Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Osteoporosis is a chronic bone disease characterized by a decrease in bone mass resulting from accelerated osteoclast-mediated bone resorption relative to formation (Baron, 1996). Therefore, inhibition of osteoclast-driven bone resorption should prove beneficial in the treatment of this disease. For resorption to occur, osteoclasts must first attach to the bone. They then form a tightly sealed extracellular compartment beneath the cell (Baron, 1996) into which they secrete protons and matrix-degrading proteinases. Tissue distribution studies (Clover et al., 1992; Helfrich et al., 1992; Nesbitt et al., 1993; Shinar et al., 1993) have shown that _v3 is the predominate integrin on the osteoclast cell surface. Neutralizing antibodies directed against this receptor (Davies et al., 1989; Horton et al., 1991; Crippes et al., 1996) as well as Arg-Gly-Asp (RGD)-containing peptides (Davies et al., 1989; Fisher et al., 1993), inhibit osteoclast adhesion. In addition, osteoclast-mediated bone resorption in vitro as well as in the acute thyroparathyroidectomized (TPTX) rat model of bone resorption also is inhibited by several of these molecules (Fisher et al., 1993; Yamamoto et al., 1993; Crippes et al., 1996). Together, these data indicate that _v3 is the functionally important integrin on the osteoclast surface.

Recently, a nonorally bioavailable, RGD-peptide mimetic, SC56631, which binds to _v3, _v5, and _IIb3, was reported to inhibit osteoclast-mediated bone resorption in vitro as well as in vivo in the TPTX rat model (Engleman et al., 1997). In addition, SC56631 also appeared to prevent the estrogen deficiency-induced loss in trabecular structure in the ovariectomized (ovx) rat model after 6 weeks of continuous, high-dose i.v. infusion. However, these studies still leave open the question as to whether an _v3 antagonist could be developed for oral administration in a reasonable dosing regime. This is highly desirable for a compound to treat chronic bone loss in osteoporosis.

In this article, we report the pharmacological characterization of SB 265123, a potent, orally bioavailable, nonpeptide _v3 antagonist that inhibits _v3-mediated cell adhesion as well as osteoclast-mediated bone resorption in vitro. SB 265123 is efficacious at inhibiting acute bone resorption in vivo in the TPTX rat model. When dosed twice a day at 30 mg/kg orally, SB 265123 also is efficacious at preventing ovariectomy-induced bone loss. These data support the hypothesis that a nonpeptide _v3 antagonist can be developed for pharmacologically acceptable dosing for the treatment of postmenopausal osteoporosis.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Integrin-Binding Assays. To determine the affinity of compounds for various integrins, binding assays were established as described previously (Wong et al., 1996). Integrins _v3, _v5, and ₅₁ were purified from human placenta and _IIb3 was purified from human platelets (Wong et al., 1996). Receptor-binding assays were established with the RGD-containing cyclic peptide [³H]SK&F 107260 as the ligand. The K_i values represent the means of values determined in two or three separate experiments.

Inhibition of _v3-Mediated Cell Adhesion. Inhibition of _v3-mediated cell adhesion was monitored with human embryonic kidney (HEK)-293 cells cotransfected with recombinant human _v and ₃ (Kumar et al., 1997). Ninety-six-well plates were coated overnight with 0.15 µg of human vitronectin in 0.1 ml of RPMI 1640 medium. The plates were then washed once with RPMI 1640 medium and blocked with 3.5% BSA for 1 h at room temperature. Transfected cells were suspended in RPMI 1640 supplemented with 20 mM HEPES (pH 7.4), 0.1 M MnCl₂, and 0.1% BSA at a density of 0.5 × 10⁶ cells/ml. A 0.1-ml aliquot of cells was added to each well and incubated for 1 h at 37°C in the presence or absence of inhibitor. After the incubation, 0.025 ml of 10% formaldehyde solution (pH 7.4) was added, and the cells were fixed at room temperature for 10 min. The plates were then washed three times with RPMI 1640, and the adherent cells were stained with 0.1 ml of 0.5% toluidine blue for 20 min at room temperature. Excess stain was removed by extensive washing with deionized water and the cell-associated toluidine blue was eluted by the addition of 0.1 ml of 50% ethanol containing 50 mM HCl. Toluidine blue, as a readout of cell number (Wong et al., 1996), was quantified at an optical density of 600 nm on a microtiter plate reader (Tulereck Multiskan, Sterling, VA). The potency of the compounds was evaluated with multidose titrations to determine IC₅₀ values.

