Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

The parathyroid hormone (PTH) 2 receptor from various species (human, rat, and zebrafish) is potently activated by a recently identified neuropeptide, tuberoinfundibular peptide of 39 residues (TIP39) (Usdin et al., 1999; Hoare et al., 2000b). The physiological role of this new peptide-receptor system is currently being investigated. The distribution of the PTH2 receptor in spinal cord and its presence in dorsal root ganglion neurons suggests a role in pain perception (Usdin et al., 1999; Wang et al., 2000). The receptor is abundantly expressed in the median eminence and periventricular nucleus of the hypothalamus, suggesting PTH2 receptor involvement in the regulation of pituitary function (Usdin et al., 1999; Wang et al., 2000). The receptor belongs to the type II family of G protein-coupled receptors, which respond to peptide modulators such as secretin, glucagon, calcitonin, vasoactive intestinal polypeptide, and corticotropin-releasing hormone.

The PTH2 receptor and TIP39 form part of an extended family of related receptors and related ligands that also includes the PTH1 receptor, PTH, and PTH-related protein (PTHrP) (Hoare and Usdin, 2001). The ligands and receptors have presumably evolved to selectively mediate different physiological functions. In this regard, the PTH2 receptor is potently activated by TIP39 but not by PTHrP (Hoare and Usdin, 2001) and the PTH1 receptor is activated by PTH and PTHrP but not by TIP39 (Jüppner et al., 1991; Hoare and Usdin, 2001). The human PTH2 receptor was initially identified as a receptor for PTH based on potent PTH-stimulated cAMP accumulation via the receptor expressed in transfected cells (Usdin et al., 1995). In contrast the rat PTH2 receptor is poorly activated by PTH; the peptide is a low-potency (20-100 nM) partial agonist for stimulation of cAMP production in transfected cells (Hoare et al., 1999; Usdin et al., 1999). The cAMP response to PTH is weak or undetectable for the rat PTH2 receptor expressed endogenously in F-11 cells, a dorsal root ganglion neuron-like cell line (Usdin et al., 1999, 2000). These findings suggest that PTH is not a physiologically significant activator of the PTH2 receptor in rats.

The human PTH2 receptor has been shown to couple to increases of intracellular calcium concentration ([Ca²⁺]_i) and inositol phosphates in response to PTH (Behar et al., 1996; Takasu et al., 1998) but the responsiveness of this signaling pathway to TIP39 has not been investigated. In addition, the possible coupling of the rat PTH2 receptor to increases of [Ca²⁺]_i has not been examined. This is important because the rat and human PTH2 receptors differ in their pharmacological profiles, and the investigation of the PTH2 receptor's physiological roles is likely to be performed in rodents. The binding profile of the rat PTH2 receptor also has not been examined (prior to the isolation of TIP39, a radioligand has not been available for this receptor). It is not known whether the weak PTH activation of the rat PTH2 receptor is in part due to a low binding affinity. The first aim of this study was therefore to extend the comparison of human and rat PTH2 receptors by measuring the [Ca²⁺]_i response and ligand binding affinity for PTH and TIP39.

Previous studies have used differences of ligand pharmacology between closely related receptors to investigate the functional role of molecular elements in ligand recognition and receptor activation (Holtmann et al., 1995; Stroop et al., 1995; Bergwitz et al., 1996, 1997; Couvineau et al., 1996; Turner et al., 1996, 1998; Clark et al., 1998; Dautzenberg et al., 1998; Hoare et al., 2000a). For the TIP39-PTH2 receptor interaction, the large extracellular N-terminal domain (N-domain) of the receptor is involved in TIP39 binding but not receptor activation. The membrane embedded "juxtamembrane" domain (J-domain) is a determinant of receptor activation by TIP39 and also contributes to its binding affinity (Hoare et al., 2000a). Previous studies have also identified regions and sequences within the human PTH2 receptor that limit its interaction with PTHrP (Bergwitz et al., 1997; Clark et al., 1998; Turner et al., 1998). However, the functional role of PTH2 receptor regions in PTH binding and signaling has not been investigated. We used chimeric human/rat PTH2 receptors to identify PTH2 receptor regions involved in PTH binding and signaling.

