Defining the molecular target for strontium in osteoblasts is important for understanding the anabolic effects of this cation on bone. The current studies demonstrate that a G-protein-mediated response to strontium persists in osteoblasts that lack CASR, suggesting a predominant role for a novel cation-sensing receptor in mediating the osseous response to strontium.
Introduction: Strontium has anabolic effects on bone and is currently being developed for the treatment of osteoporosis. The molecular target for strontium in osteoblasts has not been determined, but the existence of CASR, a G-protein-coupled receptor calcium-sensing receptor, raises the possibility that strontium actions on bone are mediated through this or a related receptor.
Materials and Methods: We used activation of a transfected serum response element (SRE)-luciferase reporter in HEK-293 cells to determine if CASR is activated by strontium. In addition, we examined strontium-mediated responses in MC3T3-E1 osteoblasts and osteoblasts derived from wild-type and CASR null mice to determine if other cation-sensing mechanisms are present in osteoblasts.
Results and Conclusions: We found that strontium stimulated SRE-luc activity in HEK-293 cells transfected with full-length CASR but not in cells expressing the alternatively spliced CASR construct lacking exon 5. In contrast, we found that MC3T3-E1 osteoblasts that lack CASR as well as osteoblasts derived from CASR null mice respond to millimolar concentrations of strontium. The response to strontium in osteoblasts was nonadditive to a panel of extracellular cations, including aluminum, gadolinium, and calcium, suggesting a common mechanism of action. In contrast, neither the CASR agonist magnesium nor the calcimimetic NPS-R568 activated SRE activity in osteoblasts, but the response to these agonists was imparted by transfection of CASR into these osteoblasts, consistent with the presence of distinct cation-sensing mechanisms. Co-expression of the dominant negative Gαq(305–359) minigene also inhibited cation-stimulated SRE activity in osteoblasts lacking known CASR. These findings are consistent with strontium activation of a novel Gαq-coupled extracellular cation-sensing receptor in osteoblasts with distinct cation specificity.
Currently available pharmaceutical treatments for osteoporosis either inhibit bone resorption (e.g., bisphosphonates, estrogen receptor modulators, and calcitonin)(1) or stimulate bone formation (e.g., PTH analogs).(2) A more ideal treatment of osteoporosis might have net anabolic effects on the skeleton by simultaneously stimulating osteoblastic-mediated bone formation and inhibiting osteoclastic-mediated bone resorption.(3) Extracellular calcium-sensing mechanisms in osteoblasts and osteoclasts participate in an endogenous regulatory pathway for maintaining skeletal homeostasis.(4) A model has been proposed in which the release of calcium by osteoclasts inhibits osteoclast-mediated bone resorption and promotes osteoblast-mediated bone formation through potentially redundant extracellular calcium-sensing receptor-like mechanisms.(1) Theoretically, pharmacologic activation of these putative cation-sensing mechanisms in osteoblasts and osteoclasts could achieve the goal of simultaneously stimulating bone formation and inhibiting bone resorption, leading to increases in bone mass in subjects with osteoporosis.
Recent in vitro and in vivo studies indicate that the divalent cation strontium can stimulate osteoblast-mediated bone formation and inhibit osteoclast-mediated bone resorption, leading to increased bone mass.(5–8) Moreover, recent clinical trials show that the administration of strontium ranelate to postmenopausal women increases bone mass, increases biomarkers of bone formation, and reduces vertebral deformities.(9) The trivalent cation aluminum also stimulates osteoblastic proliferation in culture(10–13) and osteoblast-mediated bone formation in dogs.(3,14-16) Collectively, these studies suggest that certain extracellular divalent and trivalent cations can exert net anabolic effects on bone, possibly through common mechanisms of mediated by the sensing of extracellular cations.
