Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

The parathyroid hormone-1 (PTH-1) receptor is activated by two ligands, PTH and the related peptide PTHrP. The major physiological role of PTH is the regulation of calcium homeostasis (Potts et al., 1995). PTHrP is believed to act as an autocrine factor in many tissues and has been shown to play a role in bone development (Amizuka et al., 1994). The PTH-1 receptor is being studied as a therapeutic target in two prevalent disorders: osteoporosis and hypercalcemia of malignancy (Dempster et al., 1993; Morley et al., 1998). The receptor belongs to the type II family of G protein-linked receptors (Jüppner et al., 1991), which share little sequence homology with type I receptors (typified by the ₂ adrenergic receptor) or type III receptors (typified by metabotropic glutamate receptors) (Potts et al., 1995). This family includes the receptors for glucagon, secretin, calcitonin, and calcitonin gene-related peptide.

The mechanisms by which agonist binding leads to G protein activation have been studied extensively (Kenakin, 1996). However, current models are based predominantly on data generated for small molecule receptors of the type I family, such as the ₂ adrenergic receptor (DeLean et al., 1980; Costa et al., 1992; Samama et al., 1993). Ligand-binding and functional data have been used to produce models of receptor-ligand interaction and receptor activation. A useful and commonly accepted mechanism is the extended ternary complex model of Samama et al. (1993). In this model the receptor exists in a spontaneous equilibrium between an inactive state (R), and an active state (R^*) that couples to G proteins (forming R^*G). Agonism can result from preferential binding of ligand to R^* over R and/or to R^*G over R^* (Kenakin, 1996).

The mechanisms of ligand binding and G protein activation are not well understood for the PTH-1 receptor or type II receptors in general. Other type II receptors have been studied in more detail. For glucagon and calcitonin gene-related peptide receptors, binding of agonist seems to involve an initial interaction of receptor and ligand followed by a postbinding conformational change of the complex (Wyborski et al., 1988; Chatterjee and Fisher, 1991; Post et al., 1992). The interaction and conformational change occur in the absence and presence of guanine nucleotides. A structural basis for these interactions has not been proposed. In addition, a simple explanation for the various observed effects of guanine nucleotides (Teitelbaum et al., 1982; Wyborski et al., 1988; Chatterjee and Fisher, 1991; Post et al., 1992; Van Rampelbergh et al., 1996) has remained elusive, so that the mechanism that couples agonist binding to G protein activation remains unclear.

The location of ligand-binding sites on the PTH-1 receptor has been studied in detail. Type II receptors possess a large N-terminal extracellular domain (Potts et al., 1995) and seven membrane-spanning -helices. Studies of chimeric receptors have suggested a "two-site" mode of binding (Stroop et al., 1995; Bergwitz et al., 1996). The N-terminal domain is believed to act as a high-affinity trap for the ligand. Regions in the body of the receptor have been implicated in binding interactions that activate the receptor; the binding loci have been proposed to form an "activation domain". Photoreactive ligands have been used to map potential binding loci within these domains. Amino acids 23 to 40 of the N-terminal domain of the rat receptor are functionally involved in ligand binding (Mannstadt et al., 1998). Within the body of the receptor, a photoreactive PTH analog cross-links to a region incorporating transmembrane domain six and part of the third extracellular loop (Biselo et al., 1998).

Site-directed receptor mutation studies suggest particular residues on PTH receptors that are involved in ligand recognition (Lee et al., 1995; Turner et al., 1996; Gardella et al., 1996a; Mannstadt et al., 1998; Biselo et al., 1998). Receptor mutation studies support the two-site model for the wild-type PTH-1 receptor, but the possibility has been raised that mutations may alter an equilibrium between postulated R and R^* conformations of the receptor (Lee et al. 1995; Schipani et al., 1995; Gardella et al. 1996a,b). These studies have used whole-cell-binding assays that do not allow receptor states to be identified or G protein interaction evaluated so it has not been possible to unambiguously discriminate between effects at a binding site and other conformational effects (Lee et al., 1995).

The aim of this study was to develop radioligand-binding assays for the PTH-1 receptor to examine the properties of ligands, the nature of receptor states, and the effects of GTPS. The data were used to evaluate potential models of receptor-ligand interaction.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Reagents and Peptides. All peptides were purchased from either Bachem (Torrance, CA) or Peninsula Laboratories (Belmont, CA). For the sake of brevity, peptide formulae are abbreviated in the text as follows: [Nle^8,21, Tyr³⁴]rPTH(1-34), rPTH(1-34); [Tyr³⁶]PTHrP(1-36), PTHrP(1-36); [Nle^8,18, Tyr³⁴]bPTH(3-34), bPTH(3-34); [Nle^8,18, Tyr³⁴]bPTH(1-34), bPTH(1-34). The letters b, r, and h designate the species as bovine, rat, and human, respectively. [1,2-³H(N)]myoinositol (47 Ci/mmol) was obtained from NEN (Boston, MA). Na¹²⁵I (2000 Ci/mmol) and [-³²P]ATP (800 Ci/mmol) were from ICN Biomedicals (Costa Mesa, CA). [2,8-³H]cAMP was purchased from Amersham (Arlington Heights, IL). ¹²⁵I-Labeled peptides were prepared using chloramine T as catalyst followed by purification by HPLC, as previously described (Clark et al., 1998). Monoiodinated [¹²⁵I][Tyr³⁶]PTHrP(1-36) (2000 Ci/mmol) and di-iodinated PTH-based radioligands (4000 Ci/mmol) were used in binding experiments. Radioligands were typically used within 3 weeks of the labeling reaction. Cell culture supplies were obtained from Life Technologies (Frederick, MD) except for fetal bovine serum, which was obtained from Sigma (St. Louis, MO).

