Purification and Characterization of a Receptor for Human Parathyroid Hormone and Parathyroid Hormone-related Peptide*

The human parathyroid hormone (PTH) receptor (hPTH1R), containing a 9-amino acid sequence of rhodopsin at its C terminus, was transiently expressed in COS-7 cells and solubilized with 0.25% n-dodecyl maltoside. Approximately 18 μg of hPTH1R were purified to homogeneity per mg of crude membranes by single-step affinity chromatography using 1D4, a monoclonal antibody to a rhodopsin epitope. The N terminus of the hPTH1R is Tyr23, consistent with removal of the 22-amino acid signal peptide. Comparisons of hPTH1R by quantitative immunoblotting and Scatchard analysis revealed that 75% of the receptors in membrane preparations were functional; there was little, if any, loss of functional receptors during purification. The binding affinity of the purified hPTH1R was slightly lower than membrane-embedded hPTH1R (K d = 16.5 ± 1.3versus 11.9 ± 1.9 nm), and the purified receptors bound rat [Nle8,21,Tyr34]PTH-(1–34)-NH2(PTH-(1–34)), and rat [Ile5,Trp23,Tyr36]PTHrP-(5–36)-NH2with indistinguishable affinity. Maximal displacement of125I-PTH-(1–34) binding by rat [α−aminoisobutyric acid (Aib)1,3,Nle8,Gln10,Har11,Ala12,Trp14,Arg19,Tyr21]PTH-(1–21)-NH2and rat [Aib1,3,Gln10,Har11,Ala12,Trp14]PTH-(1–14)-NH2of 80 and 10%, respectively, indicates that both N-terminal and juxtamembrane ligand binding determinants are functional in the purified hPTH1R. Finally, PTH stimulated [35S]GTPγS incorporation into Gαs in a time- and dose-dependent manner, when recombinant hPTH1R, Gαs-, and βγ-subunits were reconstituted in phospholipid vesicles. The methods described will enable structural studies of the hPTH1R, and they provide an efficient and general technique to purify proteins, particularly those of the class II G protein-coupled receptor family.

The parathyroid hormone (PTH) 1 /PTH-related protein receptor (PTH1R) is a member of the class II G protein-coupled receptor (GPCR) family (1). It binds PTH, the major regulator of blood calcium and PTHrP, an autocrine/paracrine factor (1), and an important developmental regulator, particularly of the skeleton (2,3). The molecular mechanisms that underlie PTH1R function are of high interest because of their physiological and pathophysiological importance, and because the PTH1R is a target for drugs to treat osteoporosis. Hyperparathyroidism, a disease due to overproduction of PTH, leads to hypercalcemia and bone loss, but PTH, when administered to both animals and humans at specified doses and schedules, increases bone mass (4 -6) and reduces the risk of bone fracture (6). Therefore, PTH is a promising therapy for reversing, not merely halting, bone loss. Secretion of PTHrP by cancers is the most common paraneoplastic syndrome, frequently leading to morbidity and mortality due to hypercalcemia (7)(8)(9).
The PTH1R, like all GPCRs, has seven hydrophobic segments that probably form membrane-spanning ␣-helices. The class II subfamily of GPCRs bind peptide ligands of intermediate length, such as secretin, glucagon, calcitonin, corticotropinreleasing hormone, vasoactive intestinal polypeptide, and pituitary adenylyl cyclase-activating polypeptide (PACAP) (10). Ligand binding domains of many of these receptors reside in the both the N-terminal ectodomain and the juxtamembrane region (Refs. 11-17 and for review see Ref. 18). Activation of the PTH1R stimulates multiple effectors, including adenylyl cyclase, phospholipase C, and phospholipase D (19 -21).
Here, we report the solubilization and purification of the human (h)PTH1R from COS-7 cells in high yield and to apparent homogeneity by single-step affinity chromatography. Purified hPTH1Rs retain high affinity binding, and PTH treatment of the purified receptors promotes incorporation of GTP␥S into G s dose-dependently, when they are reconstituted in phospholipid vesicles with recombinant G␣ s -and ␤␥-subunits.
