Residues in the Membrane-spanning and Extracellular Loop Regions of the Parathyroid Hormone (PTH)-2 Receptor Determine Signaling Selectivity for PTH and PTH-related Peptide*

The parathyroid hormone (PTH)-2 receptor displays strong ligand selectivity in that it responds fully to PTH but not at all to PTH-related peptide (PTHrP). In contrast, the PTH-1 receptor (PTH/PTHrP receptor) responds fully to both ligands. Previously it was shown that two divergent residues in PTH and PTHrP account for PTH-2 receptor selectivity; position 23 (Trp in PTH and Phe in PTHrP) determines binding selectivity and position 5 (Ile in PTH and His in PTHrP) determines signaling selectivity. To identify sites in the PTH-2 receptor involved in discriminating between His5 and Ile5, we constructed PTH-2 receptor/PTH-1 receptor chimeras, expressed them in COS-7 cells, and tested for cAMP responsiveness to [Trp23] PTHrP-(1–36), and to the nondiscriminating peptide [Ile5,Trp23]PTHrP-(1–36) (the Phe23 → Trp modification enabled high affinity binding of each ligand to the PTH-2 receptor). The chimeras revealed that the membrane-spanning/loop region of the receptor determined His5/Ile5 signaling selectivity. Subsequent analysis of smaller cassette substitutions and then individual point mutations led to the identification of two single residues that function as major determinants of residue 5 signaling selectivity. These residues, Ile244 at the extracellular end of transmembrane helix 3, and Tyr318 at the COOH-terminal portion of extracellular loop 2, are replaced by Leu and Ile in the PTH-1 receptor, respectively. The results thus indicate a functional interaction between two residues in the core region of the PTH-2 receptor and residue 5 of the ligand.

The parathyroid hormone (PTH)-2 receptor displays strong ligand selectivity in that it responds fully to PTH but not at all to PTH-related peptide (PTHrP). In contrast, the PTH-1 receptor (PTH/PTHrP receptor) responds fully to both ligands. Previously it was shown that two divergent residues in PTH and PTHrP account for PTH-2 receptor selectivity; position 23 (Trp in PTH and Phe in PTHrP) determines binding selectivity and position 5 (Ile in PTH and His in PTHrP) determines signaling selectivity. To identify sites in the PTH-2 receptor involved in discriminating between His 5 23 3 Trp modification enabled high affinity binding of each ligand to the PTH-2 receptor). The chimeras revealed that the membrane-spanning/loop region of the receptor determined His 5 /Ile 5 signaling selectivity. Subsequent analysis of smaller cassette substitutions and then individual point mutations led to the identification of two single residues that function as major determinants of residue 5 signaling selectivity. These residues, Ile 244 at the extracellular end of transmembrane helix 3, and Tyr 318 at the COOH-terminal portion of extracellular loop 2, are replaced by Leu and Ile in the PTH-1 receptor, respectively. The results thus indicate a functional interaction between two residues in the core region of the PTH-2 receptor and residue 5 of the ligand.

and Ile 5 , we constructed PTH-2 receptor/ PTH-1 receptor chimeras, expressed them in COS-7 cells, and tested for cAMP responsiveness to [Trp 23 ] PTHrP-(1-36), and to the nondiscriminating peptide [Ile 5 ,Trp 23 ]PTHrP-(1-36) (the Phe
The parathyroid hormone (PTH 1 )-2 receptor, a recently identified PTH receptor subtype, responds fully to PTH but not at all to PTHrP (1). This ligand selectivity profile of the PTH-2 receptor is dramatically different from that of the PTH-1 receptor (PTH/PTHrP receptor) which elicits a robust increase in cAMP formation in response to either ligand. The amino acid sequences of the two receptors are 51% identical, and each is a member of the subfamily of G protein-coupled receptors that bind peptide hormones of intermediate size including calcitonin, secretin, glucagon, vasoactive intestinal peptide, and sev-eral other peptides (2), in addition to PTH and PTHrP. These receptors are characterized by a relatively large amino-terminal extracellular domain of 100 -200 amino acids, which contains six highly conserved cysteine residues, a "core" region with seven hydrophobic transmembrane helices and connecting loops, and a carboxyl-terminal tail of 150 -200 amino acids.