The molecular targets in bone cells that mediate the response to strontium and other extracellular cations are uncertain.(17) Some authors have found evidence for expression of the classical calcium-sensing receptor CASR in osteoblasts(18) and osteoclasts(19) as well as other cells in bone marrow.(20) CASR is a G-protein-coupled receptor (GPCR) that responds to changes in extracellular calcium concentrations and belongs to the class C metabotrophic glutamate/pheromone family that is predominately located in the parathyroid gland, kidney, and thyroidal C-cells, where it regulates PTH secretion, renal calcium and water handling, and calcitonin secretion, respectively. Calcium and other polycations such as magnesium and gadolinium interact with the agonist binding sites of CASR in the large extracellular domain,(21) whereas calcimimetic allosteric modulators, such as the phenylalkylamine NPS R-568, bind to distinct sites to activate CASR in the presence of calcium.(22) CASR's actions are mediated through its coupling to Giα- and Gqα-dependent signaling pathways in response to small changes in the concentration of its physiological ligand, ionized calcium. Recent studies indicate that strontium can activate endogenous CASR in parathyroid cells and in heterologous cells transfected with CASR, but with lower efficacy and potencies than calcium.(23)
Whether CASR is expressed in osteoblasts and osteoclasts and mediates the response to extracellular cations in bone, however, is controversial. Recent studies in CASR null mice failed to identify a skeletal phenotype resulting from the loss of CASR in bone and cartilage.(24) Rather, the skeletal features in CASR null mice were caused by elevated parathyroid hormone (PTH) levels resulting from the loss of CASR function in the parathyroid glands.(24,25) To date, no studies have shown a nonredundant and essential role of CASR in regulating bone remodeling.(24) In addition, CASR has not been uniformly identified in osteoblast or osteoclast cell lines,(26–28) and there is evidence for the presence of a novel cation-sensing mechanism in osteoblasts.(10–13) In this regard, primary osteoblasts derived from CASR null mice retain responsiveness to extracellular calcium. The putative osteoblastic cation-sensing receptor also show ligand specificities that distinguish if from CASR (namely activation by aluminum but not magnesium)(10) and responds to calcium levels of 8-40 mM,(29) whereas the known extracellular calcium sensing receptor in the parathyroid gland responds to more physiological calcium concentrations of calcium. There is similar controversy regarding whether CASR accounts for the response of osteoclasts to extracellular calcium,(4,17) where the concentrations of calcium required to inhibit osteoclast activity are typically higher than that required for activation of CASR.(30) The ion selectivity of the osteoclastic cation-sensing mechanisms also differs from CASR and from the putative osteoblastic cation-sensing receptor.(31) The role of these other cation-sensing mechanisms in mediating skeletal effects of strontium has not been investigated.
In this study, we determined whether strontium activates CASR in a heterologous cell system, reaffirmed the presence of a novel calcium sensing mechanism distinct from CASR in osteoblasts, and demonstrated that strontium activates this novel cation-sensing mechanism. The novel osteoblastic cation-sensing receptor sensor may be the important target for strontium in osteoblasts.
MATERIALS AND METHODS
All culture reagents were purchased from Life Technologies (Rockville, MD, USA). Aluminum chloride (AlCl3 · 6 H2O) was obtained from Fisher (Springfield, NJ, USA). Gadolinium chloride hexahydrate was purchased from Aldrich Chemical Co. (Milwaukee, WI, USA). Calcium chloride, magnesium chloride, strontium chloride, strontium acetate, and nifedipine were purchased from Sigma Chemical (St Louis, MO, USA). Bovine serum albumin (BSA; faction V) was obtained from Roche Applied Science (Indianapo-lis, IN, USA). The calcimimetic NPS-R568 N-(3-[2-chlorophenyl]propyl)-(R)-α-methyl-3-methoxybenzylamine and its inactive isomer NPS-S568 were provided by Amgen (Thousand Oaks, CA, USA).
