ⫹⫹⫹

⫹⫹ ⫹ ⫹ ⫹⫹

a Relative ligand binding affinity ranges (for illustrative purposes): ⫹, IC50 greater than 125 nM; ⫹⫹, IC50 of 25–125 nM; ⫹⫹⫹, IC50 of 5–25 nM; ⫹⫹⫹⫹, IC50 of 1–5 nM; ⫹⫹⫹⫹⫹, IC50 less than 1 nM. b Human ⫹ catfish ECL3 and human ⫹ catfish ECL1 ⫹ 3 as shown in Table 1.

prising observation with antagonist 135-18. Instead of decreasing binding affinity for antagonist 135-18, this configuration appears to increase binding affinity at chimeric receptors, compared with that at the wild-type human receptor. It would appear that this receptor conformation is very well suited to the binding of antagonist 135-18, which binds at the wild-type human receptor with higher affinity than at the wild-type catfish receptor (Table 4). On the other hand, it is particularly unsuitable for binding (D-Lys6)GnRH II, which binds at the wild-type catfish receptor with higher affinity than at the wild-type human receptor (Table 4). These observations provide evidence that (D-Lys6)GnRH II and antagonist 135-18 select for different receptor conformations and that, as observed for GnRH II, ECL2/3 configuration influences configuration of the binding sites. ECL1’s high amino acid conservation likely to limit effect of substitution

The lack of effect of substituting ECL1 should not be interpreted as meaning this loop has a minor role in configuring the binding pocket. Indeed, it is likely to have a critical role as at least three residues at the TM2/ECL1 boundary are important for ligand binding: Asp2.61(98), Trp2.64(101), and Asn2.65(102) (5) (Fig. 1). Furthermore, Lys3.32(121), which differentiates between agonist and antagonist binding (41), and Met3.43(132), which differentiates between GnRH and GnRH II binding (50), are both present in TM3 (Fig. 1). The importance of ECL1 is reflected in its high amino acid conservation from the human to catfish I receptor (Fig. 1), which is also likely to result in a relatively conserved secondary structure. Therefore, the combination of conserved key residues and conformation would explain the limited impact of this loop’s substitution. ECL conformation as a whole can overcome effect of localized point mutations

Interestingly, a previous study showed that substitution of the Pro-Asp-Tyr motif in ECL3 (as seen in the catfish receptor I) into the human receptor in place of Ser-Asp-Pro resulted in a decreased affinity for both GnRH and a constrained GnRH analog [(His5,D-Tyr6)GnRH] but not GnRH II (22). Furthermore, this substitution abolished inositol phosphate signaling by both GnRH and GnRH II. In the present study, substitution of the entire catfish ECL3, including the Pro-Asp-Tyr motif, did not significantly decrease the affinity or potency of another constrained GnRH analog, (D-Trp6)GnRH. Furthermore, both GnRH- and GnRH II-induced inositol phosphate signaling was observed, with the potency of GnRH II being the same as when activating the wild-type human receptor (Table 2). In combi-

