Lipopolysaccharide Promotes the Survival of Osteoclasts Via Toll-Like Receptor 4, but Cytokine Production of Osteoclasts in Response to Lipopolysaccharide Is Different from That of Macrophages ¹

Lipopolysaccharide, the major component of the outer membrane of Gram-negative bacteria, induces the production of proinflammatory cytokines such as IL-1, TNF-α, and IL-6 in macrophages, lymphocytes, and endothelial cells (1, 2, 3). These cytokines activate immune systems to defend the host from bacterial infection (4, 5). Recent studies have provided new insights into the receptor system for LPS in the innate immune system (6, 7, 8, 9). CD14 is a membrane-anchored glycoprotein that functions as a member of the LPS receptor system (10). Recently, Toll-like receptor 4 (TLR4)³ was identified as the signal-transducing receptor for LPS (11, 12, 13). The binding of LPS to CD14 initiates signal transduction through TLR4, which results in the release of proinflammatory cytokines and the induction of the systemic inflammatory response.

Recent studies have also revealed that the signaling cascade of TLR4 is quite similar to that of IL-1Rs (6, 7, 8, 9). Both types of receptors use TNFR-associated factor 6 (TRAF6) as a common signaling molecule (14, 15). LPS and IL-1 induce the degradation of I-κB in the target cells to activate NF-κB (16). Then NF-κB translocates from the cytosol to the nucleus, and regulates the expression of the target genes that regulate immune and inflammatory responses. Recent studies also demonstrated that C3H/HeJ mice, which show extremely low responsiveness to LPS, have a point mutation in the intracellular domain of TLR4 (11, 12, 13). These results further support the notion that TLR4 is a signaling receptor responsible for LPS-induced inflammatory responses.

Osteoclasts, the multinucleated giant cells that resorb bone, develop from hemopoietic cells of the monocyte/macrophage lineage (17, 18, 19, 20). Osteoblasts or bone marrow stromal cells have been shown to be involved in osteoclastogenesis (20, 21). Studies of M-CSF-deficient op/op mice have shown that M-CSF produced by osteoblasts/stromal cells is an essential factor for osteoclastogenesis (22, 23). Recently, the gene for another essential factor for osteoclastogenesis, receptor activator of NF-κB ligand (RANKL), was cloned (24, 25, 26, 27). RANKL is a new member of the TNF ligand family, and is expressed by osteoblasts/stromal cells in response to many bone-resorbing factors. Osteoclast precursors express RANK, a TNFR family member, recognize RANKL expressed by osteoblasts/stromal cells, and differentiate into osteoclasts in the presence of M-CSF (28, 29). Osteoprotegerin (OPG), which is produced by many types of cells including osteoblasts/stromal cells, is a soluble decoy receptor for RANKL, and blocks osteoclastogenesis by inhibiting RANKL-RANK interaction (30, 31).

The cytoplasmic tail of RANK has been shown to interact with TRAF1, TRAF2, TRAF3, TRAF5, and TRAF6 (32, 33, 34). Among these TRAFs, TRAF6 appears to play important roles in osteoclast differentiation and function. TRAF6 knockout mice develop severe osteopetrosis (14, 15), and overexpression of TRAF6 in osteoclast progenitors induces their differentiation into osteoclasts (35). We previously reported that IL-1 stimulated the survival, fusion, and bone resorption activity of osteoclasts (36, 37), and that TNF-α directly stimulated the differentiation of osteoclasts (38). The stimulatory effects of TNF-α and IL-1 on osteoclast differentiation and function appear to be independent of the RANKL-RANK interaction, because these TNF-α- and IL-1-induced effects are not inhibited by the simultaneous addition of OPG (38). TNF type I receptor p55 (TNFR1) and type II receptor p75 (TNFR2) have been shown to use TRAF2 as a common signal transducer in the target cells (39, 40). These results suggest that TRAF-mediated signals play central roles in the regulation of osteoclast differentiation and function. Abu-Amer et al. (41) reported that LPS stimulated osteoclast formation in vivo and in vitro, and LPS-induced osteoclastogenesis is mediated by TNFR1.

