The final data of this manuscript were presented at the 22nd Annual Meeting of the Japanese Society for Bone and Mineral Research in Osaka, Japan, August 4-7, 2004, and an abstract was published.
The authors have no conflict of interest.
We examined the direct effect of SC-19220, an EP1 prostaglandin (PG) E2 receptor antagonist, on osteoclastogenesis induced by RANK/RANKL signaling in mouse cell cultures. We found that SC-19220 inhibited RANKL-induced osteoclastogenesis by suppression of the RANK/RANKL signaling pathway in osteoclast precursors.
Introduction: Bone growth is accomplished by a dynamic equilibrium between formation by osteoblasts and resorption by osteoclasts, which are regulated by many systemic and local osteotropic factors that induce osteoclast formation from hematopoietic precursors through RANK/RANKL signaling. There are four subtypes of prostaglandin E (PGE) receptors, EP1, EP2, EP3, and EP4, and PGE2 facilitates bone resorption by a mechanism mediated by EP2/EP4. It is well known that SC-19220 is an EP1-specific antagonist. We previously found that SC-19220 inhibited osteoclastogenesis induced by osteotropic factors, including PGE2; however, the inhibitory mechanism is not clear. In this study, we investigated the inhibitory effects of SC-19220 on osteoclastogenesis induced by RANK/RANKL signaling in mouse cell cultures and analyzed the mechanism involved.
Materials and Methods: A bone marrow culture system and bone marrow macrophages were used to examine the effects of SC-19220 on PGE2-, 11-deoxy-PGE1-, and RANKL-induced osteoclastogenesis. We analyzed RANKL expression in osteoblasts induced by PGE2 using RT-PCR. We also examined the effects of SC-19220 on the macrophage-colony-stimulating factor (M-CSF) receptor (c-Fms) and RANK expression in osteoclast precursors as well as RANK/RANKL signaling using RT-PCR and Western blotting analyses.
Results and Conclusion: SC-19220 dose-dependently inhibited osteoclast formation induced by PGE2, 11-deoxy-PGE1, and RANKL in the mouse culture system; however, it had no influence on RANKL expression in osteoblasts induced by PGE2. Furthermore, the expression of RANK and c-Fms in osteoclast precursors was decreased by SC-19220 at the mRNA and protein levels. In RANK signaling networks, SC-19220 inhibited c-Src and NFAT2 expression. Our findings indicated that SC-19220 inhibits RANKL-induced osteoclastogenesis through the suppression of RANK, c-Fms, c-Src, and NFAT2, suggesting that this EP1-specific antagonist inhibits osteoclast formation induced by RANKL from the early stage of osteoclastogenesis.
BONE IS A DYNAMIC tissue in which bone resorption and formation proceed in a regulated manner under the control of systemic hormones and local factors. Osteoblasts are bone-forming cells that differentiate from mesenchymal cells. Osteoclasts, multinucleated cells of monocyte-macrophage lineage that form by fusion of mononuclear precursors, are differentiated from hematopoietic stem cells and play crucial roles in bone remodeling. This multistep differentiation process is under the control of the microenvironment, which includes osteoblasts and local factors. Stimulation by osteotropic factors such as 1,25(OH)2D3, prostaglandin E2 (PGE2), parathyroid hormone (PTH), interleukin (IL)-11, and IL-6 induces RANKL expression on the surface of bone marrow stromal cells and osteoblasts, and RANKL binds to its cognate receptor, RANK, on the osteoclast precursors.(1-5) RANK activation by RANKL is followed by its interaction with TNF receptor-associated family (TRAF) members, and activation of NF-κB, c-fos, c-Src, and MAP kinases, such as p38, JNK, and p44/42.(6,7)
PGE2 is a major product that stimulates bone resorption and is produced in bone mainly by osteoblasts, whose action is mediated by rhodopsin-type receptors specific to prostaglandins. There are four subtypes of PGE receptors, designated EP1, EP2, EP3, and EP4, which are encoded by different genes and expressed differently in various type of tissue.(8-10) In recent studies, PGE2 was found to stimulate bone resorption by a mechanism involving cAMP and RANKL, which is mediated mainly by EP4 and partially by EP2.(11-15)
We previously reported that SC-19220, an EP1 receptor antagonist, inhibited osteoclast formation by 1,25(OH)2D3, PGE2, PTH, IL-11, and IL-6 in mouse bone marrow cultures.(16) These osteotropic factors transmit different signals in osteoblasts and subsequently induce RANKL expression; thus, they consequently induce osteoclast formation through RANK/RANKL signaling.