nation, these observations highlight the importance of considering the conformation of an ECL as a whole and this is further illustrated by our results investigating ECL2 mutant GnRH receptors. Mutation of Gln5.35(208) to Glu in the human receptor (as seen in the rat receptor) significantly reduced the binding affinity of GnRH II but not (D-Lys6)GnRH II (6). However, the findings of the present study imply that this detrimental effect is overcome by substitution with the entire catfish receptor ECL2, despite this substitution including Gln5.35Glu. Therefore, it would appear that the conformation of this ECL in its entirety alters the positioning of Glu5.35 so that it is no longer detrimental to GnRH II binding. Interestingly, the Gln5.35Glu mutation does not appear to influence GnRH binding, again implying that GnRH and GnRH II stabilize distinct receptor conformations. It is also intriguing that Glu5.35 has been evolutionarily tolerated in the rat, but it should be noted that evidence for endogenous GnRH II expression in this animal is lacking. Indeed, GnRH II is not expressed in the mouse (54). The His5.34(207)Glu/Gln5.35(208)Glu double mutation effectively abolished specific binding of 125I-(His5,D-Tyr6)GnRH and inositol phosphate production after treatment with GnRH II or (D-Lys6)GnRH II. Regardless of whether this resulted from lack of receptor expression, trafficking, or ligand binding, receptor conformation is likely to be compromised by these mutations. Interestingly, mutation of the adjacent conserved His5.33(206) to Ala had a similar effect (46). Notably, His5.34 and Gln5.35 are not completely conserved. Indeed, the Glu5.34/Glu5.35 combination is present in the chicken GnRH receptor and substitution of the entire chicken receptor ECL2 into the human receptor restored binding of GnRH II and (D-Lys6)GnRH II (the latter to a level not significantly different from that at the wild-type human receptor). Again, it would appear that the conformation of this ECL in its entirety alters the positioning of the C terminus of ECL2 so that the presence of these residues is tolerated. The Glu5.34/Glu5.35 combination of two negatively charged residues is not present in any other GnRH receptor cloned to date, and further investigation is required to ascertain what features of chicken receptor ECL2 confer this tolerance. However, a point of departure may be the positively charged Arg5.30, which is not present in the human or catfish I receptors (Fig. 1). Substantial conformational differences between receptors in the nonmammalian type I group are likely

Substitution of chicken receptor ECL3 into the human receptor, alone or in combination with chicken receptor

3128 Endocrinology, June 2008, 149(6):3118 –3129

ECL2, reduced binding of 125I-(His5,D-Tyr6)GnRH to below the limits of detection (K. D. G. Pfleger, unpublished observations), whereas some inositol phosphate signaling was observed after activation with GnRH II or (D-Lys6)GnRH II (Pawson, A. J., and K. D. G. Pfleger, unpublished observations). Therefore, although these chimeric receptors traffic to the plasma membrane, it appears that expression levels are severely compromised. These results are surprising considering that substitution with catfish ECL3 is well tolerated in the present study. However, it illustrates that, although type I GnRH receptors are increasingly categorized as mammalian or nonmammalian (5), there are still likely to be substantial conformational differences between receptors within the nonmammalian type I group, particularly between fish, amphibians, and birds. This is further exemplified by our finding that antagonist 135-18 does not act at the catfish I receptor to stimulate inositol phosphate production, despite acting as an agonist at the Xenopus I and chicken GnRH receptors (10, 12). Concluding remarks

In conclusion, this study has presented further evidence for the emerging concept that GnRH, GnRH II, and GnRH antagonists stabilize distinct conformations of the GnRH receptor. This has important consequences for understanding the physiological roles of GnRH and GnRH II in humans, in whom only a single functional GnRH receptor appears to be present. We have described examples of GnRH receptor gain-of-function mutations that appear to improve agonist potency independently of affinity. This has provided further evidence for a role of ECLs in stabilizing the inactive receptor conformation and demonstrated the importance of ECL conformation for configuring the entire receptor, including the intracellular binding sites of G proteins and presumably other molecules involved in signaling, phosphorylation, desensitization, and internalization. We have demonstrated the importance of considering entire domains when drawing conclusions from mutagenesis studies because the effects of localized mutations can be overcome by altering larger domains, such as entire ECLs. Finally, our findings imply that the configuration of ECL2 and -3 in combination differentiate GnRH analog binding modes. As a consequence, we conclude by proposing that this study, in simplistic terms, is consistent with a binding model whereby conserved residues in the TM1/ECL1/TM2 region are largely responsible for agonist binding, whereas the ECL2/TM5/TM6/ECL3/TM7 region plays more of a role in differentiating binding modes. This is in keeping with the concept of ligand-induced selective signaling (43) that is likely to be critical for GnRH receptor function in humans (44). Acknowledgments The authors acknowledge the expert technical assistance of R. Sellar, N. Miller, P. Taylor, and G. Pfleger. We are very grateful to R. Roeske for providing antagonist 135-18 and T. Ott, J. Bogerd, and B. Fromme for providing cDNA constructs. We also acknowledge Z. Lu for his comments on the completed manuscript. Received January 3, 2008. Accepted March 7, 2008.


