Further Insights in the Mechanisms of Interleukin-1β Stimulation of Osteoprotegerin in...

Further Insights in the Mechanisms of Interleukin-1� Stimulation ofOsteoprotegerin in Osteoblast-Like Cells

Cécile Lambert,1 Cécile Oury,2 Emmanuel Dejardin,3 Alain Chariot,2 Jacques Piette,3 Michel Malaise,1

Marie-Paule Merville,2 and Nathalie Franchimont1

ABSTRACT: The mechanisms of IL-1� stimulation of OPG were studied in more detail. Whereas p38 andERK activation was confirmed to be needed, NF-�B was not necessary for this regulation. We also found thatOPG production after IL-1� stimulation was not sufficient to block TRAIL-induced apoptosis in MG-63 cells.

Introduction: Osteoprotegerin (OPG) plays a key role in the regulation of bone resorption and is stimulatedby interleukin (IL)-1�. Herein, we defined the mechanisms of IL-1� stimulation of OPG focusing on thepotential involvement of MAPK and NF-�B. We also examined whether OPG production in response toIL-1� influences TRAIL-induced apoptosis in MG-63 cells.Materials and Methods: OPG mRNA levels in MG-63 cells were quantified by real-time RT-PCR and proteinlevels of OPG and IL-6 by ELISA. Cell viability was assessed using the methyltetrazidium salt (MTS)reduction assay. The role of the MAPK pathway was studied by both Western blotting and the use of specificchemical inhibitors. NF-�B function was studied using BAY 11-7085 and by siRNA transfection to inhibit p65synthesis. Transcription mechanisms were analyzed by transiently transfecting MG-63 cells with OPG pro-moter constructs. Post-transcriptional effects were examined by using cycloheximide and actinomycin D.Results: MG-63 cells treatment with IL-1� resulted in the phosphorylation of c-Jun NH2-terminal kinase(JNK), p38, and extracellular signal-regulated kinase (ERK). The use of the specific inhibitors showed that p38and ERK but not JNK were needed for IL-1�–induced OPG production. In contrast, NF-�B was not essentialfor IL-1� induction of OPG. We also showed a small transcriptional and a possible post-transcriptional ortranslational regulation of OPG by IL-1�. Exogenous OPG blocked TRAIL-induced apoptosis, but IL-1�induction of OPG did not influence TRAIL-induced cell death.Conclusions: IL-1� stimulates OPG production by mechanisms dependent on p38 and ERK. In contrast,NF-�B was not essential for this regulation. Although the relevance of IL-1� stimulation of OPG is still notfully understood, our data showed that IL-1� stimulation of OPG does not modify TRAIL-induced cell death.J Bone Miner Res 2007;22:1350–1361. Published online on May 14, 2007; doi: 10.1359/JBMR.070508

Key words: osteoprotegerin, interleukin-1�, NF-�B, osteoblast, apoptosis

INTRODUCTION

BONE RESORPTION IS mainly controlled by the RANK/RANKL/osteoprotegerin (OPG) pathway.(1) RANKL

is expressed by osteoblasts, T cells, and several tumor celltypes.(1) When bound to its receptor RANK, RANKL en-hances the development, the activity, and the survival ofosteoclasts.(2–5) RANKL/RANK interaction results in theactivation of several signaling cascades and the activation ofnuclear factors such as NF-�B or c-fos, which are needed

for osteoclastogenesis.(6–9) However, in the presence ofOPG, a decoy receptor of the TNF receptor family,RANKL can not bind RANK, and the development, acti-vation, and survival of osteoclasts are prevented.(10) There-fore, the local balance between RANKL and OPG regu-lates the level of bone resorption in physiological conditionsand in inflammatory diseases such as rheumatoid arthritisor in cancer-related bone loss.

Cytokines, local growth factors and hormones that con-trol bone resorption, generally act on RANKL and/or OPGsynthesis in osteoblasts and osteoblast-like cells to influenceosteoclastogenesis.(8,11) Glucocorticoids, for example, up-regulate RANKL and downregulate OPG synthesis in os-teoblasts.(12) Surprisingly, the pro-inflammatory cytokinesinterleukin (IL)-1 and TNF increase both RANKL and

Dr Franchimont served as a consultant for Eli Lilly and Co. andMSD and is an employee of Amgen, Europe. This work was notperformed at Amgen. The authors did not receive any materials orfunding from Amgen. All other authors state that they have noconflicts of interest.

1Department of Rheumatology, Center for Biomedical Integrative Genoproteomics (CBIG), University of Liège, Liège, Belgium;2Laboratory of Medical Chemistry and Human Genetics, Center for Biomedical Integrative Genoproteomics (CBIG), University ofLiège, Liège, Belgium; 3Virology and Immunology Unit, Center for Biomedical Integrative Genoproteomics (CBIG), University of Liège,Liège, Belgium.

JOURNAL OF BONE AND MINERAL RESEARCHVolume 22, Number 9, 2007Published online on May 14, 2007; doi: 10.1359/JBMR.070508© 2007 American Society for Bone and Mineral Research

1350

JO608522 1350 1361 September

OPG synthesis in osteoblasts.(13–15) The authors proposedthat OPG synthesis is a counter-regulation process as bothcytokines are commonly associated with an increase inbone destruction.(14) However, in the context of tumor cellsor even rheumatoid arthritis, OPG might also be a survivalfactor.(16–18) Indeed, OPG binds to TRAIL, but with alower affinity compared with RANKL, and subsequentlyprevents tumor cell apoptosis.(16) Whether this process ismostly happening using a high concentration of OPG orwhether it could also occur in physiological conditions isstill controversial.

We and others have shown that IL-1� stimulation ofOPG was dependent on the p38 MAPK pathway.(19,20)

However, the molecular mechanisms of OPG stimulationby IL-1� are not fully characterized. In particular, the roleof NF-�B, a nuclear factor classically activated in responseto IL-1� and necessary for osteoclastogenesis, has not yetbeen studied in OPG gene regulation.(6,21,22)

In this paper, we studied the mechanisms of IL-1� stimu-lation of OPG from gene transcription to protein release.We also examined the potential activation of MAPK andNF-�B, two pathways classically activated during inflamma-tory processes or after IL-1� stimulation. Additionally, todetermine whether the autocrine release of OPG in re-sponse to IL-1� was sufficient to prevent TRAIL-inducedMG-63 cell death, we compared the effects of both IL-1�-induced OPG release and exogenous OPG added to thecultures.

MATERIALS AND METHODS

Cell culture

MG-63 human osteogenic sarcoma cells (CRL-1427;American Type Culture Collection, Rockville, MD, USA)were maintained in MEM (Life Technologies, Grand Is-land, NY, USA) supplemented with L-glutamine (200 mM),penicillin (100 IU/ml), streptomycin (100 �g/ml), 10% FBS(Cambrex, Petit-Rechain, Belgium), and nonessentialamino acids 1% (Life Technologies) and cultured in a hu-midified incubator with 5% CO2 at 37°C. For OPG ELISAassays, RNA analysis, and Western blotting, MG-63 cellswere grown to confluence (∼50,000 cells/cm2) before serumdeprivation for 16–24 h. MG-63 cells were treated with theindicated drugs and further incubated for 1–24 h in serum-free condition. For all NF-�B experiments, cells were main-tained in 10% FBS-supplemented MEM medium to pre-vent changes in cell viability.(23) For transient transfectionexperiments, MG-63 cells were seeded at a density of300,000 cells per well in 6-well plates (VWR Scientific Prod-ucts, Leuven, Belgium) in 10% FBS-supplemented MEMmedium and transfected 24 h later in a serum-free MEMculture medium for 4 h. MG-63 cells were allowed to re-cover in 10% FBS-supplemented MEM medium for 24 h.