Cell Culture. Human embryonic kidney (HEK293) cells expressing the human PTH-1 receptor were cultured as described (Usdin, 1997). Cell stocks were maintained in G418 (500 µg/ml), whereas large scale cultures for preparation of cell membranes were grown in the absence of the selective agent. COS-7 cells for transient expression of PTH receptors were grown as previously described (Clark et al., 1998).

Transient Receptor Expression in COS-7 Cells. COS-7 cells, grown in 24-well plates, were transfected with plasmid DNA encoding the human PTH-1 receptor as previously described (Clark et al., 1998). As a control in most transfection experiments, two wells of cells were transfected with 0.5 µg/well of a plasmid encoding -galactosidase. Transfection efficiency was approximately 30% based on visual inspection of cells following histochemical detection of -galactosidase.

Preparation of Cell Membranes. Confluent monolayers of HEK293 cells grown in 15-cm plates were washed with PBS and dislodged using 4 mM EDTA in PBS. Cells were centrifuged at 1000g for 10 min, and the cell pellet was suspended in lysis buffer [25 mM Tris, 2 mM EDTA, 6 mM MgCl₂, and 100 µM (4-(2-aminoethyl))-benzenesulfonylflouride (AEBSF), pH 7.5] using 40 ml of lysis buffer for five confluent plates of cells. After 20 min at 4°C, cells were homogenized by 30 strokes with a Dounce homogenizer. The homogenate was then centrifuged at 1000g for 10 min to remove unbroken cells and larger debris. The supernatant was centrifuged at 40,000g at 4°C for 30 min (Beckman J2-21 centrifuge, JA-17 rotor; Beckman Instruments, Berkeley, CA). To ensure the removal of guanine nucleotides from the cell membrane preparation, the resulting pellet was resuspended in 30 ml of lysis buffer and centrifuged again before final suspension in assay buffer (20 mM HEPES, 100 mM NaCl, 1 mM EDTA, 3 mM MgSO₄, pH 7.5). Membrane protein was quantified using the copper bicinchoninic acid method (Pierce, Rockford IL) with BSA as the standard. Cell membranes were stored at 80°C before use. Membranes prepared from HEK293 cells stably expressing the human PTH-1 receptor are referred to as 293PTH-1 membranes in the text.

Measurement of Cellular Levels of cAMP. Transfected COS-7 cells were washed with cAMP assay buffer (Dulbecco's modified Eagle's medium supplemented with 0.1% BSA, 30 µM Ro 20-1724 (RBI, Natick, MA), a cAMP phosphodiesterase inhibitor, and 1 µg/ml bacitracin. After a 10-min incubation in cAMP assay buffer at room temperature, the buffer was replaced with 0.15 ml of the same solution containing varying concentrations of the test peptides. Cells were then incubated at room temperature for 30 min before assay termination by the addition of 0.15 ml 0.1 N HCl, 0.1 mM CaCl₂. cAMP was quantified using an enzyme-linked immunosorbent assay.

Measurement of cAMP Generation by 293PTH-1 Membranes. The assay was performed in two steps. In the first stage, ligands were incubated with membranes (40-60 µg) in a volume of 90 µl assay buffer supplemented with 0.3% nonfat dried milk powder (Super G, Inc., Landover, MD), 100 µM AEBSF, 1 µg/ml bacitracin, and 100 µM GTP. Following a 90-min incubation at 21°C, the second step of the procedure was performed; reactants for generation of cAMP were added in a volume of 10 µl. This solution contained 20 mM HEPES, 3 mM MgSO₄, 1 mM EDTA, 100 mM NaCl, 100 µM GTP, 60 mM phosphoenol pyruvate (potassium salt), 1 mM ATP, 1 mM isobutylmethylxanthine, 0.1 mg/ml pyruvate kinase, and [-³²P]ATP (4 × 10⁵ cpm/10 µl). Following an additional 30-min incubation, the reaction was terminated by addition of 0.9 ml of 0.25% SDS, 5 mM ATP, and 0.175 mM cAMP supplemented with [³H]cAMP (3000 cpm/0.9 ml). The amount of [³²P]cAMP generated was measured as described (Johnson and Salomon, 1991). The time course of [³²P]cAMP generation was linear over 60 min.

