Journal of Bone and Mineral Metabolism

, Volume 27, Issue 5, pp 555–561

Nicotine modulates bone metabolism-associated gene expression in osteoblast cells

Authors

    • Department of Orthopaedic Surgery ARambam Health Care Campus
    • The Bruce & Ruth Rappaport Faculty of Medicine, Technion-Israel Institute of Technology
  • Lilah Rothem
    • Gastroenterology and Nutrition Research LaboratoryRambam Health Care Campus
  • Michael Soudry
    • Department of Orthopaedic Surgery ARambam Health Care Campus
    • The Bruce & Ruth Rappaport Faculty of Medicine, Technion-Israel Institute of Technology
  • Aviva Dahan
    • Gastroenterology and Nutrition Research LaboratoryRambam Health Care Campus
  • Rami Eliakim
    • Gastroenterology and Nutrition Research LaboratoryRambam Health Care Campus
    • The Bruce & Ruth Rappaport Faculty of Medicine, Technion-Israel Institute of Technology
Original Article

DOI: 10.1007/s00774-009-0075-5

Cite this article as:
Rothem, D.E., Rothem, L., Soudry, M. et al. J Bone Miner Metab (2009) 27: 555. doi:10.1007/s00774-009-0075-5

Abstract

Smoking has a broad range of physiological effects, such as being a risk factor in osteoporosis, bone fracture incidence, and increased nonunion rates. Recent studies showed that nicotine has effects at the cellular level in human osteoblast cells. To identify possible mechanisms underlying nicotine-induced changes in osteogenic metabolism, we defined changes in proliferation and osteocalcin, type I collagen, and alkaline phosphatase gene expression after treating human osteosarcoma cells (MG63), with various concentration of nicotine. Nicotine affects cell proliferation in a biphasic manner, including toxic and antiproliferative effects at high levels of nicotine and stimulatory effects at low levels. Moreover, low levels of nicotine upregulated osteocalcin, type I collagen, and alkaline phosphatase gene expression. The increased cell proliferation and gene upregulation induced by nicotine were inhibited by addition of the nicotinic receptor antagonist d-tubocurarine. High nicotine concentrations downregulated the investigated genes. Our results demonstrate, for the first time, that the addition of nicotine concentrations analogous to those acquired by a light to moderate smoker yields increased osteoblast proliferation and bone metabolism, whereas the addition of nicotine concentrations analogous to heavy smokers leads to the opposite effect. The inhibition of these effects by d-tubocurarine suggests that nicotine acts via the nicotinic acetylcholine receptor (nAChR).

Keywords

Bone metabolismNicotineProliferationGene expressionOsteoblast

Introduction

Smoking has multiple systemic effects. Of particular interest to the orthopedic surgeon are the effects smoking has on the regulation of bone metabolism. In vivo studies have demonstrated that smoking is involved in many skeletal diseases. Smoking delays the rate of fracture healing and callus formation [16]. Not only do smokers have 4–5% lower bone mineral density (BMD) values [7], they also have a higher rate of pseudarthrosis in the lumbar spine after surgery [8]. Smoking is associated with a higher rate of hindfoot fusion nonunion [9] and delayed union of tibial fractures [10].

Cigarettes contain more than 150 known toxic compounds, of which nicotine is the major component, known to interact with nicotinic acetylcholine receptor (nAChR) in many cell types. Several studies have addressed the question of whether nicotine is responsible for the effects of smoking on bone metabolism. In rabbits, increased incidence of nonunion following midshaft tibia osteotomies occurred with nicotine administration [11]. A high concentration of nicotine profoundly inhibits oxidative metabolism and collagen synthesis in cultured chick bones [12] and glycosaminoglycan and collagen synthesis by nucleus pulposus cells [13]. Walker et al. [14] demonstrated the presence of nAChR subunit α4 in primary human bone cell cultures and human bone biopsies and reported that nicotine increases the proliferative rate and c-fos transcript expression in these cells via nAChR. Nicotine increases alkaline phosphatase activity and decreases cellular proliferation in a dose-dependent fashion in osteosarcoma cells [15].

