- Email: [email protected]
Demonstration of Direct Neurite–Osteoclastic Cell Communication In Vitro via the Adrenergic Receptor
J Pharmacol Sci 112, 184 – 191 (2010)
Journal of Pharmacological Sciences ©2010 The Japanese Pharmacological Society
Full Paper
Demonstration of Direct Neurite–Osteoclastic Cell Communication In Vitro via the Adrenergic Receptor Satoko Suga1,2, Shigemi Goto2, and Akifumi Togari1,* 1
Department of Pharmacology, 2Department of Orthodontics, School of Dentistry, Aichi-Gakuin University, Nagoya 464-8650, Japan
Received October 13, 2009; Accepted December 2, 2009
Abstract. There is currently great interest in the bone metabolism induced by the sympathetic nerve system. Recently, direct neurite–osteoblastic cell communication was demonstrated using an in vitro co-culture model comprising neurite-sprouting murine superior cervical ganglia and MC3T3-E1 osteoblast-like cells. In the present study, we examined whether the direct nerve–osteoclastic cell communication was present in an in vitro co-culture model comprising cultured murine superior cervical ganglia and mouse osteoclast-like cells. RAW264.7 cells treated with receptor activator of NF-κB ligand were used as osteoclast-like cells. We found that the addition of scorpion venom (SV) elicited neurite activation via intracellular Ca2+ mobilization and, after a lag period, osteoclastic Ca2+ mobilization in the co-culture. SV did not have any direct effect on the osteoclastic cells in the absence of the neurites. The addition of an α1-adrenergic receptor (AR) antagonist, prazosin, concentration-dependently prevented the osteoclastic activation that resulted as a consequence of neural activation by SV. We also found that α1-adrenergic receptor agonists evoked transient Ca2+ mobilization and gene expression of interleukin-6 in osteoclastic cells. These results demonstrate that osteoclastic activation occurs via α1-AR in osteoclastic cells as a direct response to neuronal activation. Keywords: osteoclast, neurite, co-culture, cell communication, adrenergic receptor Introduction
noradrenaline (NA) have been demonstrated to be located in the vicinity of bone tissue by immunohistochemical studies (11, 12). In the studies of the neuro-immune system, the nerve–mast cell relationship has served as a prototype of cell–cell associations and has provided substantial evidence for bi-directional communication between nerves and immune cells via membrane–membrane contact (13 – 17). In a previous experiment (18), to increase our understanding of these events, we demonstrated direct nerve–osteoblastic cell communication using an in vitro co-culture model comprising mouse osteoblast-like cells, MC3T3-E1, and neurite-spouting mouse superior cervical ganglia (SCG), with calcium imaging by confocal laser scanning microscopy (CLSM). The results showed clearly that nerve–osteoblastic cell communication occurs in the absence of intermediary transducing cells and that NA, at least in part, operating via α1-adrenergic receptors (ARs), is an important mediator of this communication. However, whether a direct neuro-osteogenic network for functional communication between nerve and osteoclastic cell exists or whether an
Bone remodeling is a physiological process involving regulation of the balance between the formation of new bone by osteoblasts and the resorption of old bone by osteoclasts. Recent studies demonstrated that bone cells are equipped with functional receptors for several neural factors and that they constitutively express diffusible axon guidance molecules known to function as chemoattractants and/or chemorepellents for growing nerve fibers (1 – 7). Therefore, it has been proposed that signaling molecules in the nervous system participate in the control of bone metabolism and that, consequently, a neuroosteogenic network is formed, similar to the previously proposed networks for neuro-immune and neuro-immune–endocrine interactions (8 – 10). Nerve fibers containing different neuropeptides and *Corresponding author. [email protected] Published online in J-STAGE on January 22, 2010 (in advance) doi: 10.1254/jphs.09283FP
184
Neuro-osteogenic Communication
intermediary cell is required is still unclear. Osteoclasts are multinucleate cells responsible for the resorption of bone. The proliferation and differentiation of mononuclear preosteoclasts and the activity of the differentiated osteoclasts are controlled by numerous systemic hormones, cytokines, growth factors, and the nervous system. Recently, the sympathetic nervous system has been implicated in bone metabolism (3, 6, 7, 19). We previously reported that the α1B-, α2B-, and β2-ARs; substance P receptor (SP-R); vasoactive intestinal peptide receptor (VIP-R); and calcitonin gene–related peptide receptor (CGRP-R) are expressed in human osteoclastic cells and that a β-adrenergic agonist was able to directly stimulate bone-resorbing activity in human osteoclastlike cells (3, 20). These findings strongly suggested the existence of a direct neuro-osteogenic network for functional communication between nerve and osteoclastic cells. In this study, we examine direct nerve–osteoclastic cell communication using an in vitro co-culture model comprising mouse osteoclast-like cells and neuritesprouting murine SCG. Our findings demonstrate that osteoclastic activation, as judged by intracellular Ca2+ mobilization, occurs as a direct consequence of contact with a specific activated nerve fiber. Moreover, we also show that this osteoclastic activation was mediated, at least in part, by NA acting through α1-ARs in osteoclastlike cells and discuss the physiological importance of α1-ARs in osteoclastic cells. Materials and Methods Nerve and osteoclastic cell co-culture The osteoclastic cells used were receptor activator of NF-κB ligand (RANKL)–induced tartrate-resistant acid phosphatase (TRAP)–positive multinucleate cells of the murine macrophage-like cell line RAW264.7. To obtain these cells, we cultured RAW 264.7 cells at a density of 3.0 × 103 cells/well in collagen-coated 35-mm diameter glass dishes at 37°C in 5% CO2 in Dulbecco’s modified Eagle’s medium (DMEM; Gibco-BRL, Grand Island, NY, USA) supplemented with 10% fetal bovine serum (FBS; BioWhittaker, Walkersville, MD, USA), 100 IU/ ml penicillin, and 100 μg/ml streptomycin in the presence of 100 ng/ml soluble RANKL (sRANKL; PeproTech, London, UK). Primary cultures of SCG neurons were established by following a previously published protocol (21 – 23). Briefly, SCG were dissected from newborn (0 – 48-h-old) BALB/c mice (Japan SLC, Shizuoka) and rinsed in HBSS containing 10 mM HEPES (pH 7.4). Each ganglion was divided into 2 – 4 pieces and incubated for 60 min at 37°C in 2 ml of HEPES containing 0.125% trypsin (grade II; Sigma, St. Louis, MO, USA).
185
For the co-culture experiments, SCG neurons at a density of 0.5 – 1 × 104 cells were added to RAW264.7 cells that had been treated with sRANKL for 72 h. These co-cultured cells were grown in DMEM supplemented with 10% FBS, 50 ng/ml murine nerve growth factor (NGF, 2.5 S; Upstate Biotechnology, Lake Placid, NY, USA), and 100 ng/ml sRANKL. After 3 days in co-culture, the cells were treated with culture medium containing the calcium fluorophore Fluo-3, and then nerve–osteoclastic cell units were monitored for intracellular Ca2+ mobilization using CLSM. Cellular activation Intracellular Ca2+ mobilization and the activation of fluorophores were used as indices of cellular activation and were assessed by CLSM (16 – 18, 24, 25). After 72 h in co-culture, the cells were incubated for 30 min at 37°C in culture medium containing Fluo-3-AM (2.5 μM; Dojindo, Kumamoto) and then washed 3 times with HEPES buffer. The cells were then observed with a CLSM (Zeiss, Oberkochen, Germany; LSM-510, argon ion laser at 488 nm), and images were captured and analyzed using IBM compatible computer software. Neurite activation was evoked by the addition of SV (Leiurus quinquestriatus hebraeus, 5 ng/ml; Sigma) to the cocultures. Pharmacological analysis of neurite-to-osteoblastic cell communication We examined whether Ca2+ mobilization in osteoclastic cells could be evoked by adding SV (10 ng/ml), adrenaline (AD, 1 μM; Sigma), NA (1 μM, Sigma), phenylephrine (PE, 1 μM; Sigma), isoprenaline (ISO, 1 μM; Sigma), neuropeptide Y (NPY, 0.1 μM; Sigma), SP (1 μM, Sigma), VIP (1 μM, Sigma), or CGRP (1 μM, Sigma). Furthermore, AR stimulation was examined as a possible neurite-derived mediator responsible for osteoclastic activation in the co-culture system. In additional studies, following 48 h of co-culture, the selective α1-AR antagonist prazosin (0.01 – 10 μM, Sigma), or the selective α2-AR antagonist yohimbine (10 μM, Sigma) was added to neurite–osteoclastic cell co-cultures 30 min before stimulation with SV, and cellular responses were then measured microscopically. Analysis of adrenergic and neuropeptide receptor mRNA expression by RT-PCR RAW264.7 cells were cultured in the absence or presence of 100 ng/ml sRANKL for 6 days, and these sRANKL-treated RAW264.7 cells differentiated into multinuclear osteoclasts positive for TRAP, an osteoclast marker. Total RNA was extracted from these osteoclastic cells and treated with DNase with the use of a spin-vac-
186
S Suga et al
Table 1. Nucleotide sequences of PCR primers Sense
Antisense
Size (bp)
Reference or Accesion No.