The molecular basis by which the peptide hormone receptors engage their respective receptors and trigger receptor activation is still largely unknown. This problem has been approached through strategies involving the construction of receptor chimeras and other types of receptor mutants (3)(4)(5)(6)(7)(8)(9)(10)(11)(12). The data emerging from these studies suggest that multiple segments of the ligand and receptor contribute to the interaction. Recent studies with chimeric ligands acting on chimeric receptors suggest that the carboxyl-terminal portion of the ligand interacts with the amino-terminal extracellular domain, whereas the amino-terminal portion of the hormone interacts with the membrane-spanning/loop region of the receptor (8). However, as yet, there are only limited data on the specific amino acids, either in the ligand or in the receptor, that contribute to the interaction.
In other receptor systems, the availability of receptor subtypes that exhibit distinct pharmacological profiles for different ligands has facilitated the identification of receptor residues involved in ligand recognition (13). The pronounced difference in the ligand selectivity profiles of the PTH-1 and PTH-2 receptors suggested that these two receptors could be used in such an analysis, since the difference in selectivity can most easily be explained by structural differences in the receptors at sites (or a site) that are involved in ligand recognition or ligandinduced receptor activation. Further, it suggests that the two ligands differ at residues that interact with these divergent receptor residues. Recently, two such residues in PTH and PTHrP that can account for their altered selectivity for the PTH-2 receptor were identified: residue 23, Phe in PTHrP and Trp in PTH, modulates ligand binding (14); and residue 5, His in PTHrP and Ile in PTH, modulates ligand-induced receptor activation (14,15). Thus, the weak binding of PTHrP to the PTH-2 receptor can be explained by the presence of Phe 23 , and the weak signaling activity at this receptor can be explained by the histidine at position 5.
In the present paper we use a receptor chimera and mutagenesis approach to search for sites in the PTH-2 receptor that are involved in His 5 /Ile 5 signaling selectivity. The results reveal two divergent amino acids in the membrane-spanning and extracellular loop portion of the receptor that contribute strongly to this effect. 36)amide was described previously (14). Herein, these two peptides are referred to as [Trp 23 ]PTHrP-(1-36) and [Ile 5 ,Trp 23 ]PTHrP-(1-36), respectively. These PTHrP analogs, and other peptides used in the study, were prepared by the biopolymer synthesis facility at Massachusetts General Hospital (Boston, MA), as were the DNA oligonucleotides used in receptor mutagenesis experiments. The PTH analog [Nle 8,21 ,Tyr 34 ]rPTH-(1-34)amide was radioiodinated by the chloramine-T procedure, and the product was purified by reverse phase high performance liquid chromatography (16). 125 I-Na (2,000 Ci/mmol) was purchased from NEN Life Science Products. Dulbecco's modified Eagle's medium, EGTA/trypsin, and concentrated antibiotic mixture (10,000 units/ml penicillin G and 10 mg/ml streptomycin) were from Life Technologies, Inc.; fetal bovine serum was from HyClone Laboratories (Logan, UT). DNA modifying reagents were from United States Biochemical Corp. (Cleveland, OH) or New England Biolabs, Inc. (Beverly, MA).
