Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

The cell surface Ca²⁺ receptor is the primary molecular entity regulating the secretion of parathyroid hormone (PTH). Activation of this receptor by increased levels of extracellular Ca²⁺ inhibits PTH secretion, whereas its presumed inactivation by lowered extracellular Ca²⁺ leads to an increase in PTH secretion. PTH regulates systemic Ca²⁺ homeostasis by acting on target cells in both the kidney and the skeleton to increase plasma levels of Ca²⁺ (Brown and MacLeod, 2001). In bone, PTH increases bone turnover but the resulting overall effect on bone mineral density is highly dependent on the temporal changes in the circulating levels of PTH. Thus, sustained elevations in plasma PTH levels, such as occur in hyperparathyroidism, result in increased bone resorption and a net decrease in bone mass at most skeletal sites (Antonsen and Sherrard, 1995; Silverberg et al., 1995). In contrast, temporary increases in plasma levels of PTH or its 1-34 fragment, achieved by the daily (or near daily) injection of either peptide, stimulate new bone formation in animal models of osteopenia (Dempster et al., 1993; Kimmel et al., 1993; Ejersted et al., 1995; Li and Wronski, 1995; Fox et al., 1997) and in clinical studies of osteoporotic patients (Reeve, 1996; Hodsman et al., 1997; Lindsay et al., 1997; Lane et al., 1998; Fujita et al., 1999, Rittmaster et al., 2000). The results of these studies using peptides administered intermittently demonstrate that PTH is a potent anabolic agent that increases bone mineral density and bone strength to a greater extent than that achieved by antiresorptive therapies (Mosekilde et al., 1994; Hodsman et al., 1997; Lindsay et al., 1997; Lane et al., 1998; Fujita et al., 1999; Rittmaster et al., 2000). The profound stimulatory effect of these peptides on bone formation has generated interest in the use of PTH or its fragments as a novel anabolic therapy for osteoporosis. However, the therapeutic use of these peptides is compromised by the need for systemic administration of a costly biological agent.

An alternative approach that might overcome these drawbacks, and yet achieve similar anabolic effects on bone, is based on the use of small, orally active compounds that regulate plasma levels of endogenous PTH (Nemeth, 1996; Fox et al., 1997). This hypothesis holds that blocking Ca²⁺ receptor activity with small molecules will stimulate PTH secretion. With the appropriate pharmacokinetic profile, such compounds would be expected to cause a marked but transient increase in circulating PTH levels, sufficient to stimulate new bone formation. While this hypothesis is in line with conventional thinking, it remains untested because no ligand has been found that blocks activation of the Ca²⁺ receptor. In contrast, a wide variety of inorganic or organic polycations and certain phenylalkylamines have been shown to act as agonists or allosteric activators of this receptor (Nemeth and Fox, 1999). Like G protein-coupled receptor agonists in general, those that activate the Ca²⁺ receptor exhibit marked tissue selectivity (Lavigne et al., 1998; Fox et al., 1999) and are not ideal ligands to study Ca²⁺ receptor function, particularly in those tissues that are not involved in systemic Ca²⁺ homeostasis and that often express much lower levels of the Ca²⁺ receptor than classic "calcemic tissues" such as the parathyroid glands and kidney. In contrast, G protein-coupled receptor antagonists typically do not show profound tissue selectivity. Because of this, receptor antagonists are more valuable tools to study receptor function in a variety of different tissues. Such compounds, if capable of stimulating secretion of PTH, might also provide structures for novel drugs capable of transiently increasing levels of plasma PTH and stimulating new bone formation.

The need for such an anabolic therapy is underscored by the serious health problem posed by osteoporosis, the incidence of which is increasing as the general population ages. Already there are nearly 6 million women and about 2 million men with osteoporosis in the United States and a far greater number of individuals with osteopenia or low bone mineral density. While currently available antiresorptive therapies, such as estrogen or bisphosphonates prevent further bone loss, they cause relatively small increases in new bone formation. The ability to stimulate new bone formation and thereby increase bone mass to levels approaching those in young adults, would constitute a significant advance in the treatment of osteoporosis.

This report describes the salient pharmacodynamic properties of NPS 2143, a small molecule (Fig. 1) that blocks the parathyroid gland Ca²⁺ receptor and stimulates PTH secretion in vitro and in vivo. This compound, which is one member of a family of structurally similar compounds, is the first substance, either ionic or molecular, shown to possess inhibitory activity at the Ca²⁺ receptor.

