Reciprocal Role of ERK and Nf-κb Pathways in Survival and Activation of Osteoclasts

Osteoclasts, multinucleated giant cells responsible for bone resorption, are terminally differentiated cells with a short life span. Recent findings suggest that the survival of osteoclasts is precisely regulated through interactions with supporting cells as well as by the action of several cytokines ( Fuller et al. 1993; Jimi et al. 1995; Hughes et al. 1996; Selander et al. 1996; Kameda et al. 1997). However, the detailed molecular events implicated in these processes still remain elusive.

The small GTP-binding protein Ras is ubiquitously found in eukaryotic organisms and functions as an inducer of various intracellular signaling pathways including cell proliferation and differentiation ( Bourne et al. 1990; Bollag and McCormick 1991; Kaziro et al. 1991). This cascade can be activated by a variety of receptors ( Chambard et al. 1987; L'Allemain et al. 1991; Ahn et al. 1992) and leads to activation of mitogen-activated protein kinase (MAPK) cascades. Upon stimulation, Ras assumes its active GTP-bound form, and recruits the protein kinase Raf-1 to the plasma membrane, where the kinase can be activated ( Leevers et al. 1994; Stokoe et al. 1994). Raf-1 stimulates MAP kinase kinases (MEK1 and MEK2), which in turn activate ERK1 (p44MAPK) and ERK2 (p42MAPK) ( Kyriakis et al. 1992). Although the Ras/ERK pathway is reported to be an important component of cell survival signal in hematopoietic cells ( Kinoshita et al. 1997), its exact role in osteoclasts has not been characterized.

Nuclear factor kappa B (NF-κB) is a transcription factor that regulates the expression of many genes involved in immune and inflammatory responses ( Siebenlist et al. 1994; Baldwin 1996). Activation of NF-κB is controlled by sequential phosphorylation, ubiquitination, and degradation of its inhibitory subunit, IκB. A large multiprotein complex, the IκB kinase (IKK) signalsome, was found to contain a cytokine-inducible IκB kinase activity that phosphorylates IκB-α and IκB-β. The functional IKK complex contains IKK1 (IKKα), IKK2 (IKKβ), and IKKγ, and both IKK1 and IKK2 appear to make an essential contribution to IκB phosphorylation. Mercurio et al. 1997 reported that Ser177 and Ser181 to Ala mutant of IKK2 inhibited NF-κB activation induced by TNF-α, whereas Ser177 and Ser181 to Glu mutant induced constitutive activation of NF-κB. It was recently reported that mice deficient in both the p50 and p52 subunits of NF-κB fail to generate mature osteoclasts and exhibit severe osteopetrosis ( Franzoso et al. 1997; Iotsova et al. 1997). Jimi et al. 1998 also reported the involvement of NF-κB pathways in the survival of osteoclasts promoted by interleukin 1 (IL-1). These results suggest that the NF-κB pathway is important for both the differentiation of osteoclast precursors and the survival of mature osteoclasts.

We recently reported that adenoviral vectors are useful for modulating osteoclast function by transducing foreign genes into osteoclasts ( Tanaka et al. 1998). Using this system as well as specific inhibitors, we investigated the role of ERK pathways and NF-κB pathways in the survival and activation of osteoclasts, and found that the ERK activation is essential in maintaining osteoclast survival as is the NF-κB pathway for upregulation of bone-resorbing activity of osteoclasts.

Newborn ddY mice and 8-wk-old male ddY mice were purchased from Shizuoka Laboratories Animal Center. α-Modified minimum essential medium (αMEM) and DME were purchased from GIBCO BRL and Life Technologies Inc., and FBS was purchased from Cell Culture Laboratory. Bacterial collagenase and 1α,25-dihydroxyvitamin D₃ (1α,25(OH)₂D₃) were purchased from Wako Pure Chemical Co., and dispase from Godo Shusei Co. MEK inhibitor PD98059, anti–phospho-ERK and anti–IκB-α antibodies were purchased from New England Biolabs, Inc., anti–v-src antibody (clone 327) from Oncogene Research Products, anti–Flag antibody from Sigma Chemicals, and anti–NF-κB (p65) antibody from Santa Cruz Biotechnology. Anti–Ras, anti–MEK1, and anti–ERK antibodies were obtained from Transduction Laboratories. Other chemicals and reagents used in this study were of analytical grade.