The rat CASR cDNA was obtained from Drs Snowman and Snyder(32) and subcloned in the mammalian expression vector pcDNA 3 (Invitrogen, Carlsbad, CA, USA) as previously described.(11) The pSV.SPORT-rCASR was created by digesting pcDNA3.rCASR DNA with HindIII and XbaI and ligating the released fragment into a modified expression vector pSV.SPORT (Life Technologies). The pRK5.Gαq(305-359) minigene construct that corresponds to the COOH-terminal peptide sequence of Gαq residues 305-359 was kindly provided by Dr Robert J Lefkowitz (Howard Hughes Medical Institute, Duke University Medical Center, Durham, NC, USA).(33) The cDNA of Gαq(305-359) minigene was excised using EcoRI and XbaI from pRK5.Gαq(305-359) and inserted into pluvM to make pLuvM.Gαq(305-359). The SRE-Luc plasmid DNA construct was generated as previously described.(34) The constructs of pcDNA3.1.hCaR and pcDNA3.1.AltCaR, respectively containing the full-length human CASR and its spliced variant lacking exon 5, were a generous gift from Dr Bikle (Department of Medicine, Veterans Affairs Medi-cal Center, San Francisco, CA, USA).(35) pSV.SPORT-calcyclin and antisense calcyclin were generated as previously described.(36)
Human embryonic kidney HEK-293 cells were obtained from American Type Culture Collection (Rockville, MD, USA). HEK-293 cells stably expressing rat CASR were created as previously described.(11) For osteoblasts cell lines, we used the previously characterized MC3T3-E1 osteoblast cell line that lacks CASR(37) and immortalized calvaria-derived osteoblasts derived from wildtype and CASR null mice.(10) Cells were grown in α-MEM supplemented as above in humidified atmosphere of 5% CO2/95% air at 37°C. Osteoblast cell lines were subcultured every 3-5 days using 0.001% pronase.
RT-PCR analysis of CASR transcripts in mouse osteoblast cell lines
RT-PCR was done using the Titan One Tube RT-PCR kit purchased from Boehringer Mannheim (Indianapolis, IN, USA). The RT reaction was performed using 2.0 μg of total RNA derived from either wildtype MC3T3-E1 cells or MC3T3-E1 cells stably transfected with pSV.SPORT.rCASR. Total RNA was treated with DNAse I (Stratagene, La Jolla, CA, USA) and incubated at 45°C for 60 minutes, and the template was denatured at 94°C for 2 minutes. We used primers 640F: cagcgagcccaaaagaa; and 1553R: cttcagaccgaacccaatgg, which recognize both human and mouse CASR. PCR was performed with thermal cycling parameters of 94°C for 30 s, 55°C for 30 s, and 68°C for 45 s for 10 cycles. This was followed by an additional 25 cycles with thermal cycling parameters of 94°C for 30 s, 55°C for 30 s, 68°C for 45 s, and an additional 5 s with each cycle. The reaction was completed with a final extension at 68°C for 7 minutes. Amplification products were resolved by electrophoresis in a 1.5% agarose gel and visualized by ethidium bromide staining.
Plasmids were prepared using the QIAGEN Plasmid Mixi kit (QIAGEN, Valencia, CA, USA). Transient transfections were preformed using 105 cells plated in the 6-well plates and incubated overnight at 37°C. A DNA-liposome complex was prepared by mixing DNA and TransFast transfection reagent (1:2 DNA:TransFast transfection reagent; Promega, Madison, WI, USA) in OPTI-MEM I reduced serum medium (Life Technologies) and incubating the mixture at room temperature for 15 minutes. The mixtures of SRE-Luc, plasmid DNA, and TransFast transfection reagent were diluted with OPTI-MEM I reduced serum medium and added to HBSS rinsed cells (final concentration: 5 μg SRE-Luc plasmid DNA/well). After 1 h of incubation at 37°C, 2 ml α-MEM containing 10% FBS and 1% P/S were added to the medium overlying the transfected cells and incubated for an additional 2 days before treatment.
Assessment of agonist-stimulated SRE activity
Quiescence of transfected cells was achieved in subconfluent cultures by incubation for 24 h in serum-free DMEM/F12 containing 0.1% BSA. Quiescent cells were treated with vehicle or stimulated for 8 h with various agonists (Al3+ [25 μM], Ca2+ [5 mM], Gd3+ [60 μM], Mg2+ [10 mM], Sr2+ [1-50 mM], and NPS-R568 and NPS-S568 [1 μM]) as indicated. Luciferase activity was assessed in cell extracts using the luciferase assay system (Promega) and a BG-luminometer (Gem Biomedical, Hamden, CT, USA).
We evaluated differences between groups by one-way ANOVA. All computations were performed using the Statgraphics Plus Version 5.0 software system (STSC, Rockville, MD, USA).
