Minimization of Parathyroid Hormone

The amino-terminal and carboxyl-terminal portions of the 1–34 fragment of parathyroid hormone (PTH) contain the major determinants of receptor activation and receptor binding, respectively. We investigated how the amino-terminal signaling portion of PTH interacts with the receptor by utilizing analogs of the weakly active fragment, rat (r) PTH(1–14)NH2, and cells transfected with the wild-type human PTH-1 receptor (hP1R-WT) or a truncated PTH-1 receptor which lacked most of the amino-terminal extracellular domain (hP1R-delNt). Of 132 mono-substituted PTH(1–14) analogs, most having substitutions in the (1-9) region were inactive in assays of cAMP formation in LLC-PK1 cells stably expressing hP1R-WT, whereas most having substitutions in the (10-14) region were active. Several substitutions (e.g.Ser3 → Ala, Asn10 → Ala or Gln, Leu11 → Arg, Gly12 → Ala, His14→ Trp) enhanced activity 2–10-fold. These effects were additive, as [Ala3,10,12,Arg11,Trp14] rPTH(1–14)NH2 was 220-fold more potent than rPTH(1–14)NH2 (EC50 = 0.6 ± 0.1 and 133 ± 16 μm, respectively). Native rPTH(1–11) was inactive, but [Ala3,10,Arg11]rPTH(1–11)NH2achieved maximal cAMP stimulation (EC50 = 17 μm). The modified PTH fragments induced cAMP formation with hP1R-delNt in COS-7 cells as potently as they did with hP1R-WT; PTH(1–34) was 6,000-fold weaker with hP1R-delNt than with hP1R-WT. The most potent analog, [Ala3,10,12,Arg11,Trp14]rPTH(1–14)NH2, stimulated inositol phosphate production with hP1R-WT. The results show that short NH2-terminal peptides of PTH can be optimized for considerable gains in signaling potency through modification of interactions involving the regions of the receptor containing the transmembrane domains and extracellular loops.

In mammals, parathyroid hormone (PTH) 1 plays a vital role in regulating blood calcium concentrations, and PTH-related peptide (PTHrP) plays a critical role in the development of the fetal skeleton (1). The biological actions of both of these peptides are mediated by the PTH/PTHrP receptor (or PTH-1 receptor) (2), a family B G protein-coupled receptor (3) that strongly activates the adenylyl cyclase/protein kinase A-signaling cascade (2), and more weakly the phospholipase C protein kinase C-signaling pathway (4). The mechanisms by which parathyroid hormone and PTHrP bind to the PTH-1 receptor and induce receptor activation are poorly understood but appear to involve multiple sites of intermolecular interaction. Early studies of PTH fragment analogs assigned the major determinants of receptor-binding affinity and cAMP-stimulating potency to the COOH-terminal and NH 2 -terminal portions of the fully active PTH  peptide, respectively (5,6). PTH -based analogs with NH 2 -terminal deletions, such as PTH  and PTH , bind efficiently to the receptor and are severely defective in stimulating a cAMP response; such peptides thus function as PTH-1 receptor antagonists (7)(8)(9). The dominant role that the NH 2 -terminal residues of PTH and PTHrP play in receptor activation is further reflected by their high level of evolutionary conservation.
Cell Culture-LLC-PK1-derived and COS-7 cells were cultured at 37°C in T-75 flasks (75 mm 2 ) in Dulbecco's modified Eagle's medium supplemented with fetal bovine serum (10%), penicillin G (20 units/ml), streptomycin sulfate (20 g/ml), and amphotericin B (0.05 g/ml) in a humidified atmosphere containing 5% CO 2 . ROS 17/2.8 cells were cultured as above except that Ham's F-12 medium was used instead of Dulbecco's modified Eagle's medium and fetal bovine serum was at 5%. Stock solutions of EGTA/trypsin and antibiotics were from Life Technologies, Inc.; fetal bovine serum was from HyClone Laboratories (Logan, UT). Cells were subcultured in 24-well plates and, when confluent, were treated with fresh media and shifted to 33°C for 12-24 h prior to assay. This shift to 33°C was included as a means to potentially maximize cell surface expression of the PTH receptors and thus optimize signal sensitivity, since we found previously that the reduced temperature incubation resulted in small (10 -50%) increases in surface expression of wild-type and mutant PTH receptors (15). The ability of lower temperatures to improve the surface expression of the lutropin receptor and other membrane proteins has been discussed previously (22). The HKRK-B7 cell line (23) was derived by stable transfection of LLC-PK1 porcine kidney cells with the hPTH-1 receptor cDNA and express these receptors at a density of ϳ950,000 receptors/cell. ROS 17/2.8 cells, a rat osteoblast-like cell line (24), express endogenous rat PTH-1 receptors at a density of ϳ70,000 receptors/cell (25).