We hypothesize that nicotine acts through binding to nAChR, leading to upregulation or downregulation of osteoblast regulatory genes, and thus suppresses osteogenesis, promotes bone resorption, and delays osteoblast differentiation. The aim of this study was to find the genes that regulate bone processes of which the expression is affected by nicotine. Therefore, in the present study we explored whether nicotine affects cell proliferation and osteocalcin, type I collagen, and alkaline phosphatase gene expression associated with osteogenesis in the human osteoblast-like cell MG63. The results showed a biphasic effect: toxic and antiproliferative effects at a high concentration of nicotine and stimulatory effects at low levels of nicotine. Finally, we demonstrated the inhibition of nicotine effect by d-tubocurarine suggests that nicotine acts via nAChR. These findings suggest that nicotine might have critical effects on osteogenesis through gene expression.

Materials and methods

Tissue culture

MG-63, a human osteoblast-like cell line obtained from Sigma-Aldrich, was used as the osteoblasts in this study. The cells were maintained in a medium consisting of Dulbecco’s modified Eagle’s medium (DMEM; Biological Industries, Beth Haemek, Israel), supplemented with 10% (v/v) heat-inactivated fetal calf serum (FCS), 2 mM glutamine, 100 units/ml penicillin G, and 100 μg/ml streptomycin sulfate (growth medium) at 37°C with a humidified atmosphere of 5% CO2. Upon reaching 70% confluence, cells were split using 0.05% trypsin–ethylenediaminetetraacetic acid (EDTA).

Nicotine and d-tubocurarine treatment

For nicotine and d-tubocurarine (dTC) (purchased from Sigma-Aldrich, Israel) experiments, MG-63 cells were seeded for each experiment onto 100-mm tissue culture plates at a density of 7 × 105 cells/cm2. After overnight incubation, the cells were cultured in a time-dependent manner for 1, 5, 24, 48, or 72 h with DMEM containing a low serum condition (1% FCS) to induce cell-cycle arrest and differentiation. In addition, various concentrations of nicotine were added to the medium (0, 0.01, 0.1, 1, 10, 100, 1,000, and 10,000 μM) according to the work of Tipton and Dabbous [16]. The nicotinic receptor antagonist dTC was also used; 1 and 10 mM dTC was added to the cell culture 1 h before various concentrations of nicotine were added, as described by Gotti et al. [17].

Cell proliferation assay

For cell proliferation assays, MG63 cells were cultured in a flat 96-well plate with 0, 0.01, 0.1, 1, 10, 100, 1,000, or 10,000 μM nicotine, and the cell numbers were determined after 24, 48, or 72 h by cell proliferation assay using tetrazolium salt (XTT) reagent (Biological Industries, Kibbutz Beth Haemek, Israel). The data shown are the means ± SD for five separate experiments.

RNA purification

Cells (5 × 106) from various experiments were harvested for RNA extraction by centrifugation and washed with phosphate-buffered saline (PBS); total RNA was isolated using the TRI REAGENT kit according to the instructions of the manufacturer (Sigma-Aldrich). Briefly, 1 ml TRI REAGENT was added directly on the culture dish. The cell lysates were passed by pipetting to an Eppendorf tube and incubated for 5 min at room temperature; 0.2 ml chloroform per 1 ml TRI REAGENT was added, and tubes were vortexed for an additional 15 min and incubated for 15 min at room temperature (RT). The resulting mixture was centrifuged at 12,000g for 15 min. The aqueous phase was transferred to a fresh tube and 0.5 ml isopropanol was added. Samples were incubated for 10 min at RT and centrifuged at 12,000g for 15 min. Finally, the supernatant was removed, and the RNA pellet was washed twice with 1 ml 75% ethanol and once with 1 ml 100% ethanol.

Semiquantitative reverse transcription-polymerase chain reaction (RT-PCR) of type I collagen, alkaline phosphatase, osteocalcin, and β-actin genes