All illustrated experiments were performed at least threetimes. The number of observations eventually pooled fromindependent experiments is indicated in the figure legends.

Reagents

Human recombinant IL-1� (R&D Systems, Minneapolis,MN, USA) was dissolved in PBS (Cambrex) supplemented

with 10% FBS to obtain a 10 �g/ml stock solution. SB203580 (SB; 270-179-M005), a p38 kinase inhibitor, PD98059 (PD; 385-023-M005), a selective inhibitor of MAPK/extracellular signal-regulated kinase (MEK), and BAY 11-7085 (BAY; 270-220-M010), an NF-�B inhibitor, werepurchased from Alexis Biochemicals (Läufelfingen, Swit-zerland).(24–26) SP 600125 (SP; 129-56-6), a c-Jun NH2-terminal kinase (JNK) inhibitor, was purchased fromBiomol (Plymouth, PA, USA).(27) SB, PD, and SP weredissolved in dimethyl sulfoxide (DMSO) and BAY in etha-nol to obtain 20 mM stock solutions. Actinomycin D (ActD; 1410, 1 mg/ml stock solution in ethanol) and cyclohexi-mide (CHX; 20 mM aqueous stock solution) were pur-chased from Sigma-Aldrich (Bornem, Belgium). OPG (185-05; R&D Systems) and human TRAIL (TRAIL, 375-TEC-010; R&D Systems) were dissolved in PBS containing 10%FBS to 100 and 10 �g/ml, respectively. Human sRANKL(sRANKL; 310-01; Peprotech, London, UK) was dissolvedin PBS containing 1% FBS to 100 �g/ml. An equal amountof appropriate vehicles were added to the control culturesfor all experiments.

OPG and IL-6 ELISA

The amounts of human OPG (Dual set; R&D Systems)and human IL-6 (Cytoset; R&D Systems) produced byMG-63 cells in the culture medium were analyzed byELISA according to the manufacturer’s instructions. Plateswere read at 450 nm using an ELISA plate reader (iEMSReader MF; Labsystems, Helsinki, Finland). OPG and IL-6levels were corrected for cell viability when MAPK or NF-�B inhibitors were used because these inhibitors may influ-ence cell viability.(23,28) Results were expressed as the per-cent of levels observed in control cultures as indicated inthe legends of the figures.

MTS reduction assay

Cell viability after IL-1� treatment in the presence orabsence of inhibitors or TRAIL was assessed by reductionof methyltetrazolium salt (MTS) to the formazan product inviable cells (cellTiter 96Aqueous; Promega, Madison, WI,USA) as described previously.(29) Results were expressedas the percent of the viability measured in untreated cells.

Real-time RT-PCR

Total RNA was extracted using RNeasy columns fromQiagen (Valencia, CA, USA), according to the manufac-turer’s recommendations. After DNase I treatment (Roche,Vilvoorde, Belgium), RNA was eluted and quantified byspectrometry (Gene Quant; Pharmacia, Peapack, NJ,USA). Five hundred nanograms of RNA was reverse tran-scribed using the First Strand cDNA Synthesis kit for RT-PCR (AMV; Roche). Quantitative PCR was performedwith SYBR Green PCR mix buffer using Light Cycler PCRtechnology (Roche). The number of cycles was selected toallow linear amplification of the tested cDNA. The primersequences were as follows: OPG: 5�-CAGCGGCACAT-TGGAC-3� and 5�-CGTGCATTAGGCCCTT-3� ;GAPDH: 5�-GTCGGAGTCAACGGAT-3� and 5�-CCACGACGTACTCAGC-3�.

IL-1� STIMULATES OPG 1351

Protein extraction and Western blotting

Western blot analysis was performed as described previ-ously.(29) Briefly, proteins (20 �g) were fractionated byelectrophoresis on a 10% polyacrylamide gel and trans-ferred onto an Immobilon-P membrane (PVDF; MilliporeCorp., Bedford, MA, USA). Membranes were blocked withTris-buffered saline-Tween (20 mM Tris, pH 7.5, 500 mMNaCl, 0,2% Tween) plus 5% dry milk before incubation for3 h with the primary antibody. The rabbit polyclonal anti-p38 (9212; 1/1000), rabbit polyclonal phospho-specific anti-p38 (P-p38; 9211S; 1/1000), rabbit polyclonal anti-MEK1/2(9122; 1/1000), rabbit polyclonal phospho-specific anti-MEK1/2 (P-MEK; 9121; 1/1000), rabbit polyclonal phos-pho-specific anti-CREB (P-CREB; 9191S; 1/1000), and rab-bit polyclonal phospho-specific anti-MSK-1 (P-MSK-1;9595S; 1/1000) antibodies were purchased from Cell Signal-ing Technology (Beverly, MA, USA). Rabbit polyclonalanti-extracellular signal-regulated kinase (ERK) (sc-94;1/1000), mouse monoclonal phospho-specific anti-ERK (P-ERK; sc-7383; 1/1000), rabbit polyclonal anti-JNK (sc-571;1/1000), mouse monoclonal phospho-specific anti-JNK (P-JNK; sc-6254; 1/1000), rabbit polyclonal anti-c-Jun (sc-45;1/1000), goat polyclonal phospho-specific anti-c-Jun (sc-16312; 1/1000), rabbit polyclonal anti-p65 (sc-109; 1/1000),and mouse monoclonal anti-� tubulin (sc-8035; 1/1000) an-tibodies were purchased from Santa Cruz Biotechnology(Santa Cruz, CA, USA). Mouse monoclonal anti-� actinwas purchased from Sigma-Aldrich (A1804; 1/1000). Horse-radish peroxidase–linked anti-rabbit, anti-goat, or anti-mouse IgG antibodies (DAKO, Glostrup, Denmark) wereused as secondary antibodies. The reaction was revealedwith the enhanced chemiluminescence detection reagentaccording to the instructions of the manufacturer (ECL kit;Amersham Pharmacia Biotech).

Kinase assay

Endogenous I�B kinase (IKK) activity was studied inMG-63 cells by kinase assay as previously described.(30)

Briefly, MG-63 cells were pretreated with BAY 11-7085, anNF-�B inhibitor, for 1 h before treatment with IL-1� for 10min. Cells were subsequently lysed in the lysis buffer (25mM HEPES, pH 7.5, 150 mM NaCl, 1% Triton X-100, 10%glycerol, 5 mM EDTA, 2 mM dithiothreitol, 1 mMNa3VO4, 1 mM NaF, 25 mM �-glycerophosphate, and“Complete Protease Inhibitor” mixture; Roche MolecularBiochemicals). Cellular extracts were incubated with ananti-IKK� antibody (sc-8330; Santa Cruz) for 2 h at 4°C.Protein A-coupled Sepharose beads were added and incu-bated under gentle agitation at 4°C overnight. After threewashes in lysis buffer, the kinase activity was assayed using1 �g of GST-I�B� as a substrate.