Receptor Mutagenesis and COS-7 Cell DNA Transfection-The cDNAs encoding the human PTH-1 (PTH/PTHrP) receptor (17) and the human PTH-2 receptor (1) were carried on the expression vectors pcDNA-1 and pcDNAI/Amp (InVitrogen, San Diego, CA), respectively. The 1E2 and 2E1 receptor chimeras were constructed by utilizing the unique, naturally occurring EcoRI site in the PTH-2 receptor plasmid that overlaps codons 139/140. A matching EcoRI site was introduced into the human PTH-1 receptor by oligonucleotide-directed mutagenesis at codons 182/183 to generate the receptor plasmid pHK-FE. The mutagenic oligonucleotide used to make pHK-FE also changed Val 183 and Asp 185 to Phe and Glu, respectively, which correspond to the amino acid sequence of the PTH-2 receptor. The HK-FE and WT PTH-2 receptor plasmids were cleaved with BamHI, which cuts in the 5Ј polylinker region, and EcoRI, and then the appropriate EcoRI-BamHI DNA fragments were gel-purified and religated to yield the desired chimeras. The chimera 1E2 has residues 1-182 of the P1R joined to residues 140 -550 of the P2R, and chimera 2E1 has residues 1-142 of the P2R joined to residues 186 -593 of the P1R. All other cassette and point mutations were introduced by oligonucleotide-directed mutagenesis (18). COS-7 cells were cultured at 37°C in Dulbecco's modified Eagle's medium supplemented with fetal bovine serum (10%), penicillin G (20 units/ml), and streptomycin (20 g/ml) in a humidified atmosphere containing 5% CO 2 . Cells were transfected in 24-well plates using plasmid DNA (200 ng/well) that was purified by cesium chloride/ ethidium bromide gradient centrifugation, except for the initial screening of the cassette mutants, in which phenol-extracted miniprep DNA was used. This miniprep DNA was quantified by ethidium bromide staining of agarose gels, and was transfected at a concentration of 100 ng/well. 48 h after transfection, the cell medium was replenished and the plates were shifted to a 33°C humidified incubator for an additional 24 -48 h, by which time the cell density reached 500,000 Ϯ 100,000 cells/well. This shift to a lower temperature resulted in a general 10 -50% increase in the number of receptors on the cell surface, as compared with cells maintained at 37°C 2 as has been found for other G protein-coupled receptors (19). The cells were then used for binding and cAMP stimulation assays.
Radioligand-receptor Binding-Binding reactions were performed as described previously (18). Each well (final volume ϭ 300 l) contained 26 fmol of 125 I-[Nle 8,21 ,Tyr 34 ]rPTH-(1-34)NH 2 (100,000 cpm) and various amounts (0.4 -300 pmol) of unlabeled competitor ligand; peptides were diluted in binding buffer (50 mM Tris-HCl, pH 7.7, 100 mM NaCl, 5 mM KCl, 2 mM CaCl 2 , 5% heat-inactivated horse serum, 0.5% heatinactivated fetal bovine serum). Incubations were at room temperature for 2 h, except for experiments performed for Scatchard analysis, which were performed at 4°C for 6 h. At the end of the binding reactions the cells were rinsed 3 times with 0.5 ml of binding buffer, lysed with 0.5 ml of 5 M NaOH, and the entire lysate was counted. Intracellular Cyclic AMP-Transfected COS-7 cells were rinsed with 500 l of binding buffer, and 200 l of IBMX buffer (Dulbecco's modified Eagle's medium containing 2 mM 3-isobutyl-1-methylxanthine, 1 mg/ml bovine serum albumin, 35 mM Hepes-NaOH, pH 7.4) and 100 l of binding buffer or binding buffer with various amounts of peptide added. The plates were incubated for 60 min at room temperature; the buffer was then withdrawn and the cells were lysed by adding 0.5 ml of 50 mM HCl and freezing. The diluted lysate (1:30 in distilled H 2 0) was analyzed for cAMP content by radioimmunoassay. For the initial screening of the mutants, we compared the cAMP response of each receptor to a maximum dose (1 M) of [Trp 23 ]PTHrP-(1-36) to its response to the nonselective analog [Ile 5 ,Trp 23 ]PTHrP-(1-36) also at 1 M. Dose-response analyses yielded EC 50 values (ligand dose resulting in 50% of maximum response (E max ) attained by that ligand), which were calculated from plots of log(E/E max Ϫ E) versus log [ligand], where E is the cAMP response measured at the corresponding dose of ligand (6).

RESULTS
The difference in the ligand selectivities of the PTH-1 receptor and the PTH-2 receptor can be seen in the cAMP response profiles shown in Fig. 1, panels A   . The ligand selectivity of the PTH-2 receptor is not due to a difference in binding affinities because the two analogs exhibited comparable potencies in their ability to inhibit the binding 125 I-[Nle 8,21 ,Tyr 34 ]rPTH-(1-34)amide (Fig. 1F). The high apparent binding affinity that the two PTHrP analogs displayed for the PTH-2 receptor in these experiments is due primarily to the Phe 23 3 Trp modification; this substitution of a PTHrP residue by the corresponding PTH residue markedly enhances binding potency at the PTH-2 receptor without affecting cAMP signaling (14). A small improvement in binding potency was also seen for the His 5 3 Ile modification, however, this improvement was not receptor specific (Fig. 1, E and F) and was much smaller in magnitude than the effect of the substitution on PTH-2 receptor signaling.