A portion of total RNA (2 μg in a total volume of 25 μl) was reverse transcribed using a Moloney Murine Leukemia Virus (MMLV) First strand cDNA Synthesis Kit (Gene Choice, Frederick, MD, USA). The reaction buffer contained random hexamer primers and deoxynucleotide triphosphates (dNTPs). Portions of cDNA (~50 ng) synthesized from cells were amplified using 10 pmol of each primer in 2 × ReddyMix PCR master mix reaction buffer according to the manufacturer’s instruction (ABgene, Epson, Surrey, UK). For each gene a set of primers was designed, and annealing temperature determined for the semiquantitative RT-PCR (which included 20–30 cycles). The PCR products were resolved on 2% agarose gel. The agarose gels were photographed for densitometry analysis using the Tina 2.0 computer program. Levels of target gene transcripts were normalized to transcript levels of reference gene (β-actin) by dividing the resulting mRNA values by the value for β-actin mRNA isolated from the same cDNA. All values were expressed as mean ± standard deviation (SD). The primers used for human type I collagen alpha 1 were 5′-AAGATGTGCCACTCTGACTG-3′ (upstream) and 5′-ATAGGTGATGTTCTGGGAGG-3′ (downstream); those used for human alkaline phosphatase were 5′-CATCTGGAACCGCACGGAAC-3′ (upstream) and 5′-GCCTGGTAGTTGTTGTGAGC-3′ (downstream); those for human osteocalcin were 5′-CTCACACTCCTCGCCCTATT-3′ (upstream) and 5′-CAACTCGCACAGTCCGGAT-3′ (downstream); and those for human β-actin primers were 5′-CCACACCTTCTACAATGAGC-3′ (upstream) and 5′-CTCGTAGATGGGCACAGTGT-3′ (downstream).

Statistical analysis

All experiments were performed in triplicate at least. Each value represents the means ± SD. The significance of differences was determined using Fisher’s exact test. Data were considered statistically significant at P < 0.05.

Results

Effect of nicotine and dTC on proliferation of osteoblast-like cells MG63 in culture

MG63 cells were treated with various concentration of nicotine (0.01–10,000 μM) in a dose-dependent manner, and the effect on proliferation was analyzed using XTT reagent (Fig. 1). Addition of nicotine at concentrations of 0.01–100 μM significantly increased the osteoblast proliferation rate by as much as 1.32-fold compared with control cells (0.01 μM nicotine, 118%, P < 0.05; 0.1 μM nicotine, 122%; 1 μM nicotine, 132%, P < 0.01; 10 μM nicotine, 125%; and 100 μM nicotine, 120%, P < 0.01, relative to control cells) (Fig. 1). Incubation of cells with 1,000 μM and 10,000 μM nicotine concentrations resulted in a 1.1- to 2.2-fold significant decrease in cell proliferation compared with control cells, respectively (1,000 μM nicotine, 90%, P < 0.01; 10,000 μM nicotine, 45% relative to control cells), and ultimately leads to cell death (Fig. 1). At the time points of 48 and 72 h, similar changes in cell proliferation were observed (data not shown).
https://static-content.springer.com/image/art%3A10.1007%2Fs00774-009-0075-5/MediaObjects/774_2009_75_Fig1_HTML.gif
Fig. 1

Effect of various doses of nicotine (ranging from 0.01 μM to 10 mM) on proliferation of human osteoblast-like cells, MG63. MG63 cells were cultured with 0, 0.01, 0.1, 1, 10, 100, 1,000, and 10,000 μM nicotine, and cell numbers were determined after 24 h by cell proliferation assay with tetrazolium salt (XXT) reagent. Data shown are the means ± SD for five separate experiments. *P < 0.05, **P < 0.01

The addition of a high concentration (10 mM) of dTC, a nicotinic receptor inhibitor, to nicotine-stimulated cells (1 and 100 μM nicotine) for 24 h significantly reversed the proliferative effect gained by nicotine treatment alone (Fig. 2). In contrast, the addition of a low concentration (10 nM or 10 μM) of dTC to cells treated with nicotine (1 or 100 μM nicotine) for 24 h did not reverse the effect of nicotine on proliferation (Fig. 2). Nicotine significantly increased cell proliferation at 1 μM (123%; P < 0.05) and 100 μM (125%; P < 0.05) relative to control cells, whereas the addition of 10 mM dTC significantly reversed cell proliferation to 98% (P < 0.01) and 90% (P < 0.01), respectively (Fig. 2). In contrast, the addition of a low concentration (10 nM or 10 μM) of dTC to cells treated with nicotine (1 or 100 μM) for 24 h did not reverse the effect of nicotine on proliferation (Fig. 2). However, 10,000 μM nicotine displayed cell death (43% relative to control cells), and the addition of various concentrations of dTC (10 nM, 10 μM, or 10 mM) did not inhibit the effect significantly (Fig. 2). Low concentrations of dTC (10 nM or 10 μM) without nicotine treatment caused increased cell proliferation (127 and 131%, respectively) in comparison to MG63 control cells; whereas 10 mM dTC caused a moderate reduction in cell proliferation, to 87% compared to control cells (Fig. 2).
https://static-content.springer.com/image/art%3A10.1007%2Fs00774-009-0075-5/MediaObjects/774_2009_75_Fig2_HTML.gif
Fig. 2