α1a-AR
TCCGTATCCACCGTAAAAATGTC
TGGATTCGCAGCACATTCTG
331
NM_013461
α1b-AR
CTGGTGATGTCTAGGTGTGTT
GGAATGGCCTTGTCTATAGTT
396
NM_007416
α1d-AR
GTATCCAGCCATTATGACAGA
CTACTCTGTGTCCCTGGATTT
363
NM_013460
α2a-AR
ATGGGTTACTGGTACTTTGGT
CTGGTAAATACGCACGTAGAC
378
NM_007417
α2b-AR
TCATCTACACCATCTTCAACC
AGGTATTCTAATCAGCCTTGG
360
BC066862
α2c-AR
AACTCGGGTAGACAGAGAGAC
GCAAACAATCTCAGTAACCAG
390
NM_007418
β1-AR
CGGATCGCCTCTTCGTCTTCT
GCCTGGCTCTCTACACCTTGG
381
L10084
β2-AR
CCTCATCCCTAAGGAAGTTTA
TAGGCACAGTACCTTGACAGT
299
NM_006420
β3-AR
TGGTGGCGTGTAGGGGCAGAT
TGAAGGCGGAGTTGGCATAGC
496
A26182
PAC1
GCAAGATGTCAGAACTATCCACCA
AAGTAACGGTTCACCTTCCAGC
256
NM_007407
VPAC1
TCTGCATCATCCGAATCCTG
CTGCACCTCGCCATTGAG
251
NM_011703
VPAC2
TCCCAGCAGGTGTTTCCTGGCCTAC
CGAGCCTCTTGTACTGTGACTGGTC
273
NM_009511
NPY-Y1R
CTTCTTCTCTGCCCTTTGTGA
ATGATGTTGATTCGCTTGGTC
289
D63819
NPY-Y2R
CTTTCTCCTACACCCGTATCT
GACTTCCAGGTTCTTTTTAGC
389
NM_008731
NPY-Y5R
TATCAAAGCAGACTTGAGAGC
ACACCGAAGACACTCAACTTA
304
NM_016708
NPY-y6R
GCCTTAGATAGTCACCAGGAT
ATGTAGCCTTGGGAATAGAAC
362
U58367
NK1-R
CCATCATCCACCCTCTTCAGC
TGTATGCATAGCCAATCACCA
252
NM_009313
CRLR
GAAGGAAAGGTTGCAGAGGA
TCCTGGATGCTTTTTCCATT
275
NM_018782
RAMP1
ACTCTCATCCAGGAGCTGTG
CTGGGATACCTACACGATGC
357
NM_016894
IL-6
GAAATGAGAAAAGAGTTG
ATTGGAAATTGGGGTAGGAAG
324
X54542
GAPDH
ACCACAGTCCATGCCATCAC
TCCACCACCCTGTTGCTGTA
452
NM_002046
AR, adrenergic receptor; PAC1, pituitary adenylate cyclase–activating peptide (PACAP) receptor type 1; VPAC, VIP/PACAP receptor; NK1-R, neurokinin 1 receptor; CRLR, calcitonin receptor–like receptor; RAMP, receptor activity–modifying protein; IL-6, interleukin-6; GAPDH, glyceraldehyde 3-phosphate dehydrogenase.
uum total RNA isolation kit (Promega, Madison, WI, USA). RT-PCR was performed by standard methods. Briefly, cDNA was synthesized from 1 μg of total RNA with the use of an oligo (dT)12–18 primer (Gibco-BRL) and Moloney murine leukemia virus reverse transcriptase (Gibco-BRL), before PCR amplification using specific primers. The oligonucleotides used as primers for the PCR were described in Table 1. PCR amplification was performed using the GeneAmp PCR System (Applied Biosystems, Foster City, CA, USA) under the following conditions: denaturation at 94°C for 15 s, annealing at 55°C for 30 s, and elongation at 72°C for 30 s for 25 or 32 cycles. The PCR products were electrophoresed on a 2% agarose gel, stained with ethidium bromide, and detected with a fluoroimage analyzer (Amersham–Pharmacia Biotech, Sunnyvale, CA, USA). Detection of α1-AR by real-time RT-PCR and Western blot analysis Total RNA was extracted from RAW264.7 cells cul-
tured with or without 100 ng/ml sRANKL and treated with DNase with the use of a spin-vacuum total RNA isolation kit. cDNA was produced from 1 μg of total RNA with the use of an oligo (dT)12–18 primer (GibcoBRL) and Moloney murine leukemia virus reverse transcriptase (Gibco-BRL), and α1a-AR and glyceraldehyde3-phosphate dehydrogenase (GAPDH) transcripts were quantified on the 7900HT-Fast Real-Time PCR System (Applied Biosystems) using a primer/probe kit (TaqMan Gene Expression Assays, Applied Biosystems). The expression of mRNA was normalized to the relative abundance of GAPDH. For Western blot analysis, RAW264.7 cells cultured with sRANKL were lysed in cell lysis buffer. Then, the proteins were separated by SDS-PAGE and electroblotted onto Hybond-P (PVDF) membranes (Bio-Rad, Hercules, CA, USA). The membranes were blocked with 5% non-fat milk at room temperature for 1 h, before being incubated overnight at 4°C with a specific primary antibody against α1-AR (Santa Cruz Biotech, Santa Cruz,
Neuro-osteogenic Communication
CA, USA) and then washed 3 times with TBS-T. Antibodies were detected with the corresponding anti-rabbit goat polyclonal secondary antibodies, and the membranes were developed using the Enhanced Chemiluminescent (ECL) system (GE Healthcare UK, Little Chalfont, Buckinghamshire, UK). Statistical analysis Student’s t-test was used to calculate the significance of difference. The results were expressed as the means ± S.E.M. and differences with P-values <0.05 were considered statistically significant. Results Effects of nerve stimulation on Ca2+ mobilization in osteoclastic cells The photomicrographs in Fig. 1A were obtained by CLSM. Figure 1Aa is a representative differential interference contrast (DIC) image of a typical neurite–osteoclastic cell co-culture. We added the calcium fluorophore Fluo-3 to this co-culture and then observed the cells for fluorescence (Fig. 1A: b – d). The addition of SV to a 3-day co-culture resulted in a transient Ca2+ increase in osteoclastic cells associated with neurites. When SV was added to the neurite–osteoclastic cell co-cultures, the expected increase in neurite activation (i.e., fluorescence, Fig. 1Ac) was followed by an increase in osteoclastic fluorescence (Fig. 1Ad). The osteoclastic fluorescence remained elevated for at least 40 s after the addition of SV to the co-culture (Fig. 1B). SV added to Fluo-3–loaded osteoclastic cells alone did not evoke any Ca2+ response as shown in Fig. 3B. Thus, SV did not affect the osteoclasts directly but indirectly through the activation of nerve cells. Expression of mRNA for AR and neuropeptide-receptor subtypes in RAW264.7 cells treated with or without sRANKL In order to identify the neurite-derived mediators involved in the functional communication between the neurite and osteoclastic cells, we used RT-PCR to examine the expression of genes for the adrenergic and neuropeptide receptors in RAW264.7 cells treated with or without sRANKL. After stimulation with sRANKL for 6 days, RAW264.7 cells differentiated into multinuclear osteoclasts positive for TRAP, an osteoclast marker (Fig. 2I). These osteoclasts exhibited mRNA expression of α1a-, α2b-, and β2-AR; NPY-R (Y1, Y2, Y5, y6); NK1-R as an SP-R; and CRLR as a CGRP-R. VPAC1 as VIP-Rs and RAMP1-R as a CGRP-R were weakly detected. These receptors have been already expressed in RAW264.7 cells, precursor cells of osteoclasts. However,
187
there is a different magnitude of expression between osteoclasts and precursor cells as shown in Fig. 2. Effects of neurotransmitters on Ca2+ mobilization in osteoclastic cells Next, we examined osteoclastic cells for alterations in their intracellular Ca2+ level induced by adrenergic agonists or neuropeptides. As shown in Fig. 3A, treatment with an α1-AR agonist, PE (1 μM), evoked transient Ca2+ mobilization in Fluo-3–loaded osteoclastic cells. We also observed the same response after adding AD (1 μM) or NA (1 μM), both of which have the capacity to stimulate α1-ARs, to osteoclastic cell cultures (Fig. 3B). However, none of the other agents tested gave a response. SV (10 ng/ml) did not affect the Ca2+ mobilization of the osteoclasts (Fig. 3B). Inhibitory effect of α-AR antagonists on neurite–osteoclastic cell communication in the co-culture system To further study whether these receptors are involved in neuro-osteoclast communication, we investigated osteoclasts for their response to neurite stimulation in the presence of AR antagonists. We also examined the ability of α-AR antagonists to abrogate neurite–osteoclastic cell communication. Pretreatment of co-cultures with the selective α1-AR antagonist prazosin did not affect the neurite activation evoked by SV, but did significantly inhibit the subsequent increase in osteoclastic activation (Fig. 4: A and B). The percentages of osteoclasts responding to neurite activation were 70.3% (n = 37) and 15.0% (n = 20) without and with pretreatment with prazosin. In contrast, yohimbine (10 μM), a selective α2-AR antagonist, affected neither the SV-stimulated neurite activity nor the subsequent Ca2+ mobilization in osteoclastic cells (Fig. 4B). Expression of α1a-AR and the effect of an α1-AR agonist on interleukin (IL)–6 expression in RAW264.7 cells treated with sRANKL The above finding led us to study the physiological properties of α1a-AR in osteoclastic cells. As shown in Fig. 5A, both the gene and protein expression of α1a-AR mRNA in RAW264.7 cells was significantly increased by sRANKL treatment, suggesting that α1-AR expression increased during osteoclastogenesis. To investigate the physiological responses of osteoclastic cells mediated via the α1-receptor, we examined the effect of the αagonist PE on IL-6 synthesis in osteoclastic cells. As shown in Fig. 5B, PE (1 μM) significantly increased the expression of IL-6 mRNA in RAW264.7 cells treated with sRANKL.