Mouse primary osteoblastic cells were obtained from 1-d-old ddY mouse calvaria by enzymatic digestion, and bone marrow cells were from tibiae of 8-wk-old male ddY mice ( Takahashi et al. 1988). Osteoblastic cells (5 × 10⁵ cells/dish) were cocultured with bone marrow cells (10⁷ cells/dish) on 10-cm dishes or collagen gel-coated dishes in the presence of 10 nM 1α,25(OH)₂D₃ and 1 μM PGE₂ as previously reported ( Akatsu et al. 1992). For protein analysis, osteoclast-like cells (OCLs) were purified after a modification of the method originally reported by Tezuka et al. 1992. In brief, the crude osteoclast preparation placed on plastic dishes was washed with αMEM, and treated with 5 ml of αMEM containing 0.1% collagenase and 0.2% dispase for 10 min to remove osteoblastic cells. The purity of OCLs and their precursors was >90% at the final preparation ( Tanaka et al. 1995).

The recombinant adenovirus vector carrying either dominant negative Ras (Ser17 to Asn, AxRas^DN), constitutively active MEK1 (Ser218 and Ser222 to Glu, AxMEK^CA), dominant negative IKK2 (also called IKKβ) (Ser177 and Ser181 to Ala, AxIKK^DN), or constitutively active IKK2 (Ser177 and Ser181 to Glu, AxIKK^CA) gene under the control of the CAG (cytomegalovirus IE enhancer + chicken β-actin promoter + rabbit β-globin poly (A)⁺ signal) promoter was constructed by homologous recombination between the expression cosmid cassette and the parental virus genome in 293 cells as described ( Niwa et al. 1991; Miyake et al. 1996); the control virus (Ax1w1) was provided by Dr. Izumu Saito (The University of Tokyo, Bunkyo-ku, Tokyo, Japan). Titers of virus stocks were determined by endpoint cytopathic effect assay with the following modifications. 50 μl of DME/10%FBS was dispensed into each well of a 96-well tissue culture plate, and then eight rows of threefold serial dilutions of the virus starting from 10⁻⁴ dilutions were prepared. 3 × 10⁵ 293 cells in 50 μl of DME/10%FBS was added to each well. The plate was incubated at 37°C in 5% CO₂ in air, and 50 μl of DME/10%FBS was added to each well every 3 d. 12 d later, the endpoint of the cytopathic effect was determined by microscopy, and the 50% tissue culture infectious dose (TCID₅₀) was calculated. One TCID₅₀/ml approximately corresponds to one plaque forming unit (PFU)/ml ( Kanegae et al. 1994). The efficiency of infection is affected not only by the concentration of viruses and cells, but also by the ratio of viruses to cells, the multiplicity of infection (MOI). In this study, MOI is expressed as a measure of titer how many PFU are added to every cell. Infection of adenovirus vectors to OCLs was carried out following the method previously described ( Tanaka et al. 1998). In short, mouse cocultures on day 4, when OCLs began to appear, were incubated with a small amount of αMEM containing the recombinant adenoviruses for 1 h at 37°C at an indicated MOI. The cells were washed twice with PBS and further incubated with αMEM/10%FBS at 37°C. Experiments were performed 24 h after the infection.

All extraction procedures were performed at 4°C or on ice. Cells were washed with ice-cold PBS, and then lysed by adding TNE buffer (1% NP-40, 10 mM Tris-HCl, pH 7.8, 150 mM NaCl, 1 mM EDTA, 2 mM Na₃VO₄, 10 mM NaF, and 10 μg/ml aprotinin). The lysates were clarified by centrifugation at 15,000 rpm for 20 min. An equal amount (20 μg) of protein was subjected to 8% SDS-PAGE under a reducing condition, transferred electrophoretically onto a nitrocellulose membrane, and probed sequentially with an appropriate primary antibody followed by secondary antibodies coupled with HRP (Promega Corp.). Immunoreactive proteins were visualized using ECL Western blotting detection reagents (Amersham Co.) following the procedures recommended by the supplier. The blots were stripped by incubating for 20 min in stripping buffer (2% SDS, 100 mM 2-mercaptoethanol, 62.5 mM Tris-HCl, pH 6.7) at 50°C and reprobed by other antibodies.