Human Platelet Aggregation Assay. Effects of the _v3 antagonist on human platelet aggregation were determined as described previously (Nichols et al., 1994). Briefly, human platelet-rich plasma was stimulated with 10 µM ADP and aggregation monitored with impedance aggregometry in the presence of various concentrations of compound.

Pharmacokinetics in Rat. All animal procedures reported in this study were reviewed and approved by the Animal Care and Use Committee at SmithKline Beecham Pharmaceuticals. Compound pharmacokinetics were evaluated in adult male rats with a crossover design on two separate study days. The animals had femoral vein catheters surgically implanted for infusion of compound. On day 1, the animals received 2 µmol/kg target dose as a 30-min i.v. infusion in 5% polyethylene glycol 300 containing 0.5% dimethyl sulfoxide, pH = 3.5 to 4.0. On day 2, the animals received 2 µmol/kg target dose by oral gavage in 5% polyethylene glycol 300 containing 0.5% dimethyl sulfoxide, pH = 4.0. Blood samples were collected from a lateral tail vein. The concentrations of compound in the plasma were quantified by liquid chromatography tandem mass spectrometry (limit of detection = 10 ng/ml). Noncompartmental methods were used for pharmacokinetic analysis of plasma concentrations versus time data.

Human Osteoclast-Mediated Bone Resorption Assay. The isolation of disaggregated human osteoclasts from fresh osteoclastoma tissue (James et al., 1996) and the in vitro human osteoclast resorption assay have been described (James et al., 1999). Briefly, human osteoclasts were seeded onto bovine cortical bone slices with compound or vehicle for 48 h at 37°C. The culture media were then removed and the levels of the carboxyl-terminal peptide of the 1 chain of human type I collagen were quantified as a biochemical readout of resorption, with a competitive binding enzyme-linked immunosorbant assay (Foged et al., 1996) (Osteometer A/S, Rodovre, Denmark). The results are expressed as percentage of inhibition of resorption compared with supernatants derived from osteoclasts cultured in vehicle without inhibitor. IC₅₀ values are determined from the resultant dose-response curves.

TPTX Rat Model of Bone Resorption. TPTX male Sprague-Dawley rats were received from the vendor (Taconic Farms, Inc., Germantown, NY) and blood-ionized calcium levels were quantitated to confirm the surgery (Chiron Diagnostics Ca²⁺/pH Analyzer model 634). All animals received deionized water and normal chow ad libitum and were given a maintenance injection of thyroid hormone (2 µg/animal s.c.), three times weekly. Rats, anesthetized with isofluorane, were surgically implanted with a femoral i.v. polyurethane-anchored catheter and a femoral intra-arterial polyethylene catheter. Both catheters were exteriorized via the nape of the neck. Animals were individually housed in standard wire rack cages and tubing was protected with a commercially available tether/swivel/jacket system.

Ovariectomized Rat Model. Virgin female Sprague-Dawley rats (Charles River Breeding Laboratories, Inc., Wilmington, MA) were used at the age of 7 months following an acclimation period of at least l month. Immediately before either sham operation or ovariectomy, proximal tibial bone mineral density (BMD) was determined. BMD was determined by dual energy X-ray absorptiometry with a Hologic QDR-4500 (Hologic Inc., Waltham, MA) equipped with high-resolution scanning software. In regional high-resolution mode, the spacing and point resolution were each 0.31 mm. Animals were maintained anesthetized with isofluorane while placed prone on the scan surface. BMD was calculated by dividing the bone mineral content by the projected bone area. The rats were then segregated into groups (n = 10) that did not differ in their mean values for proximal tibial BMD. After surgery, groups of ovx rats received, by oral gavage, a twice-daily dose of either dosing vehicle [1% aqueous solution (w/v) of carboxymethyl cellulose], or compound suspended in vehicle. A group of sham-operated rats was dosed with vehicle as a control. Proximal tibial BMDs were determined 4 weeks after the initiation of the study and analyzed as percentage of change from baseline. The differences between the four ovx-treated groups were analyzed with ANOVA (Milliken and Johnson, 1984) followed by Dunnett's test (Dunnett, 1965). In addition, the dose-related trend in the percentage of change was assessed with a trend test. This was done by with pairwise one-sided t tests on the mean percentage of change values for the four groups. To adjust for multiple testing, bootstrap P values (Westfall and Young, 1993) are reported.