188
S Suga et al
B
A a
b
180
osteoclast nerve
Arbitrary Fluorescence Units
160
0s
c
d
140 120 100 80 60 40 20 0 0
10
20
30
40
Time (s)
8s
20 s
20 µm
50
60
Fig. 1. Ca2+ mobilization in SCG neurons and osteoclastic cells after the addition of SV to a coculture system, as monitored with Fluo-3 fluorescence: a representative result. A: After the cocultured cells had been loaded with Fluo-3, they were placed under observation through CLSM and stimulated with SV. Subsequently, Fluo-3 fluorescence was detected every 2 s. Images taken at the indicated time points are shown. As shown in the upper left panel (a DIC image), nerve cells are indicated by open arrowheads and osteoclastic cells are indicated by a closed arrowhead. The fluorescence intensity is displayed as a 256-color spectrum, with red indicating a greater intensity than blue. B: Fluo-3 fluorescence traces in the neurites (open circles) and osteoclastic cells (closed circles) shown in panel A. The arrow indicates the time point of SV addition.
Fig. 2. Expression of adrenergic and neuropeptide receptor mRNAs in osteoclastic cells. RAW264.7 cells were cultured in the absence or presence of 100 ng/ ml sRANKL for 6 days. Total RNA was then extracted and subjected to RT-PCR analysis. A, α1-ARs; B, α2-ARs; C, β-ARs; D, SP-R; E, GAPDH; F, NPY-Rs; G, RAMP1, CGRP-R; H, VIP/PACAP-Rs. I: RAW264.7 cells were cultured for 6 days in the absence (a) or presence (b) of sRANKL. The cells efficiently differentiated into TRAP-positive osteoclasts in the presence of sRANKL (b). DNA size markers (φX174/HaeIII digest) are shown in the left lane (s), and “−” and “+” refer to sRANKL. The housekeeping gene GAPDH (25 cycles of PCR amplification) was used as an internal standard. Thirty-two PCR cycles were performed for all receptors.
Discussion By using an in vitro co-culture model comprising RAW264.7 cells treated with sRANKL and neuritesprouting murine SCG, we showed the existence of a direct neuro-osteogenic network for functional communication between nerve and osteoclastic cells. The present findings demonstrate that osteoclastic activation, as judged by Ca2+ mobilization, occurs as a consequence of contact with a specific activated nerve fiber and that this communication is mediated, at least in part, by NA acting through α1-AR. Pharmacological activation of α1-AR in
osteoclastic cells increases the mRNA expression of IL6, which directly affects differentiation and facilitates the proliferation of osteoclast progenitors (26 – 28). The presence of several peptidergic and catecholaminergic neurotransmitters has been revealed by histological approaches in osseal nerve fibers (11, 12), and the presence of peripheral nerve axons coursing through the marrow adjacent to bone trabeculae and osteoblastic cells was also demonstrated electron microscopically (19). However, whether direct and functional communication occurs between nerve and bone cells is unclear. In the neuro-immune system, recent studies using an in vitro
Neuro-osteogenic Communication
A
B b
00s s
cc
d
Fig. 3. Effect of adrenergic agonists and neuropeptides on Ca2+ mobilization in osteoclastic cells as monitored with Fluo-3 fluorescence: a representative result. A: osteoclastic cells loaded with Fluo-3 were observed by CLSM. A DIC image is shown in the upper left panel (a). Two fluorescent images are shown, one taken before (b) and the other taken after (c, d) stimulation with PE (1 μM). B: Fluo-3 fluorescence traces of the osteoclastic cells stimulated by adrenergic agonists, neuropeptides, and scorpion venom (as shown in upper right box). The arrow indicates the time point of agonist addition. AD: adrenaline, NA: noradrenaline, PE: phenylephrine, ISO: isoprenaline, SP: substance P, CGRP: calcitonin gene–related peptide, NPY: neuropeptide Y, VIP: vasoactive intestinal peptide, SV: scorpion venom.
AD
150
Arbitrary Fluorescence Units
a
189
NA PE ISO 100
SP CGRP NPY VIP SV
50
0 0
30 s
A
B a
b
0s
c
6s
15
50 µm
20
25
30
35
200
40
osteoclast nerve
150
100
50
↓
0 0
b
d
50s
10
Time (s)
Arbitrary Fluorescence Units
a
5
50 µm
Arbitrary Fluorescence Units
6s
10
20
30
40
50
60
Time (s)
200
osteoclast nerve
150
100
↓
50
0 0
10
co-culture approach and calcium imaging by CLSM (16, 17, 24, 25) elegantly demonstrated direct and functional neurite–mast cell communication. So we decided to examine the possibility of direct nerve–osteoblastic cell communication. At first, we used mouse osteoblast-like MC3T3-E1 cells instead of mast cells and found that the addition of SV, which is a mixture of various active substances containing a neurotoxin that opens Na+ channels, to neurite–osteoblastic cell co-cultures evoked the expected increase in neurite activation (as indicated by Ca2+ mobilization), and after a lag time, an increase in the intracellular Ca2+ level in osteoblastic cells (18). In this study, instead of osteoblastic cells, we used mouse osteoclast-like cells, sRANKL-treated RAW264.7 cells, that differentiated into multinuclear osteoclasts positive for TRAP and demonstrated the direct and functional interaction between nerve and osteoclastic cells (Fig. 1). Our previous studies reported that human osteoclastic
20
30
Time (s)
40
50
60
Fig. 4. Inhibitory effect of an α-AR antagonist on Ca2+ mobilization in osteoclastic cells after SV addition to the co-culture system, as monitored with Fluo-3 fluorescence. A: Images taken at the indicated time points are shown. As shown in the upper left panel (a DIC image), nerve cells are indicated by arrowheads. B: A representative result for 30-min pretreatment with 1 μM prazosin (a), a selective α1-AR antagonist, and 10 μM yohimbine (b), a selective α2-AR antagonist. Neurites (open circles) but not osteoclasts (closed circles) were activated in response to SV. The arrow indicates the time point of SV addition.
cells and osteoblastic cells are equipped with AR and neuropeptide receptors and that they also constitutively express diffusible axon guidance molecules known to function as chemoattractants and/or chemorepellents for growing nerve fibers (1 – 3). Furthermore, we demonstrated that a β-adrenergic agonist was able to directly stimulate the bone-resorbing activity of mature human osteoclasts (20). As shown in Fig. 2, RAW264.7 cells treated with sRANKL became TRAP-positive and multinuclear, which are characteristics of mature osteoclasts; and they expressed AR subtypes such as α1a-AR, α2b-AR, and β2-AR, which were already expressed in the precursor cells. These findings suggest that RAW264.7 cells treated with sRANKL become mature osteoclastic cells equipped with the ability to receive adrenergic stimulation. Interestingly, the mRNA level of α1a-AR in the osteoclasts was up-regulated in comparison to the level in RAW264.7 cells, which might have directly affected the
190
S Suga et al
Fig. 5. Physiological features of α1-AR in osteoclastic cells. A: Increased expression of α1-AR during osteoclastogenesis. Levels of α1a-AR mRNA (a) and α1a-AR protein (b) expressions were assessed in RAW264.7 cells cultured for 2, 4, or 6 days in the presence of sRANKL by real-time RT-PCR and Western blot analysis, respectively. B: Increased expression of IL-6 mRNA after the activation of α1-AR in osteoclastic cells. The effect of an α1-AR agonist, PE, on the expression of IL-6 mRNA was assessed in RAW264.7 cells cultured for 6 days in the presence of sRANKL. PE was treated for 1 h. The housekeeping gene GAPDH (25 cycles of PCR amplification) was used as an internal standard. Thirty-five PCR cycles were performed for IL-6. The mean values are shown and bars illustrate the S.E.M. (n = 4). *P < 0.01, significantly different according to the Student’s t-test.