In the present study, we explored the role of LPS in the function of osteoclasts in comparison with its role in bone marrow macrophages. We found that both osteoclasts and bone marrow macrophages expressed TLR4 and CD14 mRNAs, and induced degradation of I-κB in response to LPS. LPS strongly supported the survival of osteoclasts via TLR4-mediated signals. LPS enhanced the production of proinflammatory cytokines such as IL-1, TNF-α, and IL-6 in bone marrow macrophages and peritoneal macrophages. In contrast, the production of these cytokines in osteoclasts was not stimulated by LPS. Thus, osteoclasts respond to LPS through TLR4, but the responsiveness to LPS is quite different between osteoclasts and macrophages, the precursors of osteoclasts, in terms of proinflammatory cytokine production.

Five-week-old male and newborn ddY mice were obtained from Sankyo Laboratory Animal Center (Tokyo, Japan) and Saitama Experimental Animals (Saitama, Japan), respectively. TNFR1 knockout mice (C57BL/6J strain) were obtained from The Jackson Laboratory (Bar Harbor, ME). C3H/HeJ mice that have a point mutation in the TLR4 gene and normal control C3H/HeN mice were obtained from Sankyo Laboratory Animal Center. LPS was purified in our laboratory from Escherichia coli strain K235, as described previously (42). Human rM-CSF (Leukoprol) was obtained from Kyowa Hakko Kogyo (Tokyo, Japan). Anti-mouse CD14 goat polyclonal Abs and anti-mouse IL-1β goat polyclonal Abs were purchased from Santa Cruz Biotechnology (Santa Cruz, CA) and Genzyme Techne (Minneapolis, MN), respectively. Mouse IL-1R antagonist (IL-1Ra) was obtained from R&D Systems (Minneapolis, MN). ELISA kits for mouse IL-1β and TNF-α were obtained from Genzyme Techne, and that for mouse IL-6 was obtained from Endogen (Woburn, MA). ¹²⁵I-labeled human calcitonin (sp. act., 74 TBq/mmol) was purchased from Amersham (Buckinghamshire, U.K.). 1,25-Dihydroxyvitamin D₃ was purchased from Wako Pure Chemicals (Osaka, Japan). Recombinant mouse TNF-α, IL-1α, and IL-1β were obtained from Genzyme Techne. Pronase E was obtained from Calbiochem (La Jolla, CA).

Primary osteoblasts were prepared from calvariae of newborn ddY, C3H/HeJ, or C3H/HeN mice, as described (21, 43). Osteoblasts and freshly prepared bone marrow cells were cocultured in αMEM supplemented with 10% FBS (JRH Biosciences, Lenexa, KS) and 1,25-dihydroxyvitamin D₃ (10⁻⁸ M) in 100-mm-diameter dishes precoated with collagen gel, as described previously (43). Osteoclasts were formed within 6 days in the cocultures. All the cells in the cocultures were recovered from the dishes by treatment with αMEM containing 0.2% collagenase (43). The purity of osteoclasts in the crude preparation was ∼5%. To purify osteoclasts, the crude osteoclast preparation was plated in culture dishes (24-well dishes). After cells were cultured for 4 h, osteoblasts were removed by treatment with PBS containing 0.001% pronase E. Some cultures were then fixed and stained for tartrate-resistant acid phosphatase (TRAP, a marker enzyme of osteoclasts). The purity of osteoclasts in this preparation was ∼95% (36, 43). Purified osteoclasts were further incubated for the indicated periods in the presence or absence of IL-1, RANKL, TNF-α, or LPS, and stained for TRAP. TRAP-positive multinucleated cells (MNCs) with more than three nuclei were counted as living osteoclasts.