View larger version (7K): [in this window] [in a new window]   Fig. 1.   Chemical structure of NPS 2143: N-[(R)-2-hydroxy-3-(2-cyano-3-chlorophenoxy)propyl]-1,1-dimethyl-2-(2-naphthyl)ethylamine. The compound was used as the monohydrochloride salt. Shown is the R-enantiomer which, depending on the assay, is 10- to 100-fold more potent than the corresponding S-enantiomer.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Assays for Assessing Potency and Selectivity of Compounds on Ca²⁺ Receptor. HEK 293 cells engineered to express the human parathyroid Ca²⁺ receptor have been described in detail previously (Nemeth et al., 1998). This clonal cell line, referred to as HEK 293 4.0-7 cells, has been used in a high-throughput screening format to detect agonists and allosteric activators (calcimimetics) of the Ca²⁺ receptor (Nemeth et al., 1998). Changes in the concentration of cytoplasmic Ca²⁺ ([Ca²⁺]_i) provide a quantitative and functional assessment of Ca²⁺ receptor activity in these cells and the results using this assay parallel those obtained using a homologous expression system of bovine parathyroid cells. On-line continuous measurements of fluorescence in fluo-3- or fura-2-loaded HEK 293 4.0-7 cells were obtained using a custom-built spectrofluorimeter (Racke and Nemeth, 1993) or a fluorescence imaging plate reader instrument (Molecular Devices, Sunnyvale, CA). Test compounds were incubated with cells for 1 min before increasing the concentration of extracellular Ca²⁺ from 1.0 to 1.75 mM. Compounds were tested individually at a concentration of 100 µg/ml (20-80 µM) and those causing more than a 40% inhibition of the control response were considered to be biologically active.

Assays of Ca²⁺ Receptor Activity Using Bovine Parathyroid Cells. The effect of NPS 2143 on Ca²⁺ receptor-dependent regulation of PTH secretion and cyclic AMP formation was assessed using primary cultures of dissociated bovine parathyroid cells. Following overnight culture, the cells were removed from the flasks by decanting and washed with buffer containing 126 mM NaCl, 4 mM KCl, 1 mM MgSO₄, 0.5 mM CaCl₂, 0.7 mM K₂HPO₄/KH₂PO₄, 20 mM Na-HEPES, pH 7.45, 1 mg/ml glucose, and 0.1% bovine serum albumin (ICN Biomedicals, Cleveland, OH). Portions (0.2 ml) of this cellular suspension were added to 12- × 75-mm polystyrene tubes with or without NPS 2143 and/or varying concentrations of CaCl₂. Incubations (in triplicate) at 37°C for 20 or 30 min were terminated by placing the tubes on ice. Cells were pelleted by centrifugation (500g for 10 min at 4°C) and 0.1 ml of supernatant was immediately assayed for PTH content. A portion of the cellular suspension was left on ice during the incubation period and then processed in parallel with the other tubes. The amount of PTH in the supernatant from the tubes maintained on ice was defined as "basal release" and was subtracted from all other samples. PTH levels were quantified using a rat PTH(1-34) immunoradiometric assay kit, which also detects bovine PTH (Immutopics, San Clemente, CA). For each experiment, results were expressed as picograms of PTH released/10⁶ cells and then normalized to PTH released in 0.5 mM Ca²⁺. Cell numbers were determined by counting nuclei in a hemocytometer after lysing the cells and staining the nuclei with cresyl violet.

Plasma Levels of PTH in Rats. Normal male Sprague-Dawley rats (250-275 g) with unrestricted access to commercial rodent chow (Teklad 8640; Harlan Teklad, Madison, WI) and tap water were used. The animals were anesthetized by intraperitoneal injection of ketamine/xylazine (90:7 mg/kg) and chronic indwelling catheters were implanted in the inferior vena cava (for compound infusion) and in the abdominal aorta (for blood sampling) accessed by the femoral vein and artery, respectively. Following catheterization, the rats were housed individually and allowed to recover for at least 3 days before study. The protocol was approved by the Institutional Animal Care and Use Committee of NPS Pharmaceuticals, Inc. (Salt Lake City, UT).