SC-19220 has been shown to act as a competitive antagonist toward PGE2-induced smooth muscle contractions in guinea pig ileum and stomach specimens,(17,18) and also functions as a PGE2 antagonist in EP1 receptor-mediated contraction by guinea pig trachea cells.(19) However, it is known that SC-19220 has no affinity with the EP1 receptor,(20) because in general, it is thought to function as an EP1 receptor antagonist that blocks functional activation.(9, 10, 21) We previously found that SC-19220 inhibited osteoclast formation induced by not only PGE2, but also by 1,25(OH)2D3, PTH, IL-11, and IL-6. Therefore, we considered that SC-19220 was likely to inhibit osteoclastogenesis induced by RANK/RANKL signaling, which is a common pathway for osteoclast formation induced by various osteotropic factors. In this study, we examined the precise mechanism by which SC-19220 inhibits osteoclast formation induced by RANKL.
MATERIALS AND METHODS
Mice, antibodies, and reagents
Six-week-old female ddY mice were obtained from Seac Yoshitomi (Fukuoka, Japan). Recombinant human soluble RANKL was purchased from Pepro Tech EC (London, UK), and PGE2 and 11-deoxy-PGE1 were purchased from Cayman Chemical (Ann Arbor, MI, USA). Human macrophage-colony-stimulating factor (M-CSF; Leukoprol) was purchased from KYOWA HAKKO KOGYO (Tokyo, Japan). SC-19220 was provided by Searle & Co. (Skokie, IL, USA). PGE2 and 11-deoxy-PGE1 were dissolved in ethanol at 1 mM. SC-19220 was dissolved in ethanol at 5 mg/ml. The antibodies against TRAF6, M-CSF receptor (c-Fms), RANK, NFAT2, and actin antibodies were purchased from Santa Cruz Biotechnology (Santa Cruz, CA, USA).
Culture of primary mouse osteoblastic cells
Primary osteoblastic cells were isolated from 2-day-old mice calvariae after routine sequential digestion five times with 0.1% collagenase (Wako Pure Chemical Industries, Osaka, Japan) and 0.2% dispase (Godo Shusei, Tokyo, Japan), as previously described.(15) Osteoblastic cells collected from fractions 3-5 were combined and cultured in α-MEM (ICN Biomedicals, Aurora, OH, USA) containing 10% FCS, penicillin (100 U/ml), and streptomycin (100 μg/ml) at 37°C with 5% CO2 in air.
Mouse bone marrow cultures
Bone marrow cells (2 × 106 cells/well) were suspended in α-MEM containing 10% FCS and antibiotics and seeded into 24-well flat-bottomed culture plates. After being cultured for 7 days, with the medium replaced every 2 or 3 days, the cells were subjected to staining for TRACP. TRACP+ cells were considered to be osteoclast-like cells and counted. The results are expressed as the mean ± SD of four cultures.
M-CSF-dependent bone marrow macrophages
Bone marrow cells were suspended in α-MEM containing 10% FCS and antibiotics and cultured in 48-well plates (1.5 × 105 cells/0.3 ml/well) in the presence of M-CSF (100 ng/ml). After being cultured for 3 days, nonadherent cells were completely removed from the cultures by pipetting, and adherent cells were used as bone marrow macrophages.(22) Bone marrow macrophages were further cultured for 6 days with soluble RANKL (20 or 50 ng/ml) and M-CSF (50 ng/ml) in various concentrations of SC-19220.
For semiquantitative RT-PCR, total RNA was prepared from cells using ISOGENE (Nippon Gene, Tokyo, Japan). Total RNA was treated with DNase (Takara Biochemicals, Shiga, Japan) at 37°C for 18 h, and first-strand DNA was synthesized from DNase-treated total RNA with random primers and Superscript II (Invitrogen, Carlsbad, CA, USA). Reverse transcribed total RNA was subjected to PCR amplification with AmpliTaq GOLD DNA polymerase (Applied Biosystems, Foster, CA, USA) using specific PCR primers. Oligonucleotide primers used in the PCR amplification are indicated in Table 1. RT-PCR was performed under the following conditions: 1 cycle at 95°C for 9 minutes, 94–95°C for 40 s, and 60–63°C for 40–60 s for 22–45 cycles. The amplified DNA fragments were subjected to 8% polyacrylamide gel electrophoresis and visualized with ethidium bromide staining.