Bone marrow cells obtained from tibiae of 5- to 8-wk-old ddY mice were suspended in αMEM containing 10% FBS and cultured in 48-well plates (1.5 × 10⁵ cells/0.3 ml/well) in the presence of M-CSF (50 ng/ml) (38). After cells were cultured for 4 days, nonadherent cells were completely removed from the cultures by pipetting. Almost all of the adherent cells expressed macrophage-specific Ags such as Mac-1, Moma-2, and F4/80 (38). These macrophages were further cultured overnight with vehicle (control) or LPS (1 μg/ml). Peritoneal macrophages were obtained from peritoneal exudate cells by peritoneal lavage with cold αMEM medium 4 days after i.p. injection of thioglycolate (2 ml; Remel, Lenexa, KS). Peritoneal exudate cells were seeded at 1.5 × 10⁵ cells in 300 μl αMEM containing 10% FBS/well of 48-well plates, and were allowed to adhere 3 h at 37°C in 5% CO₂. Nonadherent cells were removed by washing in medium to provide cultures routinely comprising >95% adherent macrophages. Peritoneal macrophages were activated with 1 μg/ml LPS, as indicated below. Some cultures were subjected to immunostaining with anti-mouse IL-1β Abs. The concentrations of IL-1β, TNF-α, and IL-6 in the conditioned medium were determined using the respective ELISA kits.

For semiquantitative RT-PCR analysis, total cellular RNA was extracted from bone marrow macrophages and purified osteoclasts using TRIzol solution (Life Technologies, Grand Island, NY). First strand cDNA was synthesized from the total RNA with random primers and subjected to PCR amplification with EX Taq polymerase (Takara Biochemicals, Shiga, Japan) using the following specific PCR primers: mouse TLR2, 5′-TCGCTTTTTCCCAATCTCAC-3′ (forward, nt 743–762) and 5′-TGTAACGCAACAGCTTCAGG-3′ (reverse, nt 1123–1142); mouse TLR4, 5′-AATTCCTGCAGTGGGTCAAG-3′ (forward, nt 1178–1197) and 5′-AGGCGATACAATTCCACCTG-3′ (reverse, nt 2359–2378); mouse CD14, 5′-ACATCTTGAACCTCCGCAAC-3′ (forward, nt 454–473) and 5′-AGGGTTCCTATCCAGCCTGT-3′ (reverse, nt 934–953); mouse GAPDH, 5′-ACCACAGTCCATGCCATCAC-3′ (forward, nt 566–585) and 5′-TCCACCACCCTGTTGCTGTA-3′ (reverse, nt 998-1017). The PCR products were separated by electrophoresis on a 2% agarose gel, and visualized by ethidium bromide staining and UV light illumination. The sizes of PCR products for mouse TLR2, TLR4, CD14, and GAPDH are 400, 601, 500, and 452 bp, respectively.

For TRAP staining, cultures were fixed with PBS containing 10% Formalin and treated with methanol-acetone (50:50, v/v) for 1 min (43). Cells were then incubated in an acetate buffer (0.1 M sodium acetate, pH 5.0) containing naphthol AS-MX phosphate as a substrate and red violet LB as a stain in the presence of 50 mM sodium tartrate, as described (43). TRAP-positive cells appeared as dark-red cells. For immunohistochemical staining, cells were washed twice with PBS, fixed with cold methanol-acetone (50:50, v/v) for 10 min, and incubated for 90 min with anti-mouse CD14 Ab and anti-mouse IL-1β goat polyclonal Abs as the first Abs. The bound Abs were visualized using biotinylated second Abs, avidin-biotin-conjugated peroxidase, and an 3-amino-9-ethylcarbazole substrate kit (Histofine, Nichirei, Tokyo, Japan) (44). The positive cells appeared as dark-brown cells.