To localize the region of the PTH-2 receptor involved in His 5 /Ile 5 signaling selectivity, we constructed a pair of chimeras in which the amino-terminal extracellular domains of the human PTH-1 and PTH-2 receptors were reciprocally interchanged and tested the chimeras in cAMP stimulation assays for responsiveness to [Trp 23 ]PTHrP-(1-36) and to the nondiscriminating control peptide [Ile 5 ,Trp 23 ]PTHrP-(1-36). The 1E2 receptor chimera, which has the amino-terminal extracellular domain of the PTH-1 receptor connected (via an EcoRI site) to the mid-and carboxyl-terminal region of the PTH-2 receptor, discriminated between the two ligands ( Fig. 1C), whereas the reciprocal chimera 2E1 responded fully to each ligand (Fig. 1D). These results indicated that the membrane-spanning and loop portion of the receptor determines His 5 /Ile 5 signaling selectivity.
To further localize the sites involved in residue 5 selectivity, we replaced most of the divergent residues in the membranespanning helices and extracellular connecting loops of the PTH-2 receptor with the corresponding residues of the PTH-1 receptor. As shown in Fig. 2, these residues were replaced either by cassette substitution or, for two of the sites (Ala 325 and Gly 327 ), by single residue point mutation. These three cassettes are predicted to be located in the aminoterminal portion of extracellular loop 1 (P2R-Csst#5), the extracellular end of helix 3 (P2R-Csst#8), and at the COOHterminal end of extracellular loop 2 (P2R-Csst#13) (Fig. 2). For each of these three receptor mutants the maximum binding of 125 I-[Nle 8,21 ,Tyr 34 ]rPTH-(1-34)amide was elevated by a factor of 1.8 -3.2, in comparison to the binding observed for the WT PTH-2 receptor (Table I) (Table I). We  (Fig. 3B). Within cassette region 13, one other substitution Tyr 318 3 Ile (YI-318), also enhanced responsiveness to the PTHrP analog. Dose response analysis of these two mutant receptors demonstrated that each mutation by itself could account for a substantial component of the signaling selectivity inherent to the PTH-2 receptor (Fig. 4). In fact, with the IL-244 receptor, the efficacy of [Trp 23 ]PTHrP-(1-36) was equal to, if not slightly greater than, that of [Ile 5 ,Trp 23 ]PTHrP-(1-36) (Fig. 4B and Table II). The effect of the YI-318 mutation was not as pronounced, and with this mutant [Trp 23 ]PTHrP-(1-36) functioned as a partial agonist (Fig. 4C). These two point mutations did not affect the ability of either PTHrP analog to inhibit the binding of 125 I-[Nle 8,21 ,Tyr 34 ]rPTH-(1-34)amide (Fig. 4 D-F). Scatchard analyses indicated that the number of PTH-(1-34) binding sites on the surface of COS-7 cells transfected with either mutant receptor did not differ significantly from the number of binding sites on cells expressing the WT PTH-2 receptor (Table III).
To examine whether the ability to discriminate between His and Ile at position five could be conferred to the PTH-1 receptor, we introduced the reciprocal point mutations at the corresponding loci in the PTH-1 receptor and tested for effects on [Trp 23 ]PTHrP-(1-36) signaling. Neither mutation, Leu 289 3 Ile (LI-289) nor Ile 363 3 Tyr (IY-363), had any effect on His 5 signaling selectivity, as the two mutant receptors exhibited full responsiveness to [Trp 23 ]PTHrP-(1-36) (Fig. 5, A-C). Interestingly, a position 5-specific effect of the IY-363 mutation was apparent in competition binding studies; the two PTHrP analogs displayed nearly equal potency in binding to the IY-363 mutant receptor ( Fig. 5F and Table IV) Table IV).