Moderate-affinity antagonists have been developed for the human PTH2 receptor by modification of the N-terminal region of PTH or TIP39 (Behar et al., 1996; Gardella et al., 1996; Hoare et al., 2000a) [e.g., PTH(7-34) and TIP(7-39)]. A high-affinity antagonist for the rat receptor would be very useful for investigating the physiological role of the PTH2 receptor. In this study we compared antagonist binding affinity for human and rat PTH2 receptors and found that the antagonists tested bound the rat receptor with low affinity. We used chimeric human/rat PTH2 receptors to identify the molecular basis of the weakened interaction of antagonist ligands with the rat PTH2 receptor. In addition, the receptor determinants for agonist binding [PTH(1-34) and TIP39] were compared with those for the corresponding N-terminally truncated antagonists [PTH(7-34) and TIP(7-39)], to examine the possible changes of receptor-ligand interaction involved in the loss of activation produced by truncation of the ligand. These studies provided insight into the molecular basis of antagonism of rat and human PTH2 receptors, which should contribute to the development of a high-affinity antagonist for the rat PTH2 receptor.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Reagents and Peptides. All peptides were purchased from Bachem (Torrance, CA), Peninsula Laboratories (Belmont, CA), or Anaspec (San Jose, CA), except for mTIP39 (Biomeasure, Milford, MA) and mTIP(7-39) (Biomolecules Midwest, Waterloo, IL). Peptides were dissolved in 10 mM acetic acid at a concentration of 1 mM. Aliquots of 3 µl were stored at 80 C and used once. The letters b, h, m, and r designate the peptide sequence as bovine, human, mouse, and rat, respectively. The peptides used in this study were hPTH(1-34), rPTH(1-34), bTIP39, bTIP(7-39), mTIP39 amide, mTIP(7-39) amide PTHrP(1-34), [D-Tryp¹², Tyr³⁴]bPTH(7-34) amide, and [Tyr³⁴]PTHrP(1-21)/hPTH(22-34) amide. Cell culture supplies were obtained from Invitrogen (Carlsbad, CA), except for DMEM, which was from Mediatech (Herndon, VA). ¹²⁵I-cAMP was obtained from PerkinElmer Life Science Products (Boston, MA). ¹²⁵I-labeled hTIP39 and mTIP(7-39) was prepared using the lactose peroxidase method followed by high-performance liquid chromatography purification (Hoare et al., 2000a). The aequorin cDNA in pCDM8 (Button and Brownstein, 1993) was kindly provided by Dr. Don Button (Roche Pharmaceuticals, Palo Alto, CA).

Identification of Human and Mouse TIP39 Amino Acid Sequences. Genomic sequence encoding mouse (accession no. AC073740) and human (accession no. AC068670) TIP39 were identified by searching National Center for Biotechnology Information high throughput gene sequence databases with the program BLAST-P (Altschul et al., 1990) with the peptide sequence of purified bovine TIP39.

Plasmid Constructions. Chimeric human/rat PTH2 receptors were constructed by transferring restriction digest fragments between human and rat PTH2 receptors (Fig. 1). The restriction sites used were BamHI at 758 base pairs of the human receptor and BstZ17I at 441 base pairs. A BstZ17I site was engineered into the rat PTH2 receptor by mutation of C438 to thymidine by using the GeneEditor Site-Directed Mutagenesis system (Promega, Madison, WI) according to the manufacturer's protocol. To simplify some of the exchanges, the human PTH2 receptor was subcloned as a KpnI/BglII digestion product from pcDNAI/Amp (Usdin et al., 1995) into KpnI/BamHI-digested pcDNA3.1/Amp (+) (Invitrogen).

View larger version (57K): [in this window] [in a new window]   Fig. 1.   Topographical representation of the PTH2 receptor. The amino acid sequence of the human PTH2 receptor is shown, with rat receptor residues that differ from the human sequence indicated by the squares. Putative membrane-spanning -helices are shown, determined by visual inspection of hydrophobicity plots. The locations of the two restriction sites used to construct the chimeric receptors (BstZ17I and BamHI) are indicated by the solid bars.

Cell Culture and Transient Expression in COS-7 Cells. COS-7 cells were grown and transfected as previously described, using 10-cm tissue culture dishes and 10 µg of plasmid DNA (Clark et al., 1998). For assays of intracellular calcium cells were cotransfected with 10 µg of receptor plasmid DNA and 10 µg of pCDM8.AEQ (Button and Brownstein, 1993) (encoding the jellyfish calcium-sensing protein aequorin). The following day, for cAMP and radioligand binding assays, cells were transferred after trypsinization to 96-well plates at a density of 50,000 cells/well.

