Background
Acetylcholine (ACh) was the first neurotransmitter to be discovered [
1]. It is an essential mediator of both the central and the peripheral nervous system [
2,
3]. ACh acts on two main types of receptors: nicotinic acetylcholine receptors (nAChRs), which are ligand-gated ionotropic receptors, and muscarinic acetylcholine receptors (mAChRs), which are metabotropic G-protein coupled receptors [
4]. Neuronal nAChRs are pentamers composed of either homomeric or heteromeric combinations of twelve different nicotinic receptor subunits: α2 through α10 and β2 through β4 [
5]. In addition to its role in the nervous system, ACh is increasingly recognized as an important modulator in various non-neuronal tissues [
6‐
8].
nAChRs have also been identified on cells known to be essential for the maintenance of bone homeostasis, such as osteoblasts and osteoclasts [
9‐
12]. In addition, other components involved in cholinergic transmission, including the vesicular ACh transporter, but also the enzyme choline acetyltransferase needed for the synthesis of Ach, have been reported to be present in murine osteoblasts [
13]. Moreover, recent studies have revealed the presence of cholinergic innervation in bone marrow, suggesting a potential regulatory role of ACh on all bone marrow cells including osteoclasts, but possibly also osteoblasts [
12,
14,
15].
Osteoclasts are formed by the fusion of hematopoetic mononuclear cells [
16]. This process requires the presence of macrophage colony-stimulating factor (M-SCSF) and receptor activator of nuclear factor kappa B ligand (RANKL) [
16]. Recently, several studies have implicated nAChRs and mAChRs in regulating osteoclastogenesis; however, their effects on osteoclastogenesis have not yet been fully elucidated [
17‐
20]. nAChR agonists, such as nicotine or carbamylcholine, have been shown to reduce the formation of osteoclasts in vitro [
12,
17], while other studies have reported a stimulating effect of nAChR agonists on osteoclastogenesis [
18,
21,
22]. Moreover, nicotine was also shown to affect osteoblast differentiation and function [
23‐
25]. Mice lacking the α2 nAChR subunit or the M2 mAChR have increased bone resorption and low bone mass due to increased osteoclastogenesis [
12,
19], while female α7 knockout mice exhibited significantly increased bending stiffness and cortical thickness, as well as reduced gene expression of the osteoclast marker cathepsin K [
20].
The potential relevance of cholinergic regulation in bone homeostasis under pathological conditions, such as chronic inflammation, was further highlighted by investigations into α7-deficient mice which exhibit increased bone loss as compared to wild-type (WT) mice in an animal model of arthritis [
26].
Our aim was to investigate the role of nAChRs in osteoclasts in detail, and to examine the effect of these receptors on osteoclastogenesis and bone turnover in WT and knockout mice lacking individual nAChR subunits.
Methods
Mice
Experiments were performed on C57Bl/6 (WT) mice and mice with deletions of the nAChR subunit genes α7 [
27], β2 [
28] and β4 [
29]. Mice used in this study were backcrossed onto C57Bl/6 background for 6 (β4), 7 (α7) or 12 (β2) generations after germ line transmission and kept at the Center for Brain Research, Medical University of Vienna, Austria. All data were generated from sex- and age-matched littermates. The Institutional Animal Care and Use Committee of the Medical University Vienna approved all mouse procedures.
Osteoclastogenesis from bone marrow-derived macrophages
Bone marrow cells obtained from the femurs and tibias of wild-type and knockout mice were cultured in alphaMEM medium (Gibco) containing 10 % fetal calf serum and 1 % penicillin-streptomycin. For in vitro osteoclast formation, we used mouse bone marrow cells stimulated in Petri dishes for 3 days with M-CSF 10 ng/ml (bone marrow macrophages; BMMs). BMMs were subsequently plated on 96-well plates and supplemented with 30 ng/ml M-CSF and 50 ng/ml RANKL with media/cytokine changes every 2 days. After 5–6 days under these osteoclastogenic conditions, we performed staining for tartrate-resistant acidic phosphatase (TRAP) and imaged the plates using an inverted microscope. The number of osteoclasts (OCs), defined as TRAP-positive cells containing ≥3 nuclei (TRAP-positive multinucleated cells), was counted manually in each well, while the number of cells containing <3 nuclei (TRAP-positive or TRAP-negative mononucleated cells) were counted in three power fields in each well and expressed as percent of control from at least three independent cultures. All agonists and antagonists were administered to BMMs together with M-CSF and RANKL unless otherwise noted.