Effect of nicotine receptor antagonist (dTC) on proliferation of nicotine-stimulated human osteoblast-like cells, MG63. MG63 cells were treated with 10 nM, 10 μM, or 10 mM d-tubocurarine without or with 1, 100, or 10,000 μM nicotine. Cell numbers were determined after 24 h by cell proliferation assay with XXT reagent. Data shown are the means ± SD for three independent experiments. *P < 0.05, **P < 0.01

Effect of nicotine on type I collagen, alkaline phosphatase, and osteocalcin mRNA expression in MG63 cell line

Cells were treated with nicotine in a dose-dependent (0.1, 1, 10, 100, 1,000, or 10,000 μM) and time-dependent manner (24 and 72 h) (Fig. 3a, b). Thereafter, the expression of type I collagen, alkaline phosphatase, and osteocalcin was evaluated by RT-PCR. Type I collagen and alkaline phosphatase gene expression was upregulated (~140 and ~130%, respectively) after incubation of MG63 cells with nicotine (0.1–100 μM) (Fig. 3a). Similarly, 0.1–1,000 μM nicotine induced upregulation of osteocalcin gene expression (~175%) after 24 h compared with control cells (Fig. 3). For osteocalcin, incubation with 10,000 μM nicotine led to minor downregulation (87%) of the gene expression compared to control cells (Fig. 3a). Downregulation of type I collagen and alkaline phosphatase was observed after incubation with 1 mM nicotine for 24 h (Fig. 3a). However, the expression of type I collagen and alkaline phosphatase was downregulated by 10,000 μM nicotine to 63% (1.6-fold) and 40% (2.5-fold), respectively, compared with untreated cells (Fig. 3a). After 48 (data not shown) and 72 h incubation, mRNA expression of all three genes was dramatically downregulated in a dose-dependent manner by nicotine (0.1–10,000 μM) (Fig. 3b). In contrast, no changes in type I collagen, alkaline phosphatase, and osteocalcin expression were observed relative to control cells after 1 or 5 h incubation with the various nicotine concentrations (data not shown).
https://static-content.springer.com/image/art%3A10.1007%2Fs00774-009-0075-5/MediaObjects/774_2009_75_Fig3_HTML.gif
Fig. 3

Reverse transcription-polymerase chain reaction (RT-PCR) of osteocalcin, type I collagen, and alkaline phosphatase mRNA expression in MG63 in the presence of various concentrations of nicotine. Portions of cDNA (~50 ng) were synthesized from total RNA isolated from parental cells and cells that were treated for 24 h (a) or 72 h (b) with 0.1, 1, 10, 100, 1,000, or 10,000 μM nicotine. Type I collagen (filled circle), alkaline phosphatase (inverted triangle), and osteocalcin (filled square) genes were PCR amplified using oligonucleotide primers targeting these genes, as detailed in the “Materials and methods”. Data shown are means ± SD for three independent experiments. control untreated cells

A 24-h incubation of confluent osteoblast cells with 100 μM nicotine resulted in upregulation of osteocalcin (143%), type I collagen (132%), and alkaline phosphatase (118%) gene expression compared with control (Figs. 3a, 4a). Incubation of MG63 cells with 100 μM nicotine and 1 mM dTC resulted in an inhibition of the nicotine effect by ~22% osteocalcin, ~29% type I collagen, and 36% alkaline phosphatase (Fig. 4a, b) gene expression compared to cells treated with 100 μM nicotine alone. In the presence of 100 μM nicotine and 10 mM dTC, nicotine effect was reversed by ~45% osteocalcin, ~57% type I collagen, and 38% alkaline phosphatase (Fig. 4a, b) gene expression compared to cells treated with 100 μM nicotine alone. The addition of 1 or 10 mM dTC to control cells did not have a significant effect on type I collagen, alkaline phosphatase, and osteocalcin expression (Fig. 4a).
https://static-content.springer.com/image/art%3A10.1007%2Fs00774-009-0075-5/MediaObjects/774_2009_75_Fig4_HTML.gif
Fig. 4