Measurement of Cellular Levels of cAMP. After removal of medium, transfected COS-7 cells were treated at 37°C with 50 µl/well cAMP assay buffer [DMEM containing 25 mM HEPES supplemented with 0.1% bovine serum albumin, 30 µM Ro 20-1724 (Sigma/RBI, Natick, MA), 100 µM 4-(2-aminoethyl)benzenesulfonyl fluoride (AEBSF), and 1 µg/ml bacitracin]. After 40 min this buffer was removed and replaced with 40 µl of the same supplemented buffer. Test agents were added in a volume of 10 µl and the cells incubated for an additional 40 min at 37°C. The assay was then terminated by the addition of 50 µl 0.1 N HCl, 0.1 mM CaCl₂. cAMP was quantified using a radioimmunoassay as previously described (Clark et al., 1998).

Measurement of Intracellular Calcium. Intracellular calcium was measured using aequorin luminescence (Button and Brownstein, 1993). Three days after transfection, COS-7 cells coexpressing aequorin and receptor cDNA were washed once with phosphate-buffered saline then loaded for 2 h with 2.5 µM coelenterazine hcp (Molecular Probes, Eugene, OR) in DMEM supplemented with 0.1% fetal bovine serum, 30 µM glutathione (reduced form) and 25 mM HEPES. Cells were then washed with Dulbecco's phosphate-buffered saline containing 1 mM Ca²⁺ and 1 mM Mg²⁺ supplemented with 1% bovine serum albumin (DPBS buffer) and then dislodged in a 10-ml volume of the same buffer by gentle pipetting. The cell number was adjusted to 20,000 to 50,000 cells/ml and 0.2 ml cells added to 12 × 75-mm tubes. Test agents were made up in DPBS buffer supplemented with 1 µg/ml bacitracin and 100 µM AEBSF. Baseline luminescence of the cells was measured for approximately 10 s in a EG&G Berthold Lumat LB9507 luminometer followed by manual addition of 50 µl test agent and immediate initiation of luminescence measurement. After a further 60 s (during which time luminescence for all the experimental conditions returned to baseline) 50 µl of 0.05% Triton X-100 was added, using an automatic injector, to permeabilize the cells. The remaining aequorin activity was measured for 50 s, during which time luminescence returned to baseline. Measurements of relative light units (RLU) were recorded every second. In each assay duplicate time courses were measured for each test agent. The calcium response was quantified as the percentage of total RLU, the summed RLU in response to ligand in 60 s divided by the total RLU (that produced by ligand added to that produced by saturation of aequorin with Ca²⁺ for 50 s after addition of Triton X-100). We initially attempted this assay with HEK293T cells but observed a substantial increase of [Ca²⁺]_i in nontransfected HEK293T cells in response to PTH and PTHrP but not TIP39, suggesting the presence of an endogenous PTH1 receptor.

Whole-Cell Radioligand Binding Assays. Binding of unlabeled ligands was measured by displacement of ¹²⁵I-hTIP39 binding to PTH2 receptor-expressing COS-7 cells in 96-well plates. Cells were washed once with 100 µl of binding buffer (50 mM Tris, 100 mM NaCl, 5 mM KCl, 2 mM CaCl₂, pH 7.5 with HCl, supplemented with 5% heat-inactivated horse serum, 0.5% fetal bovine serum, 1 µg/ml bacitracin, and 100 µM AEBSF). To each well was added sequentially 65 µl of binding buffer, 10 µl of unlabeled ligand diluted in binding buffer, and 25 µl of radioligand diluted in binding buffer (approximately 50,000-100,000 cpm/well). Total binding was defined in the absence of unlabeled ligand and nonspecific binding was measured in the presence of 1 µM TIP39. Cells were incubated at 15°C for 3 h. The assay plates were placed on ice for 10 min and then washed twice with 100 µl/well binding buffer. Cell-associated radioactivity was extracted with 200 µl of 1.0 N NaOH. Samples were transferred to tubes and radioactivity measured in a Wallac 1470 Wizard gamma counter. Nonspecific binding was 3 to 5% of the total counts added. Total binding was less than 20% of added radioactivity.

Data Analysis. Concentration dependence data was analyzed using the following four-parameter logistic equation with Prism 2.01 (GraphPad Software, San Diego, CA):
where X is the logarithm of the ligand concentration and n is Hill slope. For cAMP accumulation data y is the amount of cAMP accumulated at a given peptide concentration, min is the cAMP level in the absence of ligand, max is the maximum level produced, and K is the logEC₅₀. For intracellular calcium, y is the percentage of total RLU produced by a given ligand concentration, min is the percentage of total RLU produced in the absence of ligand, max is the maximal ligand-stimulated response, and K is the logEC₅₀. For inhibition of radioligand binding, y is the cpm bound at a given unlabeled ligand concentration, min is nonspecific binding (measured in the presence of 1 µM hTIP39), max is total binding (measured in the absence of unlabeled ligand), and K is the logIC₅₀.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Comparison of Mouse and Human/Bovine TIP39. The sequence of bovine TIP39 has been previously described (Usdin et al., 1999). Database searching revealed that the amino acid sequence of human TIP39 (hTIP39) is identical to that of bovine TIP39 and that mouse TIP39 (mTIP39) differs at four positions in the carboxyl-terminal portion of the peptide. In mTIP39, Arg replaces His24 of the human/bovine sequence, Asp replaces Asn27, Gln replaces His 31, and Leu replaces Val35. In cAMP accumulation assays mTIP39 was equivalently active to hTIP39 for both the human (Fig. 3A) and rat (Fig. 3B) PTH2 receptors (see legend to Fig. 3).