proliferation, differentiation, and/or function of the osteoclasts (Fig. 2A). Similar to a previous study (18), we examined the ability of α-AR antagonists to abrogate neurite–osteoclastic cell communication, in order to identify the neurite-derived mediators involved in osteoclastic activation. Pretreatment of the co-cultures with the α1-AR antagonist prazosin did not affect the neurite activation evoked by SV, but did inhibit the subsequent increase in osteoclastic activation (Fig. 4A: a – c). The addition of SV, which leads to depolarization of the membrane, induces the release of large amounts of one or more neurotransmitters from nerve endings, which then bind to their receptors on osteoclastic cells, resulting in Ca2+ mobilization in these cells. In addition, previous studies demonstrated that VIP and SP increased the intracellular Ca2+ level in rat and rabbit primary osteoclasts, respectively (29, 30). Although we showed the presence of these neuropeptide receptors, in terms of mRNA expression in RAW264.7 cells and/or RAW264.7–derived osteoclastic-like cells, their agonists did not induce intracellular Ca2+ mobilization in osteoclastic cells (Fig. 3). We do not have any
data to clarify the reason for this discrepancy, so further investigation is needed. However, these results suggest that the nervous system participates in the control of bone metabolism in bone tissues, in which membrane– membrane contacts might be formed between nerve and osteoclastic cells. α1-ARs activation has been previously shown to mediate significant up-regulation of IL-6 expression in cardiomyocytes and fibroblasts (31). In the present study, we demonstrated that α1-AR expression was increased along with osteoclastogenesis and that the activation caused increases in IL-6 expression in osteoclastic cells (Fig. 5). IL-6 is known to play a positive regulatory role in osteoclast differentiation by inducing the expression of RANKL on the surface of osteoblasts. Then, RANKL interacts with RANK expressed on osteoclast progenitors, inducing osteoclast differentiation via the RANK signaling pathway. However, recent studies described the direct effect of IL-6 on osteoclasts. Yoshitake et al. (26) reported that IL-6 acts directly on osteoclast progenitors to suppress their differentiation and that the suppressive effect of IL-6 on osteoclastic cells is mediated through reductions in both c-Jun NH2-terminal kinase activation and IκB degradation. Duplomb et al. (27) also showed that IL-6 inhibits osteoclast differentiation by specifically suppressing the RANKL signaling pathways. On the contrary, there is data suggesting a positive rather than a negative effect of IL-6 on osteoclasts (28). These data suggest that the α1-ARs expressed in osteoclastic cells induce a regulatory effect on osteoclastic activity via upregulated-IL-6 synthesis. At this time, the physiological significance and cellular mechanism of α1-AR regulation of IL-6 expression in well-differentiated osteoclastic cells remains to be established. In conclusion, we suggest that our in vitro co-culture model comprising mouse osteoclast-like cells RAW264.7 cells and neurite-sprouting murine SCG is a very useful tool for further detailed studies on nerve–osteoclastic cell communication. Furthermore, we clearly demonstrated that nerve–osteoclastic cell communication occurs without intervening cells and that this osteoclastic activation is mediated, at least in part, by α1-ARs. These results strongly implicate the direct action of the peripheral sympathetic nerve system in not only bone formation by osteoblasts but also bone resorption by osteoclasts during bone metabolism. Acknowledgments This work was partly supported by a Grant-in-Aid for Scientific Research (20592193 to A.T.) from Japan Society for the Promotion of Science and by a Grant-in Aid from Strategic Research AGU-Platform Formation (2008-2012).
Neuro-osteogenic Communication
References 1 Togari A, Arai M, Mizutani S, Mizutani S, Koshihara Y, Nagatsu T. Expression of mRNAs for neuropeptide receptors and βadrenergic receptors in human osteoblasts and human osteogenic sarcoma cells. Neurosci Lett. 1997;233:125–128. 2 Togari A, Mogi M, Arai M, Yamamoto S, Koshihara Y. Expression of mRNA for axon guidance molecules, such as semaphorinIII, netrins and neurotrophins, in human osteoblasts and osteoclasts. Brain Res. 2000;878:204–209. 3 Togari A. Adrenergic regulation of bone metabolism: possible involvement of sympathetic innervation of osteoblastic and osteoclastic cells. Microsc Res Tech. 2002;58:77–84. 4 Spencer GJ, Hitchcock IS, Genever PG. Emerging neuroskeletal signaling pathways: a review. FEBS Lett. 2004;559:6–12. 5 Chenu C. Role of innervation in the control of bone remodeling. J Musculoskelet Neuronal Interact. 2004;4:132–134. 6 Togari A, Arai M, Kondo A. The role of the sympathetic nervous system in controlling bone metabolism. Expert Opin Ther Targets. 2005;9:931–940. 7 Togari A, Arai M. Pharmacological topics of bone metabolism: the physiological function of the sympathetic nervous system in modulating bone resorption. J Pharmacol Sci. 2008;106: 542–546. 8 Opp MR, Imeri L. Sleep as a behavioral model of neuro-immune interactions. Acta Neurobiol Exp. 1999;59:45–53. 9 Aller MA, Arias JL, Lorente L, Nava MP, Duran HJ, Arias J. Neuro-immune-endocrine functional system and vascular pathology. Med Hypotheses. 2001;57:561–569. 10 Chesnokova V, Melmed S. Neuro-immuno-endocrine modulation of the hypothalamic-pituitary-adrenal (HPA) axis by gp130 signaling molecules. Endocrinology. 2002;143:1571–1574. 11 Bjurholm A, Kreicbergs A, Terenius L, Goldstein M, Schultzberg M. Neuropeptide Y-, tyrosine hydroxylase- and vasoactive intestinal polypeptide-immunoreactive nerves in bone and surrounding tissues. J Auton Nerv Syst. 1988;25:119–125. 12 Hill EL, Elde R. Distribution of CGRP-, VIP-, DβH-, SP-, and NPY-immunoreaetive nerves in the periosteum of the rat. Cell Tissue Res. 1991;264:469–480. 13 Stead RH, Tomioka M, Quinonez G, Simon GT, Felten SY, Bienenstock J. Intestinal mucosal mast cells in normal and nematode-infected rat intestines are in intimate contact with peptidergic nerves. Proc Natl Acad Sci U S A. 1987;84:2975–2979. 14 Stead RH, Bienenstock J. Cellular interactions between the immune and peripheral nervous systems: a normal role for mast cells? In: Burger MM, Sordat B, Zinkernagel RM, editors. Cell to cell interaction. Basel: Karger; 1990. p. 170–187. 15 McKay DM, Bienenstock J. The interaction between mast cells and nerves in the gastrointestinal tract. Immunol Today. 1994; 15:533–538. 16 Suzuki R, Furuno T, Mckay DM, Wolvers D, Teshima R, Nakanishi M, et al. Direct neurite-mast cell communication in vitro occurs via the neuropeptide substance P. J Immunol. 1999;163:2410–2415. 17 Furuno T, Ma D, van der Kleij HP, Nakanishi M, Bienenstock J. Bone marrow-derived mast cells in mice respond in co-culture to scorpion venom activation of superior cervical ganglion neurites
18
19
20
21
22
23
24
25
26
27
28
29
30
31
191
according to level of expression of NK-1 receptors. Neurosci Lett. 2004;372:185–189. Obata K, Furuno T, Nakanishi M, Togari A. Direct neurite-osteoblastic cell communication, as demonstrated by use of an in vitro co-culture system. FEBS Lett. 2007;581:5917–5922. Takeda S, Elefteriou F, Levasseur R, Liu X, Zhao L, Parker KL, et al. Leptin regulates bone formation via the sympathetic nervous system. Cell. 2002;111:305–317. Arai M, Nagasawa T, Koshihara Y, Yamamoto S, Togari A. Effects of β-adrenergic agonists on bone-resorbing activity in human osteoclast-like cells. Biochim Biophys Acta. 2003;1640: 137–142. Blennerhassett MG, Bienenstock J. Apparent innervation of rat basophilic leukaemia (RBL-2H3) cells by sympathetic neurons in vitro. Neurosci Lett. 1990;120:50–54. Blennerhassett MG, Tomioka M, Bienenstock J. Formation of contacts between mast cells and sympathetic neurons in vitro. Cell Tissue Res. 1991;265:121–128. Janiszewski J, Bienenstock J, Blennerhassett MG. Picomolar doses of substance P trigger electrical responses in mast cells without degranulation. Am J Physiol. 1994;267:C138–C145. Mori N, Suzuki R, Furuno T, McKay DM, Wada M, Teshima R, et al. Nerve-mast cell (RBL) interaction: RBL membrane ruffling occurs at the contact site with an activated neurite. Am J Physiol. 2002;283:C1738–C1744. Suzuki A, Suzuki R, FurunoT, Teshima R, Nakanishi M. N-cadherin plays a role in the synapse-like structures between mast cells and neuritis. Biol Pharm Bull. 2004;27:1891–1894. Yoshitake F, Itoh S, Narita H, Ishihara K, Ebisu S. Interleukin-6 directly inhibits osteoclast differentiation by suppressing receptor activator of NF-κB signaling pathways. J Biol Chem. 2008;283:11535–11540. Duplomb L, Baud’huin M, Charrier C, Berreur M, Trichet V, Blanchard F, et al. Interleukin-6 inhibits receptor activator of nuclear factor κB ligand-induced osteoclastogenesis by diverting cells into the macrophage linage: key role of serine727 phosphorylation of signal transducer and activator of transcription 3. Endocrinol. 2008;149:3688–3697. Axmann R, Bohm C, Kronke G, Zwerina J, Smolen J, Schett G. Inhibition of interleukin-6 receptor directly blocks osteoclast formation in vitro and in vivo. Arthritis Rheum. 2009;60: 2747–2756. Lundberg P, Lie A, Bjurholm B, Lehenkari PP, Lerner UH, Ransjo M, et al. Vasoactive intestinal peptide regulates osteoclast activity via specific binding site on both osteoblasts and osteoclasts. Bone. 2000;27:803–810. Mori T, Ogata T, Okumura H, Shibata T, Nakamura Y, Kataoka K. Substance P regulates the function of rabbit cultured osteoclast; increase of intracellular free calcium concentration and enhancement of bone resorption. Biochem Biophys Res Commun. 1999;262:418–422. Perez DM, Papay RS, Shi T. α1-Adrenergic receptor stimulates interleukin-6 expression and secretion through both mRNA stability and transcriptional regulation: involvement of p38 mitogenactivated protein kinase and nuclear factor-κB. Mol Pharmacol. 2009;76:144–152.