Effect of dTC on osteocalcin, type I collagen, and alkaline phosphatase gene expression on nicotine-stimulated cells. a Effect of 1 and 10 mM d-tubocurarine on type I collagen (black columns), alkaline phosphatase (light gray columns), and osteocalcin (dark gray columns) mRNA levels in nicotine-stimulated cells compared to control was determined by semiquantitative RT-PCR, as detailed under “Materials and methods”. Gene expression was determined in the presence or absence of 100 μM nicotine 1 h after adding 1 mM or 10 mM dTC to the medium culture; control untreated cells. *P < 0.05, **P < 0.01. b Type I collagen, alkaline phosphatase, and osteocalcin gene expression was determined by RT-PCR analysis as detailed under “Materials and methods”. PCR of β-actin was used as the control of a housekeeping gene. The H2O group represents a negative PCR control in which no cDNA was present. The 239-, 381-, 525-, and 240-bp osteocalcin, type I collagen, alkaline phosphatase, and β-actin, respectively, were resolved on 2% agarose gels containing ethidium bromide. MW molecular weight, control untreated cells

Discussion

Smoking, and nicotine, the major ingredient of cigarette smoke, have been shown to have a major effect on skeletal remodeling. Recent reports have described the presence of nAChR in human primary osteoblasts, human trabecular bone biopsies from tibial fracture repair patients, and the osteosarcoma cell line MG63 [14]. We speculated that nicotine acts through binding to nAChR, leading to up- or downregulation of osteoblast regulatory genes, thus suppressing osteogenesis, promoting bone resorption, or delaying osteoblast differentiation. To identify possible mechanisms underlying nicotine-induced changes in osteogenic metabolism, we defined changes in proliferation and in osteocalcin, type I collagen, and alkaline phosphatase gene expression after treating human osteosarcoma cells (MG63) with various concentrations of nicotine.

In the present study, we describe the effects of nicotine on gene expression associated with osteogenesis in osteoblast-like cells (MG63). The levels of nicotine used are those concentrations of nicotine observed in the blood circulation of habitual cigarette smokers (60–1,200 μM) and saliva levels of long-term snuff users (600–9,600 μM) [18]. In the MG63 cell line, there was a bimodal effect on cell proliferation: low levels of nicotine (0.01–10 μM) increased cell proliferation, whereas high levels of nicotine (100–10,000 μM) resulted in a dramatic decrease in cell proliferation and, ultimately, cell death. These observations are supported by Fang et al. [15], who studied the effect of high levels of nicotine on proliferation of osteoblast-like cells and showed that incubation of cells with levels of nicotine >1 μM resulted in decreased proliferation. In vascular endothelial cells [19], nicotine has been shown to stimulate DNA synthesis and proliferation in a bimodal manner similar to that described here. Furthermore, nicotine has been shown to inhibit alkaline phosphatase activity, but to stimulate DNA synthesis (suggesting increased cell turnover), in an osteoblast cell line [20].