PTH Receptor Mutagenesis and COS-7 Cell Expression-The pCDNA-1-based plasmid encoding the intact hPTH-1 receptor (HKrk in Ref. 26 and herein called hP1R-WT) was used for studies in COS-7 cells. The truncated human PTH-1 receptor (hP1R-delNt) was constructed from the hP1R-WT plasmid by oligonucleotide-directed mutagenesis (27). This mutant receptor is deleted for residues 24 -181. The predicted signal peptidase cleavage of this receptor between Ala 22 and Tyr 23 (28) generates Tyr 23 as the NH 2 -terminal residue, which is joined directly to Glu 182 located at or near the boundary of the first transmembrane domain. A similarly truncated rat PTH receptor containing an NH 2terminal epitope tag (rP1R-delNt-HA) was described by us previously and shown by antibody binding experiments to be expressed at approximately 60% the level of the intact wild-type receptor (14,29). Transient transfections of COS-7 cells were performed using DEAE-dextran and 200 ng of cesium chloride-purified plasmid DNA per well of a 24-well plate, as described previously (15).
cAMP Stimulation-Stimulation of cells with peptide analogs was performed in 24-well plates. Cells were rinsed with 0.5 ml of binding buffer (50 mM Tris-HCl, 100 mM NaCl, 5 mM KCl, 2 mM CaCl 2 , 5% heat-inactivated horse serum, 0.5% fetal bovine serum, adjusted to pH 7.7 with HCl) and treated with 200 l of cAMP assay buffer (Dulbecco's modified Eagle's medium containing 2 mM 3-isobutyl-1-methylxanthine, 1 mg/ml bovine serum albumin, 35 mM Hepes-NaOH, pH 7.4) and 100 l of binding buffer containing varying amounts of peptide analog (final volume ϭ 300 l). The medium was removed after incubation for 1 h at room temperature, and the cells were frozen (Ϫ80°C), lysed with 0.5 ml of 50 mM HCl, and refrozen (Ϫ80°C). The cAMP content of the diluted lysate was determined by radioimmunoassay. Where possible, EC 50 and corresponding maximum response values (E max ) were calculated using nonlinear regression (see below). For inhibition studies, the hPTH(3-34) antagonist peptide was added to the rinsed cells in 100 l of binding buffer immediately prior to the addition of 100 l of cAMP assay buffer and 100 l of cAMP assay buffer containing varying amounts of agonist peptide (final volume ϭ 300 l); the cells were then incubated for 60 min at room temperature and processed as described above.
Stimulation of Inositol Phosphate Production-COS-7 cells transfected as above with hP1R-WT were treated with serum-free, inositolfree Dulbecco's modified Eagle's medium containing 0.1% bovine serum albumin and myo-[ 3 H]inositol (NEN Life Science Products) (2 Ci/ml) for 16 h prior to assay. At the time of the assay, the cells were rinsed with binding buffer containing LiCl (30 mM) and treated with the same buffer with or without a PTH analog. The cells were then incubated at 37°C for 40 min, after which the buffer was removed and replaced by 0.5 ml of ice-cold 5% trichloroacetic acid solution. After 3 h on ice, the lysate was collected and extracted twice with ethyl ether. The lysate was then applied to an ion exchange column (0.5-ml resin bed) and the total inositol phosphates were eluted as described previously (30), and counted in liquid scintillation mixture.