Antibodies-The monoclonal antibody, 1D4, which recognizes the rhodopsin epitope, TETSQVAPA (82), was purchased from the National Cell Culture Center (Minneapolis, MN). G48, a polyclonal antiserum raised in a sheep, contains antibodies that recognize epitopes within the sequences of the rat PTH1R, YPESKENKDVPTGSRRRGRPC, FCNG-EVQAEIRKSWSRWTLAL and SGLDEEASGSARPPPLLQEGWET-VM. The first sequence is located in the N-terminal ectodomain, and the last two sequences are located in the intracellular C-terminal tail of the receptor. For some experiments recognition of G48 was restricted to epitopes in the C-terminal tail by immuno-depletion using YPESKENKDVPTGSRRRGRPC coupled to CNBr-activated Sepharose 4B, following a protocol provided by the manufacturer.
Construction of the hPTH1R Plasmid-Mutations were introduced into the hPTH1R by site-directed mutagenesis (QuickChange Site-directed Mutagenesis Kit, Stratagene, La Jolla, CA). Five amino acids in exon 2 of the hPTH1R were replaced with the corresponding amino acids from the rat PTH1R (E92K, D94N, E96D, A97V, and Y103R) to allow efficient detection by G48 of hPTH1Rs expressed on the cell surface. The 27-mer encoding the epitope (TETSQVAPA), recognized by 1D4, was inserted in-frame immediately upstream of the stop codon at the 3Ј-end of the hPTH1R sequence. The modified hPTH1R cDNA was subcloned at BamHI and XbaI sites in a pcDNA3 expression vector (Invitrogen). The accuracy of the modified receptor construct was confirmed by sequencing the entire receptor (Tufts University, Department of Physiology, Sequencing Core Facility, Boston).
Transient Transfection of COS-7 Cell-All plasmid DNAs were prepared using the EndoFree plasmid Maxi kit (Qiagen Inc., Valencia, CA) and stored at Ϫ20°C in sterile buffer containing 10 mM Tris-HCl and 1 mM EDTA, pH 7.5. Transfection of COS-7 cells with purified plasmid containing the modified hPTH1R cDNA was performed using FuGENE 6 transfection reagents. In brief, 10 g of DNA in 1 ml of Dulbecco's modified Eagle's medium was mixed with 30 l of FuGENE 6 for 20 min at 21°C, and the mixture was added to 60 -75% confluent COS-7 cells in a 15-cm dish. Cells were harvested after 72 h for preparation of membranes. For studies of hPTH1R on intact cells, COS-7 cells were grown in 24-well plates and transfected with 250 ng of cDNA and 0.75 l of FuGENE 6 per well. Cells were studied 72 h after transfection.
Preparation of the Membrane Fraction from COS-7 Cells-Confluent monolayers of COS-7 cells, grown in 15-cm dishes, were washed with phosphate-buffered saline and harvested by scraping with a Teflon policeman in 5 ml of Buffer A (10 mM Tris-HCl and 4 mM EDTA, pH 7.4), containing proteinase inhibitors (10 g/ml pepstatin, 2 g/ml aprotinin, 10 g/ml chymostatin, and 10 g/ml leupeptin) (Sigma). Then, cell lysates were homogenized by 30 strokes with a Duall tissue grinder (Kontes, Vineland, NJ). The homogenates were centrifuged at 700 ϫ g (Beckman J2-21 centrifuge, JLA rotor, Beckman Instruments, Berkeley, CA) for 10 min at 4°C to remove nuclei and debris. Supernatants were centrifuged at 40,000 ϫ g for 30 min at 4°C, and the supernatants were aspirated, and the membrane-enriched pellets were suspended in Buffer B (50 mM Tris-HCl, 0.15 M NaCl, 2 mM CaCl 2 , 5 mM KCl, 5 mM MgCl 2 , 4 mM EDTA, 20% glycerol, pH 7.5) with the same proteinase inhibitors at a protein concentration of 2 mg/ml and stored at Ϫ80°C. Receptors in these crude membrane preparations were stable for at least 1 month under these conditions.
Solubilization and Purification of hPTH1R-The membrane proteins were diluted to a concentration of 1 mg/ml with Buffer B and were then solubilized by incubation with an equal volume of 0.5% DM in Buffer B (45 min, 4°C). A small amount of insoluble material was separated from the clear supernatant by ultracentrifugation at 100,000 ϫ g for 30 min at 4°C (Beckman L8 -55M, 70.1 TI rotor, Beckman Instruments, Berkeley, CA). The supernatant, diluted 2-fold with Buffer B, was added to 1 ml of Sepharose 4B to which 1D4 had been coupled using CNBr and was placed on a rocker platform for 1.5 h at 4°C. The solution and resin then were transferred to Bio-Rad Econo column (0.7 ϫ 5 cm, Bio-Rad), and the resin was washed with 50 ml of Buffer B containing 0.05% DM, and the hPTH1R was eluted by competition with the 9-amino acid rhodopsin peptide (100 M, 2 ml of Buffer B containing 0.05% DM).