Journal of Pharmacological Sciences ©2010 The Japanese Pharmacological Society
Full Paper
Demonstration of Direct Neurite–Osteoclastic Cell Communication In Vitro via the Adrenergic Receptor Satoko Suga1,2, Shigemi Goto2, and Akifumi Togari1,* 1
Department of Pharmacology, 2Department of Orthodontics, School of Dentistry, Aichi-Gakuin University, Nagoya 464-8650, Japan
Received October 13, 2009; Accepted December 2, 2009
Abstract. There is currently great interest in the bone metabolism induced by the sympathetic nerve system. Recently, direct neurite–osteoblastic cell communication was demonstrated using an in vitro co-culture model comprising neurite-sprouting murine superior cervical ganglia and MC3T3-E1 osteoblast-like cells. In the present study, we examined whether the direct nerve–osteoclastic cell communication was present in an in vitro co-culture model comprising cultured murine superior cervical ganglia and mouse osteoclast-like cells. RAW264.7 cells treated with receptor activator of NF-κB ligand were used as osteoclast-like cells. We found that the addition of scorpion venom (SV) elicited neurite activation via intracellular Ca2+ mobilization and, after a lag period, osteoclastic Ca2+ mobilization in the co-culture. SV did not have any direct effect on the osteoclastic cells in the absence of the neurites. The addition of an α1-adrenergic receptor (AR) antagonist, prazosin, concentration-dependently prevented the osteoclastic activation that resulted as a consequence of neural activation by SV. We also found that α1-adrenergic receptor agonists evoked transient Ca2+ mobilization and gene expression of interleukin-6 in osteoclastic cells. These results demonstrate that osteoclastic activation occurs via α1-AR in osteoclastic cells as a direct response to neuronal activation. Keywords: osteoclast, neurite, co-culture, cell communication, adrenergic receptor Introduction
noradrenaline (NA) have been demonstrated to be located in the vicinity of bone tissue by immunohistochemical studies (11, 12). In the studies of the neuro-immune system, the nerve–mast cell relationship has served as a prototype of cell–cell associations and has provided substantial evidence for bi-directional communication between nerves and immune cells via membrane–membrane contact (13 – 17). In a previous experiment (18), to increase our understanding of these events, we demonstrated direct nerve–osteoblastic cell communication using an in vitro co-culture model comprising mouse osteoblast-like cells, MC3T3-E1, and neurite-spouting mouse superior cervical ganglia (SCG), with calcium imaging by confocal laser scanning microscopy (CLSM). The results showed clearly that nerve–osteoblastic cell communication occurs in the absence of intermediary transducing cells and that NA, at least in part, operating via α1-adrenergic receptors (ARs), is an important mediator of this communication. However, whether a direct neuro-osteogenic network for functional communication between nerve and osteoclastic cell exists or whether an
Bone remodeling is a physiological process involving regulation of the balance between the formation of new bone by osteoblasts and the resorption of old bone by osteoclasts. Recent studies demonstrated that bone cells are equipped with functional receptors for several neural factors and that they constitutively express diffusible axon guidance molecules known to function as chemoattractants and/or chemorepellents for growing nerve fibers (1 – 7). Therefore, it has been proposed that signaling molecules in the nervous system participate in the control of bone metabolism and that, consequently, a neuroosteogenic network is formed, similar to the previously proposed networks for neuro-immune and neuro-immune–endocrine interactions (8 – 10). Nerve fibers containing different neuropeptides and *Corresponding author. [email protected] Published online in J-STAGE on January 22, 2010 (in advance) doi: 10.1254/jphs.09283FP
184
Neuro-osteogenic Communication
intermediary cell is required is still unclear. Osteoclasts are multinucleate cells responsible for the resorption of bone. The proliferation and differentiation of mononuclear preosteoclasts and the activity of the differentiated osteoclasts are controlled by numerous systemic hormones, cytokines, growth factors, and the nervous system. Recently, the sympathetic nervous system has been implicated in bone metabolism (3, 6, 7, 19). We previously reported that the α1B-, α2B-, and β2-ARs; substance P receptor (SP-R); vasoactive intestinal peptide receptor (VIP-R); and calcitonin gene–related peptide receptor (CGRP-R) are expressed in human osteoclastic cells and that a β-adrenergic agonist was able to directly stimulate bone-resorbing activity in human osteoclastlike cells (3, 20). These findings strongly suggested the existence of a direct neuro-osteogenic network for functional communication between nerve and osteoclastic cells. In this study, we examine direct nerve–osteoclastic cell communication using an in vitro co-culture model comprising mouse osteoclast-like cells and neuritesprouting murine SCG. Our findings demonstrate that osteoclastic activation, as judged by intracellular Ca2+ mobilization, occurs as a direct consequence of contact with a specific activated nerve fiber. Moreover, we also show that this osteoclastic activation was mediated, at least in part, by NA acting through α1-ARs in osteoclastlike cells and discuss the physiological importance of α1-ARs in osteoclastic cells. Materials and Methods Nerve and osteoclastic cell co-culture The osteoclastic cells used were receptor activator of NF-κB ligand (RANKL)–induced tartrate-resistant acid phosphatase (TRAP)–positive multinucleate cells of the murine macrophage-like cell line RAW264.7. To obtain these cells, we cultured RAW 264.7 cells at a density of 3.0 × 103 cells/well in collagen-coated 35-mm diameter glass dishes at 37°C in 5% CO2 in Dulbecco’s modified Eagle’s medium (DMEM; Gibco-BRL, Grand Island, NY, USA) supplemented with 10% fetal bovine serum (FBS; BioWhittaker, Walkersville, MD, USA), 100 IU/ ml penicillin, and 100 μg/ml streptomycin in the presence of 100 ng/ml soluble RANKL (sRANKL; PeproTech, London, UK). Primary cultures of SCG neurons were established by following a previously published protocol (21 – 23). Briefly, SCG were dissected from newborn (0 – 48-h-old) BALB/c mice (Japan SLC, Shizuoka) and rinsed in HBSS containing 10 mM HEPES (pH 7.4). Each ganglion was divided into 2 – 4 pieces and incubated for 60 min at 37°C in 2 ml of HEPES containing 0.125% trypsin (grade II; Sigma, St. Louis, MO, USA).
185
For the co-culture experiments, SCG neurons at a density of 0.5 – 1 × 104 cells were added to RAW264.7 cells that had been treated with sRANKL for 72 h. These co-cultured cells were grown in DMEM supplemented with 10% FBS, 50 ng/ml murine nerve growth factor (NGF, 2.5 S; Upstate Biotechnology, Lake Placid, NY, USA), and 100 ng/ml sRANKL. After 3 days in co-culture, the cells were treated with culture medium containing the calcium fluorophore Fluo-3, and then nerve–osteoclastic cell units were monitored for intracellular Ca2+ mobilization using CLSM. Cellular activation Intracellular Ca2+ mobilization and the activation of fluorophores were used as indices of cellular activation and were assessed by CLSM (16 – 18, 24, 25). After 72 h in co-culture, the cells were incubated for 30 min at 37°C in culture medium containing Fluo-3-AM (2.5 μM; Dojindo, Kumamoto) and then washed 3 times with HEPES buffer. The cells were then observed with a CLSM (Zeiss, Oberkochen, Germany; LSM-510, argon ion laser at 488 nm), and images were captured and analyzed using IBM compatible computer software. Neurite activation was evoked by the addition of SV (Leiurus quinquestriatus hebraeus, 5 ng/ml; Sigma) to the cocultures. Pharmacological analysis of neurite-to-osteoblastic cell communication We examined whether Ca2+ mobilization in osteoclastic cells could be evoked by adding SV (10 ng/ml), adrenaline (AD, 1 μM; Sigma), NA (1 μM, Sigma), phenylephrine (PE, 1 μM; Sigma), isoprenaline (ISO, 1 μM; Sigma), neuropeptide Y (NPY, 0.1 μM; Sigma), SP (1 μM, Sigma), VIP (1 μM, Sigma), or CGRP (1 μM, Sigma). Furthermore, AR stimulation was examined as a possible neurite-derived mediator responsible for osteoclastic activation in the co-culture system. In additional studies, following 48 h of co-culture, the selective α1-AR antagonist prazosin (0.01 – 10 μM, Sigma), or the selective α2-AR antagonist yohimbine (10 μM, Sigma) was added to neurite–osteoclastic cell co-cultures 30 min before stimulation with SV, and cellular responses were then measured microscopically. Analysis of adrenergic and neuropeptide receptor mRNA expression by RT-PCR RAW264.7 cells were cultured in the absence or presence of 100 ng/ml sRANKL for 6 days, and these sRANKL-treated RAW264.7 cells differentiated into multinuclear osteoclasts positive for TRAP, an osteoclast marker. Total RNA was extracted from these osteoclastic cells and treated with DNase with the use of a spin-vac-
186
S Suga et al
Table 1. Nucleotide sequences of PCR primers Sense
Antisense
Size (bp)
Reference or Accesion No.