Coupling of Human and Rat PTH2 Receptors to Increases of Intracellular Calcium. TIP39 and PTH peptides produced a rapid, transient elevation of [Ca²⁺]_i in COS-7 cells expressing the human PTH2 receptor (Fig. 2A). hPTH(1-34) was a partial agonist compared with the maximal effect of hTIP39. hPTH(1-34) was also less potent, with a 5.8-fold higher EC₅₀ compared with the EC₅₀ of hTIP39 (3.8 nM) (Fig. 3C; Table 1). rPTH(1-34) was less potent (EC₅₀ = 16 nM) than hTIP39 but was a near-full agonist. mTIP39 was equivalently active to hTIP39 (Fig. 3C; Table 1). PTH is therefore a weaker agonist than TIP39 for increasing [Ca²⁺]_i via the human PTH2 receptor. For the rat PTH2 receptor TIP39 (Fig. 2B) produced a rapid, transient increase of [Ca²⁺]_i, similar to that observed for the human receptor (Fig. 2A). The concentration dependence relationship for hTIP39 and mTIP39 was again indistinguishable (Fig. 3D; Table 1). However, no detectable increase of [Ca²⁺]_i was observed for either rPTH(1-34) (Figs. 2B and 3D) or hPTH(1-34) (Fig. 3D) when tested at the high concentration of 10 µM. In addition, no increase was detectable for a lower concentration of rPTH(1-34) or hPTH(1-34) (100 nM) and for 320 nM hPTH(1-84) (data not shown). Therefore, PTH does not detectably affect [Ca²⁺]_i via the rat PTH2 receptor expressed in COS-7 cells. The mTIP39-stimulated increase of [Ca²⁺]_i was blocked by the phospholipase C (PLC) inhibitor U73122 (Fig. 2C), unaffected by the inactive analog U73343 (data not shown), and unaffected by overnight preincubation of the cells with 100 ng/ml pertussis toxin (Fig. 2C). The [Ca²⁺]_i response to the PTH2 receptor therefore involves PLC and a pertussis toxin-insensitive G protein. As expected, PTHrP(1-34) did not stimulate [Ca²⁺]_i via either human or rat PTH2 receptors (Fig. 3, C and D). hTIP39 did not detectably increase [Ca²⁺]_i via the human PTH1 receptor expressed in COS-7 cells (data not shown). None of the ligands tested caused a significant elevation of [Ca²⁺]_i in COS-7 cells expressing -galactosidase (Fig. 2D).

View larger version (46K): [in this window] [in a new window]   Fig. 2.   Time course of intracellular calcium concentration in COS-7 cells expressing human and rat PTH2 receptors in response to TIP39 and PTH. Luminescence was measured as described under Materials and Methods for COS-7 cells coexpressing aequorin and the human PTH2 receptor (A), rat PTH2 receptor (B and C), or -galactosidase (D). Luminescence (RLU) was recorded at 1-s intervals, except for a 2-s delay after the time of addition of ligand or buffer alone (indicated by the arrow). The PLC inhibitor U73122 was preincubated with the cells for 20 min at room temperature prior to addition of mTIP39. Pertussis toxin (PTX, 100 ng/ml) was incubated with the cells overnight prior to the assay. Data were normalized as the percentage of the total RLU (where total RLU is the summed RLU produced by the ligand added to the summed RLU produced by permeabilization of the cells with Triton X-100). Data points are the mean ± range of measurements from duplicate time courses. Data are from representative experiments that were performed two to five times with similar results.