Three of the genes involved in bone metabolism were chosen to be examined for level of gene expression: (1) osteocalcin is a small, highly conserved molecule first identified in the mineralized matrix of bone; it has been implicated in the pathophysiology of various malignancies [21], (2) type I collagen, a fibril-forming collagen, is found in most connective tissues and is abundant in bone. Mutation of this gene is associated with osteogenesis imperfecta type I–IV, (3) alkaline phosphatase is an enzyme linked directly to skeletal defects; the proposed function of this iso-form is matrix mineralization. The expression of these extracellular matrix proteins was estimated by determining the levels of their mRNAs. Incubation of the MG63 cell line with a low nicotine concentration for 24 h resulted in an increase in both cell proliferation and upregulation of gene expression. A longer period (72 h) of incubation with the same concentrations downregulated type I collagen and alkaline phosphatase, with no change in osteocalcin expression. Incubation of MG63 cells with a high concentration of nicotine induced cell death and downregulated the three genes examined, suggesting that nicotine has a bimodal effect that is dose- and time-dependent on the selected genes. Osteocalcin, type I collagen, and alkaline phosphatase, all known to be mediators of osteogenesis and the structural constituents of bone, were directly affected by nicotine (Fig. 5). This finding is consistent with previous studies that also identified decreased alkaline phosphatase as well as type I collagen expression in the osteosarcoma cell line, Saos-2, after long-term exposure to nicotine (5, 10, and 14 days) [22]. In addition, nicotine produced a dose-dependent (1–10 mM) increase in alkaline phosphatase activity in the osteoblast-like cells UMR 106-01 [15]. Hence, these data emphasize that nicotine has a direct effect on human osteoblast cells (i.e., bone cells) in modulating upregulation and downregulation of these extracellular matrix genes, which leads to suppression of osteogenesis through a decrease in alkaline phosphatase, type I collagen, and osteocalcin production by the osteoblasts (Fig. 5). Time-dependent experiments showed an accumulated effect of nicotine. Although short-term incubation (1 and 5 h) of nicotine did not change type I collagen, alkaline phosphatase, and osteocalcin expression, longer-term incubation (48 and 72 h) of nicotine affected gene expression by downregulation. Knowing that type I collagen, alkaline phosphatase, and osteocalcin genes play a key role in bone metabolism and the osteogenesis process, the results of this study indicate that long-term nicotine incubation inhibits the formation of bone through a decrease in expression in the osteoblasts.
https://static-content.springer.com/image/art%3A10.1007%2Fs00774-009-0075-5/MediaObjects/774_2009_75_Fig5_HTML.gif
Fig. 5

Schematic model summarizing nicotine affect on osteogenesis-associated gene expression in osteoblast cells. ALPL alkaline phosphatase, COL1 type I collagen, nAChR nicotinic acetylcholine receptor, inverted triangle nicotine, L low nicotine concentration, H high nicotine concentration, light bold arrows upregulation of gene expression, dark bold arrows downregulation of gene expression

The nicotine-induced increase in cellular proliferation was reversed by a high concentration (10,000 μM) of d-tubocurarine, a nicotinic acetylcholine receptor antagonist. d-Tubocurarine is an alkaloid that binds to all nicotinic receptor subtypes [6]. The nicotine-induced elevation in type I collagen and alkaline phosphatase expression was also blocked by low levels of d-tubocurarine, suggesting the involvement of nicotinic acetylcholine receptors in the nicotine-type I collagen and alkaline phosphatase response, providing further evidence for nicotine acting at the cellular level through nicotinic receptors. In contrast, osteocalcin gene expression levels were not inhibited by d-tubocurarine, suggesting that separate mechanisms could be operating for these effects. In addition, after the addition of high levels of d-tubocurarine, gene expression was even lower than in the control group. This finding suggests that either high levels of d-tubocurarine exhibited a nonspecific effect caused by its toxic effect or that high levels of d-tubocurarine may have an effect on other pathways than the nAChR.

High levels of nicotine are found in habitual smokers and have been shown to have inhibitory proliferation effects on bone turnover in vivo [4]. Habitual smokers often suffer loss of bone mass [4, 23], increased risk of fracture [4, 5], prolonged fracture repair, and increased nonunion rates [10]. The levels of nicotine in bone have not yet been measured accurately. It is possible that circulating levels around the fracture site may exceed those measured in the blood from smokers. Habitual smokers can be divided into three groups: light (10–15 cigarettes/day), moderate (16–30/day), and heavy (>30/day) smokers. In our study, low and high nicotine concentrations for the longer term (72 h) are equivalent to those of light, moderate, and heavy habitual smokers. Our results suggest that nicotine concentrations similar to those found in habitual smokers decrease gene expression, osteogenesis, and cell proliferation, which can explain the biological and clinical influence of smoking on bone healing.

In summary, we have demonstrated that a low nicotine concentration increases osteoblast-like cell proliferation and the gene expression that mediates bone metabolism, whereas a high dose of nicotine had the opposite effect. Our results may elucidate the influence of nicotine on bone metabolism, and supply further evidence for a receptor-mediated process, providing a clue for future understanding of metabolic bone diseases. These results together with future studies will improve our understanding of the mechanisms involved in bone repair and may lead to development of treatments, and perhaps drugs, that will speed healing in smokers and in patients with bone injuries.

Acknowledgments

This work was partially supported by a Keren Ezvonot Grant from the Ministry of Health (IL) and Rambam Medical Research Foundation Grant #1304.

Copyright information

© The Japanese Society for Bone and Mineral Research and Springer 2009