Quantification of the hPTH1R-Affinity-purified hPTH1R was concentrated ϳ25-fold on Microcon YM-50 filters (Millipore Co., Bedford, MA) by centrifugation at 12,000 ϫ g for 40 min and hydrolyzed for 24 h at 110°C in 6N HCl containing 0.05% ␤-mercaptoethanol. The hydrolysate was vacuum-dried and re-dissolved in a citrate buffer, and amino acid analysis was performed on a Beckman System 6300 analyzer. Quantification was made after correction for the 9-amino acid rhodopsin epitope. The same preparation was then used to develop a standard to rapidly assess receptor concentration and recovery by Western blotting. Purified hPTH1R (1.9 g) was dissolved in 250 l of Buffer B containing 5 mg/ml of bovine serum albumin (BSA) and serially diluted over a range of 1:25 (1.5 ng) to 1:3200 (11.8 pg) in the same buffer. Standards were established by "slot-blotting" 5 l of each dilution onto nitrocellulose membranes (Bio-Dot SF apparatus, Bio-Rad) and stored for future use. Samples of solubilized crude membrane and purified hPTH1R at several dilutions in Buffer B containing 0.05% DM were then applied to the nitrocellulose membranes. The membranes were blocked by treatment with phosphate-buffered saline containing 5% dry milk and 0.2% Tween 20 and then were incubated with 1D4 antibody for 1 h at room temperature (RT). A horseradish peroxidase-conjugated anti-mouse IgG (1:10,000) (Santa Cruz Biotechnology, Inc., Santa Cruz, CA) was then added, and membranes were developed using Western Lightning Chemiluminescence Reagent Plus (PerkinElmer Life Sciences). The concentration of hPTH1R was determined by interpolation from the intensities of the serially diluted standard.
Sequence Analysis of hPTH1R-Purified hPTH1R was concentrated 25-fold on Microcon YM-50 membranes and subjected to SDS-PAGE (10%). The ϳ80-kDa protein then was electrotransfered to a polyvinylidene difluoride (PVDF) membrane (Millipore, Bedford, MA) and stained with Ponceau S. The receptor band was excised and sequenced directly using a gas-phase protein sequenator (Applied Biosystems, model 477A), and the phenylthiohydantoin derivatives of each amino acid were identified.
Radioreceptor Assay-PTH-(1-34) was labeled with 125 I using chloramine T as catalyst and purified by HPLC as described previously (25,83). Monoiodinated 125 I-PTH-(1-34) (2,000 Ci/mmol) was used within 3 weeks of labeling. Radioreceptor assays with intact COS-7 cells that express the hPTH1R were conducted as described previously (1,25). To assess binding to membrane-embedded hPTH1R, 125 I-PTH-(1-34) (5 ϫ 10 4 -10 ϫ 10 4 cpm) and unlabeled peptides were added to 0.1 ml of Buffer B, pH 7.4, with 1 mM phenylmethylsulfonyl fluoride, 10 g/ml leupeptin, and 0.3% BSA. The reaction then was initiated by addition of 2 g of membrane protein. After a 1-h incubation at 21°C, tubes were centrifuged at 16,000 ϫ g for 4 min at 21°C (Eppendorf centrifuge 5415C, Brinkmann Instrument Co., Westbury, NJ). The pellets were washed once with 200 l of Buffer B; the tips of the microcentrifuge tubes containing the membranes were excised, and radioactivity was counted in a Wallac 1470 Wizard gamma counter.
Binding to purified receptor was assessed both prior to and after dissociation from the affinity support. The solubilized receptor fraction derived from 80 g of crude membrane was added to siliconized Eppendorf tubes containing 30 l of Sepharose 4B to which 1D4 had been coupled by the CNBr method. After incubation (1 h, 4°C), the Sepharose was washed (3 times) with 250 l of Buffer B containing 0.05% DM, 125 I-PTH-(1-34) (8 -10 ϫ 10 4 cpm), and unlabeled peptides were added in 100 l of Buffer B with 10% FBS and 0.05% DM. The reactants were incubated at 21°C for varying periods. Separation of bound from free radioligand was achieved by applying samples to 25-mm PVDF filters (0.45 m, Millipore Co., Bedford, MA), which had been pre-soaked with Buffer B containing 10% FBS before mounting them on a 12-port vacuum manifold (Millipore Co., Bedford, MA). Filters were washed with 3 ml of chilled Buffer B (two times), transferred to tubes, and radioactivity counted with a gamma counter.