View larger version (43K): [in this window] [in a new window]   Fig. 3.   Agonist pharmacology of human and rat PTH2 receptors expressed in COS-7 cells. Agonist concentration dependence relationships were measured for stimulation of cAMP accumulation (A and B), increased intracellular calcium (C and D), and receptor binding (E and F). Data for the human PTH2 receptor are presented in A, C, and E, and the profile of the rat receptor is presented in B, D, and F. The assays were performed as described under Materials and Methods. A and B, cAMP was measured after a 40-min incubation with the ligands and the data expressed as a percentage of the maximal response to hTIP39, which was included as a reference at 1 µM concentration for each concentration dependence assay. The logEC50 and Emax values for hTIP39 and hPTH(1-34) are given in Table 3. For the human PTH2 receptor the logEC50 values for mTIP39 and rPTH(1-34) were, respectively, 9.32 ± 0.16 and 9.84 ± 0.08 with corresponding Emax values of 97 ± 6 and 118 ± 13%. For the rat PTH2 receptor the logEC50 values for mTIP39 and rPTH(1-34) were 9.29 ± 0.10 and 7.43 ± 0.24 with corresponding Emax values of 89 ± 6 and 38 ± 5%. C and D, [Ca2+]i was measured as the summed RLU over a 60-s time course and the data normalized and analyzed as described under Materials and Methods. E and F, binding of unlabeled agonist ligands was measured by displacement of 125I-hTIP39 binding to intact COS-7 cells expressing the receptors. In all cases, the data points are the mean ± range of duplicate measurements. The curves are fits to a four-parameter logistic equation. Data are from representative experiments that were performed three to seven times with similar results.

View this table: [in this window] [in a new window]   TABLE 1 Pharmacological characterization of PTH2 receptor coupling to increases of intracellular calcium and cAMP in COS-7 cells [Ca2+]i was measured as luminescence of the calcium-binding protein aequorin, which was coexpressed with either human or rat PTH2 receptors in COS-7 cells. Aequorin luminescence was measured and quantified as described under Materials and Methods and in the legend to Fig. 2. The values were fitted to a four-parameter logistic equation to obtain estimates of EC50 and Emax. The Emax values have been normalized to the Emax of hTIP39, which was determined in each experiment as the response to a 10 µM concentration of this ligand. The Emax value of hTIP39 was 22 ± 1% of total RLU and 18 ± 2% of total RLU for human and rat PTH2 receptors, respectively. A large variation was observed between the logEC50 values from different transfections. To determine statistical significance hTIP39 was included as a reference for each transfection and differences between the EC50 of hTIP39 and that of the other peptides tested using a paired t test. cAMP EC50 values are given in Table 3 and in the legend to Fig. 3. Data are the mean ± S.E.M. from three to five independent experiments.

Ligand Binding to Human and Rat PTH2 Receptors. For both cAMP and [Ca²⁺]_i signaling PTH activation of the rat PTH2 receptor is weaker than PTH activation of the human receptor. We measured receptor binding of the ligands, to determine the extent to which a difference of binding affinity between the receptors contributes to differential PTH responsiveness. In addition, several moderate-affinity antagonists have been described for the human PTH2 receptor. We measured their binding affinity for the rat receptor to assess their utility as antagonists for this receptor.

View larger version (27K): [in this window] [in a new window]   Fig. 4.   Antagonist binding to human and rat PTH2 receptors expressed in COS-7 cells. Binding of unlabeled antagonist ligands was measured by displacement of 125I-hTIP39 binding to intact COS-7 cells expressing the human PTH2 receptor (A) and the rat PTH2 receptor (B). Data points are the mean ± range of duplicate measurements. The curves are fits to a four-parameter logistic equation. Data are from representative experiments that were performed three to five times with similar results.

Regions of Human and Rat PTH2 Receptors Specifying Differential PTH-Stimulated cAMP Production. PTH from various species is a low-potency partial agonist for the rat PTH2 receptor, relative to TIP39, but acts as a highly potent full agonist for the human receptor (for review, see Hoare and Usdin, 2001). Chimeric rat/human PTH2 receptors were used to identify regions in the PTH2 receptor that contributed to this differential responsiveness to PTH. The amino acid sequences of human and rat PTH2 receptors display considerable homology (82%; Fig. 1). Within the predicted transmembrane domains sequence identity is particularly high (95%) and the substitutions are nearly all conservative (Fig. 1). There are differences within the predicted extracellular regions: N-domain, EL1, EL2, and EL3. The large, 40-amino acid EL1 is particularly divergent, with only 65% sequence identity (Fig. 1). We constructed chimeric receptors in which extracellular regions were exchanged between the human and rat PTH2 receptors, because these regions are most likely involved in receptor-ligand interaction. Receptors were constructed that exchanged the N-domain (hN-rPTH2 and rN-hPTH2), a region from TM1 to TM3 conserved except for EL1 (hEL1-rPTH2 and rEL1-hPTH2), and a region from TM3 to the C terminus, including EL2 and EL3 (rNTM3-hPTH2 and hNTM3-rPTH2) (for systematic purposes the nomenclature describes the insertion of the N-terminal-most receptor portion of one species homolog into the remainder of the other species homolog). For all the PTH2 receptors, chimeric and wild-type, the hTIP39 efficacy (E_max) was similar, suggesting a similar level of cell surface expression (Table 3; Fig. 5). The hTIP39 potency (EC₅₀) was also similar (Table 3; Fig. 5), suggesting that the conformation of the chimeric PTH2 receptors was not significantly disrupted.