Compounds. SB 223245 (Keenan et al., 1997) and SB 265123 (Miller et al., 1999) were synthesized in the Department of Medicinal Chemistry, SmithKline Beecham Pharmaceuticals, King of Prussia, PA.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Identification of _v3 Antagonists for In Vivo Evaluation. Previous studies led to the identification of SB 223245, a potent, nonpeptide _v3 antagonist (Keenan et al., 1997). Unfortunately, SB 223245 has low oral bioavailabliity (<10%) and a short circulating half-life (9-16 min); therefore, it was inappropriate for in vivo evaluation. Subsequent lead optimization studies led to the identification of SB 265123 (Miller et al., 1999) (Table 1), a potent _v3 antagonist with improved pharmacokinetics. SB 265123 maintained its ability to potently bind (K_i = 4 nM) _v3 (Fig. 1A) and improved its potency (IC₅₀ = 60 nM versus 145 nM for SB 223245) at inhibiting _v3-mediated cell adhesion (Fig. 1B). Furthermore, SB 265123 maintained its selectivity relative to other RGD-binding integrins. SB 265123 binds _v3 with a K_i = 4 nM, but is >1000-fold less potent at binding both _IIb3 (K_i = 9 µM) and ₅₁ (K_i = 18 µM). Consistent with its poor activity at _IIb3, SB 265123 has an IC₅₀ >200 µM at inhibiting human platelet aggregation. Neither SB 223245 nor SB 265123 is completely selective for _v3 because they also bind to the closely related _v integrin _v5 with high affinity (K_i values = 0.4 and 1.3 nM, respectively).

View this table: [in this window] [in a new window]   TABLE 1 In vitro activity of nonpeptide v3 antagonists

View larger version (17K): [in this window] [in a new window]   Fig. 1.   SB 223245 and SB 265123 bind V3 and inhibit V3-mediated cell adhesion. A, SB 223245 () and SB 265123 () were evaluated for binding to V3 in a binding assay with [3H]SK&F 107260 as the radioligand. One hundred percent binding of the ligand gave a signal of 4589 ± 156 cpm, with a background of 310 ± 33 cpm. B, SB 223245 () and SB 265123 () were evaluated for their ability to inhibit the attachment of HEK-293 cells, transfected with V3, to vitronectin-coated wells. In this assay, 100% binding of the cells gave a signal of 0.942 ± 0.095 A405 with a background of 0.087 ± 0.013 A405.

SB 265123 Inhibits Human Osteoclast-Mediated Bone Resorption In Vitro. To directly address whether SB 265123 would be effective at inhibiting osteoclast-mediated bone resorption, an in vitro assay was used in which human osteoclast-mediated bone resorption was monitored (James et al., 1999). Bone resorption was quantified biochemically with C-telopeptide fragments of type I collagen released into the culture medium as a readout. Previous studies have shown that there is a strong correlation between both osteoclast pit number (Foged et al., 1996) and pit volume relative to this C-telopeptide biochemical readout. In this assay, both SB 223245 and SB 265123 inhibit bone resorption in a concentration-dependent fashion (Fig. 2). However, SB 265123 (IC₅₀ = 48 ± 3 nM) is significantly more potent that SB 223245 (IC₅₀ = 300 ± 3 nM). Both compounds can completely inhibit bone resorption in vitro at concentrations >0.5 µM. These data indicate that both _v3 antagonists can inhibit human osteoclast-mediated bone resorption in the same concentration range that they inhibit _v3-mediated cell adhesion; however, consistent with its potency in the cell adhesion assay, SB 265123 is more potent in the resorption assay.

View larger version (15K): [in this window] [in a new window]   Fig. 2.   SB 223245 and SB 265123 inhibit human osteoclast-mediated bone resorption. The effects of SB 223245 () and SB 265123 () on human osteoclast-mediated resorption of bovine cortical bone were evaluated with the C-telopeptide of type I collagen (CTx) biochemical readout. Data are reported as percentage of inhibition (n = 3 independent assays) of the maximal CTx signal in each assay. Because the assay uses primary human osteoclasts and fresh bovine bone as a substrate, there is interassay variability in the maximal resorption signal. The 100% resorption (no compound controls) ranged from 16 to 100 nM CTx.