Small interfering RNA

MG-63 cells were plated at a density of 100,000 cells in6-well plates (VWR Scientific Products). After 24 h, MG-63cells were switched to serum-free Opti-MEM medium (LifeTechnologies), and each well was transfected with 30 nMp65 small interfering RNA (siRNA; final concentration;OR-KIT2-SIRA; Eurogentec, Seraing, Belgium) or control

siRNA (OR-0030-neg10; Eurogentec) using DMRIE-C re-agent as recommended by the manufacturer. After 24 h,MG-63 cells were transferred to MEM media containing10% FBS and treated with IL-1� for 6 h. Levels of OPGand IL-6 were determined in the conditioned media, andcellular extracts were subjected to Western blot analysisfor p65.

Transient transfections

The OPG plasmids were kindly provided by Dr JEOnyia, Eli Lilly and Co. Research, Indianapolis, IN,USA.(31) The 0.9-, 0.4-, and 0.2-kb KpnI-BglII fragments ofthe OPG promoter were cloned into pGL3-basic (Pro-mega). To study the IL-1� effects on OPG promoter activ-ity, MG-63 cells were transiently transfected with 2 �g ofeither the OPG-promoter-luciferase constructs or thepGL3-basic, using Jet PEI transfection reagent (LucronBioproducts, De Pinte, Belgium) as recommended by themanufacturer. Co-transfection of pGL4.74 (hRluc/SV40)Luciferase Reporter Vector (Promega) was used to normal-ize for transfection efficiency. Luciferase activity was deter-mined with the Dual-Luciferase Reporter Assay system(Promega) as described by the manufacturer. To furtherdetermine the effects of p50/p65 subunits of NF-�B onOPG synthesis by IL-1�, we also transfected transientlyMG-63 cells with 1 �g p50 and 1 �g p65 expression vectorsor 2 �g of empty plasmid pMT2T (Promega) using Jet PEItransfection reagent following the manufacturer‘s recom-mendation.

Statistical analysis

Results were generally expressed as mean ± SE. Experi-mental data resulting from 2 × 2 factorial designs (stimula-tion and inhibition) were analyzed by the general linearmodel (GLM) to assess the effect of each factor and theirinteraction. For most variables studied, a log-transform wasapplied to the data before the GLM to normalize theirdistribution. Likewise on graphs, mean data were plottedon the log-scale together with their SE. For comparingmean values in nonfactorial experiments, one-wayANOVA was used followed by Scheffé’s simultaneous con-fidence intervals to maintain the nominal significance level.All results were considered to be significant at the 5% criti-cal level (p < 0.05). Statistical calculations were carried outusing SAS (version 9.1 for Windows).

RESULTS

Both p38 and ERK MAPK activation contribute toIL-1�–induced OPG release by MG-63 cells

To clarify the role of the MAPK pathways for OPGstimulation by IL-1� in MG-63 cells, we first examined theeffects of IL-1� on JNK phosphorylation. As shown in Fig.1A, IL-1� stimulated JNK phosphorylation after 15 and 30min. Treatment of MG-63 cells with IL-1� resulted in en-hanced both OPG mRNA (p < 0.0001) and protein (p <0.0001) levels (Figs. 1B and 1C). The significant increase ofboth OPG mRNA and protein levels (p < 0.0001 for bothparameters) was also observed during MEK inhibition

LAMBERT ET AL.1352

(Figs. 2B and 2C) and p38 inhibition experiments (Figs. 3Band 3C). However, SP 600125 (SP) did not significantlymodify OPG mRNA levels in control or IL-1� treated cul-tures (Fig. 1B), whereas it decreased OPG protein levels inthe absence of IL-1� (Fig. 1C). As previously described, SPdecreased c-Jun phosphorylation confirming that SP effec-tively inhibits the JNK pathway in MG-63 cells (Fig.1D).(27) Altogether, these data suggest that JNK phos-phorylation is not needed for IL-1� stimulation of OPG,although JNK is phosphorylated after IL-1� stimulation.On the other hand, IL-1� stimulation also resulted in thephosphorylation of MEK1 and ERK after 5 min (Fig. 2A).Treatment with the MEK inhibitor PD 98,059 (PD) did notlead to a significant change of the OPG transcript but PDsignificantly inhibited OPG release (p � 0.0081) when MG-63 cells were treated by IL-1� (Figs. 2B and 2C). PD did notsignificantly affect OPG protein production in untreatedMG-63 cells (Fig. 2C). We also confirmed that PD de-

creases ERK phosphorylation in MG-63 cells (Fig. 2D). p38is rapidly activated in IL-1�–treated MG-63 cells with amaximum phosphorylation observed after 15 min, followedby a decrease after 30 min (Fig. 3A). We next examined theeffects of the p38 inhibitor SB203580 (SB) on both OPGtranscripts and OPG proteins. SB (20 �M) significantly re-duced OPG mRNA (p � 0.0016) and protein levels (p <0.0001) in both untreated and IL-1�–treated cells. In IL-1�–stimulated cultures, SB inhibited OPG mRNA and pro-tein levels by 60% and 70%, respectively (Figs. 3B and 3C).

To confirm that selected downstream proteins of p38were effectively inhibited in the presence of SB, we studiedthe phosphorylation of MSK-1, CREB, and ATF-1 byWestern blot analyses. As shown in Figs. 3D and 3E, phos-phorylation of MSK-1, CREB, and ATF-1 appeared after30 min of IL-1� treatment in MG-63 cells. Pretreatment ofMG-63 cells with SB for 1 h resulted in a decrease in thephosphorylation of MSK-1, CREB, and ATF-1 after IL-1�

FIG. 1. Role of JNK in OPG production by MG-63 cells. (A) JNK is phosphorylated after IL-1� stimulation. MG-63 cells were treatedwith control media or IL-1� (10 ng/ml) for 0–360 min. Total protein extracts from control or treated cultures were analyzed by Westernblotting with anti-P-JNK, anti-JNK, and anti-� tubulin antibodies. (B) Effects of SP 600125 (SP) on OPG transcripts. MG-63 cells werepretreated with SP at 20 �M for 1 h before treatment with vehicle or IL-1� (10 ng/ml) for 24 h. OPG mRNA levels were quantifiedby real-time RT-PCR. The data were normalized using GAPDH mRNA. The OPG mRNA in control cells was arbitrarily set to 100%,and data are expressed as mean percentage of control levels ± SE (n � 3–6 observations pooled from two independent experiments).aSignificant effect of IL-1�. (C) Effects of SP on OPG release. MG-63 cells were pretreated with SP at 20 �M for 1 h before treatmentwith vehicle or IL-1� (10 ng/ml) for 24 h. Cell culture media were collected, and OPG levels were measured by ELISA. OPG levelswere corrected for cell viability assessed by a colorimetric proliferation assay, and results are expressed as the percent of levels observedin control cultures. Values are means ± SE (n � 23–24 observations pooled from six independent experiments). aSignificant effect ofIL-1�. bSignificant effect of SP. (D) P-c-Jun is inhibited after treatment with SP. MG-63 cells were pretreated with SP at 20 �M for 1h before treatment with vehicle or IL-1� (10 ng/ml) for 30 min. Total protein extracts from control or treated cultures were analyzedby Western blotting with anti-P-c-Jun and anti-c-Jun antibodies.