Statistical Analyses. The potency of NPS 2143 for stimulating PTH secretion in parathyroid cells (EC₅₀) or inhibiting [Ca²⁺]_i responses in HEK 293 4.0-7 cells (IC₅₀) was determined by fitting a curve to the data from each series of experiments with the Levenberg-Marquardt algorithm using the KaleidaGraph program (Synergy Software, Reading, PA). The differences in PTH secretion and cyclic AMP accumulation by parathyroid cells treated with agonists and antagonists of the Ca²⁺ receptor were analyzed by analysis of variance followed by Fisher's least-significant difference procedure (StatView; SAS Institute, Cary, NC). The plasma PTH and Ca²⁺ levels in the in vivo studies were analyzed by repeated measures analysis of variance, followed by a t test to determine the significance of differences at each time point (StatView).

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Potency and Selectivity of NPS 2143. Increasing the concentration of extracellular Ca²⁺ from 1.0 to 1.75 mM caused a rapid and transient increase followed by a lower yet more prolonged increase in [Ca²⁺]_i in HEK 293 4.0-7 cells expressing the human Ca²⁺ receptor (Fig. 2a). Preincubation of these cells with NPS 2143 caused a concentration-dependent inhibition of the cytoplasmic Ca²⁺ response to extracellular Ca²⁺ (Fig. 2a). When NPS 2143 (300 nM) was added to cells during the prolonged phase of the response, there was an immediate fall in [Ca²⁺]_i to baseline values (Fig. 2a). Concentration-response curves, obtained using a change in agonist concentration equivalent to the EC₈₀ (a 1.0-1.75 mM increase in the concentration of extracellular Ca²⁺) yielded an IC₅₀ for NPS 2143 of 43 ± 5 nM.

View larger version (12K): [in this window] [in a new window]   Fig. 2.   NPS 2143 blocks increases in [Ca2+]i elicited by increases in extracellular Ca2+ but not those induced by ATP or thrombin in HEK 293 cells expressing the human Ca2+ receptor. Fluorescence recordings were obtained from fura-2-loaded cells in a cuvette with continuous stirring. Receptor agonists and NPS 2143 were added to the cuvette as indicated by the arrows. Breaks in the fluorescence traces at the arrows are addition artifacts resulting from opening the cuvette compartment. In the traces shown here, and in those of Figs. 3 and 6, these artifacts have been intentionally eliminated. The recordings in a are superimposed traces in the absence of NPS 2143 (top trace) or after pretreatment with 30 (middle trace) or 300 nM (bottom trace) NPS 2143 before increasing the concentration of extracellular Ca2+ from 1.0 to 1.75 mM. The recordings in b are superimposed traces showing responses to 1 U/ml thrombin in the absence or presence of 10 µM NPS 2143 and 1.0 mM extracellular Ca2+. Traces in c and d show responses to ATP (100 nM) and to extracellular Ca2+ (1.0-1.75 mM) in the presence or absence of 300 nM NPS 2143. The numbers accompanying each trace are estimates of [Ca2+]i in nanomolar concentration. Shown are representative recordings from two or three separate experiments.

View larger version (12K): [in this window] [in a new window]   Fig. 3.   Influx of extracellular Ca2+ elicited by thapsigargin is blocked by La3+ but not by NPS 2143. Fluorescence was recorded from a cuvette containing HEK 293 4.0-7 cells loaded with fura-2. Thapsigargin, NPS 2143, and La3+ were each used at a final concentration of 1 µM. The numbers accompanying each trace are estimates of [Ca2+]i in nanomolar concentration. The traces shown are from a single preparation of cells and, with the exception of b, are representative of the results obtained using a separate cell preparation. In b, basal [Ca2+]i was slightly elevated compared with those in a or c and those in another cell preparation.

View larger version (7K): [in this window] [in a new window]   Fig. 4.   NPS 2143 does not affect the activity of mGluR1. A chimeric rat mGluR1 receptor containing the human mGluR1 transmembrane domain was transiently expressed in HEK 293 cells. Recordings show the [Ca2+]i responses elicited by 3,5-dihydroxyphenylglycine (DHPG) (10 µM), a selective group 1 mGluR agonist, in the absence (a) or presence (b) of 10 µM NPS 2390, a small molecule antagonist of group I mGluRs, or 3 µM NPS 2143 (c). Experiments were performed in buffer containing 1.0 mM extracellular Ca2+ and fluorescence was recorded using the fluorescence imaging plate reader. Each trace is from a single well of a 96-well plate and is representative of that recorded in 3 to 11 other wells treated identically. The numbers accompanying each trace are arbitrary units of fluorescence and, when measured in wells from a single plate, are directly comparable. DMSO, dimethyl sulfoxide.