The affinity of purified hPTH1R also was assessed after elution from the affinity support. 96-Well plates pre-coated with protein G by the manufacturer (Pierce) were reacted with a 1:10 dilution of G48 (100 l,  hour. The contents of each well were aspirated; the wells were washed with Buffer B (100 l, two times), and bound radioactivity was recovered by incubation with 1 N NaOH (100 l, 30 min, at RT). The contents of each well were transferred to a glass tube, and the radioactivity was counted.

Measurement of Intracellular cAMP and Inositol
Phosphate Accumulation-Intracellular cAMP and inositol phosphate were measured as described previously (19,25).
Preparation of Membrane Fraction from Sf9 Cells-Recombinant G␣ s -, ␤1-, and ␥2-subunits were generated in Sf9 cells and purified by minor modifications of methods described previously (84). The concentration of each G-protein subunit was estimated by Coomassie Blue staining using BSA as standard. Purified G-protein subunits were stored in aliquots (10 -50 ng/l) at Ϫ80°C.
Reconstitution of hPTH1R in Phospholipid Vesicles-Two mg of phospholipids, 1-palmitoyl-2-oleoyl-sn-glycero-3 phosphocholine, phosphoethanolamine, and 1,2-dimyristoyl-sn-glycero-phosphate at a molar ratio of 6:3:1, were mixed, and the chloroform was evaporated overnight in the Speed-Vacuum Concentrator (SVC200H-115) (Savant Instrument Inc., Hicksville, NY), a minor modification of the methods of Mirzabekov et al. (85). The lipid was resuspended in Buffer B (1 ml) without glycerol; [ 3 H]DPPC (0.05 Ci/) was added, and the mixture was sonicated (1 min, Sonifier 450, Branson Ultrasonic Co., Danbury, CT). The sonicated solution was repeatedly passed (20 times) through a polycarbonate membrane (100 nm, pore size) mounted on Liposo Fast-Basic extrusion device (Avestin Inc., Ottawa, Canada). The purified, concentrated hPTH1R (7.6 pmol in 5 l of Buffer B containing 0.05% DM) was mixed with phospholipid vesicles (50 l of Buffer B containing 0.1% DM) and incubated for 30 min at RT. The mixture was then added gently to a polyethylene tube containing a discontinuous sucrose gradient (10, 20, 30, and 50%; 400, 400, 400, and 300 l, respectively). The tubes were centrifuged in SW60 rotor at 160,000 ϫ g for 16 h at 4°C, and incorporation of hPTH1R into phospholipid vesicles was assessed. Samples (120 -170 l) were collected by puncturing the bottom of the tube with a 21-gauge needle. Aliquots from each fraction were counted for [ 3 H], and hPTH1R was assessed by immunoblotting with 1D4 antibody. Miscellaneous Methods-Protein concentration was measured by intrinsic fluorescence using BSA as standard. Protein was excited at 280 nm, and the emission at 350 nm was recorded.

Incorporation of GTP␥S into G s after Reconstitution with hPTH1R and G Protein Subunits
Statistical Analysis-Data were calculated from 2 to 5 independent experiments and expressed as mean Ϯ S.E. of duplicate or triplicates samples. Statistical significance was determined by analysis of variance with Fisher's projected least significant difference (PLSD).