NPS-568 fails to stimulate c-fos promoter transcription in MC3T3-E1 osteoblasts: further evidence for a novel osteoblastic calcium-sensing mechanism
A characteristic of extracellular cation-sensing response in osteoblasts is the rapid induction of the c-fos promoter transcription by specific cations.(12,15) We found that Gd3+ (60 μM), Al3+ (25 μM), Ca2+ (5 mM), but not magnesium (10 mM), stimulated luciferase activity in MC3T3-E1 osteoblasts compared with unstimulated controls (Fig. 1A), confirming the ligand specificity of the response in these osteoblasts is distinct from CASR, for which magnesium, but not aluminum, is a ligand.(10,12) Next, we evaluated whether the R isomer of N-(3-[2-chlorophenyl]propyl)-(R)-α-methyl-3-methoxybenzylamine (NPS R-568), a specific allosteric modulator of CASR,(22) is capable of stimulating SRE-luciferase activity in MC3T3-E1 osteoblasts. Application of NPS R-568 at doses ranging from 0.5 to 100 μM, which we previously have shown activates SRE-luciferase activity in HEK293 cells transfected with CASR,(38) failed to stimulate luciferase activity in MC3T3-E1 osteoblasts (Figs. 1A and 1B). In addition, the inactive stereo-isomer, NPS S-568, had no effect on SRE-luciferase activity in MC3T3-E1 osteoblasts (Fig. 1B).
To eliminate the possibility that MC3T3-E1 osteoblasts are lacking downstream signaling pathways necessary for detecting a response to CASR, we re-evaluated the response of MC3T3-E1 osteoblasts to NPS R-568 and cations after transfection of these cells with a full-length rat CASR cDNA construct (Fig. 2). Consistent with our prior results, we failed to detect endogenous CASR transcripts in nontransfected MC3T3-E1 osteoblasts, but detected the correct size product in MC3T3-E1 cells transfected with rat CASR (Fig. 2A). MC3T3-E1 osteoblasts overexpressing rat CASR gained responsiveness to NPS R-568 and magnesium, consistent with the availability of the necessary signaling pathways when CASR is present in these cells (Fig. 2B). NPS R-568 at doses of 1 μM maximally activated CASR in media containing 1 mM calcium, whereas the inactive isomer NPS S-568 had no effect. The response to aluminum, calcium, and gadolinium was unchanged by the addition of CASR to MC3T3-E1 osteoblasts (Fig. 2C), suggesting that CASR and the putative osteoblast cation-sensing response may share common pathways leading to SRE activation (see below).
MC3T3-E1 cells lacking CASR and osteoblasts derived from CASR null mice respond to strontium
Next, we examined if strontium stimulates SRE-luciferase activity in MC3T3-E1 cells (Fig. 3). In MC3T3-E1 osteoblasts, both strontium chloride (Fig. 3A) and strontium acetate (Fig. 3B) at millimolar concentrations stimulated SRE-luciferase activity. Strontium displayed nonadditive effects with aluminum, calcium, and gadolinium on SRE luciferase activity in these cells, suggesting shared mechanisms of action (Fig. 3C). We have previously shown that osteoblasts from CASR null mice retains their responsiveness to a panel of extracellular cations, including calcium, aluminum, and gadolinium, but not magnesium.(10) We also found that strontium stimulates SRE luciferase activity in osteoblasts derived from wildtype mice (Fig. 4A) and that this response is unaltered in osteoblasts derived from CASR null mice (Fig. 4B).
Strontium stimulates CASR in HEK-293 cells
HEK-293 cells transfected with the full-length human CASR also responded to strontium (Fig. 5), but the response was biphasic, achieving a peak response at 10 mM and no stimulation at 20 mM. In contrast, HEK293 cells transfected with vector alone failed to respond to strontium at concentrations up to 20 mM (Fig. 5). We also evaluated the response in HEK293 cells of the alternatively spliced CASR variant lacking exon 5 into HEK-293 cells (Fig. 5). We failed to observe strontium-induced SRE-luciferase activity in HEK-293 overexpressing the alternatively spliced CASR (Fig. 5). Collectively, these findings indicate that CASR, but not the alternatively spliced form of CASR that has been detected in some tissues,(35) respond to strontium.