IL-1� STIMULATES OPG 1353

treatment, confirming that SB inhibited downstream pro-teins of the p38 pathway (Figs. 3D and 3E). These datasuggest that p38 and ERK signaling pathways are neededfor IL-1�–induced OPG release by MG-63 cells, whereasJNK pathway activation seems not to be involved.

NF-�B is not essential for IL-1� induction of OPG

Because NF-�B is a nuclear factor classically activated byIL-1�, we studied whether IL-1� stimulation of OPG couldbe dependent on NF-�B activation.(32) To analyze the roleof NF-�B, we used a chemical inhibitor of I�B� phos-phorylation, BAY 11-7085 (BAY) in the presence of 10%FBS to avoid changes in cellular viability.(26) This inhibitorwas evaluated at 5 �M against IKK activity in IL-1�–treated MG-63 cells using a kinase assay. As expected, theIKK activity measured in IL-1�–treated cells was reducedby BAY (5 �M; Fig. 4A). In the presence of serum, IL-1�stimulated OPG mRNA (p < 0.0001) and protein levels(p < 0.0001) by 2- to 4-fold (Figs. 4B and 4C). We alsofound that BAY (5 �M) significantly decreased OPG tran-script in the absence of IL-1� (p < 0.0001), whereas thiseffect was not observed after IL-1� treatment (Fig. 4B). Nosignificant effect of BAY on OPG release was observed,although a 40% decrease in OPG release was measured inthe presence of IL-1� (result not statistically significant,

p � 0.3779; Fig. 4C). No dose–response relationship couldbe assessed with BAY because concentrations >5 �M de-creased MG-63 cell viability (data not shown). To furtherstudy possible NF-�B involvement in IL-1� stimulation ofOPG gene transcription, we used the technique of RNAinterference. MG-63 cells were transiently transfected withhuman p65 siRNA or control siRNA. As shown in Fig. 4D,p65 siRNA transfection results in a p65 protein level de-crease as shown by Western blotting, whereas controlsiRNA did not change p65 expression level. Transfection ofp65 siRNA in MG-63 cells did not modify OPG proteinlevels in the presence of IL-1� (Fig. 4E). In contrast, IL-6protein levels were significantly decreased (p < 0.0001) (Fig.4F). These data suggested that NF-�B was not involved inIL-1�–induced OPG production by MG-63 cells. To con-firm these results, we transiently transfected MG-63 cellswith p50 and p65 expression vectors (1 �g of each) or withthe control plasmid pMT2T (2 �g). The efficiency of trans-fection was monitored by Western blotting. Co-transfectionof p50 and p65 expression plasmids in MG-63 cells resultedin a p65 protein level increase (Fig. 5A). However, no sig-nificant change in OPG protein level was observed (Fig.5B). These two approaches show that NF-�B is not neededfor OPG production in MG-63 cells treated with IL-1�.

FIG. 2. Role of ERK in OPG productionby MG-63 cells. (A) MEK and ERK arephosphorylated after IL-1� stimulation. MG-63 cells were treated with control media orIL-1� (10 ng/ml) for 0–360 min. Total proteinextracts from control or treated cultureswere analyzed by Western blotting with anti-P-MEK, anti-MEK, anti-P-ERK, anti-ERK,and anti-� tubulin antibodies. (B) Effects ofPD 98,059 (PD) on OPG transcripts. MG-63cells were pretreated with PD at 50 �M for 1h before treatment with vehicle or IL-1� (10ng/ml) for 24 h. OPG mRNA levels werequantified by real-time RT-PCR. Values aremeans ± SE (n � 3 from one representativeexperiment). aSignificant effect of IL-1�. (C)Effects of PD on OPG release. MG-63 cellswere pretreated with PD at 50 �M for 1 hbefore treatment with vehicle or IL-1� (10ng/ml) for 24 h. OPG levels were measuredby ELISA. Values are means ± SE (n � 29–30 observations pooled from six independentexperiments). aSignificant effect of IL-1�.bSignificant effect of PD. (D) P-ERK is in-hibited after treatment with PD. MG-63 cellswere pretreated with PD at 50 �M for 1 hbefore treatment with vehicle or IL-1� (10ng/ml) for 30 min. Total protein extractsfrom control or treated cultures were ana-lyzed by Western blotting with anti-P-ERKand anti-� tubulin antibodies.

LAMBERT ET AL.1354

Mechanisms of IL-1� stimulation ofOPG expression

To better understand the mechanisms by which IL-1�regulates OPG expression, we studied IL-1� effects onOPG gene promoter activity. In this context, we performedtransient transfection of MG-63 cells with several OPG pro-moter constructs (from 0.9 to 0.2 kb) fused to the luciferasereporter gene. We analyzed OPG promoter ability to driveluciferase expression in the presence or absence of IL-1�(Fig. 6A). After 6 h, IL-1� significantly reduced the OPGpromoter activity in the pGL3-Basic control vector (p <0.0001), whereas it induced a 40% increase in 0.9-kb OPGpromoter activity (p < 0.0001) and an increase of 12% in0.4-kb OPG promoter activity (p � 0.0007). No significantstimulatory effect of IL-1� was observed with the 0.2-kbOPG construct. No additional stimulatory effect was ob-

served after 16 and 24 h (data not shown). Because IL-1�weakly affected OPG promoter activity, we next examinedwhether IL-1� influences OPG transcripts in a post-transcriptional manner. For this purpose, MG-63 cells weretreated with IL-1� for 6 h in the presence or absence of theprotein synthesis inhibitor CHX. As shown in Fig. 6B, treat-ment with IL-1� resulted in a significant increase in OPGmRNA synthesis (p < 0.0001). Moreover, addition of 20 �MCHX significantly stimulates OPG mRNA levels in bothcontrol and IL-1�–treated cells (p < 0.0001). These datasuggest that blockage of protein translation strongly in-creases the levels of OPG transcript. We next studied OPGmRNA decay by treating MG-63 cells with 5 �g/ml of ActD. OPG mRNA levels corrected for GAPDH mRNA lev-els were significantly decreased after 10 and 16 h of Act Dtreatment in nonstimulated cells (p � 0.0024). The de-crease of OPG transcripts by Act D was abolished in the

FIG. 3. Role of p38 in OPG production by MG-63 cells. (A) p38 is phosphorylated after IL-1� stimulation. MG-63 cells were treatedwith control media or IL-1� (10 ng/ml) for 0–360 min. Total protein extracts from control or treated cultures were analyzed by Westernblotting with anti-P-p38, anti-p38, and anti-� tubulin antibodies. (B) Effects of SB 203580 (SB) on OPG transcripts. MG-63 cells werepretreated with SB at 20 �M for 1 h before treatment with vehicle or IL-1� (10 ng/ml) for 24 h. OPG mRNA levels were quantifiedby real-time RT-PCR. Values are means ± SE (n � 5–6 observations pooled from two independent experiments). aSignificant effectof IL-1�. bSignificant effect of SB. (C) Effects of PD on OPG release. MG-63 cells were pretreated with SB at 20 �M for 1 h beforetreatment with vehicle or IL-1� (10 ng/ml) for 24 h. OPG levels were measured by ELISA. Values are means ± SE (n � 23–24observations pooled from six independent experiments). aSignificant effect of IL-1�. bSignificant effect of SB. (D) SB inhibits MSK-1phosphorylation induced by IL-1�. MG-63 cells were pretreated with SB at 20 �M for 1 h before treatment with vehicle or IL-1� (10ng/ml) for 30 min. Total protein extracts from control or treated cultures were analyzed by Western blotting with anti-P-MSK-1 andanti-� tubulin antibodies. (E) SB inhibits CREB phosphorylation induced by IL-1�. MG-63 cells were pretreated with SB at 20 �M for1 h before treatment with vehicle or IL-1� (10 ng/ml) for 30 min. Total protein extracts from control or treated cultures were analyzedby Western blotting with anti-P-CREB and anti-� tubulin antibodies.