Treated cell cultures were washed twice with ice-cold PBS and lysed in lysis buffer (20 mM Tris-HCl, pH 7.5, 150 mM NaCl, 1% Triton X-100, 12 mM β-glycerophosphate, 5 mM EGTA, 0.5% deoxycholate, 3 mM DTT, 10 mM NaF, 1 mM Na₃VO₄, 2 μM leupeptin, 20 μg/ml aprotinin, and 1 mM PMSF). After 30 min on ice, cell lysates were cleared by centrifugation at 12,000 g for 20 min. The protein concentration in each sample was quantified by the Bradford method, and immunoprecipitation was performed by incubating 200 μl of lysate with 2 μg of anti–ERK2 (p42) antibody (Santa Cruz Biotechnology) for 1 h, and then adding 20 μl of protein A–agarose. After incubation for 1 h at 4°C with end-over-end mixing, the immunocomplex was recovered by centrifugation and washed with washing buffer (20 mM Tris-HCl, pH 7.5, 150 mM NaCl, 5 mM EGTA, 2 mM DTT, and 1 mM PMSF) twice. Kinase activity was assayed for 20 min at 37°C in the presence of 6 μg substrate (MBP), 30 μM ATP, and 20 μCi γ-[³²P] ATP in 55 μl assay buffer (20 mM Tris-HCl, pH 7.5, and 20 mM MgCl₂). After completion of kinase assays, the protein was resolved by SDS-PAGE, and the gels were dried and subjected to autoradiography. The relative activity of ERK2 was quantified by measuring the radioactivity of phosphorus-32 incorporated into MBP.

DNA was prepared and analyzed by gel electrophoresis according to the method described previously ( Jimi et al. 1998). In brief, purified OCLs were lysed by incubating at 60°C overnight in a digestion buffer containing 150 mM NaCl, 25 mM EDTA, 100 μg/ml proteinase K, and 0.2% SDS. The DNA was extracted twice with phenol/chloroform/isoamylalcohol and once with chloroform, and precipitated in ethanol with 150 mM CH₃COONa, pH 5.2. The DNA was dissolved in TE buffer (10 mM Tris-HCl, pH 8.0, and 1 mM EDTA) and treated with 20 μg/ml RNase A. The procedure for DNA extraction and precipitation were repeated. 2 μg of DNA was separated by electrophoresis on a 1.5% agarose gel and visualized by ethidium bromide staining with UV light illumination.

Cells undergoing apoptosis were identified by means of the TdT-mediated dUTP-dioxigenin nick-end labeling (TUNEL) method, which specifically labels the 3′-hydroxyl terminal of DNA strand breaks. For the TUNEL procedure, all agents, including buffers, were part of a kit (apoptosis in situ detection kit; Wako Pure Chemical Co.); the staining procedure was carried out according to the manufacturer's recommendation. Negative controls included omission of TdT. Positive controls included treatment of the samples with DNase I. Apoptotic cells were recognized by their dark nuclear staining (TUNEL-positive), and nuclei of nonapoptotic cells were visualized by staining with methyl green.

The survival rate was measured as reported ( Jimi et al. 1998). OCLs were purified 24 h after the infection and some of the cultures were subjected to tartrate-resistant acid phosphatase (TRAP) staining. Cell viability/survival is expressed as morphologically intact TRAP-positive multinucleated cells. Other cultures were further incubated for the indicated times, and then the number of living OCLs was counted. The number of viable cells remaining at the different time points is shown as a percentage of the cells at time zero.

For immunofluorescence analysis, cells were plated on sterile FBS-coated glass coverslips and purified by treatment with αMEM containing 0.1% collagenase and 0.2% dispase. After purification, OCLs were preincubated for 20 min at 37°C in αMEM without FBS and further incubated in αMEM with IL-1α (10 ng/ml) for indicated times, fixed in PBS containing 4% paraformaldehyde for 15 min, permeabilized with 0.2% Triton X-100 in PBS for 5 min, and then blocked with 5% skim milk in PBS for 20 min at room temperature. Cells were sequentially stained with anti–p65 antibody followed by Cy5-conjugated anti-rabbit IgG (Amersham Co.) as previously reported ( Isshiki et al. 1998). The subcellular localization of Cy5-labeled p65 was analyzed by a cooled CCD camera (Roper Scientific).