Osteoclast preparations obtained from mouse cocultures were plated on coverslips in 24-well plates, and incubated with 0.2 nM ¹²⁵I-labeled human calcitonin in αMEM containing 0.1% BSA for 1 h at room temperature (43, 45). The cells were then washed twice with PBS and fixed for 5 min in 0.1 M sodium cacodylate buffer (pH 7.4) containing 1% formaldehyde and 1% glutaraldehyde. The specimens were then subjected to immunocytochemistry for IL-1β. Thereafter, the coverslips were mounted on glass slides and dipped in NR-M2 emulsion (Konica Photo Film, Tokyo, Japan). They were then stored for 14 days in the dark and developed in Rendol (Fuji Photo Film, Tokyo, Japan). The expression of calcitonin receptors was detected by microscopic examination as a dense accumulation of grains due to ¹²⁵I-labeled calcitonin binding.

Bone marrow macrophages and purified osteoclasts were incubated for various periods in the presence of LPS (1 μg/ml). Cells were washed twice with ice-cold PBS and then resuspended in a lysis buffer (10 mM HEPES-NaOH (pH 8.0), 1.5 mM MgCl₂, 10 mM KCl, 200 mM sucrose, 1% Triton X-100, 0.5 mM DTT, 0.5 mM PMSF, 10 mg/ml leupeptin) (36). Cell lysates (20 μg of protein) were resolved by 10% SDS-PAGE and transferred to nitrocellulose membranes (Hybond ECL; Amersham Pharmacia Biotech, Little Chalfont, U.K.). The membranes were treated with 5% skim milk in PBS containing 0.5% Tween 20 (PBS-T), incubated with anti-I-κBα Abs (1/1000 dilution), and then incubated with HRP-conjugated anti-rabbit IgG. ECL reagents (Amersham Pharmacia Biotech) were used for detecting the immunoreactive bands (36).

We first examined whether osteoclasts express LPS receptors using the RT-PCR technique. Expression of TLR2, which appears to be involved in recognition of Gram-positive bacterial cell walls such as peptidoglycan and lipoproteins, was also examined. Bone marrow macrophages were prepared as osteoclast precursors. Both osteoclasts and bone marrow macrophages expressed TLR2, TLR4, and CD14 mRNAs (Fig. 1,A). Some reports have indicated that in contrast to macrophages, osteoclasts do not express CD14 protein in the cell membrane (46, 47). To examine the expression of CD14 protein in osteoclasts, immunostaining with anti-CD14 Abs was performed on purified osteoclasts formed in vitro and also on bone marrow macrophages as the positive control (Fig. 1,B). Osteoclasts weakly but detectably expressed CD14 protein. These results suggest that osteoclasts as well as bone marrow macrophages express TLR2, TLR4, and CD14. We previously reported that IL-1 rapidly and transiently induced activation of NF-κB in osteoclasts, concomitantly with the degradation of I-κBα (36). Treatment of osteoclasts and bone marrow macrophages with LPS for 30 min resulted in almost complete disappearance of I-κB in both types of cell lysate (Fig. 1,C). I-κB reappeared in osteoclasts and bone marrow macrophages after the cells were treated with LPS for 60 min (Fig. 1 C).

We previously reported that purified osteoclasts spontaneously died due to apoptosis, and cytokines such as RANKL, IL-1α, and TNF-α promoted the survival of purified osteoclasts (28, 37, 38). The number of purified osteoclasts was decreased in a time-dependent manner, and almost all the osteoclasts disappeared within 36 h (Fig. 2,A). LPS as well as RANKL significantly reduced the spontaneous apoptosis of osteoclasts (Fig. 2, A and B). Dose-response experiments showed that LPS at 100 ng/ml significantly stimulated the survival of osteoclasts (data not shown). LPS at 1 μg/ml also stimulated the survival of bone marrow macrophages. When bone marrow macrophages were further cultured in the absence of M-CSF, only 16.7 ± 1.0% (the mean ± SEM of four cultures) of bone marrow macrophages survived after culture for 48 h. LPS at 1 μg/ml increased the survival rate of bone marrow macrophages up to 46.3 ± 1.2%.