Osteoblast cation-sensing response involves a G-protein-coupled mechanism
To determine if the response in osteoblasts is mediated through G-protein-coupled mechanisms, we attempted inhibition of Gαq-mediated signaling in MC3T3-E1 osteoblasts with a dominant negative Gαq minigene, Gαq(305-359) (Fig. 6A). This minigene consists of the COOH-terminal peptide residues 305-359 and has previously been shown to uncouple Gαq-coupled receptors without affecting coupling to other classes of Gα subunits.(33) Overexpression of Gαq(305-359) in MC3T3-E1 osteoblasts inhibited gadolinium-, aluminum-, and strontium-mediated activation of SRE-luciferase (Fig. 6A). Because c-fos mRNA expression can be by the voltage-sensitive Ca2+ channels in other systems,(39) we tested the ability of the calcium channel blocker nifedipine to inhibit cation-stimulated SRE activity in osteoblasts. Treatment of MC3T3-E1 osteoblasts with 20 μM nifedipine failed to inhibit response to aluminum, calcium, gadolinium, and strontium (Fig. 6B).
Calcyclin is not involved in strontium-stimulated SRE signaling
We recently isolated calcyclin from a MC3T3-E1 osteoblast library using a functional expression cloning strategy and demonstrated that calcyclin transfected into HEK-293 cells imparted sensing to extracellular Ca2+, Mg2+, Al3+, and Gd3+. To evaluate whether calcyclin also mediates the effects of strontium, we overexpressed calcyclin into HEK-293 cells. Similar to previous reports,(36) HEK-293 cells transfected with calcyclin gained responsiveness to calcium and gadolinium, but not to strontium chloride (Fig. 7). In addition, antisense calcylcin that was stably transfected into MC3T3-E1 osteoblasts failed to inhibit strontium-stimulated SRE activity (data not shown).
The molecular mechanisms through which calcium and other cations exert actions on bone cells are controversial. Some studies attribute the response to CASR,(23) whereas others indicate the presence of a novel cation-sensing receptor mechanism distinct from CASR.(4,10,17,24,26) Our findings suggest that strontium can act on osteoblasts in vitro through the novel cation-sensing mechanism that is distinct from CASR. In this regard, both MC3T3-E1 cells, which do not express CASR transcripts but express genes that characterize them as osteoblasts (Fig. 2A),(26) as well as osteoblasts derived from CASR-null mice,(10) display a dose-dependent response to strontium (Figs. 3 and 4B). In addition, strontium responses in osteoblasts derived from CASR-null mice were similar to that observed in wildtype osteoblasts (Fig. 4A). The dose of strontium required to activate osteoblasts is in the millimolar range, similar to that required for calcium.(10)
Because certain tissues from CASR null mice can express a spliced variant of CASR that lacks exon 5,(35,40) concerns have been raised that the response observed in osteoblasts derived from CASR-null mice might be caused by this alternatively spliced form of CASR.(40) However, additional studies indicate the alternatively spliced CASR construct does not respond to strontium (Fig. 5), similar to other reports showing a lack of function of this alternatively spliced receptor.(35,40) Thus, the alternatively splice CASR receptor, even if it could be documented to be expressed in osteoblasts,(10) does not account for the response to strontium and other cations observed in CASR null osteoblasts. Nor does the lack of essential mediators of CASR responsiveness explain the lack of response of osteoblasts to CASR agonists, magnesium and calcimimetic NPS R-568, because transfection of CASR into MC3T3-E1 osteoblasts results in gain-of-function to these ligands (Fig. 2). These findings support the lack of the classical CASR in these cultured osteoblasts.
Previous studies also have shown that strontium stimulates proliferation and collagen synthesis in primary rat osteoblast cultures and bone formation in rat calvaria organ cultures.(5) In addition, both strontium chloride and strontium ranelate also stimulate bone formation in vivo.(7,23) Strontium ranelate is being developed for the treatment of osteoporosis.(9) We found that both strontium chloride and acetate stimulated osteoblasts in vitro (Figs. 3A and 3B). Aluminum, which also purportedly activates the novel osteoblast cation-sensing mechanism but not CASR, also has been shown to stimulate de novo bone formation and inhibit resorption in some animal models.(3,14) The nonadditive effects of these agents (Fig. 3C) indicate that aluminum and strontium might be acting through the same signal transduction pathways in osteoblasts involving a novel cation-sensing receptor-like mechanism.