Nuclear extracts were prepared as described previously ( Jimi et al. 1998). For samples of purified OCLs, cells were scraped off each dish and put in ice-cold PBS. The cells were washed in ice-cold PBS, lysed for 10 min in ice-cold hypotonic buffer (buffer A: 10 mM Hepes, pH 7.9, 1.5 mM MgCl₂, 10 mM KCl, 0.5 mM DTT, and 0.5 mM PMSF), and lysed for 10 min on ice in buffer A containing 0.1% NP-40. The lysates were centrifuged for 20 s at 12,000 g. The pelleted nuclei were resuspended in a high salt lysis buffer (20 mM Hepes, pH 7.9, 420 mM NaCl, 25% glycerol, 1.5 mM MgCl₂, 0.2 mM EDTA, 0.5 mM PMSF, and 0.5 mM DTT), incubated at 4°C for 20 min, supplemented with 5 vol of storage buffer (20 mM Hepes, pH 7.9, 100 mM KCl, 20% glycerol, 0.2 mM EDTA, 0.5 mM PMSF, and 0.5 mM DTT), and vortexed for 10 s. Lysed nuclei were centrifuged for 10 min at 12,000 g, and the resulting supernatants were referred to as nuclear extracts. Oligonucleotide probes containing consensus sequences for NF-κB was provided by Dr. Jun-Ichiro Inoue (The Institute of Medical Science, The University of Tokyo, Minato-ku, Tokyo, Japan). The DNA binding reaction was performed at room temperature in a volume of 20 μl, which contained the binding buffer (25 mM Tris-HCl, pH 7.5, 1.5 mM EDTA, 7.5% glycerol, 75 mM NaCl, 1.5 mM DTT, 0.3% NP-40, and 1 mg/ml BSA), 1 μg of poly (dI-dC), 10⁵ cpm of ³²P-labeled probe, and 8 μg of nuclear proteins. After incubation for 30 min, the samples were fractionated on a 4% polyacrylamide gel and visualized by exposing dried gel to film.

Functionally active OCLs formed in cocultures were obtained by a collagen gel culture system as described by Akatsu et al. 1992. In brief, primary osteoblastic cells and bone marrow cells were cocultured in the presence of 10 nM 1α,25(OH)₂D₃ and 1 μM PGE₂ on culture dishes (6 cm) coated with 0.2% collagen gel matrix. OCLs 24 h after the infection were recovered by digesting the collagen gel with 0.2% collagenase for 20 min at 37°C. The cells released from the dishes were collected by centrifugation at 250 g for 5 min and resuspended in 5 ml of αMEM containing 10% FBS. An aliquot of the crude OCL preparation was transferred onto dentine slices and cultured for an additional 12 h. After 12 h of incubation, the medium was removed and 1 M NH₄OH was added to the wells for 30 min. Adherent cells were removed from the dentine slices by ultrasonication, and the resorption pits were visualized by staining with 1% toluidine blue. The resorbed area was measured using an image analysis system (SYSTEM SUPPLY) linked to a light microscope (Nikon).

Each series of experiments was repeated at least three times. The results obtained from a typical experiment were expressed as the means ± SD. Significant differences were determined using factorial analysis of variance (ANOVA).

Recently, we have reported that a replication-deficient adenovirus vector that contains a reporter gene encoding β-galactosidase efficiently infected human OCLs derived from human giant cell tumors and mouse OCLs formed in vitro ( Tanaka et al. 1998). The proportion of β-gal–positive OCLs increased in accordance with the number of recombinant adenoviruses inoculated in the culture, and >85% of OCLs became strongly positive for β-gal activity at an MOI of 100, with no obvious morphological change or toxic effects. As shown in Fig. 1 A, the expression of Ras^DN was clearly induced in infected OCLs in an MOI-dependent manner. The bottom panel shows that an equal amount of protein was applied in each lane as determined by the Western blotting of c-Src. We next examined the effect of Ras^DN expression on the survival of OCLs. After the treatment with 0.1% collagenase and 0.2% dispase to remove osteoblastic cells, noninfected OCLs and OCLs infected with Ax1w1 or AxRas^DN at an MOI of 100 were further incubated in αMEM containing 10% FBS to determine their survival. After the removal of osteoblastic cells, ∼70% of OCLs infected with or without Ax1w1 died within 18 h ( Fig. 1D and Fig. E). In contrast, almost all the Ras^DN-expressed OCLs disappeared within 18 h ( Fig. 1D and Fig. E), indicating a critical role of Ras in the survival of osteoclasts.