View larger version (42K): [in this window] [in a new window]   Fig. 5.   hPTH(1-34) and hTIP39-stimulated cAMP accumulation via chimeric human/rat PTH2 receptors expressed in COS-7 cells. cAMP was measured as described under Materials and Methods. Data were normalized as a percentage of the maximal response to hTIP39, which was included as a reference at 1 µM concentration for each hPTH(1-34) concentration dependence assay. Data points are the mean ± range of duplicate measurements. The curves are fits to a four-parameter logistic equation. Data are from representative experiments that were performed three to seven times with similar results.

Regions of Human and Rat PTH2 Receptors Specifying Differential hPTH(1-34) and hTIP39 Binding. hPTH(1-34) binds with considerably lower affinity to the rat PTH2 receptor (450 nM) than the human receptor (18 nM; Table 2). It was also found that hTIP39 bound with slightly lower affinity to the rat receptor (3.4-fold; Table 2). In the whole-cell binding assay used, ligand binding to the high-affinity active state of the receptor is unlikely to contribute significantly to the specific binding signal. Due to the presence of high intracellular concentrations of guanine nucleotides, which break down the receptor-G protein complex, this state probably represents a transient intermediate (Gilman, 1987). We have argued that the receptor states identified in this assay are the G protein-uncoupled receptor or desensitized or internalized receptor states (Hoare and Usdin, 2001). We investigated the molecular determinants of ligand binding to these inactive states of the PTH2 receptor by using this assay. [We previously measured ligand affinity for defined active (G protein-coupled) and inactive receptor states by using a cell membrane-based binding assay (Hoare et al., 2001)]. Unfortunately, none of the radioligands tested [¹²⁵I-hTIP39, ¹²⁵I-mTIP39, ¹²⁵I-rPTH(1-34)] had a useable signal-to-background ratio for the rat PTH2 receptor in this assay.]

View larger version (46K): [in this window] [in a new window]   Fig. 6.   Agonist and antagonist binding to chimeric human/rat PTH2 receptors expressed in COS-7 cells. Binding of unlabeled ligands was measured by displacement of 125I-hTIP39 binding to intact COS-7 cells expressing the receptors. Binding was measured for the agonists hPTH(1-34) (A-C) and hTIP39 (D-F), and for the antagonists [D-Tryp12]bPTH(7-34) (G-I) and hTIP(7-39) (J-L). The chimeric receptors were hN-rPTH2 and rN-hPTH2 (A, D, G, and J), hEL1-rPTH2, and rEL1-hPTH2 (B, E, H, and K), and hNTM3-rPTH2 and rNTM3-hPTH2 (C, F, I, and L). Ligand binding to the wild-type human PTH2 receptor is indicated by the short-dashed lines (----) and binding to the wild-type rat receptor indicated by the long-dashed lines (- - - -). Data points are the mean ± range of duplicate measurements. The curves are fits to a four-parameter logistic equation. Data are from representative experiments that were performed three to five times with similar results.

Regions of Human and Rat PTH2 Receptors Specifying Differential bPTH(7-34) and hTIP(7-39) Binding. The rat PTH2 receptor binds available antagonist ligands with considerably lower affinity than the human receptor (Fig. 4; Table 2). We investigated the molecular basis of this affinity difference for [D-Tryp¹²]bPTH(7-34) and hTIP(7-39) by using the chimeric human/rat PTH2 receptors described above.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Structure-activity and photochemical cross-linking studies have identified a common "two-site" orientation of ligand binding to PTH2 and PTH1 receptors (Bergwitz et al., 1996, 1997; Zhou et al., 1997; Bisello et al., 1998; Clark et al., 1998; Mannstadt et al., 1998; Turner et al., 1998; Behar et al., 1999; Hoare et al., 2000a, 2001) and other type II G protein-coupled receptors (Holtmann et al., 1995; Stroop et al., 1995; Bergwitz et al., 1996; Couvineau et al., 1996; Dautzenberg et al., 1998). The extracellular N-terminal domain of the receptor (N-domain) binds the C-terminal portion of the ligand (N-interaction) and the J-domain binds the ligand's amino-terminal portion (J-interaction). In this study we compared rat and human PTH2 receptor pharmacology by measuring ligand-stimulated increases of intracellular calcium and ligand binding. Included in this evaluation were mouse TIP39 and the antagonist analog mTIP(7-39). We then used chimeric human/rat PTH2 receptors to examine the functional role of receptor regions in PTH binding and signaling, and in receptor interaction with antagonist ligands.