α1a-AR
TCCGTATCCACCGTAAAAATGTC
TGGATTCGCAGCACATTCTG
331
NM_013461
α1b-AR
CTGGTGATGTCTAGGTGTGTT
GGAATGGCCTTGTCTATAGTT
396
NM_007416
α1d-AR
GTATCCAGCCATTATGACAGA
CTACTCTGTGTCCCTGGATTT
363
NM_013460
α2a-AR
ATGGGTTACTGGTACTTTGGT
CTGGTAAATACGCACGTAGAC
378
NM_007417
α2b-AR
TCATCTACACCATCTTCAACC
AGGTATTCTAATCAGCCTTGG
360
BC066862
α2c-AR
AACTCGGGTAGACAGAGAGAC
GCAAACAATCTCAGTAACCAG
390
NM_007418
β1-AR
CGGATCGCCTCTTCGTCTTCT
GCCTGGCTCTCTACACCTTGG
381
L10084
β2-AR
CCTCATCCCTAAGGAAGTTTA
TAGGCACAGTACCTTGACAGT
299
NM_006420
β3-AR
TGGTGGCGTGTAGGGGCAGAT
TGAAGGCGGAGTTGGCATAGC
496
A26182
PAC1
GCAAGATGTCAGAACTATCCACCA
AAGTAACGGTTCACCTTCCAGC
256
NM_007407
VPAC1
TCTGCATCATCCGAATCCTG
CTGCACCTCGCCATTGAG
251
NM_011703
VPAC2
TCCCAGCAGGTGTTTCCTGGCCTAC
CGAGCCTCTTGTACTGTGACTGGTC
273
NM_009511
NPY-Y1R
CTTCTTCTCTGCCCTTTGTGA
ATGATGTTGATTCGCTTGGTC
289
D63819
NPY-Y2R
CTTTCTCCTACACCCGTATCT
GACTTCCAGGTTCTTTTTAGC
389
NM_008731
NPY-Y5R
TATCAAAGCAGACTTGAGAGC
ACACCGAAGACACTCAACTTA
304
NM_016708
NPY-y6R
GCCTTAGATAGTCACCAGGAT
ATGTAGCCTTGGGAATAGAAC
362
U58367
NK1-R
CCATCATCCACCCTCTTCAGC
TGTATGCATAGCCAATCACCA
252
NM_009313
CRLR
GAAGGAAAGGTTGCAGAGGA
TCCTGGATGCTTTTTCCATT
275
NM_018782
RAMP1
ACTCTCATCCAGGAGCTGTG
CTGGGATACCTACACGATGC
357
NM_016894
IL-6
GAAATGAGAAAAGAGTTG
ATTGGAAATTGGGGTAGGAAG
324
X54542
GAPDH
ACCACAGTCCATGCCATCAC
TCCACCACCCTGTTGCTGTA
452
NM_002046
AR, adrenergic receptor; PAC1, pituitary adenylate cyclase–activating peptide (PACAP) receptor type 1; VPAC, VIP/PACAP receptor; NK1-R, neurokinin 1 receptor; CRLR, calcitonin receptor–like receptor; RAMP, receptor activity–modifying protein; IL-6, interleukin-6; GAPDH, glyceraldehyde 3-phosphate dehydrogenase.
uum total RNA isolation kit (Promega, Madison, WI, USA). RT-PCR was performed by standard methods. Briefly, cDNA was synthesized from 1 μg of total RNA with the use of an oligo (dT)12–18 primer (Gibco-BRL) and Moloney murine leukemia virus reverse transcriptase (Gibco-BRL), before PCR amplification using specific primers. The oligonucleotides used as primers for the PCR were described in Table 1. PCR amplification was performed using the GeneAmp PCR System (Applied Biosystems, Foster City, CA, USA) under the following conditions: denaturation at 94°C for 15 s, annealing at 55°C for 30 s, and elongation at 72°C for 30 s for 25 or 32 cycles. The PCR products were electrophoresed on a 2% agarose gel, stained with ethidium bromide, and detected with a fluoroimage analyzer (Amersham–Pharmacia Biotech, Sunnyvale, CA, USA). Detection of α1-AR by real-time RT-PCR and Western blot analysis Total RNA was extracted from RAW264.7 cells cul-
tured with or without 100 ng/ml sRANKL and treated with DNase with the use of a spin-vacuum total RNA isolation kit. cDNA was produced from 1 μg of total RNA with the use of an oligo (dT)12–18 primer (GibcoBRL) and Moloney murine leukemia virus reverse transcriptase (Gibco-BRL), and α1a-AR and glyceraldehyde3-phosphate dehydrogenase (GAPDH) transcripts were quantified on the 7900HT-Fast Real-Time PCR System (Applied Biosystems) using a primer/probe kit (TaqMan Gene Expression Assays, Applied Biosystems). The expression of mRNA was normalized to the relative abundance of GAPDH. For Western blot analysis, RAW264.7 cells cultured with sRANKL were lysed in cell lysis buffer. Then, the proteins were separated by SDS-PAGE and electroblotted onto Hybond-P (PVDF) membranes (Bio-Rad, Hercules, CA, USA). The membranes were blocked with 5% non-fat milk at room temperature for 1 h, before being incubated overnight at 4°C with a specific primary antibody against α1-AR (Santa Cruz Biotech, Santa Cruz,
Neuro-osteogenic Communication
CA, USA) and then washed 3 times with TBS-T. Antibodies were detected with the corresponding anti-rabbit goat polyclonal secondary antibodies, and the membranes were developed using the Enhanced Chemiluminescent (ECL) system (GE Healthcare UK, Little Chalfont, Buckinghamshire, UK). Statistical analysis Student’s t-test was used to calculate the significance of difference. The results were expressed as the means ± S.E.M. and differences with P-values <0.05 were considered statistically significant. Results Effects of nerve stimulation on Ca2+ mobilization in osteoclastic cells The photomicrographs in Fig. 1A were obtained by CLSM. Figure 1Aa is a representative differential interference contrast (DIC) image of a typical neurite–osteoclastic cell co-culture. We added the calcium fluorophore Fluo-3 to this co-culture and then observed the cells for fluorescence (Fig. 1A: b – d). The addition of SV to a 3-day co-culture resulted in a transient Ca2+ increase in osteoclastic cells associated with neurites. When SV was added to the neurite–osteoclastic cell co-cultures, the expected increase in neurite activation (i.e., fluorescence, Fig. 1Ac) was followed by an increase in osteoclastic fluorescence (Fig. 1Ad). The osteoclastic fluorescence remained elevated for at least 40 s after the addition of SV to the co-culture (Fig. 1B). SV added to Fluo-3–loaded osteoclastic cells alone did not evoke any Ca2+ response as shown in Fig. 3B. Thus, SV did not affect the osteoclasts directly but indirectly through the activation of nerve cells. Expression of mRNA for AR and neuropeptide-receptor subtypes in RAW264.7 cells treated with or without sRANKL In order to identify the neurite-derived mediators involved in the functional communication between the neurite and osteoclastic cells, we used RT-PCR to examine the expression of genes for the adrenergic and neuropeptide receptors in RAW264.7 cells treated with or without sRANKL. After stimulation with sRANKL for 6 days, RAW264.7 cells differentiated into multinuclear osteoclasts positive for TRAP, an osteoclast marker (Fig. 2I). These osteoclasts exhibited mRNA expression of α1a-, α2b-, and β2-AR; NPY-R (Y1, Y2, Y5, y6); NK1-R as an SP-R; and CRLR as a CGRP-R. VPAC1 as VIP-Rs and RAMP1-R as a CGRP-R were weakly detected. These receptors have been already expressed in RAW264.7 cells, precursor cells of osteoclasts. However,
187
there is a different magnitude of expression between osteoclasts and precursor cells as shown in Fig. 2. Effects of neurotransmitters on Ca2+ mobilization in osteoclastic cells Next, we examined osteoclastic cells for alterations in their intracellular Ca2+ level induced by adrenergic agonists or neuropeptides. As shown in Fig. 3A, treatment with an α1-AR agonist, PE (1 μM), evoked transient Ca2+ mobilization in Fluo-3–loaded osteoclastic cells. We also observed the same response after adding AD (1 μM) or NA (1 μM), both of which have the capacity to stimulate α1-ARs, to osteoclastic cell cultures (Fig. 3B). However, none of the other agents tested gave a response. SV (10 ng/ml) did not affect the Ca2+ mobilization of the osteoclasts (Fig. 3B). Inhibitory effect of α-AR antagonists on neurite–osteoclastic cell communication in the co-culture system To further study whether these receptors are involved in neuro-osteoclast communication, we investigated osteoclasts for their response to neurite stimulation in the presence of AR antagonists. We also examined the ability of α-AR antagonists to abrogate neurite–osteoclastic cell communication. Pretreatment of co-cultures with the selective α1-AR antagonist prazosin did not affect the neurite activation evoked by SV, but did significantly inhibit the subsequent increase in osteoclastic activation (Fig. 4: A and B). The percentages of osteoclasts responding to neurite activation were 70.3% (n = 37) and 15.0% (n = 20) without and with pretreatment with prazosin. In contrast, yohimbine (10 μM), a selective α2-AR antagonist, affected neither the SV-stimulated neurite activity nor the subsequent Ca2+ mobilization in osteoclastic cells (Fig. 4B). Expression of α1a-AR and the effect of an α1-AR agonist on interleukin (IL)–6 expression in RAW264.7 cells treated with sRANKL The above finding led us to study the physiological properties of α1a-AR in osteoclastic cells. As shown in Fig. 5A, both the gene and protein expression of α1a-AR mRNA in RAW264.7 cells was significantly increased by sRANKL treatment, suggesting that α1-AR expression increased during osteoclastogenesis. To investigate the physiological responses of osteoclastic cells mediated via the α1-receptor, we examined the effect of the αagonist PE on IL-6 synthesis in osteoclastic cells. As shown in Fig. 5B, PE (1 μM) significantly increased the expression of IL-6 mRNA in RAW264.7 cells treated with sRANKL.