We next examined the signaling pathways working downstream of Ras in osteoclasts. As shown in the top and middle panels of Fig. 1 C, overexpression of Ras^DN clearly inhibited the basal activity of ERK in OCLs and ERK activity was ∼90% downregulated in Ras^DN-expressed OCLs at an MOI of 100. To determine whether or not modulating ERK activity could affect the survival of OCLs, we used PD98059, a specific inhibitor of ERK activation that does not affect the activation of other MAPKs ( Alessi et al. 1995; Favata et al. 1998). PD98059 treatment severely diminished the survival of OCLs ( Fig. 1 D). Conversely, upregulation of ERK activity in OCLs using recombinant adenovirus encoding constitutively active MEK1 (AxMEK^CA) remarkably enhanced their survival ( Fig. 1, B–E).

To examine whether or not ERK is the main signaling pathway working downstream of Ras in osteoclasts, we investigated the survival of OCLs that were coinfected with AxRas^DN and AxMEK^CA. Coinfected OCLs expressed Ras^DN and MEK^CA at the similar level with the cells infected with either Ras^DN or MEK^CA ( Fig. 2 A), and the deteriorative effect of AxRas^DN on OCLs was completely rescued by coinfection with AxMEK^CA ( Fig. 2 B). Nor did the treatment with PD98059, which suppresses the intrinsic MEK activation, inhibit the survival of MEK^CA-expressed OCLs ( Fig. 2 B). We also found that the expression of Ras^DN did not alter the activity of JNK/SAPK or p38 in OCLs (data not shown). Taken together, these findings clearly demonstrate that the Ras/ERK pathway is an important component of survival signal in osteoclasts.

We next determined if the reduction of OCL survival by expression of Ras^DN was due to apoptosis. After purification, uninfected OCLs or OCLs infected with Ax1w1, AxRas^DN, or AxMEK^CA at an MOI of 100 were cultured for 12 or 24 h. The cells were harvested, and genomic DNA was isolated and resolved on 1.5% agarose gel. After 12 h of culturing, DNA fragmentation was hardly detectable in uninfected OCLs or OCLs infected with Ax1w1, whereas OCLs infected with AxRas^DN clearly showed DNA fragmentation ( Fig. 3 A). This fragmentation was observed even in uninfected OCLs or OCLs infected with Ax1w1 after 24 h of culture, and the expression of MEK^CA markedly inhibited the fragmentation ( Fig. 3 A). Almost all the nuclei of MEK^CA-expressed OCLs, 24 h after the purification, were stained uniformly with Hoechst 33258, which indicates that the cells were alive and the nuclei were intact, whereas expression of Ras^DN caused condensation of >80% of nuclei 12 h after purification ( Fig. 3 B). We further determined that ∼80% of the nuclei infected with AxRas^DN underwent apoptosis 12 h after the purification as shown in Fig. 3 C by TUNEL staining. In contrast, positive TUNEL staining was hardly observed in MEK^CA-expressed OCLs even 24 h after the purification ( Fig. 3 C). These findings suggest that Ras/ERK pathways contribute greatly to preventing apoptosis of osteoclasts.

IL-1α is a proinflammatory cytokine that strongly promotes the survival of osteoclasts by preventing their apoptosis ( Jimi et al. 1998). Therefore, we examined the effect of IL-1α treatment on ERK activity in OCLs. After purification, Ax1w1-infected OCLs and Ras^DN-expressed OCLs were preincubated for 20 min at 37°C in αMEM without FBS and further incubated in αMEM with IL-1α (10 ng/ml) for various periods. As shown in the top panel of Fig. 4 A, IL-1α stimulated a rapid and transient increase in ERK activity in OCLs. ERK activation in response to IL-1α was transient, reaching the highest level after 15–30 min of stimulation and returning to the basal level after 60 min. Stimulation of ERK activity was not accompanied by a change in the expression level of ERK protein ( Fig. 4 A, middle). As previously reported by Jimi et al. 1998, IL-1α treatment strongly promoted the survival of OCLs ( Fig. 4 B). Interestingly, the expression of Ras^DN in OCLs only partially inhibited the ERK stimulation induced by IL-1α, as shown in the top panel of Fig. 4 A, and partially suppressed the promotion of their survival induced by IL-1α ( Fig. 4 B).