Measurement of Cellular Levels of [³H]Inositol Phosphates. Transfected COS-7 cells in 24-well plates were labeled with 0.5 µCi/well [³H]inositol, in inositol-free Dulbecco's modified Eagle's medium, at approximately 48 h post-transfection. After 15 h, cells were washed twice at 37°C in minimum essential medium supplemented with 20 mM LiCl and 0.1% BSA. Various concentrations of test peptides were then added in a volume of 0.5 ml. Following a 40-min incubation at 37°C, the assay was terminated by addition of 0.5 ml of 1 M KOH, 18 mM sodium borate, 3.8 mM EDTA, and 7.6 mM NaOH. The solution was then neutralized using 0.5 ml of 2.0 N HCl before chromatography of inositol phospholipid metabolites. The assay mixture was applied to a 1-ml column of AG1-XA resin (Bio-Rad, Melville, NY) previously equilibrated with water. Following stepwise elution of free inositol with 10 ml of water and glycerophosphoinositides with 8 ml of 5 mM sodium borate, 60 mM sodium formate, inositol phosphates were eluted with 3 ml of 100 mM formic acid/2.0 M ammonium formate and collected into scintillation vials. After addition of 10 ml of FloScint 4 (Packard) sample cpm values were determined by scintillation counting.

Radioligand-Binding Assays. A centrifugation assay was developed to allow accurate measurement of radioligand-binding parameters for PTH- and PTHrP-based peptides to cell membranes. Centrifugation was chosen as the method rather than filtration to permit the detection of rapidly dissociating binding states (Hulme, 1992). ¹²⁵I-Labeled peptides (2 × 10⁵ to 4 × 10⁵ cpm) and unlabeled ligands (diluted in siliconized glass tubes) were added to 1.7-ml siliconized microfuge tubes in 0.3 ml of assay buffer (20 mM HEPES, 100 mM NaCl, 1 mM EDTA, 3 mM MgSO₄, pH 7.5) containing 0.3% nonfat dried milk powder, 100 µM AEBSF, and 1 µg/ml bacitracin. Membranes (0.2 ml, 40-50 µg) were added to initiate the binding reaction. Following incubation for 90 min at 21°C, the tubes were centrifuged at 15,000g for 10 min at 4°C (Eppendorf Centrifuge 5403, 16 T60-11 rotor; Brinkmann Instrument Co., Westbury NY) (in preliminary time course experiments, this incubation period was found to be sufficient to allow complete association of the radioligands to the PTH-1 receptor). Under these conditions the effective sedimentation time is approximately 2 min (Hulme, 1992). The pellets were then gently washed as described (Hulme, 1992) four times with assay buffer, followed by draining of remaining buffer, excision of the microfuge tube tip, and counting in a Wallac 1470 Wizard gamma counter. An aliquot of the radioligand was also counted to calculate the amount added to each assay tube. The summed radioactivity of the supernatant and the pellet was found to be equal to that added to the tube, indicating that binding of radioligand to the assay tube was negligible. Total binding was less than 20% of the added radioactivity. Although this level of total binding is not ideal (the standard is usually considered to be less than 10%), it does not appreciably affect measurements of radioligand-binding parameters (Hulme and Birdsall, 1992). Analysis by HPLC indicated that 75 to 80% of the added radioactivity eluted in the same fraction as the starting material after incubation of [¹²⁵I]rPTH(1-34) or [¹²⁵I]PTHrP(1-36) with 293 PTH-1 membranes for 90 min at 21°C. These experiments indicate that the total concentration of radioligand was approximately equal to the free concentration and that this amount did not change appreciably for the duration of the assay. Binding of [¹²⁵I]rPTH(1-34) to whole cells was compared using 0.3% nonfat milk or serum to block nonspecific binding. [The latter has been used extensively in studies of ligand binding at PTH receptors (e.g., Bergwitz et al., 1996; Clark et al., 1998)] Estimates of K_d or B_max were indistinguishable for the two blocking reagents (data not shown). Nonspecific binding of radioligand was measured in the presence of a saturating concentration (300 nM) of the corresponding unlabeled ligand. In some experiments, GTPS was included at a concentration of 10 µM. The assay volume was increased to 1.0 ml for experiments using [¹²⁵I]bPTH(1-34) without adjusting the amount of membranes added (50 µg), to reduce the [bound radioligand]/[total radioligand] ratio to below 20%. In experiments designed to saturate G_S with GTPS, membranes (2-3 mg in 1 ml of assay buffer) were preincubated with 10 µM GTPS for 2 h at 30°C. As a control, membranes were treated in the same way but without GTPS.