188
S Suga et al
B
A a
b
180
osteoclast nerve
Arbitrary Fluorescence Units
160
0s
c
d
140 120 100 80 60 40 20 0 0
10
20
30
40
Time (s)
8s
20 s
20 µm
50
60
Fig. 1. Ca2+ mobilization in SCG neurons and osteoclastic cells after the addition of SV to a coculture system, as monitored with Fluo-3 fluorescence: a representative result. A: After the cocultured cells had been loaded with Fluo-3, they were placed under observation through CLSM and stimulated with SV. Subsequently, Fluo-3 fluorescence was detected every 2 s. Images taken at the indicated time points are shown. As shown in the upper left panel (a DIC image), nerve cells are indicated by open arrowheads and osteoclastic cells are indicated by a closed arrowhead. The fluorescence intensity is displayed as a 256-color spectrum, with red indicating a greater intensity than blue. B: Fluo-3 fluorescence traces in the neurites (open circles) and osteoclastic cells (closed circles) shown in panel A. The arrow indicates the time point of SV addition.
Fig. 2. Expression of adrenergic and neuropeptide receptor mRNAs in osteoclastic cells. RAW264.7 cells were cultured in the absence or presence of 100 ng/ ml sRANKL for 6 days. Total RNA was then extracted and subjected to RT-PCR analysis. A, α1-ARs; B, α2-ARs; C, β-ARs; D, SP-R; E, GAPDH; F, NPY-Rs; G, RAMP1, CGRP-R; H, VIP/PACAP-Rs. I: RAW264.7 cells were cultured for 6 days in the absence (a) or presence (b) of sRANKL. The cells efficiently differentiated into TRAP-positive osteoclasts in the presence of sRANKL (b). DNA size markers (φX174/HaeIII digest) are shown in the left lane (s), and “−” and “+” refer to sRANKL. The housekeeping gene GAPDH (25 cycles of PCR amplification) was used as an internal standard. Thirty-two PCR cycles were performed for all receptors.
Discussion By using an in vitro co-culture model comprising RAW264.7 cells treated with sRANKL and neuritesprouting murine SCG, we showed the existence of a direct neuro-osteogenic network for functional communication between nerve and osteoclastic cells. The present findings demonstrate that osteoclastic activation, as judged by Ca2+ mobilization, occurs as a consequence of contact with a specific activated nerve fiber and that this communication is mediated, at least in part, by NA acting through α1-AR. Pharmacological activation of α1-AR in
osteoclastic cells increases the mRNA expression of IL6, which directly affects differentiation and facilitates the proliferation of osteoclast progenitors (26 – 28). The presence of several peptidergic and catecholaminergic neurotransmitters has been revealed by histological approaches in osseal nerve fibers (11, 12), and the presence of peripheral nerve axons coursing through the marrow adjacent to bone trabeculae and osteoblastic cells was also demonstrated electron microscopically (19). However, whether direct and functional communication occurs between nerve and bone cells is unclear. In the neuro-immune system, recent studies using an in vitro
Neuro-osteogenic Communication
A
B b
00s s
cc
d
Fig. 3. Effect of adrenergic agonists and neuropeptides on Ca2+ mobilization in osteoclastic cells as monitored with Fluo-3 fluorescence: a representative result. A: osteoclastic cells loaded with Fluo-3 were observed by CLSM. A DIC image is shown in the upper left panel (a). Two fluorescent images are shown, one taken before (b) and the other taken after (c, d) stimulation with PE (1 μM). B: Fluo-3 fluorescence traces of the osteoclastic cells stimulated by adrenergic agonists, neuropeptides, and scorpion venom (as shown in upper right box). The arrow indicates the time point of agonist addition. AD: adrenaline, NA: noradrenaline, PE: phenylephrine, ISO: isoprenaline, SP: substance P, CGRP: calcitonin gene–related peptide, NPY: neuropeptide Y, VIP: vasoactive intestinal peptide, SV: scorpion venom.
AD
150
Arbitrary Fluorescence Units
a
189
NA PE ISO 100
SP CGRP NPY VIP SV
50
0 0
30 s
A
B a
b
0s
c
6s
15
50 µm
20
25
30
35
200
40
osteoclast nerve
150
100
50
↓
0 0
b
d
50s
10
Time (s)
Arbitrary Fluorescence Units
a
5
50 µm
Arbitrary Fluorescence Units
6s
10
20
30
40
50
60
Time (s)
200
osteoclast nerve
150
100
↓
50
0 0
10
co-culture approach and calcium imaging by CLSM (16, 17, 24, 25) elegantly demonstrated direct and functional neurite–mast cell communication. So we decided to examine the possibility of direct nerve–osteoblastic cell communication. At first, we used mouse osteoblast-like MC3T3-E1 cells instead of mast cells and found that the addition of SV, which is a mixture of various active substances containing a neurotoxin that opens Na+ channels, to neurite–osteoblastic cell co-cultures evoked the expected increase in neurite activation (as indicated by Ca2+ mobilization), and after a lag time, an increase in the intracellular Ca2+ level in osteoblastic cells (18). In this study, instead of osteoblastic cells, we used mouse osteoclast-like cells, sRANKL-treated RAW264.7 cells, that differentiated into multinuclear osteoclasts positive for TRAP and demonstrated the direct and functional interaction between nerve and osteoclastic cells (Fig. 1). Our previous studies reported that human osteoclastic
20
30
Time (s)
40
50
60
Fig. 4. Inhibitory effect of an α-AR antagonist on Ca2+ mobilization in osteoclastic cells after SV addition to the co-culture system, as monitored with Fluo-3 fluorescence. A: Images taken at the indicated time points are shown. As shown in the upper left panel (a DIC image), nerve cells are indicated by arrowheads. B: A representative result for 30-min pretreatment with 1 μM prazosin (a), a selective α1-AR antagonist, and 10 μM yohimbine (b), a selective α2-AR antagonist. Neurites (open circles) but not osteoclasts (closed circles) were activated in response to SV. The arrow indicates the time point of SV addition.
cells and osteoblastic cells are equipped with AR and neuropeptide receptors and that they also constitutively express diffusible axon guidance molecules known to function as chemoattractants and/or chemorepellents for growing nerve fibers (1 – 3). Furthermore, we demonstrated that a β-adrenergic agonist was able to directly stimulate the bone-resorbing activity of mature human osteoclasts (20). As shown in Fig. 2, RAW264.7 cells treated with sRANKL became TRAP-positive and multinuclear, which are characteristics of mature osteoclasts; and they expressed AR subtypes such as α1a-AR, α2b-AR, and β2-AR, which were already expressed in the precursor cells. These findings suggest that RAW264.7 cells treated with sRANKL become mature osteoclastic cells equipped with the ability to receive adrenergic stimulation. Interestingly, the mRNA level of α1a-AR in the osteoclasts was up-regulated in comparison to the level in RAW264.7 cells, which might have directly affected the
190
S Suga et al
Fig. 5. Physiological features of α1-AR in osteoclastic cells. A: Increased expression of α1-AR during osteoclastogenesis. Levels of α1a-AR mRNA (a) and α1a-AR protein (b) expressions were assessed in RAW264.7 cells cultured for 2, 4, or 6 days in the presence of sRANKL by real-time RT-PCR and Western blot analysis, respectively. B: Increased expression of IL-6 mRNA after the activation of α1-AR in osteoclastic cells. The effect of an α1-AR agonist, PE, on the expression of IL-6 mRNA was assessed in RAW264.7 cells cultured for 6 days in the presence of sRANKL. PE was treated for 1 h. The housekeeping gene GAPDH (25 cycles of PCR amplification) was used as an internal standard. Thirty-five PCR cycles were performed for IL-6. The mean values are shown and bars illustrate the S.E.M. (n = 4). *P < 0.01, significantly different according to the Student’s t-test.