Table Table 1.. Primers and RT-PCR Conditions
Western blot analysis
Bone marrow macrophages were cultured in α-MEM containing 10% FCS, antibiotics, and M-CSF (50 ng/ml) in the presence or absence of the stimulants and washed twice with ice-cold PBS and dissolved in a lysis buffer (50 mM Tris-HCl [pH 6.8], 2% SDS). The cell extracts were boiled in the presence of SDS sample buffer (0.125 M Tris-HCl [pH 6.8], 2% [w/v] SDS, 5% glycerol, 2% 2-mercaptoethanol, 0.05% [w/v] bromophenol blue) for 5 minutes and separated on SDS-polyacrylamide gels. The proteins were electro-blotted onto polyvinylidene fluoride membranes using a semidry blotter (ATTO, Tokyo, Japan), and the membranes were incubated in TBS buffer (100 mM Tris-HCl [pH 7.5], 150 mM NaCl) containing 5% nonfat dry milk to reduce nonspecific binding. The membranes were exposed to the primary antibodies overnight at 4°C. Finally, the membranes were washed three times and incubated with secondary anti-mouse or anti-rabbit IgG horseradish peroxidase-conjugated antibodies for 1 h. Detection of immuno-complexes was performed using an ECL Western detection kit (Amersham Pharmacia Biotech, Buckinghamshire, UK).
SC-19220 inhibits osteoclast formation by PGE2 and 11-deoxy-PGE1, an EP2/EP4 agonist
Both 10−6 M PGE2 and 11-deoxy-PGE1, an EP2/EP4 agonist, enhanced the formation of TRACP+ cells in mouse bone marrow cultures. SC-19220 dose-dependently decreased the number of TRACP+ cells induced by PGE2 and 11-deoxy-PGE1 (data not shown). Addition of 25 μg/ml nearly eliminated the increase of TRACP+ cells induced by the prostanoids.
SC-19220 has no effect of RANKL expression in osteoblasts
RT-PCR analysis revealed that stimulation of 10−6 M PGE2 increased RANKL expression in osteoblastic cells; however, various concentrations of SC-19220 had no effect on that expression after 24 h (Fig. 1).
Expression of RANK and c-Fms by osteoclast precursors
We examined the expression of the receptors on bone marrow macrophages using RT-PCR and Western blot analysis, because osteoclastogenesis requires two ligand signals, M-CSF and RANKL. c-Fms mRNA was decreased by the addition of SC-19220 in a dose-dependent manner after 1 day (Fig. 2A), when mouse bone marrow macrophages were co-cultured with M-CSF and RANKL. Furthermore, c-Fms protein was also diminished after 2 days (Fig. 2B). SC-19220 decreased RANK mRNA in bone marrow macrophages after 1 day in a dose-dependent manner (Fig. 2A), when mouse bone marrow macrophages were co-cultured with M-CSF and RANKL. RANK protein in osteoclast precursors also has weakly diminished by SC-19220 after 2 days (Figs. 2B and 2C). In contrast, OPG expression was not detectable in the bone marrow macrophages and was not influenced by SC-19220 (Fig. 2A).
SC-19220 inhibits osteoclast formation induced by RANKL in mouse bone marrow culture and bone marrow macrophages
To determine whether SC-19220 inhibits osteoclastogenesis induced by RANKL, bone marrow cells were exposed to 50 ng/ml of M-CSF and 50 ng/ml of RANKL in the presence of SC-19220 for 7 days. SC-19220 dose-dependently decreased the number of TRACP+ cells induced by M-CSF and RANKL in mouse bone marrow cultures (Fig. 3). Next, we examined the inhibitory effects of SC-19220 on osteoclastogenesis induced by RANKL in bone marrow macrophages. When mouse bone marrow macrophages were cultured with M-CSF and RANKL, TRACP+ cells were detected. However, SC-19220 decreased the number of TRACP+ cells induced by RANKL in a dose-dependent manner (Figs. 4 and 5), and 25 μg/ml of SC-19220 nearly eliminated the increased number of TRACP+ cells induced by RANKL (Figs. 3, 4, and 5).