Address all correspondence and requests for reprints to: Dr. Kevin D. G. Pfleger, Western Australian Institute for Medical Research, Ground Floor, B Block, QEII Medical Centre, Hospital Avenue, Nedlands, Western Australia 6009, Australia. E-mail: [email protected]. This work was supported by the U.K. Medical Research Council. K.D.G.P. is supported by a Peter Doherty Fellowship from the National Health and Medical Research Council of Australia (353709). Portions of this work were presented at the 84th Annual Meeting of The Endocrine Society, San Francisco, CA, 2002. Disclosure Statement: The authors have nothing to disclose.

References 1. Conn PM, Crowley Jr WF 1994 Gonadotropin-releasing hormone and its analogs. Annu Rev Med 45:391– 405 2. Millar RP 2003 GnRH II and type II GnRH receptors. Trends Endocrinol Metab 14:35– 43 3. Pawson AJ, Morgan K, Maudsley SR, Millar RP 2003 Type II gonadotrophinreleasing hormone (GnRH-II) in reproductive biology. Reproduction 126:271– 278 4. Millar RP 2005 GnRHs and GnRH receptors. Anim Reprod Sci 88:5–28 5. Millar RP, Lu ZL, Pawson AJ, Flanagan CA, Morgan K, Maudsley SR 2004 Gonadotropin-releasing hormone receptors. Endocr Rev 25:235–275 6. Pfleger KD, Bogerd J, Millar RP 2002 Conformational constraint of mammalian, chicken, and salmon GnRHs, but not GnRH II, enhances binding at mammalian and nonmammalian receptors: evidence for preconfiguration of GnRH II. Mol Endocrinol 16:2155–2162 7. Janovick JA, Haviv F, Fitzpatrick TD, Conn PM 1993 Differential orientation of a GnRH agonist and antagonist in the pituitary GnRH receptor. Endocrinology 133:942–945 8. Assefa D, Pawson AJ, McArdle CA, Millar RP, Flanagan CA, Roeske R, Davidson JS 1999 A new photoreactive antagonist cross-links to the N-terminal domain of the gonadotropin-releasing hormone receptor. Mol Cell Endocrinol 156:179 –188 9. Cui J, Smith RG, Mount GR, Lo JL, Yu J, Walsh TF, Singh SB, DeVita RJ, Goulet MT, Schaeffer JM, Cheng K 2000 Identification of Phe313 of the gonadotropin-releasing hormone (GnRH) receptor as a site critical for the binding of nonpeptide GnRH antagonists. Mol Endocrinol 14:671– 681 10. Sun YM, Flanagan CA, Illing N, Ott TR, Sellar R, Fromme BJ, Hapgood J, Sharp P, Sealfon SC, Millar RP 2001 A chicken gonadotropin-releasing hormone receptor that confers agonist activity to mammalian antagonists. Identification of D-Lys(6) in the ligand and extracellular loop two of the receptor as determinants. J Biol