proliferation, differentiation, and/or function of the osteoclasts (Fig. 2A). Similar to a previous study (18), we examined the ability of α-AR antagonists to abrogate neurite–osteoclastic cell communication, in order to identify the neurite-derived mediators involved in osteoclastic activation. Pretreatment of the co-cultures with the α1-AR antagonist prazosin did not affect the neurite activation evoked by SV, but did inhibit the subsequent increase in osteoclastic activation (Fig. 4A: a – c). The addition of SV, which leads to depolarization of the membrane, induces the release of large amounts of one or more neurotransmitters from nerve endings, which then bind to their receptors on osteoclastic cells, resulting in Ca2+ mobilization in these cells. In addition, previous studies demonstrated that VIP and SP increased the intracellular Ca2+ level in rat and rabbit primary osteoclasts, respectively (29, 30). Although we showed the presence of these neuropeptide receptors, in terms of mRNA expression in RAW264.7 cells and/or RAW264.7–derived osteoclastic-like cells, their agonists did not induce intracellular Ca2+ mobilization in osteoclastic cells (Fig. 3). We do not have any
data to clarify the reason for this discrepancy, so further investigation is needed. However, these results suggest that the nervous system participates in the control of bone metabolism in bone tissues, in which membrane– membrane contacts might be formed between nerve and osteoclastic cells. α1-ARs activation has been previously shown to mediate significant up-regulation of IL-6 expression in cardiomyocytes and fibroblasts (31). In the present study, we demonstrated that α1-AR expression was increased along with osteoclastogenesis and that the activation caused increases in IL-6 expression in osteoclastic cells (Fig. 5). IL-6 is known to play a positive regulatory role in osteoclast differentiation by inducing the expression of RANKL on the surface of osteoblasts. Then, RANKL interacts with RANK expressed on osteoclast progenitors, inducing osteoclast differentiation via the RANK signaling pathway. However, recent studies described the direct effect of IL-6 on osteoclasts. Yoshitake et al. (26) reported that IL-6 acts directly on osteoclast progenitors to suppress their differentiation and that the suppressive effect of IL-6 on osteoclastic cells is mediated through reductions in both c-Jun NH2-terminal kinase activation and IκB degradation. Duplomb et al. (27) also showed that IL-6 inhibits osteoclast differentiation by specifically suppressing the RANKL signaling pathways. On the contrary, there is data suggesting a positive rather than a negative effect of IL-6 on osteoclasts (28). These data suggest that the α1-ARs expressed in osteoclastic cells induce a regulatory effect on osteoclastic activity via upregulated-IL-6 synthesis. At this time, the physiological significance and cellular mechanism of α1-AR regulation of IL-6 expression in well-differentiated osteoclastic cells remains to be established. In conclusion, we suggest that our in vitro co-culture model comprising mouse osteoclast-like cells RAW264.7 cells and neurite-sprouting murine SCG is a very useful tool for further detailed studies on nerve–osteoclastic cell communication. Furthermore, we clearly demonstrated that nerve–osteoclastic cell communication occurs without intervening cells and that this osteoclastic activation is mediated, at least in part, by α1-ARs. These results strongly implicate the direct action of the peripheral sympathetic nerve system in not only bone formation by osteoblasts but also bone resorption by osteoclasts during bone metabolism. Acknowledgments This work was partly supported by a Grant-in-Aid for Scientific Research (20592193 to A.T.) from Japan Society for the Promotion of Science and by a Grant-in Aid from Strategic Research AGU-Platform Formation (2008-2012).
Neuro-osteogenic Communication
References 1 Togari A, Arai M, Mizutani S, Mizutani S, Koshihara Y, Nagatsu T. Expression of mRNAs for neuropeptide receptors and βadrenergic receptors in human osteoblasts and human osteogenic sarcoma cells. Neurosci Lett. 1997;233:125–128. 2 Togari A, Mogi M, Arai M, Yamamoto S, Koshihara Y. Expression of mRNA for axon guidance molecules, such as semaphorinIII, netrins and neurotrophins, in human osteoblasts and osteoclasts. Brain Res. 2000;878:204–209. 3 Togari A. Adrenergic regulation of bone metabolism: possible involvement of sympathetic innervation of osteoblastic and osteoclastic cells. Microsc Res Tech. 2002;58:77–84. 4 Spencer GJ, Hitchcock IS, Genever PG. Emerging neuroskeletal signaling pathways: a review. FEBS Lett. 2004;559:6–12. 5 Chenu C. Role of innervation in the control of bone remodeling. J Musculoskelet Neuronal Interact. 2004;4:132–134. 6 Togari A, Arai M, Kondo A. The role of the sympathetic nervous system in controlling bone metabolism. Expert Opin Ther Targets. 2005;9:931–940. 7 Togari A, Arai M. Pharmacological topics of bone metabolism: the physiological function of the sympathetic nervous system in modulating bone resorption. J Pharmacol Sci. 2008;106: 542–546. 8 Opp MR, Imeri L. Sleep as a behavioral model of neuro-immune interactions. Acta Neurobiol Exp. 1999;59:45–53. 9 Aller MA, Arias JL, Lorente L, Nava MP, Duran HJ, Arias J. Neuro-immune-endocrine functional system and vascular pathology. Med Hypotheses. 2001;57:561–569. 10 Chesnokova V, Melmed S. Neuro-immuno-endocrine modulation of the hypothalamic-pituitary-adrenal (HPA) axis by gp130 signaling molecules. Endocrinology. 2002;143:1571–1574. 11 Bjurholm A, Kreicbergs A, Terenius L, Goldstein M, Schultzberg M. Neuropeptide Y-, tyrosine hydroxylase- and vasoactive intestinal polypeptide-immunoreactive nerves in bone and surrounding tissues. J Auton Nerv Syst. 1988;25:119–125. 12 Hill EL, Elde R. Distribution of CGRP-, VIP-, DβH-, SP-, and NPY-immunoreaetive nerves in the periosteum of the rat. Cell Tissue Res. 1991;264:469–480. 13 Stead RH, Tomioka M, Quinonez G, Simon GT, Felten SY, Bienenstock J. Intestinal mucosal mast cells in normal and nematode-infected rat intestines are in intimate contact with peptidergic nerves. Proc Natl Acad Sci U S A. 1987;84:2975–2979. 14 Stead RH, Bienenstock J. Cellular interactions between the immune and peripheral nervous systems: a normal role for mast cells? In: Burger MM, Sordat B, Zinkernagel RM, editors. Cell to cell interaction. Basel: Karger; 1990. p. 170–187. 15 McKay DM, Bienenstock J. The interaction between mast cells and nerves in the gastrointestinal tract. Immunol Today. 1994; 15:533–538. 16 Suzuki R, Furuno T, Mckay DM, Wolvers D, Teshima R, Nakanishi M, et al. Direct neurite-mast cell communication in vitro occurs via the neuropeptide substance P. J Immunol. 1999;163:2410–2415. 17 Furuno T, Ma D, van der Kleij HP, Nakanishi M, Bienenstock J. Bone marrow-derived mast cells in mice respond in co-culture to scorpion venom activation of superior cervical ganglion neurites
18
19
20
21
22
23
24
25
26
27
28
29
30
31
191
according to level of expression of NK-1 receptors. Neurosci Lett. 2004;372:185–189. Obata K, Furuno T, Nakanishi M, Togari A. Direct neurite-osteoblastic cell communication, as demonstrated by use of an in vitro co-culture system. FEBS Lett. 2007;581:5917–5922. Takeda S, Elefteriou F, Levasseur R, Liu X, Zhao L, Parker KL, et al. Leptin regulates bone formation via the sympathetic nervous system. Cell. 2002;111:305–317. Arai M, Nagasawa T, Koshihara Y, Yamamoto S, Togari A. Effects of β-adrenergic agonists on bone-resorbing activity in human osteoclast-like cells. Biochim Biophys Acta. 2003;1640: 137–142. Blennerhassett MG, Bienenstock J. Apparent innervation of rat basophilic leukaemia (RBL-2H3) cells by sympathetic neurons in vitro. Neurosci Lett. 1990;120:50–54. Blennerhassett MG, Tomioka M, Bienenstock J. Formation of contacts between mast cells and sympathetic neurons in vitro. Cell Tissue Res. 1991;265:121–128. Janiszewski J, Bienenstock J, Blennerhassett MG. Picomolar doses of substance P trigger electrical responses in mast cells without degranulation. Am J Physiol. 1994;267:C138–C145. Mori N, Suzuki R, Furuno T, McKay DM, Wada M, Teshima R, et al. Nerve-mast cell (RBL) interaction: RBL membrane ruffling occurs at the contact site with an activated neurite. Am J Physiol. 2002;283:C1738–C1744. Suzuki A, Suzuki R, FurunoT, Teshima R, Nakanishi M. N-cadherin plays a role in the synapse-like structures between mast cells and neuritis. Biol Pharm Bull. 2004;27:1891–1894. Yoshitake F, Itoh S, Narita H, Ishihara K, Ebisu S. Interleukin-6 directly inhibits osteoclast differentiation by suppressing receptor activator of NF-κB signaling pathways. J Biol Chem. 2008;283:11535–11540. Duplomb L, Baud’huin M, Charrier C, Berreur M, Trichet V, Blanchard F, et al. Interleukin-6 inhibits receptor activator of nuclear factor κB ligand-induced osteoclastogenesis by diverting cells into the macrophage linage: key role of serine727 phosphorylation of signal transducer and activator of transcription 3. Endocrinol. 2008;149:3688–3697. Axmann R, Bohm C, Kronke G, Zwerina J, Smolen J, Schett G. Inhibition of interleukin-6 receptor directly blocks osteoclast formation in vitro and in vivo. Arthritis Rheum. 2009;60: 2747–2756. Lundberg P, Lie A, Bjurholm B, Lehenkari PP, Lerner UH, Ransjo M, et al. Vasoactive intestinal peptide regulates osteoclast activity via specific binding site on both osteoblasts and osteoclasts. Bone. 2000;27:803–810. Mori T, Ogata T, Okumura H, Shibata T, Nakamura Y, Kataoka K. Substance P regulates the function of rabbit cultured osteoclast; increase of intracellular free calcium concentration and enhancement of bone resorption. Biochem Biophys Res Commun. 1999;262:418–422. Perez DM, Papay RS, Shi T. α1-Adrenergic receptor stimulates interleukin-6 expression and secretion through both mRNA stability and transcriptional regulation: involvement of p38 mitogenactivated protein kinase and nuclear factor-κB. Mol Pharmacol. 2009;76:144–152.