IL-1� STIMULATES OPG 1355

presence of IL-1� (p � 0.009; Fig. 6C). This indicates thatIL-1� treatment increases the half-life of OPG transcripts.

IL-1� treatment does not modify TRAIL-inducedapoptosis of MG-63 cells

As shown in Table 1, treatment with TRAIL at dosesranging from 10 to 100 ng/ml significantly decreased MG-63cell viability after 24 h (p < 0.0001). We confirmed thatMG-63 cell death occurred by apoptosis by using annexin V

and propidium iodide labeling (data not shown). Overnight(16 h) pretreatment with IL-1� did not modify MG-63 cellviability in response to TRAIL, even though the level ofOPG was 404 ± 6 (SE) ng/ml (n � 5) after IL-1� stimula-tion in the absence of TRAIL (Table 1). When exogenoussRANKL (50–500 ng/ml) was added to the cell culture con-comitantly to TRAIL (10 ng/ml), a marginal increase in celldeath was observed (p � 0.0352; Table 2). Pretreatmentwith sRANKL 1–16 h before TRAIL treatment did not

FIG. 4. NF-�B is not essential for IL-1� induction of OPG by MG-63 cells. (A) IKK activity after IL-1� treatment. MG-63 cells werepretreated with 5 �M BAY for 1 h before treatment with vehicle or IL-1� (10 ng/ml) for 10 min. IKK complex was immunoprecipitated(IP) with an anti-IKK� antibody, and kinase assay was performed using a GST-I�B� substrate. The presence of IKK� in the extractsbefore the immunoprecipitation was analyzed by Western blotting. Phosphorylated I�B� was detected with an anti-P-I�B�. (B) OPGmRNA levels in MG-63 cells pretreated with BAY 11–7085 (BAY) at 5 �M for 1 h before treatment with vehicle or IL-1� (10 ng/ml)for 24 h. OPG mRNA levels were quantified by real-time RT-PCR. Values are means ± SE (n � 3 from one representativeexperiment). aSignificant effect of IL-1�. bSignificant effect of BAY. (C) OPG protein levels in supernatants of MG-63 cells pretreatedwith BAY at 5 �M for 1 h before treatment with vehicle or IL-1� (10 ng/ml) for 24 h. OPG levels measured by ELISA. Values aremeans ± SE (n � 8–9 observations pooled from three independent experiments). aSignificant effect of IL-1�. (D) p65 protein levelsin MG-63 cells transfected with p65 siRNA or control siRNA and treated for 6 h with vehicle or IL-1� (10 ng/ml). Total cellular extractsfrom control or treated cells were subjected to Western blotting. The basal level of p65 proteins in untransfected cells (−) wasdetermined in parallel. Human p65 was detected with an anti-p65 antibody. Western blotting for � actin was performed in parallel. (E)Effects of p65 siRNA on OPG protein levels in IL-1�–treated cells. MG-63 cells transfected with p65 siRNA or control siRNA weretreated for 6 h with IL-1� (10 ng/ml). Values are the mean ± SE and are expressed as the percent of IL-1�–treated cells transfectedwith siRNA negative control (n � 6 from one representative experiment). (F) Effects of p65 siRNA on IL-6 levels in identicalconditions as in E. aSignificant effect of p65 siRNA.

LAMBERT ET AL.1356

modify this observation (data not shown). The basal pro-duction of OPG in the same experiment was 59 ± 4 ng/ml(n � 12). However, exogenous OPG blocked TRAIL-induced cell death (p � 0.0002; Table 3). Altogether, thesedata show that exogenous OPG blocked TRAIL-inducedapoptosis but suggest that IL-1� induction of OPG does notinfluence TRAIL-induced cell death in MG-63 cells.

DISCUSSION

In this paper, we characterized part of the signaling path-ways involved in IL-1�–induced OPG production. IL-1�has long been known to induce bone-resorbing activity invitro, and several papers have underlined the role of IL-1 inbone resorption associated with osteoporosis, rheumatoidarthritis, and myeloma.(33–36) It is therefore not surprisingto observe that IL-1� induces RANKL expression in osteo-blasts.(13) What is more difficult to understand is the stimu-latory effect of IL-1� on OPG. We and others have con-firmed that IL-1� increases both OPG protein release andmRNA levels.(13,15,19,20) In this work, we focused our at-tention on two signaling pathways classically involved ininflammation: the MAPK and NF-�B pathways. We con-firmed that p38 is needed for both OPG protein andmRNA synthesis in IL-1�–treated MG-63 cells (Figs 3A–3C).(19,20) We also showed that PD 98059, an inhibitor ofMEK, decreased OPG protein levels after IL-1� stimula-tion (Fig. 2C). However, this inhibitor did not reduce IL-1�stimulated OPG mRNA levels (Fig. 2B). Therefore, ERKcould be considered as an important kinase for the extra-cellular release of OPG. Whereas JNK was phosphorylatedafter IL-1� stimulation in MG-63 cells (Fig. 1A), no inhibi-tory effect of SP on OPG transcripts and protein levels wasfound in the presence of IL-1� (Figs. 1B and 1C). We con-firmed the validity of SP by Western blot analysis of c-Junphosphorylation in the same conditions of culture (Fig. 1D).Altogether, these data confirmed that, among the MAPKtested in this work, p38 and ERK, but not JNK, are neces-sary for OPG release (Fig. 7).

There is a strong rationale for studying the involvementof NF-�B in IL-1� induction of OPG. First, IL-1� is known

to induce the phosphorylation of I�B�, which leads to itsdegradation and to p50 and p65 translocation into thenucleus (Fig. 7).(37) Second, there is an interaction betweenthe MAPK pathways and the p65 subunit of NF-�B. Forexample, MSK-1 was identified as a nuclear kinase for p65,because MSK-1 associates with p65 in response to TNF andphosphorylates p65 on Ser276.(38) Third, NF-�B is criticalfor osteoclastogenesis as shown by the osteopetrotic phe-notype of the p50/p52 double knockout mice.(6) Last, acti-vation of NF-�B is involved in the survival of osteoclastspromoted by IL-1.(39) Although we confirmed that IL-1�increased I�B� phosphorylation (Fig. 4A) and enhancesNF-�B binding activity in MG-63 cells (data not shown),BAY, a chemical inhibitor of I�B� phosphorylation, failedto decrease OPG mRNA and protein levels in the presenceof IL-1� (Figs. 4B and 4C). We also tested the effect of p65siRNA on OPG protein release (Fig. 4E) and showed thatOPG levels, in contrast to IL-6 levels (Fig. 4F), were notmodified by transfecting p65 siRNA in MG-63 cells (Figs.4E and 4F). In addition, transfection of MG-63 cells withp50 and p65 expressing vector did not cause further releaseof OPG (Figs. 5A and 5B), indicating that NF-�B is notessential for IL-1�–induced production of OPG in MG-63cells. This is in contrast with the requirement of NF-�B, butalso JNK, for RANKL/RANK signaling activation in os-teoclast precursors and osteoclasts.(8)