Chem 276:7754 –7761 11. Millar R, Lowe S, Conklin D, Pawson A, Maudsley S, Troskie B, Ott T, Millar M, Lincoln G, Sellar R, Faurholm B, Scobie G, Kuestner R, Terasawa E, Katz A 2001 A novel mammalian receptor for the evolutionarily conserved type II GnRH. Proc Natl Acad Sci USA 98:9636 –9641 12. Ott TR, Troskie BE, Roeske RW, Illing N, Flanagan CA, Millar RP 2002 Two mutations in extracellular loop 2 of the human GnRH receptor convert an antagonist to an agonist. Mol Endocrinol 16:1079 –1088 13. Probst WC, Snyder LA, Schuster DI, Brosius J, Sealfon SC 1992 Sequence alignment of the G-protein coupled receptor superfamily. DNA Cell Biol 11:1–20 14. Zhou W, Flanagan C, Ballesteros JA, Konvicka K, Davidson JS, Weinstein H, Millar RP, Sealfon SC 1994 A reciprocal mutation supports helix 2 and helix 7 proximity in the gonadotropin-releasing hormone receptor. Mol Pharmacol 45:165–170 15. Awara WM, Guo CH, Conn PM 1996 Effects of Asn318 and Asp87Asn318 mutations on signal transduction by the gonadotropin-releasing hormone receptor and receptor regulation. Endocrinology 137:655– 662 16. Flanagan CA, Zhou W, Chi L, Yuen T, Rodic V, Robertson D, Johnson M, Holland P, Millar RP, Weinstein H, Mitchell R, Sealfon SC 1999 The functional microdomain in transmembrane helices 2 and 7 regulates expression, activation, and coupling pathways of the gonadotropin-releasing hormone receptor. J Biol Chem 274:28880 –28886 17. Ballesteros J, Kitanovic S, Guarnieri F, Davies P, Fromme BJ, Konvicka K, Chi L, Millar RP, Davidson JS, Weinstein H, Sealfon SC 1998 Functional microdomains in G-protein-coupled receptors. The conserved arginine-cage motif in the gonadotropin-releasing hormone receptor. J Biol Chem 273:10445– 10453 18. Blomenrohr M, Bogerd J, Leurs R, Schulz RW, Tensen CP, Zandbergen MA, Goos HJ 1997 Differences in structure-function relations between nonmammalian and mammalian gonadotropin-releasing hormone receptors. Biochem Biophys Res Commun 238:517–522 19. Flanagan CA, Becker, II, Davidson JS, Wakefield IK, Zhou W, Sealfon SC, Millar RP 1994 Glutamate 301 of the mouse gonadotropin-releasing hormone receptor confers specificity for arginine 8 of mammalian gonadotropin-releasing hormone. J Biol Chem 269:22636 –22641 20. Fromme BJ, Katz AA, Roeske RW, Millar RP, Flanagan CA 2001 Role of aspartate7.32(302) of the human gonadotropin-releasing hormone receptor in stabilizing a high-affinity ligand conformation. Mol Pharmacol 60:1280 –1287