We next examined whether LPS stimulates the survival of osteoclasts through TLR4. Purified osteoclasts were prepared in cocultures of osteoblasts and bone marrow cells obtained from C3H/HeJ mice or from normal C3H/HeN mice (Fig. 3). IL-1α similarly supported the survival of osteoclasts derived from mice of both strains. However, LPS prolonged the survival of osteoclasts derived from C3H/HeN mice, but not from C3H/HeJ mice.

LPS stimulates the target cells to produce proinflammatory cytokines such as IL-1 and TNF-α, which have been shown to support the survival of osteoclasts (28, 37, 38). We therefore examined the possibility that the effect of LPS on the survival of osteoclasts is mediated by such cytokines (Fig. 4). In this experiment, we used IL-1β instead of IL-1α, because bone marrow macrophages produce large amounts of IL-1β in response to LPS (see Fig. 5). IL-1β as well as IL-1α markedly stimulated the survival of osteoclasts (Fig. 4,A). The survival of osteoclasts supported by LPS was not inhibited by adding mouse IL-1Ra, which strongly inhibited IL-1β-induced survival of osteoclasts (Fig. 4,A). OPG, a decoy receptor of RANKL, completely suppressed the survival of osteoclasts supported by RANKL, but had no inhibitory effect on LPS-prolonged survival of osteoclasts (Fig. 4,B). TNF-α failed to promote the survival of osteoclasts derived from TNFR1 knockout mice, but LPS and RANKL stimulated the survival of those osteoclasts to a similar extent (Fig. 4 C). These results suggest that LPS supports the survival of osteoclasts via TLR4, and that none of the cytokines examined (RANKL, IL-1, and TNF-α) was involved in LPS-enhanced survival of osteoclasts.

We next examined whether osteoclasts also produce proinflammatory cytokines in response to LPS, as do macrophages. Purified osteoclasts and bone marrow macrophages were treated with increasing concentrations of LPS, and the concentration of IL-1β in the conditioned medium was measured by ELISA (Fig. 5,A). Bone marrow macrophages released a large amount of IL-1β in the conditioned medium in response to LPS in dose- and time-dependent manners. In contrast, osteoclasts failed to produce a large amount of IL-1β in the presence or absence of LPS. Immunocytochemical studies confirmed that bone marrow macrophages produced IL-1β in response to LPS (Fig. 5,B). These results indicate that bone marrow macrophages lose the ability to produce IL-1β in response to LPS during their differentiation into osteoclasts. Expression of calcitonin receptors is believed to be the most reliable marker of osteoclast differentiation (18, 43). Using the technique of double detection of calcitonin receptors and IL-1β, we determined the stage of differentiation at which the cells stop producing IL-1β (Fig. 5,C). LPS-treated multinucleated osteoclasts that expressed calcitonin receptors were completely negative for IL-1β staining (Fig. 5,C). Even small mononuclear cells expressing calcitonin receptors were also negative for IL-1β production in response to LPS (Fig. 5 C, inset).

We finally examined effects of LPS on the production of TNF-α and IL-6 in macrophages and osteoclasts (Fig. 6). In this experiment, freshly isolated peritoneal macrophages as well as bone marrow macrophages were examined, because bone marrow macrophages had been treated with M-CSF in culture. Production of TNF-α and IL-6 in bone marrow macrophages and in peritoneal macrophages was markedly increased in response to LPS. Peritoneal macrophages also produced IL-1β in response to LPS (control, 3.5 ± 2.0 ng/ml; LPS, 79 ± 25 ng/ml, the mean ± SEM of four cultures). In contrast, the concentrations of TNF-α were very low in the conditioned medium of cultures treated with or without LPS after culture for 48 h (Fig. 6). Interestingly, osteoclasts spontaneously produced a large amount of IL-6 even in the absence of LPS. Treatment of osteoclasts with LPS again failed to stimulate IL-6 production. Thus, the responsiveness of macrophages to LPS was quite different from that of osteoclasts: macrophages produced proinflammatory cytokines in response to LPS, but osteoclasts did not.