We next studied the molecular mechanisms by which IL-1� stimulates OPG gene transcription. The effects of IL-1�on OPG gene promoter activity were modest as determinedby transient transfection of MG-63 cells with a 0.9-kb frag-ment of this promoter (Fig. 6A). This fragment of the pro-moter contained several binding sites including cbfa1 andactivator protein 1 (AP-1) elements.(31) IL-1� stimulatoryeffect on OPG promoter activity remains statistically sig-nificant when transfecting MG-63 cells with the 0.4-kb frag-ment of OPG promoter (Fig. 6A). No effect of IL-1� wasobserved when the 0.2-kb fragment of OPG promoter wasused (Fig. 6A). The small effect of IL-1� on OPG promoteractivity, the discordance between the intensity of stimula-tion of IL-1� on OPG mRNA levels (1.5- to 3-fold) and onOPG protein release (4- to 20-fold), and the additive stimu-

FIG. 5. OPG production is not influenced by p50/p65 transfection. MG-63 cells were transfected with 1 �g of p50 and 1 �g of p65expression vectors or empty plasmid (pMT2T) and treated for 6 h with IL-1� (10 ng/ml). Culture media were collected for OPG, andtotal lysates were subjected to Western blotting. (A) Western blotting of p65 levels in MG-63 cells transfected with pMT2T or with p50and p65 expression plasmids. Human p65 was detected by an anti-p65 antibody. Western blotting for � actin was performed in parallel.(B) OPG levels measured by ELISA. Values are means ± SE and are expressed as the percent of cells transfected with pMT2T (n �6 from one representative experiment).

IL-1� STIMULATES OPG 1357

latory effect of cycloheximide on OPG mRNA levels (Fig.6B) indicated a post-transcriptional or translational addi-tional mechanism. A post-transcriptional effect was furthersupported by the IL-1�–mediated OPG mRNA stabiliza-tion over time in the presence of actinomycin D (Fig. 6C).Interestingly, p38 is known to stabilize other genes such asIL-3, IL-6, IL-8, c-fos, granulocyte macrophage-colonystimulating factor (GM-CSF), TNF-�, vascular endothelialgrowth factor (VEGF), and cyclooxygenase 2 (COX-2).(40)

A translational effect is also possible because ERK and p38activate the serine/threonine kinases Mnk-1 and Mnk-2,which in turn, stimulate the initiation factor eIF4E, which is

critical for translation.(41) The dissociation between PD in-hibitory effect on OPG protein and mRNA levels couldsupport this hypothesis.

Although OPG stimulation by IL-1� does not need twocritical factors involved in osteoclastogenesis (JNK and NF-�B), the remaining question is why IL-1� stimulates OPGin parallel to RANKL. Presumably, this could be a counter-regulatory process. This is in accordance to what we ob-served in patients with Crohn’s disease. Indeed, OPG levelswere found to be increased both in the serum and in theculture media of intestinal biopsies from these patients.(42)

We also showed that OPG production by the intestinal mu-

FIG. 6. IL-1� stimulates OPG mRNA levels by transcriptional and post-transcriptional mechanisms. (A) IL-1� slightly stimulatesOPG promoter activity. MG-63 cells were transiently transfected with pGL3-basic containing 0.9-, 0.4-, or 0.2-kb fragments from theOPG promoter or pGL3-basic. After 24-h recovery, MG-63 cells were treated with vehicle or IL-1� (10 ng/ml) for 6 h. pGL4.74luciferase reporter vector was co-transfected to normalize transfection efficiency. Luciferase activity was expressed as percentage of thecontrol. The results represent the mean ± SE of 6–12 pooled from two to three independent experiments. aSignificant effect of IL-1�.(B) Effect of IL-1� on OPG mRNA levels was quantified in the presence or absence of cycloheximide (CHX). MG-63 cells werepretreated with CHX at 20 �M for 1 h before treatment with vehicle or IL-1� (10 ng/ml) for 6 h. OPG mRNA levels were quantifiedby real-time RT-PCR. Values are means ± SE (n � 3 from one representative experiment). aSignificant effect of IL-1�. bSignificanteffect of CHX. (C) Effect of IL-1� on OPG mRNA levels in the presence or absence of actinomycin D (Act D). MG-63 cells werepretreated with vehicle or IL-1� (10 ng/ml) for 6 h before treatment with 5�g/ml of Act D for 10–16 h. OPG mRNA was quantifiedby real-time RT-PCR. Results are expressed as the percent of baseline value (n � 4 observations pooled from two independentexperiments). aSignificant effect of Act D in control cells. bSignificant effect of IL-1� in Act D–treated cells.

FIG. 7. Signaling pathways involved in IL-1�–induced OPG production. IL-1� stimu-lates OPG protein release by a mechanismdependent on the phosphorylation of p38and ERK. In contrast, JNK and NF-�B werenot essential for this regulation. Proteinnodes are shown in representations of theirfunctional annotation and are connected byarrows. Red bars indicate target of inhibitors.SB, a p38 kinase inhibitor(24); PD, a MEKselective inhibitor(25); SP, an inhibitor of theJNK pathway(27); BAY, an inhibitor of I�B�phosphorylation.(26)

LAMBERT ET AL.1358

Fig 7 live 4/C

cosa was correlated to the levels of pro-inflammatory cyto-kines.(42) Another possibility might be that OPG binds toTRAIL and therefore acts as a survival factor as describedfor cancer cells or even synoviocytes.(16–18) In this work, wecould not show a protective effect of IL-1� on TRAIL-induced cell death (Table 1) even though MG-63 cells pro-duce up to 400 ng/ml of OPG after IL-1� stimulation. Thisis in contrast to what has been shown in fibroblast-like sy-novial cells.(18) A possible explanation is that ELISAs avail-able currently might not be able to distinguish free OPGfrom OPG bound to RANKL, and IL-1� stimulates bothOPG and sRANKL in osteoblast-like cells.(13–15,43) Al-though we did not develop a binding assay (sRANKL as a

capture agent) to specifically answer this question, wetested the effects of sRANKL in the presence or absence ofTRAIL (Table 2). Because OPG affinity for TRAIL islower than that for RANKL, one would expect that freeOPG would bind preferably to sRANKL and that TRAIL-induced apoptosis would be increased.(16) However, wefound a marginal, although statistically significant, effect ofsRANKL on TRAIL-induced apoptosis (Table 2) indicat-ing that either OPG is bound to endogenous RANKL inresponse to IL-1� or that, for an unknown reason, OPGreleased after IL-1� might not have the same activity asexogenously added OPG. In this work, a partial inhibitionof TRAIL-induced apoptosis was observed with a 5-foldratio of exogenous OPG/TRAIL, whereas a 10-fold ratioprevented TRAIL-induced MG-63 apoptosis (Table 3),confirming previous observations in other cancer celltypes.(17,44) Extrapolation from in vitro data to in vivo hasto be made with caution. Indeed, although OPG was shownto dose-dependently inhibit TRAIL-induced apoptosis ofmyeloma cells and prostate cancer cells, OPG treatment inmice resulted in a decrease in osteoclastogenesis and tumorgrowth.(17,44–46) In future clinical development, the humanmonoclonal antibody against RANKL (Denosumab) mayoffer a better alternative than OPG as being specific toRANKL and not binding to TRAIL.(47)

In conclusion, IL-1� stimulates OPG protein release by amechanism dependent on the phosphorylation of p38 andERK. Although relevant to inflammation, the classicalpathway of NF-�B activation is not necessary for IL-1�induction of OPG transcripts. Finally, IL-1�–induced OPGrelease does not seem to influence TRAIL-induced celldeath.