21. Petry R, Craik D, Haaima G, Fromme B, Klump H, Kiefer W, Palm D, Millar R 2002 Secondary structure of the third extracellular loop responsible for ligand selectivity of a mammalian gonadotropin-releasing hormone receptor. J Med Chem 45:1026 –1034 22. Fromme BJ, Katz AA, Millar RP, Flanagan CA 2004 Pro7.33(303) of the human GnRH receptor regulates selective binding of mammalian GnRH. Mol Cell Endocrinol 219:47–59 23. Wang C, Yun O, Maiti K, Oh DY, Kim KK, Chae CH, Lee CJ, Seong JY, Kwon HB 2004 Position of Pro and Ser near Glu7.32 in the extracellular loop 3 of mammalian and nonmammalian gonadotropin-releasing hormone (GnRH) receptors is a critical determinant for differential ligand selectivity for mammalian GnRH and chicken GnRH-II. Mol Endocrinol 18:105–116 24. Blomenrohr M, ter Laak T, Kuhne R, Beyermann M, Hund E, Bogerd J, Leurs R 2002 Chimaeric gonadotropin-releasing hormone (GnRH) peptides with improved affinity for the catfish (Clarias gariepinus) GnRH receptor. Biochem J 361:515–523 25. Fernandez LM, Puett D 1996 Lys583 in the third extracellular loop of the lutropin/choriogonadotropin receptor is critical for signaling. J Biol Chem 271:925–930 26. Zhao MM, Gaivin RJ, Perez DM 1998 The third extracellular loop of the ␤2-adrenergic receptor can modulate receptor/G protein affinity. Mol Pharmacol 53:524 –529 27. Nanevicz T, Wang L, Chen M, Ishii M, Coughlin SR 1996 Thrombin receptor activating mutations. Alteration of an extracellular agonist recognition domain causes constitutive signaling. J Biol Chem 271:702–706 28. Klco JM, Wiegand CB, Narzinski K, Baranski TJ 2005 Essential role for the second extracellular loop in C5a receptor activation. Nat Struct Mol Biol 12: 320 –326 29. Parma J, Van Sande J, Swillens S, Tonacchera M, Dumont J, Vassart G 1995 Somatic mutations causing constitutive activity of the thyrotropin receptor are the major cause of hyperfunctioning thyroid adenomas: identification of additional mutations activating both the cyclic adenosine 3⬘,5⬘-monophosphate and inositol phosphate-Ca2⫹ cascades. Mol Endocrinol 9:725–733 30. Ballesteros JA, Weinstein H 1995 Integrated methods for the construction of three-dimensional models and computational probing of structure-function relations in G protein-coupled receptors. Methods Neurosci 25:366 – 428 31. Chi L, Zhou W, Prikhozhan A, Flanagan C, Davidson JS, Golembo M, Illing N, Millar RP, Sealfon SC 1993 Cloning and characterization of the human GnRH receptor. Mol Cell Endocrinol 91:R1–R6 32. Tensen C, Okuzawa K, Blomenrohr M, Rebers F, Leurs R, Bogerd J, Schulz R, Goos H 1997 Distinct efficacies for two endogenous ligands on a single cognate gonadoliberin receptor. Eur J Biochem 243:134 –140 33. Flanagan CA, Fromme BJ, Davidson JS, Millar RP 1998 A high affinity gonadotropin-releasing hormone (GnRH) tracer, radioiodinated at position 6, facilitates analysis of mutant GnRH receptors. Endocrinology 139:4115– 4119 34. Berg KA, Clarke WP, Sailstad C, Saltzman A, Maayani S 1994 Signal transduction differences between 5-hydroxytryptamine type 2A and type 2C receptor systems. Mol Pharmacol 46:477– 484 35. Pawson AJ, Maudsley SR, Lopes J, Katz AA, Sun YM, Davidson JS, Millar RP 2003 Multiple determinants for rapid agonist-induced internalization of a nonmammalian gonadotropin-releasing hormone receptor: a putative palmitoylation site and threonine doublet within the carboxyl-terminal tail are critical. Endocrinology 144:3860 –3871 36. Samama P, Cotecchia S, Costa T, Lefkowitz RJ 1993 A mutation-induced activated state of the ␤2-adrenergic receptor. Extending the ternary complex model. J Biol Chem 268:4625– 4636 37. Myburgh DB, Millar RP, Hapgood JP 1998 Alanine-261 in intracellular loop III of the human gonadotropin-releasing hormone receptor is crucial for Gprotein coupling and receptor internalization. Biochem J 331(Pt 3):893– 896 38. Spiegel AM 1996 Defects in G protein-coupled signal transduction in human disease. Annu Rev Physiol 58:143–170