PTH weakly activates the rat PTH2 receptor in assays of cAMP accumulation, whereas TIP39 is potent and efficacious. hTIP39 and mTIP39 potently increased [Ca²⁺]_i via the rat PTH2 receptor (Fig. 2B; EC₅₀ = 15 and 2.3 nM, respectively). This pathway involved PLC and was pertussis toxin-insensitive, suggesting the involvement of Gq/11 G proteins. Similarly, the PTH1 receptor has been shown to stimulate PLC via Gq/11 in COS-7 cells (Offermanns et al., 1996). However, PTH did not increase [Ca²⁺]_i via the rat PTH2 receptor (Figs. 2B and 3D). In addition PTH bound the rat PTH2 receptor with low affinity [IC₅₀ = 160 nM for rPTH(1-34)], compared with mTIP39 (7.5 nM). These findings support the hypothesis that PTH is not a physiologically significant modulator of the PTH2 receptor in the rat. In contrast, PTH acting at the human PTH2 receptor has been shown to be a potent full agonist for increasing cAMP, with a response equivalent to that of TIP39 (Hoare et al., 1999; Usdin et al., 1999; Table 3). PTH also binds the human PTH2 receptor with high affinity (Behar et al., 1996; Gardella et al., 1996; Hoare et al., 2000a; Table 2). PTH stimulated increases of [Ca²⁺]_i via the human PTH2 receptor expressed in COS-7 cells (Figs. 2A and 3C). Relative to hTIP39, hPTH(1-34) was a partial agonist and was 6-fold less potent (Fig. 3C; Table 1). The partial response to PTH may explain in part the weak coupling of the human PTH2 receptor to PLC observed previously (Takasu et al., 1998), because PTH was the only ligand available. The significance of hPTH(1-34)'s partial agonism for calcium signaling is not clear at present. It is possible that the calcium response of the human PTH2 receptor is highly sensitive to slight differences of ligand efficacy owing to the weak coupling of the receptor to the response.

PTH's activation selectivity for the human PTH2 receptor over the rat receptor was measured using ligand-stimulated cAMP accumulation. The efficacy (E_max) of PTH was specified by the receptor's J-domain and by the ligand's amino-terminal region [PTH(7-34) was an antagonist]. These findings implicate the J-interaction in receptor activation. Two regions of the J-domain contributed to the efficacy of PTH: the divergent EL1 and a region including EL2 and EL3 (Fig. 5). The EL2/EL3 region is probably involved in direct contact with the ligand: [Bpa¹]PTH cross-links to Val380 in TM6, close to EL3 (Behar et al., 1999). In a molecular model of PTH-PTH2 receptor interaction based on this cross-linking site, Val2 of PTH interacts with EL3 (Rolz et al., 1999). EL3 has also been identified as a determinant of PTH binding to the PTH1 receptor (Lee et al., 1994, 1995). The EL regions may also play a conformational role in activation. In molecular simulations of the PTH-PTH2 receptor complex EL2 and EL3 fold over the central core of the TM bundle (Rolz et al., 1999), consistent with a "closed" conformation postulated for the active state of the PTH1 receptor (Hoare et al., 2001). Interestingly, in this study the EL2/EL3 region was also a determinant of TIP39 interaction with the PTH2 receptor (Fig. 6, C and F), although this was evident as a contribution to binding affinity rather than signaling efficacy and was only evident for receptors containing the N-domain of the human PTH2 receptor.

The binding selectivity of PTH was measured using a whole-cell radioligand binding assay, which probably measures ligand affinity for inactive receptor states (Hoare and Usdin, 2001). The affinity of hPTH(1-34) for the inactive state of the PTH2 receptor was completely specified by the N-domain. The N-domain has been identified as a determinant of binding affinity for TIP39 binding to the PTH2 receptor (Hoare et al., 2000a). PTH selectivity for the active state of the receptor was evaluated using the EC₅₀ for cAMP accumulation. Although this measurement incorporates ligand affinity for the active state of the receptor, it provides only an approximate readout because other processes can contribute to the EC₅₀ (such as receptor reserve). None of the human PTH2 receptor regions restored high PTH potency when incorporated into the rat PTH2 receptor. Notably incorporation of the human N-domain (the affinity determinant) with EL1 or with EL2/EL3 (the efficacy determinants) failed to restore high PTH potency, i.e., the effect of these regions was not additive. Reciprocally, insertion of the rat N-domain, EL1, or EL2/EL3 into the human receptor decreased PTH(1-34) potency to that of the rat receptor. These findings indicate that multiple regions throughout the receptor are required in concert for the PTH-PTH2 receptor complex to form the high-affinity active state. This suggests that the global conformation of the rat PTH2 receptor prevents it from adopting the high-affinity active state when in complex with PTH.