Because activation of NF-κB was reported to be involved in the survival of osteoclasts promoted by IL-1α ( Jimi et al. 1998), we examined the effect of Ras^DN protein on the activation of NF-κB in response to IL-1α. NF-κB is present in the cytosol in an inactive state, complexed with the inhibitory IκB proteins. Activation occurs via phosphorylation of IκB-α at Ser 32 and 36, which promotes the degradation of IκB-α, resulting in the dissociation and nuclear translocation of active NF-κB ( Traenckner et al. 1995; Chen et al. 1996). As shown in the bottom panel of Fig. 4 A by Western blotting, IκB-α expression rapidly decreased upon IL-1α stimulation and returned to the basal level after 60 min of stimulation in Ax1w1-infected OCLs. The bottom panel of Fig. 4 A shows that Ras^DN expression in OCLs did not affect the kinetics of the degradation of IκB-α by IL-1α. The change in the localization of NF-κB in OCLs treated with IL-1α was also examined immunocytochemically using specific antibodies against the p65 subunit of NF-κB. Before stimulation, p65 was diffusely distributed throughout the cytoplasm. Consistent with the previous report ( Jimi et al. 1998), when purified Ax1w1-infected OCLs were treated with IL-1α for 30 min, p65 was detected in most nuclei of the cells. Overexpression of Ras^DN did not affect the kinetics of nuclear translocation of NF-κB in OCLs ( Fig. 4 D). Consistent with the results of the immunocytochemistry, electrophoretic mobility shift assay (EMSA) showed that NF-κB binding activity was not impaired in Ras^DN-overexpressed OCLs ( Fig. 4 C).

To further examine the role of NF-κB pathway in osteoclasts, we constructed adenovirus vectors encoding dominant negative IKK2 (IKK^DN) and constitutively active IKK2 (IKK^CA) with Flag tag, and investigated their effects on OCLs. Fig. 5 A demonstrates the clear induction of IKK^DN and IKK^CA in infected OCLs in an MOI-dependent manner. Fig. 5 B shows that the degradation of IκB-α induced by IL-1α treatment was strongly blocked by IKK^DN expression. Immunocytochemical studies showed that IKK^DN expression inhibited nuclear translocation of p65 in IL-1α–stimulated OCLs ( Fig. 5 D), and <25% of the nuclei of OCLs were positively stained for p65 even after 30 min of IL-1α stimulation. In contrast, >60% of the nuclei of AxIkk^CA-infected OCLs were positively stained without any stimulation ( Fig. 5 D). EMSA demonstrated that NF-κB activation was inhibited in IKK^DN-expressed OCLs treated with IL-1α. Conversely, slight constitutive NF-κB activation was observed in IKK^CA-expressed OCLs ( Fig. 5 C). As shown in Fig. 6, neither IKK^DN nor IKK^CA virus infection significantly changed the survival of OCLs with or without IL-1α stimulation.

We next examined the role of ERK pathways and NF-κB pathways on bone-resorbing activity of OCLs using the adenovirus vectors described above. This activity of OCLs was quantified by measuring the pit area formed on dentine slices as previously reported ( Tamura et al. 1993). To minimize the influence of the different survival periods on the pit formation by OCLs, we performed the pit assay after 12 h of culturing and the pit-forming activity is expressed as the area resorbed per osteoclast. As shown in Fig. 7 A, neither Ras^DN nor MEK^CA expression altered the pit-forming activity of OCLs, indicating that the ERK pathway is not involved in the bone-resorbing activity of the individual OCL. In contrast, OCL pit-forming activity was significantly inhibited by induction of IKK^DN, whereas overexpression of IKK^CA enhanced it. Because IL-1α strongly promotes not only the survival of osteoclasts, but also their bone resorbing activity ( Boyce et al. 1989), we examined the effect of IKK^DN expression on the bone-resorbing activity promoted by IL-1α. As shown in Fig. 7 B, IL-1α treatment strongly enhanced the OCL pit-forming activity, and inhibiting NF-κB activation by the IKK^DN virus significantly suppressed the pit-forming activity enhanced by IL-1α.