Mechanism of Action of NPS 2143. Concentration-response curves to increasing concentrations of extracellular Ca²⁺ were generated using data derived from studies with HEK 293 4.0-7 cells in the presence or absence of NPS 2143. The addition of NPS 2143 shifted the extracellular Ca²⁺ concentration-response relationship to the right in a concentration-dependent manner (Fig. 5). The maximal response to extracellular Ca²⁺ was not affected by treatment with NPS 2143 nor did NPS 2143 decrease [Ca²⁺]_i below that caused by lowered levels of extracellular Ca²⁺. The inhibitory potency of NPS 2143 is thus dependent on the concentration of extracellular Ca²⁺ and this compound has little effect at levels of extracellular Ca²⁺ outside the range of those occurring physiologically. In this respect, the effects of NPS 2143 parallel those of type II calcimimetic compounds.

View larger version (16K): [in this window] [in a new window]   Fig. 5.   NPS 2143 shifts the concentration-response curve to extracellular Ca2+ in HEK 293 cells expressing the human parathyroid Ca2+ receptor. HEK 293 4.0-7 cells were incubated for 40 to 60 sec in the absence () or presence of 30 (), 100 (), 300 (), or 1000 () nM of NPS 2143 before increasing the concentration of extracellular Ca2+ to the final indicated level. Each point is the mean of duplicate determinations from a single experiment and representative of the results obtained in two other separate experiments.

View larger version (14K): [in this window] [in a new window]   Fig. 6.   NPS 2143 blocks cytoplasmic Ca2+ responses to NPS R-467 in HEK 293 4.0-7 cells. Cells were preincubated with 300 nM NPS 2143 or vehicle before cumulative additions of NPS R-467 (arrowheads) to final concentrations of 0.1, 1.1, or 4.1 µM. The numbers accompanying each trace are estimates of [Ca2+]i in nanomolar concentration.

NPS 2143 Affects Functional Responses of Parathyroid Cells. The above-mentioned results showed that NPS 2143 was a potent antagonist of the Ca²⁺ receptor with some degree of selectivity, enough, at least, to test the suitability of this compound as a tool to alter Ca²⁺ receptor activity in cells normally expressing this receptor. Bovine parathyroid cells were used as a homologous expression system, and two distinct functional readouts of Ca²⁺ receptor activity were monitored: secretion of PTH and formation of cyclic AMP.

View larger version (16K): [in this window] [in a new window]   Fig. 7.   NPS 2143 stimulates PTH secretion from bovine parathyroid cells. a, parathyroid cells were suspended in buffer containing 1.25 mM Ca2+ and incubated for 20 min in the absence () or presence () of increasing concentrations of NPS 2143. b, parathyroid cells were initially suspended in buffer containing 0.5 mM Ca2+ and then transferred to tubes containing a small volume of CaCl2 to obtain the final indicated concentration of extracellular Ca2+. Incubation in the absence () or presence () of 300 nM NPS 2143 was for 20 min. Each point is the mean ± S.E. of four separate experiments each performed in triplicate.

View larger version (18K): [in this window] [in a new window]   Fig. 8.   NPS 2143 blocks the inhibitory effects of extracellular Ca2+ on isoproterenol-stimulated increases in cyclic AMP formation in bovine parathyroid cells. Dissociated parathyroid cells were incubated with or without isoproterenol in buffer containing 0.5 or 2 mM CaCl2 in the presence or absence of 300 nM NPS 2143. Control levels of cyclic AMP (in buffer containing 0.5 mM CaCl2 and isoproterenol) were 5.6 ± 1.6 pg/well (n = 3). a, p < 0.05 versus control. b, p < 0.05 versus 2 mM calcium in the absence of NPS 2143.