In vivo osteoclastogenesis
Calvarial bone resorption was induced by subcutaneous injection of lipopolysaccharide (LPS; Sigma) above the calvaria into 7-week-old C57Bl/6 mice. The heads of anesthetized mice were shaved to receive subcutaneous injections above the calvaria in the midline of the skull located between the ears and eyes. A total of 20 mice were divided into four groups: group 1 (
n = 5), injections of 500 μg/ml LPS and on alternate days an equivalent dose of phosphate-buffered saline (PBS); group 2 (
n = 5), injections of 500 μg/ml LPS together with 4 mg/kg nicotine (in a single volume of 100 μl) and on alternate days 4 mg/kg nicotine alone (100 μl); group 3 (
n = 5), daily injections of 4 mg/kg nicotine (100 μl); and group 4 (
n = 5) daily injections of PBS (100 μl) for a period of 1 week. Animals were sacrificed by cervical dislocation; calvariae were removed and suspended in phosphate-buffered formalin for 6 hours. Following 7 days of decalcification in ethylenediaminetetraacetic acid (EDTA), tissues were embedded in paraffin, and serial sections were stained for TRAP (Sigma). Osteoclast numbers and eroded areas were determined as previously described [
30] by Osteomeasure
® software.
Immunocytochemical staining
BMMs stimulated with M-CSF and RANKL were cultured on 12-mm plain glass coverslips and stained with anti-choline acetyl transferase antibody (AB144P; Millipore, 1:25) for 1 hour, followed by a secondary, biotinylated rabbit-antigoat IgG (Vector) for 30 minutes, avidin/biotin complex (Vector) for 45 minutes and diaminobenzidine substrate (Vector) for 10 minutes. Coverslips were finally stained with Mayer’s hematoxylin solution and were mounted on glass slides.
Quantitative and semi-quantitative real-time polymerase chain reaction
For investigating the presence/absence of nAChR subunits we performed semi-quantitative real-time polymerase chain reaction (RT-PCR) analysis while quantitative RT-PCR analysis was undertaken to investigate changes in RANKL-dependent genes. Nucleus interpeduncularis was homogenized using an Ultra-Turrax-Disperser. Messenger RNA was extracted from BMMs stimulated with M-CSF and RANKL as well as nucleus interpeduncularis homogenate using the RNeasy Kit (Qiagen). RNA quality was ascertained by measuring absorption at 260 and 280 nm and calculating the A260/A280 ratio, which was 1.8–2.0 for the RNA samples used in our experiments. Care was taken when creating complementary DNA (cDNA) templates to obtain libraries free of inhibitors. Reaction efficiency for the housekeeping gene glyceraldehyde-3-phosphate dehydrogenase as well as all genes of interest was ensured by the creation of standard curves using measurements obtained over a dilution range. A volume of 800 nl/1 μl cDNA was used for the semi-quantitative/quantitative PCR, respectively. For the semiquantitative RT-PCR, cDNA was amplified using REDTaq-DNA Polymerase (Sigma-Aldrich) and the primers listed in Additional file
1: Table S1 were analyzed visually on an acrylamide gel. Quantitative RT-PCR expression of mRNA was detected and quantified using SYBR Green PCR Master Mix (Applied Biosystems). The following primers were investigated (
genes): nicotinic receptor subunits: α2-7, 9, 10 as well as β2–4 (see Additional file
1: Table S1 for primer sequences), cathepsin K (
Ctsk), matrix-metalloprotease 9 (
Mmp9), tartrate-resistant acidic phosphatase (
Acp5), nuclear factor of activated T cells c1 (
Nfatc1), and receptor activator of nuclear factor kappa beta (
Tnfrsf11a).