Because CSF-1 (also called macrophage colony-stimulating factor) has been reported to prolong the survival of OCLs without upregulating their bone resorbing activity, we examined the effect of CSF-1 treatment on ERK and NF-κB activity in OCLs. As shown in the top panel of Fig. 8 A, ERK activation in response to CSF-1 was transient, reaching the highest level after 2–15 min of stimulation and returning to the basal level after 60 min. Stimulation of ERK activity was not accompanied by a change in the expression level of ERK protein ( Fig. 8 A, middle). As previously reported ( Fuller et al. 1993; Jimi et al. 1995), CSF-1 treatment strongly promoted the survival of OCLs ( Fig. 8 C). On the other hand, CSF-1 did not alter the expression level of IκB-α ( Fig. 8 A, bottom) and activate NF-κB ( Fig. 8 B). In addition, OCL bone-resorbing activity remained unchanged despite CSF-1 stimulation ( Fig. 8 D). These results strongly support our findings that ERK and NF-κB pathways are involved in osteoclast survival and activation for bone resorption, respectively.

The Ras superfamily of GTPases is implicated in a number of cellular physiological processes including cell proliferation and differentiation, and plays important roles in the organization of cellular architecture and regulation of membrane dynamism in addition to cell cycle control ( Bourne et al. 1990; Bollag and McCormick 1991; Kaziro et al. 1991). Recently, Luckman et al. 1998 reported that the nitrogen-containing bisphosphonates alendronate, ibandronate, and risedronate, which cause apoptosis in osteoclasts and suppress their bone-resorbing activity, inhibit posttranslational modification (prenylation) of proteins with farnesyl or geranylgeranyl isoprenoid groups, including the Ras superfamily of GTPases. Further studies performed by Fisher et al. 1999 demonstrated that alendronate acts directly on osteoclasts and interferes with the generation of geranylgeranyl diphosphate by inhibiting enzymes in the cholesterol biosynthesis pathway, resulting in the suppression of osteoclast function. These findings suggest the critical role of the Ras superfamily in this function.

The pleiotropic functions of the Ras protein are now explained by its capability to activate multiple downstream pathways ( Koide et al. 1993; Moodie et al. 1993; Rodriguez-Viciana et al. 1994; Van Aelst et al. 1993, Van Aelst et al. 1994; Vojtek et al. 1993). One of the most important signaling cascades lying downstream of Ras is the MAPK cascade, which is activated via the Ras–Raf interaction. This interaction induces the membrane translocation and activation of Raf, and activated Raf phosphorylates tyrosine/threonine kinase MEK, which in turn phosphorylates ERK. In the present study, we demonstrated the essential role of Ras/ERK pathways in the survival of osteoclasts. First, adenovirus vector carrying a dominant negative mutant of Ras (AxRas^DN) suppressed ERK activity in OCLs and promoted the apoptosis of OCLs. A similar effect was observed using a MEK inhibitor, PD98059. Second, expression of constitutively active MEK1-induced sustained stimulation of ERK in OCLs and dramatically promoted their survival, and the apoptosis-inducing effect of AxRas^DN was completely rescued by coinfection with AxMEK^CA. Finally, IL-1α, which strongly promotes OCL survival, rapidly stimulates ERK activation in OCLs. It should be noted that overexpression of Ras^DN only partially inhibited the ERK activation and suppressed the prolonged survival induced by IL-1α, whereas PD98059 strongly eliminated the effect of IL-1α (data not shown), suggesting that IL-1α stimulates ERK activity via Ras-independent pathways as well. These results clearly demonstrate that ERK activation is indispensable for the survival of osteoclasts, and that Ras is at least partially involved in this process.