Data Calculation-Calculations were performed using Microsoft Excel. Nonlinear regression analysis of cAMP stimulation data was performed using four parameters, defined as the minimum (Min), maximum (Max, E max ), midpoint (EC 50 ), and slope of the response curve. The predicted response (y p ) for a given dose (x) of peptide was calculated using the following equation: The initial parameter values were estimated from the primary data, and the Excel "solver function" was then used to vary the four parameters in order to minimize the differences between the predicted and actual responses (least-squares method) (31). For each experiment, the maximum was constrained to within Ϯ1 standard deviation of the maximum response observed in that experiment for rPTH  at a dose of 1 ϫ 10 Ϫ7 M. The optimized equations were used to curve-fit the data shown in the graphs and to obtain the EC 50 and corresponding maximum (E max(calc) ) values reported in the tables. The observed maximum responses (E max(obs) ) were those attained by each NH 2 -terminal fragment analog at a dose of 100 M and by each PTH  analog at a dose of 100 nM, except for studies in cells expressing hP1R-delNT where the E max(obs) for rPTH  and hPTH  was determined at a dose of 10 M and for [Ala 1,3,10,12 ,Arg 11 ,Tyr 34 ]hPTH(1-34)NH 2 at dose of 20 M. In some cases where the dose-response curves did not attain a true asymptotic maximum, as with native rPTH (1)(2)(3)(4)(5)(6)(7)(8)(9)(10)(11)(12)(13)(14), the E max(calc) values are greater than the E max(obs) values. The statistical significance between two data sets was determined using a one-tailed Student's t test assuming unequal variances for the two sets.

TABLE III cAMP responses in COS-7 cells expressing intact or truncated PTH-1 receptors
The wild-type and NH 2 -terminally truncated human PTH receptors were transiently transfected in COS-7 cells and stimulated with varying doses of the indicated peptides. Values for cAMP production (means Ϯ S.E.) were determined from three experiments, as described under "Materials and Methods" and in Table I other bioactive peptides in the data bases (analyzed using the FASTA program of the Genetics Computer Group (Madison WI) software package). Even [Ala 3,10 ,Arg 11 ]PTH(1-11) retained closest homology to PTH and PTHrP (73% and 55% identity, respectively). Consistent with this, the signaling responses induced by the PTH fragments were fully dependent on the PTH-1 receptor, as they were inactive in untransfected LLC-PK1 and COS-7 cells (data not shown). How these substitutions enhance activity is unknown. It is clear that they modify interactions with the juxtamembrane portion of the receptor containing the transmembrane helices and extracellular loops, because they improved potency on the truncated receptor by as much as they did on the intact receptor. The functional intolerance of residues in the (1-9) region, especially Val 2 , Ile 5 , and Met 8 , is at least consistent with the recently reported computer model of the complex formed between PTH  and the PTH-1 receptor (39), which predicts that these NH 2 -terminal residues of PTH interact with the transmembrane helices and/or extracellular loops of the receptor. Some of the enhancing substitutions may induce a more favorable peptide conformation (a short ␣-helix has been detected in this region of PTH(1-34)-type analogs by NMR spectroscopy (Refs. 40 -42)), while other substitutions might introduce more favorable side chain interactions with the receptor, possibly to sites that have been identified by cross-linking and mutational studies to be interaction sites for the NH 2 -terminal portion of PTH(1-34) (16 -19, 29, 43). Eventually, structural models derived from NMR studies and further mutational data, such as that reported here, should help to define the overall receptor/ligand interaction more completely.
The weak but measurable levels of agonism seen in peptides such as [Ala 3,10 ,Arg 11 ]rPTH(1-11)NH 2 or [Ala 3 ,Gln 10 ]rPTH(1-10)NH 2 , in transfected cells and even in an osteoblast-like cell indicate that a peptide based only on the first 10 amino acids of native PTH (84 amino acids in humans) can be sufficient for specific PTH receptor activation. Such peptides should be suitable as scaffolds for further rounds of optimization, as our current data show that considerable improvements in agonist efficacy are attainable in comparatively small ligands. The use of minimized PTH ligands and truncated receptors, as we have reported here, should help in determining the essential components of the cAMP and inositol phosphate activation mechanisms in the PTH-1 receptor. The resultant information may prove useful in the analysis of other family B G protein-coupled receptors that interact with peptide hormones of similar size to PTH , and could potentially aid in the rational design of even simpler activating molecules.