Data Analysis. Displacement of [¹²⁵I]bPTH(3-34) by unlabeled peptides was analyzed using Prism 2.01 (GraphPad). Data were fitted to models that assume inhibition of radioligand binding at one affinity state or two independent affinity states. To compare fits to these models, a partial F test was performed to compare the residual sum of squares for the different fits according to the "extra sum of squares principle". The F value was calculated using the following equation:
where SS1 and SS2 are the residual sum of squares for the fit to the simpler equation (fewer parameters) and more complex equation (more parameters), respectively, and where dF1 and dF2 refer to the corresponding degrees of freedom for the residuals. Prism 2.01 was also used to analyze homologous displacement data with a four-parameter logistic equation. An equation was derived to analyze homologous displacement data using a saturation isotherm without the requirement to initially convert the value for cpm bound into a concentration value. Using Prism 2.01, raw data (cpm bound versus peptide concentration) were fitted to the following equation:
where [L] is the total ligand concentration ([unlabeled ligand] + [radioligand]), which was assumed to be approximately equal to the free ligand concentration. c represents nonspecific binding expressed as a constant fraction of the ligand concentration. An equation that assumes ligand saturation at two independent sites was also used. Comparison of the goodness of fit for the two equations was performed using the partial F test described above.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Characterization of bPTH(3-34) As an Antagonist of the Human PTH-1 Receptor. We attempted to identify an antagonist ligand that could be used as a radioligand for the PTH-1 receptor. Antagonist radioligands have been useful at other G protein-coupled receptors for comparison with radiolabeled agonists, to enable detection of agonist low-affinity states in competition assays and to compare binding of competing ligands in the presence and absence of guanine nucleotides (DeLean et al., 1980). bPTH(3-34) has previously been shown to bind the human PTH-1 receptor with high affinity and to cause little or no stimulation of second messenger pathways (reviewed in Potts et al., 1995). We confirmed these characteristics of bPTH(3-34) at the recombinant human PTH-1 receptor. The ligand did not produce a detectable increase in cellular cAMP levels in COS-7 cells transiently expressing the PTH-1 receptor or in membranes prepared from HEK293 cells stably expressing the receptor (Fig. 1, A and C). The agonists hPTH(1-34) and PTHrP(1-34) produced a 16 ± 5-fold and 14 ± 9-fold increase in cAMP accumulation in COS-7 cells (with EC₅₀ values of 1.9 ± 0.6 nM and 1.2 ± 0.6 nM, respectively). For 293PTH-1 membranes hPTH(1-34) and PTHrP(1-34) produced a 7.1 ± 0.6-fold and 6.4 ± 0.4-fold increase in cAMP generation, respectively, with corresponding EC₅₀ values of 1.3 ± 0.1 nM and 9.3 ± 2.3 nM. In addition, bPTH(3-34) did not stimulate the accumulation of [³H]inositol phosphates in transiently transfected COS-7 cells, whereas the agonists hPTH(1-34) and PTHrP(1-34) produced a concentration-dependent increase (Fig. 1B, EC₅₀ values of 16 ± 3 nM and 27 ± 8 nM, respectively).

View larger version (14K): [in this window] [in a new window]   Fig. 1.   Pharmacological characterization of bPTH(3-34). Second messenger responses were measured for [Nle8,18, Tyr34]bPTH(3-34) (), hPTH(1-34) (), and PTHrP(1-34) (). Data points are the mean ± range of duplicate measurements. Where error bars are not apparent, they are smaller than the symbols. Representative experiments are presented, and all experiments were repeated twice with similar results. A, cAMP accumulation in COS-7 cells transiently expressing the PTH-1 receptor. Data are total cAMP measured per well of a 24-well plate. B, [3H]inositol phosphate accumulation in COS-7 cells transiently expressing the PTH-1 receptor. Total inositol phosphate accumulation was measured, and data are total cpm measured per well of a 24-well plate. C, cAMP generated by 293PTH-1 membranes. Data are measurements of the total [32P]cAMP accumulation per milligram of membrane protein per minute.

GTPS Sensitivity of Radioligand Binding at the PTH-1 Receptor. A cell membrane-based radioligand-binding assay was developed for the PTH-1 receptor. This procedure allowed accurate measurement of ligand-binding parameters without the limitations that are often associated with the use of peptide ligands (such as high nonspecific binding or ligand instability). The radioligands [¹²⁵I]rPTH(1-34), [¹²⁵I]bPTH(1-34), [¹²⁵I]PTHrP(1-36), and [¹²⁵I]bPTH(3-34) bound to 293PTH-1 membranes with typical total binding: nonspecific binding ratios of 4:1, 4:1, 4:1, and 2.5:1, respectively. No specific binding was observed for membranes prepared from nontransfected HEK293 cells (data not shown). The guanine-nucleotide sensitivity of radioligand binding was measured in the presence of a range of concentrations of GTPS. The nucleotide reduced binding of the agonist radioligands [¹²⁵I]rPTH(1-34) and [¹²⁵I]PTHrP(1-36) in a concentration-dependent fashion (Fig. 2) but only to a limited extent; 100 µM GTPS reduced [¹²⁵I]rPTH(1-34) binding by 32 ± 5% and [¹²⁵I]PTHrP(1-36) by 51 ± 3%. This reduction was specific to the agonist ligands; binding of the antagonist [¹²⁵I]bPTH(3-34) was insensitive to GTPS at even the highest concentrations tested (Fig. 2).

View larger version (21K): [in this window] [in a new window]   Fig. 2.   Binding of radiolabeled ligands to 293PTH-1 membranes in the presence of GTPS. The concentration dependence of GTPS-mediated inhibition was measured for modulation of the binding of [125I][Nle8,18, Tyr34]bPTH(3-34) (), [125I][Nle8,21, Tyr34]rPTH(1-34) (), and [125I][Tyr36]PTHrP(1-36) (). The radioligand concentration in these assays was between 100 pM and 200 pM. Specific binding was calculated by subtraction of the nonspecific binding value (obtained in the presence of a 300 nM concentration of the corresponding unlabeled peptide). Total specific binding is specific binding in the absence of GTPS. Data points are the mean ± S.E.M. values from three independent experiments.