We have shown in this study that osteoclasts as well as bone marrow macrophages, precursors of osteoclasts, express TLR4 and CD14. This receptor system for LPS appeared to be involved in LPS-induced signaling in osteoclasts. Degradation of I-κB was induced in osteoclasts as well as bone marrow macrophages in response to LPS within 30 min. The time course of changes in I-κB degradation in osteoclasts after LPS stimulation was very similar to that induced by IL-1. LPS stimulated the survival of osteoclasts derived from normal mice, but not that of osteoclasts from C3H/HeJ mice, which possess a missense mutation in the TLR4 gene. These results suggest that TLR4-mediated signals are essentially involved in osteoclast function supported by LPS.

We previously reported that cytokines such as IL-1, TNF-α, and RANKL enhanced the survival of osteoclasts (28, 37, 38). The survival of osteoclasts supported by LPS was not mediated by any of those cytokines. Moreover, LPS-supported survival of osteoclasts was not inhibited by adding either OPG, a decoy receptor of RANKL, or IL-1Ra. The survival of osteoclasts derived from TNFR1-deficient mice was also supported by the addition of LPS. M-CSF, which is mainly produced by osteoblasts/stromal cells, strongly stimulated the survival of osteoclasts (28, 48). Anti-c-Fms (M-CSF receptor) Abs did not inhibit the LPS-induced survival of osteoclasts (data not shown). These results suggest that LPS directly supports the survival of osteoclasts through a mechanism independent of the production of proinflammatory cytokines.

The signaling cascade of TLR4 is quite similar to that of IL-1Rs (6, 7, 8, 9). Both receptors have been shown to use TRAF6 and MyD88 as common signaling molecules (9). We previously reported that IL-1 induced activation of NF-κB and promoted the survival of osteoclasts (36). When osteoclasts were pretreated with antisense oligodeoxynucleotides to p65 and p55 of NF-κB, the expression of the respective mRNAs by osteoclasts was suppressed, and the IL-1-enhanced survival of osteoclasts was inhibited concomitantly (36). In addition, IL-1β as well as RANKL stimulated pit-forming activity of purified osteoclasts cultured on dentine slices. Suda et al. (49) recently reported that LPS induced the degradation of I-κB in prefusion mononuclear osteoclasts, and stimulated their fusion and pit-forming activity. It was also shown that mice deficient in TRAF6 developed osteopetrosis, and that TRAF6 deficiency resulted in defects in the signaling of not only IL-1, but also LPS (14, 15). These results suggest that, like IL-1, LPS directly stimulates osteoclast function through signals mediated by TRAF6 and NF-κB activation. Additional experiments using MyD88-defcient mice will provide more conclusive evidence for the direct action of LPS on macrophages and osteoclasts.

Mitogen-activated protein kinases (MAPK) and phosphatidylinositol 3 (PI-3) kinase have been implicated in the survival and function of cells. Miyazaki et al. (50) first reported that extracellular signal-regulated kinase (ERK) activity relegates the osteoclast apoptosis. Therefore, we examined effects of PD98059, a specific inhibitor of ERK activation, on the survival of osteoclasts in the presence and absence of LPS (1 μg/ml). PD98059 (10⁻⁶ M) failed to inhibit the survival of osteoclasts supported by LPS (data not shown). SB203580, a specific inhibitor of p38 MAPK, had no inhibitory effects on LPS-enhanced survival of osteoclasts (51). These results suggest that neither ERK-mediated signals nor p38 MAPK-mediated signals are involved in the LPS-induced survival of osteoclasts. In contrast, wortmannin (10⁻⁸ M), a specific inhibitor of PI-3 kinase, inhibited the survival of osteoclasts induced by LPS (data not shown). This suggests that the PI-3 kinase/Akt (protein kinase B) pathway regulates the survival of osteoclasts.