ACKNOWLEDGMENTS

The authors thank Dr JE Onyia and Eli Lilly and Co. forproviding the OPG plasmids; Aline Desoroux, Simone Gas-par, and Miguel Lopez for expert technical assistance; andthe Department of Biostatistics (Prof A Albert) for statis-tical assistance. This work was supported by grants from theF.R.I.A. (Fund for Training in Research in Industry andFund for Training in Research in Agriculture), the BelgianNational Fund for Scientific Research (FNRS), and theLéon Frédéricq Foundation. NF, M-PM, AC, ED, and COare research associates, and JP is senior research associateat the FNRS.

REFERENCES

1. Walsh MC, Choi Y 2003 Biology of the TRANCE axis. Cyto-kine Growth Factor Rev 14:251–263.

2. Burgess TL, Qian Y, Kaufman S, Ring BD, Van G, CapparelliC, Kelley M, Hsu H, Boyle WJ, Dunstan CR, Hu S, Lacey DL1999 The ligand for osteoprotegerin (OPGL) directly activatesmature osteoclasts. J Cell Biol 145:527–538.

3. Lacey DL, Tan HL, Lu J, Kaufman S, Van G, Qiu W, RattanA, Scully S, Fletcher F, Juan T, Kelley M, Burgess TL, BoyleWJ, Polverino AJ 2000 Osteoprotegerin ligand modulates mu-rine osteoclast survival in vitro and in vivo. Am J Pathol157:435–448.

4. Lacey DL, Timms E, Tan HL, Kelley MJ, Dunstan CR, Bur-gess T, Elliott R, Colombero A, Elliott G, Scully S, Hsu H,

TABLE 1. IL-1� DOES NOT MODIFY TRAIL-INDUCED

CELL DEATH

TRAIL(0 ng/ml)

TRAIL(10 ng/ml)

TRAIL(50 ng/ml)

TRAIL(100 ng/ml)

Control 100 ± 3%) 40 ± 3%* 26 ± 1%* 26 ± 1%*IL-1�

(10 ng/ml) 107 ± 2% 43 ± 5%* 20 ± 1%* 20 ± 1%*

MG-63 cells were pretreated with IL-1� (10 ng/ml) for 16 h before treat-ment with vehicle or TRAIL (10–100 ng/ml) for 24 h. Cell viability wasmeasured with the MTS reduction assay. Results are expressed as the per-cent of the viability observed in the control cultures. Values are means ± SE(n � 12 observations pooled from two independent experiments).

* Significant effect of TRAIL.

TABLE 2. SRANKL BARELY INCREASES TRAIL-INDUCED

CELL DEATH

TRAIL 0 ng/ml TRAIL 10 ng/ml

sRANKL (0 ng/ml) 100 ± 4% 37 ± 3%*sRANKL (50 ng/ml) 100 ± 4% 32 ± 2%*†

sRANKL (100 ng/ml) 100 ± 2% 32 ± 2%*†

sRANKL (500 ng/ml) 100 ± 3% 28 ± 1%*†

MG-63 cells were treated with 50–500 ng/ml sRANKL in the presence orabsence of TRAIL (10 ng/ml) for 24 h. Cell viability was measured with theMTS reduction assay. Results are expressed as the percent of the viabilityobserved in the control cultures. Values are means ± SE (n � 12 observa-tions pooled from two independent experiments).

* Significant effect of TRAIL.† Significant effect of sRANKL.

TABLE 3. EXOGENOUS OPG PREVENTS TRAIL-INDUCED

CELL DEATH

TRAIL (0 ng/ml) TRAIL (10 ng/ml)

OPG (0 ng/ml) 100 ± 1% 62 ± 2%*OPG (50 ng/ml) 100 ± 2% 82 ± 4%*†

OPG (100 ng/ml) 100 ± 3% 96 ± 3%*†

OPG (500 ng/ml) 100 ± 2% 108 ± 4%*†

OPG (1000 ng/ml) 100 ± 5% 98 ± 3%*†

MG-63 cells were treated with vehicle or 10 ng/ml TRAIL in the pres-ence or absence of 50–1000 ng/ml OPG for 24 h. Cell viability was measuredwith the MTS reduction assay. Results are expressed as the percent of theviability observed in the control cultures. Values are means ± SE (n � 9observations pooled from three independent experiments).

* Significant effect of TRAIL.† Significant effect of OPG.

IL-1� STIMULATES OPG 1359

Sullivan J, Hawkins N, Davy E, Capparelli C, Eli A, Qian YX,Kaufman S, Sarosi I, Shalhoub V, Senaldi G, Guo J, DelaneyJ, Boyle WJ 1998 Osteoprotegerin ligand is a cytokine thatregulates osteoclast differentiation and activation. Cell 93:165–176.

5. O’Brien EA, Williams JH, Marshall MJ 2000 Osteoprotegerinligand regulates osteoclast adherence to the bone surface inmouse calvaria. Biochem Biophys Res Commun 274:281–290.

6. Iotsova V, Caamano J, Loy J, Yang Y, Lewin A, Bravo R 1997Osteopetrosis in mice lacking NF-kappaB1 and NF-kappaB2.Nat Med 3:1285–1289.

7. Grigoriadis AE, Wang ZQ, Cecchini MG, Hofstetter W, FelixR, Fleisch HA, Wagner EF 1994 c-Fos: A key regulator ofosteoclast-macrophage lineage determination and bone re-modeling. Science 266:443–448.

8. Boyle WJ, Simonet WS, Lacey DL 2003 Osteoclast differen-tiation and activation. Nature 423:337–342.

9. Tanaka S, Nakamura K, Takahasi N, Suda T 2005 Role ofRANKL in physiological and pathological bone resorption andtherapeutics targeting the RANKL-RANK signaling system.Immunol Rev 208:30–49.

10. Simonet WS, Lacey DL, Dunstan CR, Kelley M, Chang MS,Luthy R, Nguyen HQ, Wooden S, Bennett L, Boone T, Shi-mamoto G, DeRose M, Elliott R, Colombero A, Tan HL, TrailG, Sullivan J, Davy E, Bucay N, Renshaw-Gegg L, HughesTM, Hill D, Pattison W, Campbell P, Sander S, Van G, TarpleyJ, Derby P, Lee R, Boyle WJ 1997 Osteoprotegerin: A novelsecreted protein involved in the regulation of bone density.Cell 89:309–319.

11. Manolagas SC, Jilka RL 1995 Bone marrow, cytokines, andbone remodeling. Emerging insights into the pathophysiologyof osteoporosis. N Engl J Med 332:305–311.

12. Hofbauer LC, Gori F, Riggs BL, Lacey DL, Dunstan CR,Spelsberg TC, Khosla S 1999 Stimulation of osteoprotegerinligand and inhibition of osteoprotegerin production by gluco-corticoids in human osteoblastic lineage cells: Potential para-crine mechanisms of glucocorticoid-induced osteoporosis. En-docrinology 140:4382–4389.