Phylogenetic analysis of the PTH receptor family is consistent with a common ancestor of human and rat PTH2 receptors (Rubin and Jüppner, 1999). This suggests that the functional difference of PTH responsiveness resulted either from the human receptor maintaining a strong PTH response through evolution with a loss of this response in the rat, or that the human receptor has gained the ability to be activated by the ligand. The latter hypothesis is supported by the weak PTH efficacy at the zebrafish PTH2 receptor (Rubin and Jüppner, 1999; Hoare et al., 2000b). In this study we determined which parts of the PTH2 receptor are responsible for the differential PTH responsiveness of the human and rat receptors. As described above, at least three molecular determinants were identified, and these regions probably act in concert to produce different conformations of the two receptors when in complex with PTH. Based on this finding it is tempting to speculate that PTH activation of the human PTH2 receptor serves a physiological function.

Previously described antagonists that bind the human PTH2 receptor with moderate affinity [hTIP(7-39), [D-Tryp¹²]bPTH(7-34), and PTHrP(1-21)/PTH(22-34)] bound the rat receptor with 10- to 30-fold lower affinity. [hTIP(7-39), shown to be an antagonist in cAMP assays, also did not stimulate increases of Ca²⁺ via the human PTH2 receptor (data not shown).] A new ligand, mTIP(7-39), was the highest affinity antagonist identified for the human PTH2 receptor (IC₅₀ = 6.2 nM). ¹²⁵I-mTIP(7-39) provides the first radiolabeled antagonist for this receptor and should be useful for structure-function studies. mTIP(7-39) bound the rat receptor with low affinity (300 nM). A high-affinity antagonist for the rat PTH2 receptor would be useful for investigating the PTH2 receptor's physiological role in rodents, so we investigated the molecular basis of weak antagonist affinity for the PTH2 receptor.

For both [D-Tryp¹²]bPTH(7-34) and hTIP(7-39) human/rat PTH2 receptor selectivity was determined by the N-domain and EL1. The N-domain of the receptor probably contributes direct interactions with these antagonist ligands: position 13 of PTH cross-links to residues 138 to 147 of the human receptor, at the C-terminal end of the N-domain (Behar et al., 1999). The N-domain of the PTH1 receptor binds bPTH(7-34) (Jüppner et al., 1994). It is not presently clear whether EL1 is directly involved in binding antagonist ligands, or indirectly, for example, by affecting the receptor conformation. A recent photochemical cross-linking study identified an interaction between parabenzylbenzoyl-conjugated Lys27 of PTH and Leu261 of the PTH1 receptor, suggesting a direct role of EL1 in receptor-ligand interaction (Greenberg et al., 2000). The EL2/3 region was not involved in specifying hPTH(7-34) and hTIP(7-39) binding, in striking contrast to this region's involvement in binding full-length hPTH(1-34) and hTIP39. These considerations suggest that the loss of activation produced by amino-terminal truncation of the ligands involves an altered interaction with the J-domain, possibly a loss of interaction with the EL2/3 region of the receptor.

N-terminal truncation of hTIP39 and mTIP39 was attempted to develop a high-affinity antagonist. However, this truncation of TIP39 decreased the binding affinity for the rat PTH2 receptor (9-fold for hTIP39 and 40-fold for mTIP39) to a greater extent than for the human receptor (2.6- and 4.1-fold decrease, respectively). This suggests that interaction of the 1 to 6 portion of TIP39 with the J-domain contributes more binding energy for the rat receptor than for the human receptor. This hypothesis is supported by the increase of hTIP39 binding affinity produced by inserting the rat EL2/3 region into the human PTH2 receptor (Fig. 6F). These considerations suggest two possibilities for the development of a high-affinity rat PTH2 receptor antagonist. 1) Increase the affinity of the N-interaction by modifying residues in the C-terminal portion of TIP(7-39). However, the utility of TIP(7-39) analogs might be limited due to their high-affinity antagonism of the PTH1 receptor (Hoare and Usdin, 2000; Jonsson et al., 2001). 2) Introduce modifications into the N-terminal region of TIP39 that eliminate activation without appreciably affecting binding affinity.