Wang et al. (52) reported that alendronate induces apoptosis of osteoclasts through up-regulation of Fas expression. RT-PCR analysis revealed that osteoclasts formed in mouse cocultures expressed low levels of Fas ligand and Fas mRNAs, and LPS added to the osteoclast cultures did not significantly influence the expression of Fas ligand and Fas expression (data not shown). These results suggest that LPS-induced survival of osteoclasts is not mediated by suppression of Fas or Fas ligand expression in osteoclasts. Mitochondria play key roles in apoptosis, a central step being the release of cytochrome c into the cytoplasm. The Bcl-2 family members are shown to regulate the release of cytochrome c (53). We previously reported that most osteoclasts died during the first 24 h after purification, but expression levels of Bcl-2 and Bcl-x_L remained unchanged in purified osteoclasts for 18 h (54). Further studies will elucidate the relationship between antiapoptotic signaling pathways and mitochondrial activities in osteoclasts.

Kikkawa et al. (55) reported that osteoclasts transiently produce TNF-α in response to LPS. In agreement with their finding, the concentration of TNF-α was increased in the culture medium of purified osteoclasts after treatment with LPS at 1 μg/ml for 2–12 h (control, 11 ± 1.5 pg/ml; LPS, 74 ± 23 pg/ml at 12 h, the mean ± SEM of four cultures), but returned to the basal level after 24–48 h (control, 10.5 ± 0.87 pg/ml; LPS, 11 ± 0.51 pg/ml at 48 h). The survival of osteoclasts derived from TNFR1-deficient mice was also supported by the addition of LPS. These results suggest that TNF-α production by osteoclasts is essentially different from that by macrophages, and TNF-α transiently released from osteoclasts is not involved in the LPS-supported survival of osteoclasts.

Bone marrow macrophages differentiate into osteoclasts in the presence of RANKL and M-CSF (Fig. 7). LPS stimulates the production of proinflammatory cytokines such as IL-1β, TNF-α, and IL-6 in bone marrow macrophages and in freshly isolated peritoneal macrophages, but not in osteoclasts. Interestingly, osteoclasts constitutively produced a high level of IL-6, which was unaffected by adding LPS (Fig. 7). This finding about the IL-6 production by osteoclasts is consistent with the findings of Roodman et al. (56), who showed that human osteoclasts expressed a relatively high level of IL-6, and osteoclasts in patients with Paget’s disease produced increased levels of IL-6 compared with normal subjects. These results suggest that the characteristics of bone marrow macrophages and osteoclasts are quite different from each other in terms of cytokine production in the presence or absence of LPS.

We previously examined chronological changes in the expression of osteoclast- and macrophage-associated phenotypes during the differentiation of osteoclast precursors into multinucleated osteoclasts (44). Calcitonin receptor-positive mononuclear cells of a small size, which appeared before the emergence of multinucleated osteoclasts, expressed macrophage-associated phenotypes such as nonspecific esterase, Mac-1, and Mac-2. Nonspecific esterase and Mac-1 in calcitonin receptor-positive cells disappeared during the differentiation of these cells into multinucleated osteoclasts (44). The present study showed that even small calcitonin receptor-positive mononuclear cells failed to produce IL-1β in response to LPS. These results suggest that osteoclast precursors quit cytokine production (except for IL-6 production) in response to LPS as soon as the differentiation pathway of osteoclast precursors is determined (Fig. 7). Loss of inflammatory responsiveness to LPS in osteoclasts must be requirement for performing essential roles in physiological bone turnover. Further studies will be necessary to elucidate the mechanism of regulation of proinflammatory cytokine production in macrophages and osteoclasts.