The current studies also provide additional insights into characteristics of this putative osteoblastic cation-sensing mechanism. In this regard, we confirmed that the osteoblast receptor has different cation specificity compared with CASR, being activated by aluminum but not magnesium (Fig. 1A). In addition, we extended these observations by showing that NPS R-568, an allosteric modulator of CASR, fails to activate the osteoblastic cation-sensing mechanism linked to SRE activation (Figs. 1A and 1B). Consistent with prior studies implicating a G-protein-coupled cation-sensing receptor-like mechanism,(13) we show that the signal transduction pathway linking the response to SRE requires activation of Gαq (Fig. 6A). Indeed, cation-mediated activation of SRE in MC3T3-E1 osteoblasts is blocked by overexpression of a carboxy-terminal peptide of the α subunit Gαq (Fig. 6A). The osteoblastic cation-sensing response that is responsible for SRE activation does not seem to be mediated by an L-type calcium channel, because nifedipine failed to block the response (Fig. 6B).
The activation of a novel cation-sensing mechanism by strontium in osteoblasts, however, does not preclude additional actions of strontium to activate CASR at sites where this classical receptor is known to be expressed. Indeed, we show that transfection of CASR into HEK-293 cells results in these cells acquiring responsiveness to strontium (Fig. 5), findings consistent with prior observations that strontium inhibits cyclic adenosine monophosphate (cAMP) accumulation and stimulates inositol phosphate synthesis in dispersed bovine parathyroid cells expressing endogenous CASR.(42) Recently, we developed a mouse model to study the function of CASR outside of the parathyroid glands by backcrossing CASR-deficient mice onto the Gcm2-null mice molecularly ablate parathyroid glands in CASR-deficient mice.(24) To date, no specific bone function for CASR has been demonstrated in vivo.(24) Future studies that examine the anabolic effects of strontium in wildtype, Gcm2 null, and double CASR- and Gcm2-deficient mice will provide important information regarding the differential roles of CASR and the novel osteoblastic cation-sensing mechanisms in mediating the response to strontium in vivo. Our current studies with CASR-deficient osteoblasts, however, predict that the bone anabolic response to strontium may be preserved in the absence of CASR.
Identification of the molecular target in bone for strontium will require additional studies. The data presented here support the existence of another cation-sensing G-protein-coupled receptor in osteoblasts. Recently, expression cloning efforts have identified other cation-sensing pathways in osteoblasts involving calcyclin,(36) raising the possibility of the presence of nonreceptor cation-sensing mechanisms in osteoblasts that could mediate the anabolic responses to extracellular calcium. However, calcyclin does not play an important role in mediating strontium effects in osteoblasts (Fig. 7). The fact that calcyclin also mediates the response to magnesium, an agonist that does not activate the putative osteoblast calcium receptor, suggests that there are other molecular targets, likely G-protein-coupled receptors, in osteoblasts mediating the response to strontium. In support of the possibility of multiple cation-sensing receptors, the ovarian cancer G-protein-coupled receptor 1 (OGR1) has recently been shown to act as a proton-sensing receptor in osteoblasts.(43)
Regardless, functional responses to cations that have been reported in osteoblasts cannot be used as evidence that CASR is solely mediating this response. Given the diversity of G-protein-coupled receptors, it is likely that several other cation-sensing receptors will be discovered, one of which may be the cation-sensing receptor that mediates the response to strontium and aluminum in osteoblasts. Whatever the mechanisms, the effects of calcium may be critically important in regulating the local bone remodeling process, and this might be exploited to develop pharmacologic agents that mimic the effects of calcium on bone.
This work was supported by National Institutes of Arthritis and Musculoskeletal and Skin Disorders Grant RO1-AR73708. The authors thank Cristy McGranahan for secretarial assistance with the preparation of this manuscript.