RESULTS AND DISCUSSION
Characterization of the Modified hPTH1R-COS-7 cells were transiently transfected with either wild type hPTH1R or hPTH1R, which had been modified by substitution of five cor- responding amino acids from the rat PTH1R sequence encoded by exon 2 to improve recognition by the G48 antiserum, and by addition of the rhodopsin epitope at the C terminus. Wild type and modified hPTH1R were equally well expressed, as assessed by ligand binding (data not shown). The IC 50 values for binding of PTH- , an agonist, to the two receptors were indistinguishable (wild type, 27.3 Ϯ 7.6 nM versus modified hPTH1R, 28.8 Ϯ 14.2 nM (n ϭ 4)), as were the IC 50 values for binding of PTHrP-(5-36), an antagonist (wild type, 41.7 Ϯ 10.8 nM versus modified hPTH1R, 39.7 Ϯ 7.5 nM (n ϭ 4)) (Fig. 1, A and B). PTH stimulation of cells expressing wild type and modified hPTH1R resulted in accumulation of cAMP and generation of inositol phosphates that also were indistinguishable (cAMP (EC 50 )wild type, 0.29 Ϯ 0.04 nM versus modified hPTH1R, 0.29 Ϯ 0.05 nM, n ϭ 2, inositol phosphates (EC 50 )-wild type, 43.7 Ϯ 1.6 nM versus modified hPTH1R, 40.3 Ϯ 2.1 nM, n ϭ 2, Fig. 1, C and D).
Solubilization and Purification of the Recombinant hPTH1R-COS-7 cell membranes expressing modified hPTH1R were suspended in Buffer B and solubilized (45 min, 4°C) by addition of an equal volume of Buffer B, which contained 0.5% DM. This concentration maximized recovery of functionally intact receptors as assessed by binding of 125 I-PTH-(1-34) to 1D4-bound receptors and by Western blotting of eluted receptors using 1D4 antibody. Compared with receptors solubilized in 0.25% DM, recovery of receptors was inefficient at a final concentration of 0.125% DM, and receptors recovered after solubilization at a final concentration of 0.5% DM did not efficiently bind the radioligand (data not shown). Additionally, DM at a final concentration of 0.25% was more efficacious for solubilizing functional receptors than other detergents we assessed: digitonin, CHAPSO, n-hexyl-, n-septyl-, n-octyl-, nnonyl-, or n-decyl-␤-D-glucopyranoside (data not shown). Fig.  2A shows a Coomassie Blue-stained gel of proteins recovered from the membrane extract and after affinity chromatography; the purified hPTH1R appears as a broad ϳ80-kDa band, free of detectable contaminating proteins.
Recovery of receptor protein and functional receptors was estimated, respectively, by comparing the intensity of the purified receptor band detected by immunoblotting with 1D4 against a serially diluted standard of hPTH1R that had first been quantified by amino acid analysis and by Scatchard analysis of equilibrium ligand binding. Two preparations of crude COS-7 cell membrane extract, each harvested from 20 15-cm dishes, contained 4.85 and 5.50 mg of protein, 1.56 and 1.70 nmol of receptor protein, and 1.21 and 1.30 nmol of functional receptors, respectively (Table I). Thus, ϳ75% of the membraneembedded receptors were functional. DM (0.25%) solubilized 1.44 and 1.50 nmol of receptor protein in the two experiments, or ϳ90%. The eluate after 1D4 affinity chromatography contained 1.09 and 1.15 nmol of receptor protein (75%), giving an overall receptor recovery of about 69%. Scatchard analysis revealed minimal, if any, loss of functional receptors after solubilization and purification. Human PTH1Rs comprised about 2% of the protein in the crude membrane preparations and were purified about 42-fold, with a resultant increase in specific activity from 0.25 to 10.5 nmol/mg.
Immunoblotting with 1D4 of the SDS-PAGE, after electrotransfer to Immobilon P, showed an apparently homogeneous broad band with a molecular size of ϳ80 kDa, consistent with a glycosylated receptor protein. In some gels, an ϳ180-kDa protein band also was detected, regardless of the presence of reducing reagent, which may represent dimers (Fig. 2B).
The modified hPTH1R also was expressed in Sf9 cells. However, under a variety of detergent conditions that solubilized as little as 10 -15% of the receptors, including conditions used by Ohtaki et al. (79) for the PACAP receptor, receptors no longer bound 125 I-PTH-(1-34) (data not shown).
The SDS-PAGE-purified receptor band was electrotransfered to PVDF membrane and subjected to sequence analysis for 10 cycles, which yielded YALVDADDVM. Tyr 23 at the N terminus is consistent with algorithms that predict removal of a 22-amino acid signal peptide from the N terminus of the receptor. Signal peptide sequences are highly conserved among all cloned PTH1Rs, including opossum, rat, canine, murine, human, porcine, and zebrafish, and thus tyrosine is the likely N terminus in these mature receptors.