NPS 2143 Increases Plasma Levels of PTH in Rats. The in vitro experiments described above clearly demonstrate that NPS 2143 increases the release of PTH from parathyroid cells. However, from these experiments it is not clear whether the increased amount of PTH appearing in the medium results from either a rapid stimulation of regulated exocytotic PTH secretion or increased PTH synthesis or some combination of both. To help distinguish between these mechanisms, NPS 2143 was administered intravenously to normal rats. This route of administration was chosen to avoid pharmacokinetic features of this molecule that could limit its plasma concentration to subthreshold levels. This study, performed under physiological conditions, also provides information relevant to the therapeutic feasibility of using Ca²⁺ receptor antagonists to stimulate bone formation.

View larger version (20K): [in this window] [in a new window]   Fig. 9.   Plasma levels of PTH and Ca2+ in normal rats infused i.v. with vehicle () or NPS 2143 () (0.1 µmol/kg · min) for 2 h. Values are mean ± S.E., n = 3/group. *p < 0.05 versus vehicle-infused controls.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The present study describes the salient pharmacodynamic properties of the first compound shown to possess antagonist activity at the Ca²⁺ receptor. For most G protein-coupled receptors, potent and selective antagonists have typically preceded the discovery of corresponding agonists. The converse has been the case with the Ca²⁺ receptor and a plethora of inorganic ions and organic compounds, acting either as true agonists or allosteric activators of this receptor, have been reported (Nemeth and Fox, 1999). Many of these calcimimetics lack potency and selectivity yet a ligand possessing even such meager properties, but acting as an antagonist, has not been identified. In fact, NPS 2143 is the first substance, either atomic or molecular, shown to block Ca²⁺ receptor activity. We call such blocking ligands "calcilytics" and, together with calcimimetics, they encompass the known pharmacological means of altering Ca²⁺ receptor activity.

While G protein-coupled receptors in general are hardly ideal candidates for receptor-based drug design, the Ca²⁺ receptor is considerably more challenging because Ca²⁺ ions fail to offer much structural information for molecular ligand-based drug design. High-throughput screening of small molecule libraries seemed to offer the best chance of discovering a Ca²⁺ receptor antagonist. Although the initial hit detected was not potent and clearly lacked specificity, it was not pan-active. Importantly, its structure suggested that it might have pharmaceutically acceptable properties. The compounds that emerged from the ensuing medicinal chemistry effort have potencies that span three orders of magnitude and differ markedly in selectivity. NPS 2143 was chosen as representative of the compounds in this family based on its potency and the initial results suggesting some degree of selectivity.

We initially tested for actions that might interfere with the readout of Ca²⁺ receptor activity and/or those that might generally affect the activity of G protein-coupled receptors. Because NPS 2143 did not inhibit responses to a number of different receptors, all of which couple to the mobilization of intracellular Ca²⁺ in HEK 293 cells, NPS 2143 affects neither receptor activity nor the signal transduction pathways leading from these receptors to the increase in [Ca²⁺]_i. Likewise, the failure of NPS 2143 to alter cytoplasmic Ca²⁺ responses to thapsigargin suggests that this compound does not affect ATPase activity or mechanisms comprising capacitive Ca²⁺ influx. In this respect, NPS 2143 differs from La³⁺, which decreases the sustained increases in [Ca²⁺]_i induced by thapsigargin. Thus, NPS 2143 does not block extracellular Ca²⁺ influx under these conditions, as does La³⁺. This, in turn, suggests that the ability of NPS 2143 to rapidly lower [Ca²⁺]_i (Fig. 2, a and c) does not result from inhibition of Ca²⁺ influx. If the sole action of NPS 2143 in these experiments is the inhibition of the Ca²⁺ receptor then the rapid fall in [Ca²⁺]_i might indicate that continuous activation of the Ca²⁺ receptor is required to maintain increased levels of [Ca²⁺]_i.