Bone morphology
We performed histomorphometry using the Osteomeasure
® software as previously described [
30] in 16-week old male β4–/– and α7–/– mice as well as WT age-matched littermates. We measured bone mineral density (BMD) by peripheral quantitative computed tomography (pQCT) with an XCT Research M+ pQCT machine (Stratec Medizintechnik). We measured three slices in the proximal tibia and calculated BMD values as the mean of three slices, using a voxel size of 0.07 mm and a threshold of 400 mg/cm
3.
Flow cytometry and MTT assay
For flow cytometric analyses, we used a BD FACS Canto II (BD). We stained for CD11b (AM1/70), GR1 and F4/80 (all from BD Biosciences). We characterized osteoclast precursors as (CD11b
high)/(GR1
low) cells. For apoptosis evaluation, M-CSF- and RANKL-stimulated BMMs were stained for Annexin V (Axxora LLC) and 7-Aminoactinomycin-D (7-AAD) (Fluka Sigma-Aldrich) [
31]. We characterized live cells as Annexin V low/7-AAD low, apoptotic cells as Annexin V high/7-AAD low, and dead cells as Annexin V high/7-AAD high, respectively. To distinguish from small cellular debris, forward and right angle scatter gates have been defined by contour plot and the probability scaling method, where regions were drawn to contain 95 % of the relevant lymphoid and myeloid cell populations. Gates were limited in their forward angle scatter dimension not to contain events falling in channel numbers higher than 250,000, thereby eliminating the majority of cellular clusters or multiplets. Acquisition was stopped when 10,000 events within a separate lymphoid region had been reached. The number of myeloid cells acquired was in proportional relations between equivalent cultures, albeit slightly varying by culture conditions. Percentages were read out as precentage of parent, where the parent region comprised of lymphoid and myeloid cell populations not containing small cellular debris or large cellular clusters. Thresholds for staining positivity were determined using unstained cell preparations by just omitting the addition of fluorescein-isothiocyanate (FITC)-conjugated Annexin-V or 7-AAD as a viability dye. For each antibody, appropriate isotype-matched control staining was performed to distinguish positive staining from non-specific noise. We used the following fluorophores: Annexin V: FITC; CD11b and F4/80: Allophycocyanine; and GR-1: R-phycoerythrine. Cytotoxicity was evaluated using a standard MTT assay (Merck Millipore) by measuring absorbance on an enzyme-linked immunosorbent assay (ELISA) plate reader.
Calcium measurement
Mouse bone marrow cells were first cultured for 2 days in the presence of recombinant mouse M-CSF (10 ng/ml) (BMMs). Non-adherent cells were then plated at 2 × 105 cells/cm2 density and cultured in the presence of 50 ng/ml recombinant mouse M-CSF and 50 ng/ml mouse RANKL for 72 hours. For Ca2+ measurements, the cells were incubated with 5 μM Fura-2-AM (Tocris Bioscience) and 0.05 % pluronic F127 (Sigma-Aldrich). Fluorescence intensity was measured using excitation wavelengths of 340 and 380 nm, and the emitted fluorescence at 510 nm (fluorescence ratio 340/380 nm). Changes in the fluorescence ratio 340/380 nm are shown as a function of time. Evaluation of the images was performed with MetaFluor® software (Molecular Devices). The images were scanned and plotted with an interval of 5 seconds. Graphs are representative of four independent cells from three independent experiments.