SC-19220 suppresses c-Src and NFAT2 expression in bone marrow macrophages through RANK/RANKL signaling pathway
The expression of TRAF6 was not influenced by SC-19220 (Fig. 6), when mouse bone marrow macrophages were stimulated with M-CSF and RANKL for 3 days. c-Src expression was induced by RANKL stimulation; however, it was decreased by SC-19220 after 2 days (Fig. 6) in a dose-dependent manner. To investigate whether SC-19220 has an effect on the expression of NFAT2, the master regulator of RANK signaling in osteoclast differentiation, we performed RT-PCR and Western blotting analyses. When mouse bone marrow macrophages were stimulated with M-CSF and RANKL, NFAT2 mRNA and protein were induced. SC-19220 dose-dependently inhibited NFAT2 expression 1 day after RANKL stimulation in mouse bone marrow macrophages (Fig. 7).
We confirmed that SC-19220 inhibited osteoclastogenesis induced by PGE2 and also found that it inhibited osteoclast formation induced by 11-deoxy-PGE1, an EP2/EP4 agonist. In our previous study, SC-19220 inhibited osteoclastogenesis induced by the osteotropic factors PGE2, PTH, 1,25(OH)2D3, IL-11, and IL-6.(16) Although it has been found that EP1 receptor is expressed in osteoclast precursors,(23) there are no abnormalities in bone metabolism of EP1 knockout mice. In addition, the role of EP1 in osteoclastogenesis induced by osteotropic factors are not known. In our preliminarily experiments, other EP1 antagonists that have binding affinity with EP1 receptor, SC-51322 and SC-51089 up to 1 mg/ml, do not inhibit osteoclastogenesis induced by RANKL. Pretreatment of bone marrow macrophages with SC-19220 for the first 3 days and subsequent culture of the cells with M-CSF/RANKL but without SC-19220 did not affect osteoclastogenesis. These results suggest that inhibition of osteoclastogenesis induced by osteotropic factors, including PGE2, by SC-19220 may be not caused by the action as an EP1 antagonist.
It is well known that those osteotropic factors induce osteoclast formation through RANK/RANKL signaling. Although SC-19220 had no effect on the RANKL expression induced by PGE2 in osteoblastic cells, it inhibited c-Fms and RANK expression induced by M-CSF and RANKL in osteoclast precursors (Figs. 1 and 2). These results suggest that SC-19220 is likely to act directly on osteoclast precursors without influencing their ability to support osteoclastogenesis through RANKL on osteoblasts.
c-Fms is important to survival and formation in osteoclasts and is expressed in osteoclast precursors at the early stage of osteoclastogenesis, whereas M-CSF enhances RANK expression in osteoclast precursors and osteoclast differentiation of osteoclasts from osteoclast precursors.(24) We found that SC-19220 inhibited osteoclastogenesis by M-CSF and RANKL in mouse bone marrow cultures and bone marrow macrophages, because it decreased c-Fms and RANK expression in osteoclast precursors (Fig. 2). It is known that the expression of c-Fms is downregulated by various cytokines including granulocyte macrophage-colony-stimulating factor, IL-3, IL-4, interferon γ, and TNF-α.(25) The decrease of c-Fms expression might have been caused by production of these cytokines. In addition, downregulation of c-Fms expression may cause a decrease of RANK expression in osteoclast precursors. Therefore, the commitment to osteoclast may be suppression. Because that was inadequate to explain the SC-19220 inhibition of osteoclast formation induced by RANKL, we next investigated the effects of SC-19220 in RANK/RANKL signaling in osteoclast precursors and found that it obviously inhibited c-Src and NFAT2 expression in the RANK signaling network.
c-Src has been reported to play crucial roles in the functions of osteoclasts as a downstream target of RANK and c-Fms. After RANKL stimulation, c-Src was reported to form a complex with RANK and TRAF6, after which activation of PI3 kinase occurred, and the signal was transmitted.(6, 7, 26) Furthermore, M-CSF stimulation induces c-Src, PI3 kinase, and phospholipase Cγ (PLCγ) binding in the cytoplasmic domain of c-Fms, after which the signal is transmitted.(27-30) Osteopetrosis develops in c-src knockout mice,(31) and osteoclastogenesis is normal; however, osteoclasts are not activated.(32,33) Thus, it is known that c-Src is essential for bone resorption and actin ring formation in osteoclasts. In this study, we found that the treatment of SC-19220 markedly decreased the expression of c-Src in mouse bone marrow macrophages (Fig. 6). These results suggest that the decrease of c-Src expression might have been caused by a failure to generate multinuclear osteoclasts by SC-19220 and that downregulation of c-Fms might be decreased expression and activation of c-Src.