1

Dissociation of Radiolabeled Agonists from the PTH-1 Receptor. The complexity of radiolabeled agonist binding at the PTH-1 receptor was initially addressed by measuring radioligand dissociation. To assess the effect of G protein coupling on agonist binding, membranes were preincubated with 10 µM GTPS using conditions that should yield receptor uncoupled from G protein [a 2-h incubation in the presence of magnesium ions at 30°C (Northup et al., 1982)]. As a control, membranes were treated similarly but in the absence of GTPS.

View larger version (24K): [in this window] [in a new window]   Fig. 3.   Dissociation of agonist radioligands from 293PTH-1 membranes in the absence and presence of GTPS. Dissociation of [125I][Nle8,21, Tyr34]rPTH(1-34) (A) and [125I][Tyr36]PTHrP(1-36) (B) was measured following a 90-min equilibration incubation by the addition of a 300 nM concentration of the corresponding unlabeled ligand. Radioligand concentration was typically 100 to 200 pM. GTPS treatment () involved preincubation of membranes with 10 µM GTPS for 2 h at 30°C. GTPS was included in the equilibration and dissociation phases. For the control (), membranes were treated similarly but in the absence of GTPS. Raw data (cpm versus time) were fitted to a biexponential dissociation function, which provided a better fit to the data (p < .05) than a single-rate function in all cases. This function provided an estimate of the plateau of binding reached after complete dissociation of reversibly bound ligand. In the absence of GTPS, this value was greater than nonspecific binding in all cases, demonstrating a pseudoirreversible phase of binding. In the presence of GTPS, this value was equivalent to nonspecific binding. For normalization of presented data, specific binding at time 0 was calculated by subtracting nonspecific binding (300 nM unlabeled ligand included in the equilibration reaction) from total binding (harvested without the addition of unlabeled ligand, following the equilibration reaction). Values at each time point were calculated by subtraction of nonspecific binding followed by division by specific binding at time 0. Data are from representative experiments that were repeated twice. Data points are the mean ± S.E.M. of triplicate measurements, and in some cases error bars are enclosed by the symbols.

View this table: [in this window] [in a new window]   TABLE 1 Radioligand dissociation kinetics at the PTH-1 receptor Dissociation of [125I][Nle8,21,Tyr34]rPTH(1-34), and [125I][Tyr36]PTHrP(1-36) was measured using the procedure described for Fig. 3 involving a 2-h preincubation at 30°C in the presence or absence of GTPS. The preincubation was not used for bPTH radioligands. Data were fitted to mono- and biexponential dissociation equations, as described for Figs. 3 and 4, and the best fit was determined using a partial F test (see Materials and Methods). The analysis provided estimates of the dissociation rates (k1) and the proportion of specific binding at time 0 that dissociates at these rates. The analysis also provided an estimate of the plateau of binding reached after complete dissociation of reversibly bound radioligand. This value was greater than nonspecific binding in the absence of GTPS for the agonist radioligands, indicating the presence of a pseudoirreversible binding phase. The percentage of specific binding represented by this component was calculated. Pseudoirreversible binding was not detected (N.D.) for [125I][Nle8,18,Tyr34]bPTH(3-34), or for the agonist ligands in the presence of GTPS; the plateau value was equivalent to nonspecific binding. For analysis of data for the presence of GTPS, the lower plateau was fixed in the fitting procedure to the value for nonspecific binding. For these data, the kinetic parameters were not significantly different (p > .05) from those obtained when the lower plateau was allowed to vary in the fitting procedure. Data are presented as the mean ± S.E.M. from three experiments. Dissociation rate constants were compared in the presence and absence of GTPS using paired Student's t tests.

View larger version (26K): [in this window] [in a new window]   Fig. 4.   Dissociation of [125I][Nle8,18, Tyr34]bPTH(3-34) and dissociation of high-affinity [125I][Nle8,18, Tyr34]bPTH(1-34) binding from 293PTH-1 membranes. Dissociation of [125I][Nle8,18, Tyr34]bPTH(1-34) () and [125I][Nle8,18, Tyr34]bPTH(3-34) () was measured following a 90-min equilibration reaction by the addition of an excess (300 nM) of the unlabeled ligand. Radioligand concentrations used in this experiment were 110 pM for [125I][ Nle8,18, Tyr34]bPTH(1-34) and 160 pM for [125I][Nle8,18, Tyr34]bPTH(3-34). Data were analyzed and normalized as described for Fig. 3. Dissociation data were fitted best by a bi-exponential function for [125I][Nle8,18, Tyr34]bPTH(1-34). The plateau value was greater than nonspecific binding, indicating the presence of a pseudoirreversible phase of binding. Data for [125I][Nle8,18, Tyr34]bPTH(3-34) were fitted best by a single-rate function, and the plateau value was similar to nonspecific binding. Data are from a representative experiment that was repeated twice. Data points represent mean ± S.E.M. of triplicate determinations (in most cases, error bars are contained within the symbols). Inset, homologous displacement of [125I][Nle8,18, Tyr34]bPTH(1-34) binding to 293PTH-1 membranes. Data were normalized as described for Fig. 3. The low concentration of radioligand used (65 pM) did not label the low-affinity state observed in competition assays (Fig. 5C). Data are mean ± S.E.M. of triplicate measurements (the error bars are enclosed by the symbols). The experiment was repeated twice with similar results.