SB 265123 Inhibits Bone Resorption in the TPTX Rat Model. Because of its high affinity for _v3, its favorable pharmacokinetic characteristics in the rat, and potency in the in vitro bone resorption assays, SB 265123 was evaluated in vivo in the TPTX rat model of bone resorption. In this model, TPTX rats are rendered hypocalcemic and infused with PTH to stimulate an osteoclast-mediated calcemic response. These data indicated that the increased serum calcium had no effect on kidney PTH receptors so SB 265123 was coinfused i.v. at a rate of 2.53 mg/kg/h with PTH and the effect on serum calcium was measured. Under these conditions, SB 265123 inhibits the PTH-induced calcemic response by >80% at all time points (Fig. 3). At the end of the experiment (6 h), 85% inhibition was observed. These data clearly indicate that SB 265123 is active in vivo in an acute model of PTH-stimulated bone resorption.

View larger version (17K): [in this window] [in a new window]   Fig. 3.   SB 265123 inhibits bone resorption in the acute TPTX rat model in vivo. SB 265123 was coinfused i.v. with PTH () and compared with infusion of PTH alone () or the vehicle used to suspend SB 265123 (). Seven or eight animals per group were used for this study. The data are reported as a percentage of increase in serum-ionized calcium relative to time zero. Mean baseline-ionized calcium was 0.72 ± 0.03 mM. SB 265123 significantly inhibited the calcemic response at 2, 4, and 6 h after the initiation of the infusion (p < .001).

SB 265123 Inhibits Bone Resorption in the ovx Rat. The ovx rat provides an excellent model for the study of estrogen-deficiency induced osteopenia. In this model, aged female rats are surgically ovx and within 4 weeks a significant reduction in BMD is apparent (Fig. 4). To determine whether SB 265123 prevents this ovx-induced bone loss, animals were dosed from the time of surgery and BMD was measured 4 weeks after surgery in the proximal tibia. Because SB 265123 has high oral bioavailability and has a half-life of 3 to 6 h in the rat, compound was administered at 3, 10, and 30 mg/kg b.i.d. orally for the duration of the study. Under these conditions, SB 265123 significantly inhibited ovariectomy-induced bone loss in a dose-dependent fashion. ANOVA showed a significant difference between the four groups (p = .0196) with the SB 265123 30-mg/kg dose group being significantly different (p = .0171) from the ovx control group based on Dunnett's test. In addition, the dose-related trend seen in the mean values was confirmed by the trend test (p = .0027). No overt signs of toxicity were noted in the animals, even in the high-dose group (30 mg/kg). These results suggest that an orally bioavailable _v3 antagonist can safely and effectively prevent estrogen deficiency-induced bone loss in vivo when administered at a pharmacologically acceptable dose and route of administration.

View larger version (26K): [in this window] [in a new window]   Fig. 4.   SB 265123 inhibits ovariectomy-induced bone loss at 4 weeks in the ovx rat model in vivo. SB 265123 was dosed at 3, 10, or 30 mg/kg b.i.d. orally for 4 weeks from the time of initiation of ovariectomy. BMD was measured in the proximal tibia and is reported as percentage of change relative to time zero for each group. The BMD in compound-treated animals was compared with that quantified for sham-operated animals (Sham) and ovx animals dosed with only vehicle (ovx). Mean BMD (g/cm2) at time zero for each group was 0.2573 ± 0.006 for the sham group; 0.2532 ± 0.005 for the ovx group; 0.2538 ± 0.006 for the 3 mg/kg SB 265123 group; 0.2609 ± 0.005 for the 10 mg/kg SB 265123 group; and 2544 ± 0.004 for the 30 mg/kg SB 265123 group. Ten animals per group were used in this study.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Normal bone balance is controlled through a tightly integrated balance of osteoclast-mediated bone resorption and osteoblast-mediated bone formation. In an estrogen-depleted setting, such as is seen in postmenopausal osteoporosis, there appears to be an imbalance in this process resulting from a relative deficit in bone formation along with an increase in the rate of bone turnover. The resulting bone loss causes a significant reduction in bone strength, ultimately leading to an increased incidence of fracture. Several therapeutic approaches have been taken to try to regulate this imbalance, including estrogens and bisphosphonates. Both of these approaches have limitations, and the precise molecular mechanisms of action for both estrogen and bisphosphonates are unknown. Identification of specific molecular targets for pharmacological intervention provides opportunities for design of therapeutic modalities with improved safety and/or efficacy.