NPS 2143 did not affect the activity of several G protein-coupled receptors from class C that are structurally similar to Ca²⁺ receptors. NPS 2143 neither activated nor inhibited group I mGluRs, which like the parathyroid Ca²⁺ receptor, couple to phospholipase C and mobilize intracellular Ca²⁺. ¹ To test for actions on the GABA_B receptors, a GABA_BR1 fusion construct was used. Like the native GABA_BRs, coexpression of GABA_BR2 and the R1 fusion construct was required to obtain a response to GABA or baclofen. The requirement for coexpression implies that affecting either receptor would alter the response, and the failure of NPS 2143 to do so suggests that it affects neither the R1 nor the R2 GABA_B receptors. In the aggregate, the results show that NPS 2143 is a compound with reasonable selectivity and, when used appropriately, might be a useful compound to probe the functions of Ca²⁺ receptors in nonhuman animals.²

Despite clear evidence for potent and selective inhibitory actions of NPS 2143 on the Ca²⁺ receptor, it was not a certainty that a calcilytic compound would in fact stimulate PTH secretion. While intuitively appealing, this assumption rests solely on the finding that lowering the level of extracellular Ca²⁺ stimulates PTH secretion. This is not, however, equivalent to blocking the activity of the receptor in a normocalcemic setting. Moreover, our screening assay used measurements of [Ca²⁺]_i as a functional readout of Ca²⁺ receptor activity. In the parathyroid cell, increases in [Ca²⁺]_i are associated with an inhibition of PTH secretion. The mechanisms giving rise to this unusual relationship between [Ca²⁺]_i and hormone secretion are not understood and the possibility remains that the Ca²⁺ receptor couples to an additional or alternative intracellular signal that regulates PTH secretion. Thus, it was not a certainty that compounds detected using measures of [Ca²⁺]_i would also affect PTH secretion. It was therefore rewarding to discover that NPS 2143 stimulated PTH secretion in vitro in the absence of changes in the level of extracellular Ca²⁺ and also in normocalcemic rats.

Curiously, the effects of NPS 2143 on PTH secretion from bovine parathyroid cells mirror those of type II calcimimetic compounds. These calcimimetics are positive allosteric modulators that shift the concentration-response curve of extracellular Ca²⁺ to the left without changing the maximum or minimum responses (Nemeth et al., 1998). NPS 2143 affects the agonist concentration-response curve in a converse manner: there is a shift to the right that is unaccompanied by changes in either the maximal or minimal response. These changes in the agonist concentration-response curves suggest that NPS 2143 decreases, whereas type II calcimimetic compounds increase the sensitivity of the Ca²⁺ receptor to activation by extracellular Ca²⁺. Although the rightward shift of the concentration-response curve could indicate competitive inhibition by NPS 2143, the shifts in the concentration-response curves shown in Fig. 5 would also be produced by a noncompetitive antagonist acting on a tissue with a large receptor reserve. Additional studies are underway to understand further the molecular actions of NPS 2143.

The stimulatory effects of NPS 2143 on PTH secretion in vitro and on circulating levels in vivo underscore the key role of the Ca²⁺ receptor in controlling PTH secretion. These findings are additionally relevant for therapies, which by controlling circulating levels of endogenous PTH, seek to promote new bone growth. In this latter respect, it is instructive to compare the magnitude and rate of changes in plasma levels of PTH elicited by intravenous infusion of NPS 2143 with those elicited by subcutaneous administration of PTH in rats or humans. The 4-fold increase in plasma PTH levels following infusion of NPS 2143 falls within the range of levels produced by doses of exogenous PTH that stimulate new bone formation and increase bone mineral density in osteopenic rats (Fox et al., 1997) and osteoporotic humans (Lindsay et al., 1993, 1997). The increase in circulating levels of PTH is also rapid and indicates that NPS 2143 affects the regulated, rather than the constitutive pathway of PTH secretion. This finding mirrors that obtained with calcimimetics, which inhibit only the secretory responses sensitive to extracellular Ca²⁺ (Nemeth et al., 1998). The decrease in circulating levels of PTH following the end of infusion is likewise comparable with that seen following administration of exogenous hormone. Moreover, the elicited increase in plasma PTH levels induced the appropriate physiological response, i.e., a delayed increase in the plasma levels of Ca²⁺. Thus, NPS 2143 causes a pattern of change in circulating levels of PTH that is quite similar to that seen following doses of PTH that stimulate bone growth. Calcilytic compounds might therefore prove useful as anabolic therapies for bone diseases such as osteoporosis. Indeed, we have shown that NPS 2143 when administered daily by oral gavage to estrogen-treated osteopenic rats for 5 weeks increases plasma PTH levels and bone formation sufficiently to increase bone mineral density and trabecular bone volume (Gowen et al., 2000).