View larger version (29K): [in this window] [in a new window]   Fig. 5.   Inhibition of [125I][Nle8,18, Tyr34]bPTH(3-34) binding to 293PTH-1 membranes by agonist and antagonist ligands in the absence and presence of GTPS at a 1.5-h incubation. Inhibition of radioligand binding was measured in the absence of GTPS () and in the presence of 10 µM GTPS () for 1.5 h at 21°C. The agonist peptides were hPTH(1-34) (A), [Tyr36]PTHrP(1-36) (B), and [Nle8,18, Tyr34]bPTH(1-34) (C). The antagonist peptides were [Nle8,18, Tyr34]bPTH(3-34) (D) and PTHrP(7-34) (E). Raw data (cpm versus [ligand]) were analyzed using equations that assume inhibition of radioligand binding at one affinity state or two independent states (except for [Nle8,18, Tyr34]bPTH(3-34), where one- and two-site saturation equations were used). The best fit was assessed using a partial F test (see Materials and Methods). Specific binding is the binding value following subtraction of nonspecific binding. Total specific binding is the total binding value (the absence of unlabeled ligand) minus nonspecific binding. With one exception, the values of total and nonspecific binding used for normalizing the presented data were those provided by the curve-fitting analyses, which were in good agreement with the measured values for total binding (the absence of ligand) and nonspecific binding (the presence of 300 nM unlabeled [Nle8,18, Tyr34]bPTH(3-34)). The exception to this paradigm was nonspecific binding for [Nle8,18, Tyr34]bPTH(1-34) assays, for which the measured value was used for normalization and in the curve-fitting analysis. Data are from individual experiments that were repeated two to four times. Data points are the mean ± S.E.M. of triplicate determinations. Downward-pointing arrows indicate IC50 values for binding states in the absence of GTPS, and upward arrows indicate those values in the presence of the nucleotide.

Inhibition of [¹²⁵I]bPTH(3-34) Binding to the PTH-1 Receptor by Unlabeled Ligands. Radiolabeled agonists likely bind detectably only to high-affinity states of the receptor at the concentrations used (approximately 100 pM). This assumption was demonstrated to be correct for [¹²⁵I]bPTH(1-34); the ligand bound with only high affinity in homologous displacement experiments (Hill slope = 1.18 ± 0.04, IC₅₀ = 0.24 ± 0.02 nM) (Fig. 4, inset). To probe the existence of low-affinity states, unlabeled agonists were competed against the antagonist [¹²⁵I]bPTH(3-34). A 1.5-h incubation at 21°C was found to be long enough to allow low concentrations (100 pM) of [¹²⁵I]bPTH(3-34) and radiolabeled forms of the agonists used to closely approach the asymptotic maximum of binding (data not shown). The effects of 10 µM GTPS were examined by including the nucleotide in this incubation. It was not possible to use a preincubation with the nucleotide in these experiments because the signal/noise ratio of [¹²⁵I]bPTH(3-34) binding was too low after the treatment to allow reliable analysis of the data. However, the effect of 10 µM GTPS on radiolabeled agonist binding after the 1.5-h incubation (29 ± 5% and 42 ± 3% reduction of [¹²⁵I]rPTH(1-34) and [¹²⁵I]PTHrP(1-36) binding, respectively) was not significantly different from that obtained by including a preincubation [20 ± 3% (p = .071) and 39 ± 3% (p = .56) reduction, respectively]. These data suggest that the effect of GTPS was complete within the 1.5-h incubation.

Effect of Extending the Incubation Time in [¹²⁵I]bPTH(3-34)/Agonist Competition Experiments for the PTH-1 Receptor. A 1.5-h incubation was sufficient to allow ligands to approach the asymptotic maximum of binding in association experiments (data not shown). However, slow kinetic components that hinder the approach to steady state in these assays may not be detected by measuring radioligand association. Therefore, the incubation period in competition assays was doubled to determine whether steady state was reached in the shorter incubation.

View larger version (20K): [in this window] [in a new window]   Fig. 6.   Inhibition of [125I][Nle8,18, Tyr34]bPTH(3-34) binding to 293PTH-1 membranes by agonist ligands in the absence and presence of GTPS at a 3-h incubation. Inhibition of radioligand binding was measured in the absence of GTPS () and in the presence of 10 µM GTPS () for 3 h at 21°C. The peptides tested were hPTH(1-34) (A), [Tyr36]PTHrP(1-36) (B), and [Nle8,18, Tyr34]bPTH(1-34) (C). Raw data (cpm versus [ligand]) were analyzed using equations that assume inhibition of radioligand binding at one affinity state or two independent affinity states. The best fit was assessed using a partial F test (see Materials and Methods). Specific binding is the binding value following subtraction of nonspecific binding. Total specific binding is the total binding value (the absence of unlabeled ligand) minus nonspecific binding. The values of total and nonspecific binding used for normalizing the presented data were those provided by the curve-fitting analyses, which were in good agreement with the measured values for total binding (the absence of ligand) and nonspecific binding (the presence of 300 nM unlabeled [Nle8,18, Tyr34]bPTH(3-34)). Data are from individual experiments that were repeated one or two times. Data points are the mean ± S.E.M. of triplicate determinations. Downward-pointing arrows indicate IC50 values for binding states in the absence of GTPS, and upward arrows indicate those values in the presence of the nucleotide.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The PTH-1 receptor is being investigated as a therapeutic target for two disorders: osteoporosis, where it has been demonstrated that PTH-1 receptor agonists increase bone mass (Dempster et al., 1993), and hypercalcemia of malignancy, where PTH-1 receptor antagonists may block the effects of elevated PTHrP (Morley et al., 1998). Quantitative assessment of ligand-binding parameters will be useful in the development of these ligands.