In this study, we report that the potent _v3 antagonist SB 265123 is efficacious at inhibiting bone resorption both in vitro and in vivo in several bone resorption models. This compound inhibits both matrix degradation (type I collagen readout in the human osteoclast resorption assay) as well as demineralization (TPTX rat model). These results are consistent with the mechanism of action of SB 265123. Inhibition of the attachment and adhesion of osteoclasts, or osteoclast precursors, to the bone would be expected to prevent the formation of an acidic resorption lacuna in which mineral dissolution and matrix degradation occur. This significant effect on osteoclast function appears to translate into a biologically relevant response in vivo as measured with BMD.

It has recently been reported (McHugh et al., 1998) that knockout of _V in mice results in elimination of _V3. Osteoclasts isolated from the _V3-deficient mice have matrix attachment defects and have only 15% of the bone-resorbing capabilities of cells from the wild-type littermates. It is currently unclear what effects on bone resorption in a high-turnover state the knockout will have; however, elimination of _V3 clearly compromises osteoclastic activity and supports a role for therapeutic intervention with an _V3 antagonist.

Previously, the small RGD-peptide mimetic SC56631 (Engleman et al., 1997) was reported to inhibit bone resorption when dosed by continuous i.v. infusion at a dose of 0.5 mg/kg/min (720 mg/kg/day). Both the route of administration (continuous i.v. infusion) and the amount of compound required for efficacy are unrealistic for treatment of postmenopausal osteoporosis. Although these initial studies were encouraging, the question remained as to whether an _v3 antagonist could realistically be developed to control bone resorption in this disease. SB 265123 demonstrates that such a goal may be feasible.

SB 265123 is significantly more selective than SC56631 in that it binds weakly to the related RGD-binding integrin _IIb3 and minimally inhibits human platelet aggregation. Thus, extended chronic dosing with a compound like SB 265123 is not expected to result in complications due to inhibition of platelet aggregation. Consistent with its selectivity, no overt signs of toxicity were noted in the 4-week ovx rat study, even when the compound was dosed at 30 mg/kg twice a day. Interestingly, SB 265123 does bind _v5 with high affinity, similar to SC56631. Although the biological role of _v5 is not clear at this time, studies have shown that _v5 is expressed on osteoclast precursors (Inoue et al., 1995; Teitelbaum et al. 1997) and is subsequently replaced by _v3 as the cells mature into functionally active osteoclasts. In addition, studies suggest that _v5 may mediate the attachment of these osteoclast precursors to matrix (Inoue et al., 1995; Teitelbaum et al., 1997). Therefore, in principle, inhibition of both _v3 and _v5 could have an even more profound effect at blocking bone resorption in vivo than inhibition of only one of these receptors.

Recently, the potent ₃ ligand echistatin also was reported to have efficacy in both the ovx rat and mouse models when dosed by continuous i.v. infusion (Yamamoto et al., 1997). Like SB 265123, echistatin inhibits bone resorption in both the TPTX (Fisher et al., 1993) and ovx rat models in vivo (Yamamoto et al., 1997). In the ovx rat, echistatin-inhibited resorption as measured by BMD, was ~26 and 37% in the femur. In our study, we demonstrated ~65% prevention of bone loss with SB 265123 in the proximal tibia. Thus, SB 265123 appears as active as the extremely potent peptide inhibitor echistatin in this aggressive model of bone resorption.

Net bone loss is the result of limited bone formation as well as accelerated bone resorption. Thus, the effects of antiresorptive compounds such as SB 265123 could possibly be improved if administered in combination with an anabolic molecule, such as intermittent PTH (Reeve, 1996), which has been shown to stimulate bone formation in vivo.

In conclusion, this is the first demonstration of an orally bioavailable _v3 antagonist that inhibits bone resorption both in vitro and in vivo. This molecule inhibits ovariectomy-induced bone resorption in a dose-dependent fashion as assessed with BMD. Such a molecule could be an efficacious and safe treatment for postmenopausal osteoporosis, which appears to be driven through increased osteoclastic activity resulting from estrogen deficiency.