3129

39. Kenakin T 2003 Ligand-selective receptor conformations revisited: the promise and the problem. Trends Pharmacol Sci 24:346 –354 40. Millar RP, Pawson AJ 2004 Outside-in and inside-out signaling: the new concept that selectivity of ligand binding at the gonadotropin-releasing hormone receptor is modulated by the intracellular environment. Endocrinology 145:3590 –3593 41. Zhou W, Rodic V, Kitanovic S, Flanagan CA, Chi L, Weinstein H, Maayani S, Millar RP, Sealfon SC 1995 A locus of the gonadotropin-releasing hormone receptor that differentiates agonist and antagonist binding sites. J Biol Chem 270:18853–18857 42. Flanagan CA, Rodic V, Konvicka K, Yuen T, Chi L, Rivier JE, Millar RP, Weinstein H, Sealfon SC 2000 Multiple interactions of the Asp[2.61(98)] side chain of the gonadotropin-releasing hormone receptor contribute differentially to ligand interaction. Biochemistry 39:8133– 8141 43. Davidson JS, McArdle CA, Davies P, Elario R, Flanagan CA, Millar RP 1996 Asn102 of the gonadotropin-releasing hormone receptor is a critical determinant of potency for agonists containing C-terminal glycinamide. J Biol Chem 271:15510 –15514 44. Blomenrohr M, Kuhne R, Hund E, Leurs R, Bogerd J, ter Laak T 2001 Proper receptor signalling in a mutant catfish gonadotropin-releasing hormone receptor lacking the highly conserved Asp(90) residue. FEBS Lett 501:131–134 45. Maliekal JC, Jackson GE, Flanagan CA, Millar RP 1997 Solution conformations of gonadotropin releasing hormone (GnRH) and [Gln(8)]GnRH. S Afr J Chem 50:217–219 46. Hoffmann SH, ter Laak T, Kuhne R, Reilander H, Beckers T 2000 Residues within transmembrane helices 2 and 5 of the human gonadotropin-releasing hormone receptor contribute to agonist and antagonist binding. Mol Endocrinol 14:1099 –1115 47. Chauvin S, Berault A, Lerrant Y, Hibert M, Counis R 2000 Functional importance of transmembrane helix 6 Trp(279) and exoloop 3 Val(299) of rat gonadotropin-releasing hormone receptor. Mol Pharmacol 57:625– 633 48. Hovelmann S, Hoffmann SH, Kuhne R, ter Laak T, Reilander H, Beckers T 2002 Impact of aromatic residues within transmembrane helix 6 of the human gonadotropin-releasing hormone receptor upon agonist and antagonist binding. Biochemistry 41:1129 –1136 49. Li JH, Choe H, Wang AF, Maiti K, Wang C, Salam A, Chun SY, Lee WK, Kim K, Kwon HB, Seong JY 2005 Extracellular loop 3 (EL3) and EL3-proximal transmembrane helix 7 of the mammalian type I and type II gonadotropinreleasing hormone (GnRH) receptors determine differential ligand selectivity to GnRH-I and GnRH-II. Mol Pharmacol 67:1099 –1110 50. Lu ZL, Gallagher R, Sellar R, Coetsee M, Millar RP 2005 Mutations remote from the human gonadotropin-releasing hormone (GnRH) receptor-binding sites specifically increase binding affinity for GnRH II but not GnRH I: evidence for ligand-selective, receptor-active conformations. J Biol Chem 280:29796 – 29803 51. Soderhall JA, Polymeropoulos EE, Paulini K, Gunther E, Kuhne R 2005 Antagonist and agonist binding models of the human gonadotropin-releasing hormone receptor. Biochem Biophys Res Commun 333:568 –582 52. Pfleger KDG, Miller NA, Ott TR, Millar RP, Extracellular loop (ECL) interactions differentially affect GnRH, GnRH II, superagonist and antagonist binding to GnRH receptors. Program of the 84th Annual Meeting of The Endocrine Society, San Francisco, CA, 2002, p 182 (Abstract P1-117) 53. Caunt CJ, Hislop JN, Kelly E, Matharu AL, Green LD, Sedgley KR, Finch AR, McArdle CA 2004 Regulation of gonadotropin-releasing hormone receptors by protein kinase C: inside out signalling and evidence for multiple active conformations. Endocrinology 145:3594 –3602 54. Morgan K, Millar RP 2004 Evolution of GnRH ligand precursors and GnRH receptors in protochordate and vertebrate species. Gen Comp Endocrinol 139:191–197

Endocrinology is published monthly by The Endocrine Society (http://www.endo-society.org), the foremost professional society serving the endocrine community.