In this study, we developed a method to measure radioligand-binding parameters at the PTH-1 receptor using a membrane-based assay. General properties of agonist and antagonist binding were identified. 1) Agonist binding was complex, whereas antagonist binding conformed to a simple single-site interaction. 2) Agonist binding was affected by GTPS, whereas antagonist binding was GTPS insensitive. 3) GTPS treatment only partially reduced high-affinity radiolabeled agonist binding (Fig. 2). 4) Radiolabeled agonist dissociation was complex in the presence of GTPS (Figs. 3 and 4, Table 1). 5) The kinetics of radiolabeled agonist binding were slow (Figs. 5 and 6; Tables 2 and 3). The first two properties are typical of G protein-coupled receptors and can be explained by receptor-G protein interaction (DeLean et al., 1980; Kenakin, 1996). Complex agonist binding arises from differential affinities of the ligand for G protein-coupled and uncoupled receptor states; a fraction of the agonist-bound receptors are coupled to G proteins in an agonist-high-affinity state, and the remainder represent the lower affinity state of the uncoupled receptor. The reduction of agonist binding in the presence of GTPS may result from breakdown of the RG complex; biochemical observations are consistent with the hypothesis that GTPS uncouples receptor from G protein, resulting in a loss of high-affinity agonist binding (Gilman, 1987). At the PTH-1 receptor, GTPS-sensitive agonist binding was identified as a pseudoirreversible component of agonist binding in dissociation experiments. This phase provides a measure of the G protein-coupled receptor state. Pseudoirreversible binding represents a fraction of total agonist binding, suggesting that only a fraction of the agonist-bound receptors are coupled to G protein in the absence of guanine nucleotides. The same conclusion has been reached for a number of G protein-coupled receptors (DeLean et al., 1980; Neubig et al., 1985).

The third and fourth observations above were unusual for a G protein-coupled receptor: detection of radiolabeled agonist binding and complex agonist-binding kinetics in the presence of GTPS. In this study conditions were used that result in GTPS saturation of purified G_s (Northup et al., 1982). As a result, the receptor is likely uncoupled from G protein, suggesting that these two agonist properties are intrinsic to the uncoupled receptor. This interpretation could be complicated by coupling of the PTH-1 receptor to G_q as well as G_s (Offermanns et al., 1996). Two observations suggest that the receptor is predominantly coupled to G_s in the host cell line used (HEK293). In other cells the potency for inositol phosphate production is an order of magnitude lower than that for cAMP generation (Offermanns et al., 1996). In addition, PTH-mediated stimulation of inositol phosphate production is undetectable in HEK293 cells expressing a similar number of PTH-1 receptors (80,000 receptors/cell) to the cells used in this study (Schneider et al., 1994).

An attempt was made to explain these ligand-binding data for the PTH-1 receptor using a binding model. The ternary complex model predicts complex, GTPS-sensitive agonist binding (DeLean et al., 1980), but the model cannot account for complex agonist binding observed in the presence of the nucleotide. As described below, the two-site-binding hypothesis can account for this observation and all of the other ligand properties described above. However, other models cannot be unequivocally excluded as explanations for complex agonist binding at the G protein-uncoupled receptor. At the calcitonin gene-related peptide receptor, this phenomenon was explained by a model that assumes a spontaneous equilibrium between unliganded receptor states (Chatterjee and Fisher, 1991). An oligomeric receptor model was used to explain and quantify complex ligand-binding behavior at purified cardiac muscarinic acetylcholine receptors resolved from G protein (Wreggett and Wells, 1995). However, the two-site hypothesis requires only the known properties of the receptor to explain the data obtained in this study.