Earlier studies with the PTH-1 receptor have indicated that PTH and PTHrP bind to overlapping sites (21)(22)(23). To examine whether the sites Ile 244 and Tyr 318 in the PTH-2 receptor also have position 5-specific effects with PTH analogs, we studied the binding and signaling properties of [His 5 ]PTH-(1-34) ([His 5 ,Tyr 34 ]PTH-(1-34)amide) with the IL-244 and YI-318 mutant receptors. This PTH analog is inactive in cAMP assays with the WT PTH-2 receptor, although it binds with adequate affinity (IC 50 ϭ 250 nM) (14). Both the IL-244 and YI-318 mutations caused at least a partial rescue of the signaling defect of [His 5 ]PTH-(1-34) (Fig. 6, A-C). As was observed for the PTHrP analog, the corresponding mutations in the PTH-1 receptor had no effect on [His 5 ]PTH-(1-34) signaling (Fig. 6, D-F), and the binding potencies that [His 5 ]PTH-(1-34) displayed with the mutant PTH-2 or PTH-1 receptors were unaltered from the binding potencies seen with the corresponding WT receptor (data not shown). DISCUSSION As part of our efforts to understand the mechanisms of ligand interaction in PTH receptors, we are exploring the molecular basis for the unique ligand selectivity of the PTH-2 receptor, which responds to PTH but not PTHrP. In the first stage of these studies, we examined the ligands for amino acid residue divergences that could explain this selectivity. We identified position 23 as the major determinant of binding selectivity, and position 5 as the major determinant of signaling selectivity (14). In the present study we sought to identify the sites in the PTH-2 receptor that enable it to discriminate, on the basis of cAMP signaling, between His and Ile at position 5. The receptor residues involved in this effect were of particular interest, because it seemed likely that they would be at, or near, sites involved in triggering the signal transduction mechanism.
The two PTHrP analogs used in our analysis differed by having His or Ile at position 5, and were thus nonfunctional or functional with the WT PTH-2 receptor. Each of the two analogs contained the Phe 3 Trp modification at position 23, which enabled high affinity binding of either analog to the PTH-2 receptor (14). The role of residue 5 in signaling selectivity was also demonstrated by Behar st al. (15). Importantly, this modification does not influence cAMP signaling, as demonstrated by the ability of [Trp 23  tive histidine at position 5) to function as a PTH-2 receptorselective antagonist (14). The Trp 23 modification thus permitted a direct analysis of the effects of residue 5 on signaling without the need to correct for weak binding affinities.
The cAMP responses of the two reciprocal PTH-1/PTH-2 receptor chimeras showed that the major determinants of His 5 / Ile 5 signaling selectivity mapped to the portion of the receptor containing the membrane-spanning helices and connecting loops. Cassette mutagenesis then revealed three segments that affected this selectivity; these were located in extracellular loop 1, transmembrane helix 3, and extracellular loop 2. Scanning mutagenesis analysis of the extracellular loop 1 segment failed to reveal a single site that led to improved responsiveness to [Trp 23 ]PTHrP-(1-36), a finding that suggests that multiple residues in this region might be involved in residue 5 recognition. Point mutational analysis of the segments in transmembrane helix 3 and extracellular loop 2 successfully revealed two amino acids that strongly affected PTHrP signaling; Ile 244 near the extracellular end of transmembrane helix 3 and Tyr 318 near the carboxyl-terminal end of extracellular loop 2 (Fig. 2). Changing either site in the PTH-2 receptor to the corresponding residue of the PTH-1 receptor resulted in a pronounced gain-of-function phenotype, as demonstrated by a selective increase in responsiveness to the [Trp 23 ]PTHrP-(1-36) analog.