Purified and Membrane-embedded hPTH1R Have Similar Binding Properties-Ligand association to the high concentration of receptors in both membrane-embedded and purified hPTH1R, when still attached to the affinity gel, was rapid; half-maximal binding was achieved in less than 3 min at 21°C; equilibrium binding was achieved within 10 min (Fig. 3, A and  B). Ligand dissociation was measured after initially binding 125 I-PTH-  to the purified receptors for 1 h at RT and then measuring residual binding at various times after addition of 10 Ϫ6 M PTH- . Dissociation was rapid and biexponential; the half-time of the rapid component was ϳ3 min (Fig. 3C). This compares favorably with dissociation of ligand bound to endogenous canine renal PTH1R and also to hPTH1R overexpressed in COS-7 cells after addition of GTP␥S, 1-2 (86) and 5 min (87), respectively. In the absence of added GTP␥S, less than half of the PTH radioligand dissociated from endogenous receptors by 200 min (86), which contrasts strikingly with a half-time of only 6.5 min when radioligand is bound to exogenously expressed receptors on COS-7 cells (87).
Equilibrium binding to membrane-embedded receptors had a slightly lower affinity constant (K d ϭ 11.9 Ϯ 1.9 nM, specific binding, 10.7 Ϯ 1.4%, n ϭ 8) compared with binding to immobilized, purified hPTH1R (K d ϭ 16.5 Ϯ 1.3 nM, specific binding-18.6 Ϯ 4.7%, n ϭ 4) (Fig. 4A) and was indistinguishable from the apparent affinity, K d ϭ 17 nM, of ligand bound to canine renal membranes after treatment with GTP␥S. In contrast, the apparent K d of canine PTH1R in the absence of added GTP␥S was 2-5 nM (86,88). The data are consistent with the notion that most of the extraordinarily large number of receptors expressed in COS-7 cells are not coupled to G proteins and thus are in a low affinity state (87). Because purified receptors are not associated with G proteins, they retain binding properties closely similar to those of membrane-embedded receptors in COS-7 cells.
The affinity of solubilized receptors also was measured after they had been eluted from the affinity gel and added to wells containing immobilized G48, which had been depleted of antibodies that recognize the epitope in the N-terminal ectodomain. Specific binding at equilibrium was 7.6 Ϯ 0.5%, and the apparent K d by Scatchard analysis was 69.5 Ϯ 16.9 nM (n ϭ 3) (Fig.  4B). The higher affinity constant of these receptors may well reflect altered receptor conformation attendant to binding to the immobilized G48 antiserum. We cannot exclude, however, that receptors were slightly denatured upon elution from the 1D4 resin.
PTH Treatment of Purified hPTH1R Reconstituted in Phospholipid Vesicles with G Proteins Stimulates Incorporation of [ 35 S]GTP␥S into G s -We first determined that 0.1% DM was sufficient to solubilize the phospholipids (data not shown). Purified hPTH1R (7.6 pmol) in Buffer B containing 0.05% DM was then mixed with the solubilized phospholipids containing [ 3 H]DPPC, applied to a discontinuous sucrose gradient, and centrifuged. Fractions including the interface of the 20 -30% sucrose layers contained the highest concentration of [ 3 H]D-PPC and hPTH1R, as assessed by immunoblotting with 1D4 (Fig. 5A).
In summary, we have purified hPTH1R to apparent homogeneity and demonstrated that the signal peptide is removed, leaving Tyr 23 as the N terminus of the mature receptor. The purified receptors retain ligand-binding properties closely similar to those of membrane-embedded receptors, including functional N-terminal ectodomain and juxtamembrane ligand-interactive domains. PTH-stimulated GTP␥S incorporation into G␣ s , when receptors are reconstituted in phospholipid vesicles with recombinant G protein subunits, demonstrates that the purified receptors couple to and activate G s . In studies to be published elsewhere, we have replaced all unpaired cysteine residues with functionally neutral amino acids, mutated native amino acids to cysteine one at a time, and have begun to identify structural features of the hPTH1R, using thiol-specific reagents and site-directed spin labeling/electron paramagnetic resonance. Transient expression of hPTH1R mutants in COS-7 cells also yields sufficient receptors to readily enable analysis of protein structure using fluorescent probes. The synthesis of sufficient hPTH1R for crystallographic and NMR studies obviously will require considerable "scale-up" and may best be achieved through recovery of receptors expressed stably in mammalian cells adapted for growth in suspension culture. The use of 1D4-affinity chromatography should efficiently purify other functional class II GPCR as well as other proteins, which have been difficult to purify by more conventional methods.