Western blotting
BMMs stimulated with 30 ng/ml M-CSF and 50 ng/ml RANKL were lysed in radioimmunoprecipitation buffer (RIPA-buffer; 10 mM Tris-HCl, pH 7.5, 150 mM NaCl, 1 % NP-40, 1 % sodium deoxycholate, 0.1 % SDS, 1 mM EDTA) supplemented with protease inhibitors (Roche). Lysates were cleared by centrifugation in a microcentrifuge. Protein content was determined using the Bradford protein assay (Bio-Rad). To reduce and denature the protein, samples were boiled in Laemmli sample buffer. Lysates were resolved by SDS-PAGE. This was followed by electrotransfer onto nitrocellulose membrane, with a loading amount of 30 μg protein. After blocking with 5 % milk, membranes were incubated with primary antibodies NFATC1 (Santa-Cruz), c-FOS (Cell Signalling) and Actin (Cytoskeleton) according to the manufacturer’s protocol and then exposed to horseradish peroxidase-conjugated secondary antibodies (Jackson). Specific bands were detected with the ECL-detection kit (Pierce) on Amersham Hyperfilm ECL (GE Healthcare). Protein expression was quantified using Image J® software.
Statistical analyses
All experiments were performed at least three times (or on at least three individual mice) with comparable results. Data are presented as bar graphs showing the mean ± standard deviation (SD). Statistical analysis was performed using one-way analysis of variance with Dunnett's post-hoc test for comparisons involving more than three groups and Student’s two-population unpaired t-test for comparisons involving two groups, both performed using GraphPad Prism® version 4.00 for Windows (GraphPad Software, San Diego California, USA). Asterisks in figures indicate the following p-values: *p < 0.05; **p < 0.005; ***p < 0.001.
Discussion
The RANK receptor and its ligand, RANKL, play key roles in osteoclastogenesis both within the framework of physiological bone turnover and in various diseases characterized by inflammatory bone loss [
16]. The central role of RANKL in osteoclast formation and bone degradation is further underlined by successful therapeutic approaches targeting this pathway. Denosumab, a RANKL-specific monoclonal antibody, is not only an effective treatment for postmenopausal osteoporosis, but also reduces local bone loss in patients with rheumatoid arthritis [
37,
38]. The finding that nAChR activation can interfere with RANK signaling is supported by the following findings: 1) when nicotine was added only to the first (solely M-CSF-driven) phase of osteoclast differentiation, it had no effect on OC generation; 2) when nicotine was present during the second phase in parallel with MCSF and RANKL, it caused an inhibitory effect; 3) even when M-CSF was absent from the second phase, nicotine was able to reduce osteoclastogenesis; and 4) the relative expression of RANKL-dependent gene, including TRAP, cathepsin K, matrix metalloproteinase 9 and NFATc1, was decreased in OCs treated with nicotine.
Herein we demonstrate that nAChRs are crucially involved in RANKL-induced osteoclastogenesis by inhibiting Ca
2+ oscillations, which in turn blocks RANKL-induced activation of c-fos and NFATc1. In contrast to our findings, it has been recently suggested that nAChR agonists have no effect on osteoclastogenesis in vitro [
12]; however, the ligands were used at lower concentrations than we utilized in our study. Tanaka et al. administered nicotine at doses comparable to those used here, and observed a reduction in the number of large OCs and the resorptive index, but an increase in the number of small OCs suggesting a reduction in fusion [
17]. We were also able to observe an increase in the number of osteoclasts upon administration of low-dose nicotine and PNU282987. Thus nicotinic agonists at low concentration may favor the generation of osteoclasts, whereas high concentrations inhibit osteoclastogenesis. Interestingly it has been suggested that different nAChRs may also convey opposing effects as evidenced by recent studies, which described increased in vitro osteoclastogenesis as well as a significant, though modest, decrease in the systemic bone mass of α2–/– mice [
12] and increased bending stiffness and cortical thickness in female α7–/– mice, while male counterparts showed no difference as compared to WT [
20]. However, male α7 β2–/– exhibited a significant increase in BMD when compared to WT mice [
39]. Interestingly, we found an, albeit modest, increased bone mass in male α7–/– mice, which we could not demonstrate in female α7–/– mice. Gender disparity with regard to bone phenotype is not uncommon and has been described among others in CB1–/– or GPR55–/– mice [
40,
41]. In both of these cases males exhibited an osteopetrotic phenotype which females did not, and which was not otherwise explainable. In line with previous studies, which showed no change in gene expression of osteoblast markers, we found no change in the osteoblast phenotype in α7–/– mice [
19,
20]. Taken together it appears that high levels of nAChR stimulation may have an overall negative effect on osteoclastogenesis but actions mediated via the α7 receptor may indeed be different and positive in nature.