The molecular basis by which Ile 244 and Tyr 318 affect ligand selectivity is not known. The three-dimensional structure of the PTH-2 receptor, or any G protein-coupled receptor, with the exception of rhodopsin (24), has not been determined. The two-dimensional schematic of the core region of the PTH-2 receptor shown in Fig. 2 is based mainly on evolutionary and hydropathy analyses of the primary structure (25). Although the end points of the seven membrane-spanning helices have not been firmly established, it is predicted that Ile 244 and Tyr 318 lie at, or close to, the boundary of the extracellular fluid and the lipid membrane. These two sites could thus be in a reasonable position for interacting with the ligand. Peptidebinding sites in another class of peptide hormone receptors, the tachykinin receptors, have been mapped to similar locations (13). Our functional data do not exclude the possibility that the mutations at Ile 244 and Tyr 318 have allosteric effects on other residues, but global changes in receptor structure seem unlikely, because there was little or no effect of the mutations on  One question to consider in these studies is whether the PTH-2 receptor and the PTH-1 receptor engage their ligands in a similar fashion. In an effort to address this question we introduced reciprocal mutations, LI-289 and YI-363, into the PTH-1 receptor and tested the mutants for the ability to discriminate between analogs with His or Ile at position 5. No effect on cAMP signaling responsiveness was detected for either receptor mutation; we were thus unable to determine whether equivalent sites in the PTH-1 and PTH-2 receptors are involved in residue 5 recognition. That neither mutation was sufficient for conferring His 5 /Ile 5 signaling selectivity to the PTH-1 receptor indicates that multiple PTH-2 receptor residues are required for this effect, as was suggested by the initial cassette mutagenesis studies in which mutations at three distinct sites led to a loss of selectivity (Fig. 3).
The bioactive regions of PTH and PTHrP differ considerably in primary structure (26), yet most studies indicate that the two ligands bind to the same site in the PTH-1 receptor (21)(22)(23). To examine whether the PTH-2 receptor sites that we identified here also influence interactions with position 5 of PTH, we studied the binding and signaling properties of [His 5 ]hPTH-(1-34). This analog is inactive in cAMP assays with the PTH-2 receptor, though, it binds with adequate affinity (14). As with PTHrP, both the IL-244 and YI-318 mutations were able to rectify the signaling defect of [His 5 ]PTH-(1-34) (Fig. 6). These results suggest that the histidine at position 5 in both [Trp 23 ]PTHrP-(1-36) and [His 5 ]PTH-(1-34) is recognized by the same region of the PTH-2 receptor.
The isoleucine at position 244 is conserved in the rat and human PTH-2 receptor (1,27), and leucine is preserved at the homologous site in each PTH-1 receptor (Xenopus, rat, mouse, human, porcine, and opossum). The same pattern of evolutionary preservation is seen for the Tyr 318 site in extracellular loop 2. Residue 5 in the ligands also shows this trend; the polar histidine is found here in all PTHrP ligands, and a hydrophobic isoleucine or methionine residue is found at the corresponding site in each vertebrate PTH sequence. It may be that the two receptor sites and the cognate residue 5 of the ligands are under the same evolutionary constraints. These constraints would ensure that PTHrP specifically interacts with the PTH-1 receptor to mediate its biological actions, including the regulation of embryonic bone development (28,29) and not with the PTH-2 receptor, which is expressed in several different tissues, albeit with unknown functional consequences (27). Whether PTH is the actual ligand for the PTH-2 receptor is unknown; recent evidence suggests that the hypothalamus contains a novel peptide that selectively activates the PTH-2 receptor (30) For the PTH receptors, and other members in this same peptide hormone receptor family, the amino-terminal extracellular domain has been shown to play an important role in ligand-binding affinity and specificity (3, 5, 8 -10, 31, 32). Several other studies on these receptors have implicated the extracellular loops or transmembrane helices in ligand binding or signaling interactions (6 -8, 10, 11). Our present studies with and [Ile 5 ,Trp 23 ,Tyr 36 ]PTHrP-(1-36)NH 2 (q). Intracellular cAMP and competition binding assays were performed as described under "Experimental Procedures," and in the legend to Fig. 4. The graphs show data that were combined from seven to nine independent experiments (mean Ϯ S.E.) each performed in duplicate.  (7) 27 Ϯ 5 (7) P1R-IY-363 29 Ϯ 6 (7) 24 Ϯ 5 (7) the PTH-2 receptor provide additional information on the functional map of the ligand interaction surface of the receptor, as they identify specific residues in helix 3 and extracellular loop 2 that modulate the signaling selectivity determined by residue 5 in the ligand.