It remains elusive how ERK activation stimulates OCL survival. In fibroblasts, ERK activation is followed by phosphorylation of transcription factors and induction of immediate early gene expression. The dimeric transcription factor AP-1 (activator protein-1) is a major target of cell growth, differentiation, and stress signaling pathways ( Angel and Karin 1991; Karin and Hunter 1995; Whitmarsh and Davis 1996). AP-1 consists of various combinations of Fos and Jun family members that dimerize via a leucine zipper domain and bind to DNA via an adjacent basic region ( Landschulz et al. 1988; Gentz et al. 1989; Kouzarides and Ziff 1989). The Fos family consists of four gene products (c-Fos, FosB, Fra-1, and Fra-2), whereas the Jun family is made up of three gene products (c-Jun, JunB, and JunD). We also found that sustained stimulation of ERK in OCLs by MEK^CA expression as well as IL-1α stimulation increased c-Fos, Fra-2, c-Jun, and JunB expression (data not shown), suggesting that AP-1 activation may be involved in the survival signaling in osteoclasts. Further study will be necessary to elucidate the detail of osteoclast survival signaling.

Jimi et al. 1998 recently reported that the activation of NF-κB is involved in the survival of osteoclasts promoted by IL-1α. They clearly showed the nuclear translocation of NF-κB and degradation of IκB in OCLs in response to IL-1α treatment. The survival of OCLs supported by IL-1α was suppressed by pretreatment with proteasome inhibitors, which inhibits NF-κB activation by preventing IκB degradation, or with antisense oligonucleotides to NF-κB components ( Jimi et al. 1998). In contrast to their results, we were unable to show the effect of NF-κB on the survival of OCLs. Consistent with the previous report, IL-1α treatment induced rapid nuclear translocation of NF-κB in OCLs and degradation of IκB. However, overexpression of neither Ras^DN nor MEK^CA changed the localization of NF-κB in the presence or absence of IL-1α. In addition, either inhibition of NF-κB activity by overexpression of dominant negative IKK (IKK^DN) or sustained activation of NF-κB by transducing constitutively active IKK (IKK^CA) had little effect on the survival of OCLs with or without IL-1α stimulation. We cannot fully explain the reason for the discrepancy between our results and those of the previous report, but it is possible that proteasome inhibitors and antisense oligonucleotides may possess some undesirable effects on OCLs unrelated to NF-κB inhibition.

However, NF-κB has an important role in the bone-resorbing activity of osteoclasts because adenovirus vector-induced IKK^DN expression significantly suppressed the pit-forming activity of OCLs and IKK^CA expression enhanced it, although these two viruses had no effect on OCL survival. Members of the TNF receptor family and the IL-1 receptor associated either directly or indirectly with TNF receptor-associated factors (TRAFs), adaptor proteins that recruit and activate downstream signaling transducers ( Rothe et al. 1994). TRAF6 is a newly identified member of TRAF, which is not involved in TNF signaling but in IL-1 signaling ( Cao et al. 1996; Darnay et al. 1998; Galibert et al. 1998). Not only does TRAF6 activate NF-κB by signaling via NIK and IKK, but it also activates ERK in a Ras-independent manner ( Kashiwada et al. 1998). A recent study on TRAF6 knockout mice demonstrated the essential role of TRAF6 on the activation of osteoclasts ( Lomaga et al. 1999). Taken together, it is likely that IL-1α can promote survival and activity of osteoclasts simultaneously via ERK and NF-κB activation, both of which are mediated by TRAF6. Our present study proposes the reciprocal role of ERK and NF-κB pathways in the survival and activation of osteoclasts. Further investigation of these pathways in osteoclasts will give us new insight into the molecular mechanism regulating bone resorption.

R. Yamaguchi (Department of Orthopaedic Surgery, The University of Tokyo) provided expert technical assistance. We thank I. Saito, J. Miyazaki, and J.-I. Inoue (The Institute of Medical Science, The University of Tokyo) for gifts of the recombinant adenovirus vector system, the CAG promoter, and the NF-κB binding oligonucleotide, respectively.

This work was in part supported by a grant from the Japan Orthopaedic and Traumatology Foundation Inc. (No. 0089; Alcare Award) and Grants-in-Aid from the Ministry of Education, Science, Sports and Culture of Japan to S. Tanaka and from Bristol-Myers Squibb/Zimmer Unrestricted Research Grants and Health Science Research Grants from Ministry of Health and Welfare to K. Nakamura.