According to the two-site model (Stroop et al., 1995; Bergwitz et al., 1996), agonist ligands bind with high affinity to the N-terminal region of the receptor (forming RA_N) and with lower affinity to the extracellular loops and transmembrane regions (forming RA_B) (Fig. 7). Ligand can also be tethered to both domains after the initial binding event (to produce RA_NB). It is likely that the single site binding interaction detected for the antagonist [¹²⁵I]bPTH(3-34) represents an interaction with the N-terminal domain (Jüppner et al., 1994). The model can explain high-affinity agonist binding in the presence of guanine nucleotides (Fig. 2) as high-affinity binding is intrinsic to the receptor, composed of an interaction with the N-terminal region of the receptor (RA_N) together with ligand tethered to both domains (RA_NB). These states may represent the radiolabeled agonist binding detected on whole cells. The two phases of radiolabeled agonist dissociation observed with GTPS present can represent dissociation from the N-terminal region (k₂) and breakdown of RA_NB via RA_B (k₈ and k₄). Binding of agonists to the PTH-1 receptor was slow, but we were unable to determine the mechanism underlying this phenomenon. It is unlikely that association of ligand with receptor is rate limiting because this effect would yield inhibition curves steeper than bimolecular mass-action isotherms before equilibrium (Motulsky and Mahan, 1984). In other studies of type II receptors, slow agonist binding kinetics have been explained by a postbinding conformational change within the receptor-ligand complex (Wyborski et al., 1988; Chatterjee and Fisher, 1991; Post et al., 1992). This interaction can be explained as the formation of RA_NB from RA_N. The extremely unusual effect of GTPS on bPTH(1-34) binding after 3 h, the removal of low-affinity binding, suggests that a low-affinity state is stabilized by interaction with G protein. This low-affinity state may represent RA_B; studies of chimeric receptors have suggested that interaction of ligand with the activation domain can stimulate signaling, independent of binding to the N-terminal domain (Stroop et al., 1995; Bergwitz et al., 1996).

View larger version (16K): [in this window] [in a new window]   Fig. 7.   Two-site model for ligand interaction with the PTH-1 receptor. This model has been developed from studies of chimeric and mutant PTH-1 receptors. Two structurally distinct regions of the receptor are recognized by two complementary binding domains of agonist ligands. R and A are receptor and ligand, respectively, RAN represents ligand bound at the N-terminal region of the receptor, RAB describes ligand bound to the remaining body of the receptor, and RANB signifies ligand tethered to both binding sites. An equation was derived for inhibition of radiolabeled antagonist binding to the receptor R by ligand A. The following assumptions were made. 1) The radiolabeled antagonist (L) binds detectably only at the N-terminal domain. 2) The binding of ligand A to the body of the receptor (forming RAB) inhibits the binding of radioligand to the N-terminal region (so that RLNAB does not form). 3) Steady state is closely approached within the incubation period of the assay. The following equation can be derived: where [RL] is the concentration of radiolabeled antagonist bound to the receptor, and [RL0] is this value in the absence of competing ligand; KL is the equilibrium association constant for the radiolabeled antagonist, and KA is the macroscopic affinity constant for the agonist.

We attempted to estimate certain ligand-binding constants within the two-site model. The affinity of antagonist ligands for the N-terminal domain can be estimated from displacement of [¹²⁵I]bPTH(3-34) binding (Fig. 5). For bPTH(3-34) saturation analysis yielded the values in Table 2. Given that the concentration of radioligand used was 5- to 10-fold less than the K_d (1 nM), the IC₅₀ for competing antagonist ligands is approximately equal to the affinity for the ligand. The values for PTHrP(7-34) (Table 2) for the presence and absence of GTPS were not significantly different (p > .05), which suggests that the interaction of ligand with the N-terminal receptor region is insensitive to R-G coupling. For agonists it was not possible to model equilibrium binding in the absence of GTPS because of the presence of pseudoirreversible binding (Figs. 3 and 4). The binding of agonists in the presence of GTPS was quantified using [¹²⁵I]bPTH(3-34) displacement data obtained after a 3-h incubation (Fig. 6). The two-site model predicts an agonist binding curve analogous to a single-state inhibition isotherm (Fig. 7). After 3 h the inhibition curves for hPTH(1-34) and bPTH(1-34) conformed to a single-state model. The IC₅₀ for inhibition of [¹²⁵I]bPTH(3-34) binding was used to calculate the value of 1/K_A, the macro affinity constant that governs the interaction of agonist with the uncoupled receptor (Fig. 7). These values were 4.5 nM and 1.5 nM for hPTH(1-34) and bPTH(1-34), respectively. A kinetic constant can also be estimated using the model. The dissociation rate for [¹²⁵I]bPTH(3-34) (Table 1) represents k₂, the rate constant for ligand dissociation from RA_N (Fig. 7). The similarity of the dissociation rate for [¹²⁵I]bPTH(3-34) to the more rapid rate for [¹²⁵I]bPTH(1-34) suggests that the rapid component for the agonist represents k₂ (Table 1). The rapid rate constant for agonist dissociation was therefore used as an estimate of k₂ (Table 1). The values of this constant were not significantly different (p > .05) in the presence or absence of GTPS, supporting the hypothesis above that RA_N formation is insensitive to R-G coupling.

The two-site model for the PTH-1 receptor will require further verification to determine whether the model is sufficient to describe the binding of ligands to the receptor. The presence of an equilibrium between unliganded receptor states (e.g., R and R^*) could be probed by study of constitutively active mutant PTH-1 receptors (Samama et al., 1993). Several of the binding constants of the model remain to be estimated, the effect of G protein coupling on the constants requires further study, and the nature of the "active" state of the receptor remains to be established. If the model is correct, it is possible that small ligands targeted to the N terminus of the receptor could block binding of PTH and PTHrP. Inhibition of [¹²⁵I]bPTH(3-34) binding would provide a direct assay for the interaction of small ligands with this putative antagonist binding region of the receptor. Low molecular weight agonists could activate signal transduction by binding only to the "activation" domain in the membrane-enclosed receptor region. This study provides a previously unavailable quantitative framework to explore these possibilities.