NF-κB is important molecule for osteoclastogenesis and osteoclast survival.(6, 7, 34) It is known that interaction of RANK with TRAF6 is necessary and sufficient to activate NF-κB.(6, 7, 26) In this study, SC-19220 had no effect of on the expression of TRAF6 in osteoclast precursor cells induced by RANKL (Fig. 6). In addition, Western blot analysis revealed that treatment of SC-19220 did not affect the expression of NF-κB p50 subunit in the nuclear extract of the cells treated with RANKL (data not shown). These finding suggest that SC-19220 does not directly activate NF-κB in the osteoclast precursor cells treated with RANKL.
Recently, it was reported that RANKL stimulation increased intracellular calcium ion, which led to calcineurin-mediated activation of NFAT2.(35) NFAT2-deficient embryonic stem cells fail to differentiate into osteoclasts in response to RANKL stimulation.(35) In addition, when the expression of NFAT2 was suppressed by introducing antisense NFAT2, multinucleated cell formation was severely hampered.(35,36) Cyclosporin A and FK506, both immunosuppressants, bind calcineurin and consequently inhibit the expression of NFAT 2, thereby suppressing the activation of T-cells. Those immunosuppressants also significantly suppressed multinucleated cell formation that accompanied a reduction in nuclear localization of NFAT2.(36,37) In addition, ectopic expression of NFAT2 causes precursor cells to undergo efficient differentiation without RANKL signaling.(35) Ectopic expression of the constitutively active calcineurin-independent NFAT2 mutant in RAW264.7 cells is sufficient to induce them to express an osteoclast-specific pattern of gene expression and differentiate into morphologically distinct, multinucleated osteoclasts capable of inducing the resorption of a physiological matrix substrate.(38) These findings indicate that NFAT2 is a master switch for regulating terminal differentiation of osteoclasts that function downstream of RANKL.(6, 7, 35-38)
Recently, Koga et al.(39) reported that mice lacking immunoreceptor tyrosin-based activation motif (ITAM)-harboring adaptors, Fc receptor common γ subunit (FcRγ), and DNAX-activating protein (DAP) 12 exhibit severe osteopetrosis owing to impaired osteoclast differentiation. In osteoclast precursor cells, FcRγ and DAP12 associate with multiple immunoreceptors and activate calcium signaling through PLCγ. It has been reported that PGE2 signaling through EP1 receptor causes activation of the calcium ion channel in cell membranes mediated by G-proteins, consequently increasing calcium ion from extracellular to intracellular.(40) SC-19220 completely blocked the PGE2-stimulated increase of intracellular calcium in the renal collecting duct and human granulosa-lutein cells.(41,42) Therefore, it may be possible that SC-19220 inhibits the increase of intracellular calcium levels by RANKL stimulation in osteoclast precursors. Furthermore, the inhibition of NFAT2 expression by SC-19220 may be because of the decreased intracellular calcium levels by RANKL stimulation in osteoclast precursors. Further work is needed to elucidate the role of intracellular calcium using SC-19220 in mouse bone culture systems.
In summary, SC-19220, an EP1 antagonist, inhibited osteoclast formation induced by RANKL through inhibition of RANK and c-Fms expression in osteoclast precursors. In addition, SC-19220 was found to inhibit c-Src and NFAT2 expression in the RANK signaling network in a mouse cell culture system. Taken together, our results suggest that SC-19220 may be useful in the development of novel inhibitor reagents for treatment of osteoporotic disorders and inflammatory bone resorption.
The authors thank Searle & Co. for kindly supplying the SC-19220. This study was supported in part by Grant 12771129 for science research from Japan Society for the Promotion of Science.