Our observation that nAChR antagonists also inhibit osteoclastogenesis, coupled with our results showing reduced osteoclastogenesis in α7–/– mice, can be interpreted in the context of desensitization, a key property of nAChRs. Depending on the receptor subtype and the type of agonist, as well as the duration of exposure to nAChRs, agonists have been shown to alter the affinity state of the receptor inducing desensitization, leading to a functional blockade of the receptor [
42,
43]. In particular, nicotine and certain experimental α7-selective partial agonists are known to produce a transient activation of α7 receptors followed by a period of prolonged residual inhibition or desensitization [
44]. While we derived the dosage of nicotine from published studies [
17,
18], due to the fact that we used high doses we cannot exclude the possibility that the inhibition seen with the high-dose nAChR ligands is due to either effects on mAChRs, non-receptor mediated effects (such as the blocking of additional ion channels), or to other factors involved in the regulation of osteoclastogenesis.
Oscillating levels of intracellular calcium appear to be critical for osteoclastogenesis primarily by inducing the dephosphorylation and shuttling of NFATc1 into the nucleus, where it plays a decisive role in the regulation of osteoclastogenesis [
32‐
34]. After observing the marked acute effect of both nAChR agonists and antagonists on Ca
2+ oscillation, we could show that, indeed, the nAChR agonist nicotine inhibited RANKL-induced expression and production of NFATc1 and also had a similar effect on c-fos, an early inducer of NFATc1 [
45].
The fact that cholinergic signaling has an effect on osteoclastogenesis is also supported by studies which have demonstrated that ACh is produced by cells of the immunoinflammatory sytem, including lymphocytes [
46], and the fact that bone marrow also possesses its own sensory and autonomic innervation including cholinergic neurons [
12,
14,
15]. This suggests that, under certain circumstances, in addition to ACh released from BMMs, differentiating OCs may also be exposed to ACh released from neurons, which could then have various effects on their nAChRs.
Conclusions
Our results suggest that cholinergic agonists inhibit RANKL-induced osteclastogenesis by interfering with intracellular calcium levels and consequently with the NFATc1 signaling pathway. Of note, this complete blockade of osteoclastogenesis in vitro was neither due to toxic effects nor associated with apoptosis, and was shown to be reversible. However, while high levels of nAChR activity appear to inhibit RANKL-induced osteoclastogenesis, actions mediated by certain nAChRs, in particular the α7 homomeric receptor, may favor it. Further studies are needed to determine the gender-specific effect of α7 and other nAChRs and mAChRs on bone homeostasis.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
PM designed and performed in vitro and in vivo osteoclastogenesis, RT-PCR experiments, immunocytochemistry and bone morphology, analyzed data and drafted the manuscript. SHa and DS designed and performed in vitro osteoclastogenesis experiments and revised the manuscript. DG performed calcium measurements and revised the manuscript. TK performed Western blot experiments and revised the manuscript. PS provided nAChR–/– mice, performed RT-PCR experiments and revised the manuscript. SB performed flow cytometry and MTT experiments and revised the manuscript. SHu, GP, AM and JSS participated in evaluating data and revised the manuscript. KR directed the project, designed the experiments and drafted the manuscript. All authors read and approved the final manuscript.