13. Hofbauer LC, Lacey DL, Dunstan CR, Spelsberg TC, RiggsBL, Khosla S 1999 Interleukin-1beta and tumor necrosis fac-tor-alpha, but not interleukin-6, stimulate osteoprotegerin li-gand gene expression in human osteoblastic cells. Bone25:255–259.

14. Hofbauer LC, Khosla S, Dunstan CR, Lacey DL, Boyle WJ,Riggs BL 2000 The roles of osteoprotegerin and osteoprote-gerin ligand in the paracrine regulation of bone resorption. JBone Miner Res 15:2–12.

15. Hofbauer LC, Dunstan CR, Spelsberg TC, Riggs BL, Khosla S1998 Osteoprotegerin production by human osteoblast lineagecells is stimulated by vitamin D, bone morphogenetic protein-2, and cytokines. Biochem Biophys Res Commun 250:776–781.

16. Emery JG, McDonnell P, Burke MB, Deen KC, Lyn S, Silver-man C, Dul E, Appelbaum ER, Eichman C, DiPrinzio R,Dodds RA, James IE, Rosenberg M, Lee JC, Young PR 1998Osteoprotegerin is a receptor for the cytotoxic ligand TRAIL.J Biol Chem 273:14363–14367.

17. Shipman CM, Croucher PI 2003 Osteoprotegerin is a solubledecoy receptor for tumor necrosis factor-related apoptosis-inducing ligand/Apo2 ligand and can function as a paracrinesurvival factor for human myeloma cells. Cancer Res 63:912–916.

18. Miyashita T, Kawakami A, Nakashima T, Yamasaki S, TamaiM, Tanaka F, Kamachi M, Ida H, Migita K, Origuchi T, NakaoK, Eguchi K 2004 Osteoprotegerin (OPG) acts as an endog-enous decoy receptor in tumour necrosis factor-related apo-ptosis-inducing ligand (TRAIL)-mediated apoptosis of fibro-blast-like synovial cells. Clin Exp Immunol 137:430–436.

19. Lambert C, Ribbens C, Merville M, Bours V, Malaise M,Franchimont N 2003 Identification of the signalling pathwaysrequired for interleukin-1beta stimulation of osteoprotegerinsynthesis in osteoblastic cells. J Bone Miner Res 18:SA228.

20. Pantouli E, Boehm MM, Koka S 2005 Inflammatory cytokines

activate p38 MAPK to induce osteoprotegerin synthesis byMG-63 cells. Biochem Biophys Res Commun 329:224–229.

21. Franzoso G, Carlson L, Xing L, Poljak L, Shores EW, BrownKD, Leonardi A, Tran T, Boyce BF, Siebenlist U 1997 Re-quirement for NF-kappaB in osteoclast and B-cell develop-ment. Genes Dev 11:3482–3496.

22. Jimi E, Ghosh S 2005 Role of nuclear factor-kappaB in theimmune system and bone. Immunol Rev 208:80–87.

23. Relic B, Bentires-Alj M, Ribbens C, Franchimont N, GuernePA, Benoit V, Merville MP, Bours V, Malaise MG 2002 TNF-alpha protects human primary articular chondrocytes from ni-tric oxide-induced apoptosis via nuclear factor-kappaB. LabInvest 82:1661–1672.

24. Young PR, McLaughlin MM, Kumar S, Kassis S, Doyle ML,McNulty D, Gallagher TF, Fisher S, McDonnell PC, Carr SA,Huddleston MJ, Seibel G, Porter TG, Livi GP, Adams JL, LeeJC 1997 Pyridinyl imidazole inhibitors of p38 mitogen-activated protein kinase bind in the ATP site. J Biol Chem272:12116–12121.

25. Dudley DT, Pang L, Decker SJ, Bridges AJ, Saltiel AR 1995 Asynthetic inhibitor of the mitogen-activated protein kinase cas-cade. Proc Natl Acad Sci USA 92:7686–7689.

26. Pierce JW, Schoenleber R, Jesmok G, Best J, Moore SA, Col-lins T, Gerritsen ME 1997 Novel inhibitors of cytokine-inducedIkappaBalpha phosphorylation and endothelial cell adhesionmolecule expression show anti-inflammatory effects in vivo. JBiol Chem 272:21096–21103.

27. Han Z, Boyle DL, Chang L, Bennett B, Karin M, Yang L,Manning AM, Firestein GS 2001 c-Jun N-terminal kinase isrequired for metalloproteinase expression and joint destruc-tion in inflammatory arthritis. J Clin Invest 108:73–81.

28. Relic B, Benoit V, Franchimont N, Ribbens C, Kaiser MJ,Gillet P, Merville MP, Bours V, Malaise MG 2004 15-deoxy-delta12,14-prostaglandin J2 inhibits Bay 11-7085-induced sus-tained extracellular signal-regulated kinase phosphorylationand apoptosis in human articular chondrocytes and synovialfibroblasts. J Biol Chem 279:22399–22403.

29. Olivier S, Fillet M, Malaise M, Piette J, Bours V, Merville MP,Franchimont N 2005 Sodium nitroprusside-induced osteoblastapoptosis is mediated by long chain ceramide and is decreasedby raloxifene. Biochem Pharmacol 69:891–901.

30. Bonif M, Meuwis MA, Close P, Benoit V, Heyninck K,Chapelle JP, Bours V, Merville MP, Piette J, Beyaert R,Chariot A 2006 TNFalpha- and IKKbeta-mediated TANK/I-TRAF phosphorylation: Implications for interaction withNEMO/IKKgamma and NF-kappaB activation. Biochem J394:593–603.

31. Thirunavukkarasu K, Halladay DL, Miles RR, Yang X, GalvinRJ, Chandrasekhar S, Martin TJ, Onyia JE 2000 The osteo-blast-specific transcription factor Cbfa1 contributes to the ex-pression of osteoprotegerin, a potent inhibitor of osteoclastdifferentiation and function. J Biol Chem 275:25163–25172.

32. Pahl HL 1999 Activators and target genes of Rel/NF-kappaBtranscription factors. Oncogene 18:6853–6866.

33. Gowen M, Wood DD, Ihrie EJ, McGuire MK, Russell RG1983 An interleukin 1 like factor stimulates bone resorption invitro. Nature 306:378–380.

34. Hofbauer LC, Schoppet M 2004 Clinical implications of theosteoprotegerin/RANKL/RANK system for bone and vasculardiseases. JAMA 292:490–495.

35. Strand V, Kavanaugh AF 2004 The role of interleukin-1 inbone resorption in rheumatoid arthritis. Rheumatology (Oxf)43(Suppl 3):iii10–iii16.

36. Roodman GD 2001 Biology of osteoclast activation in cancer.J Clin Oncol 19:3562–3571.

37. Bonizzi G, Karin M 2004 The two NF-kappaB activation path-ways and their role in innate and adaptive immunity. TrendsImmunol 25:280–288.

38. Vermeulen L, De Wilde G, Van Damme P, Vanden Berghe W,Haegeman G 2003 Transcriptional activation of the NF-kappaB p65 subunit by mitogen- and stress-activated proteinkinase-1 (MSK1). EMBO J 22:1313–1324.

39. Jimi E, Nakamura I, Ikebe T, Akiyama S, Takahashi N, Suda

LAMBERT ET AL.1360

T 1998 Activation of NF-kappaB is involved in the survival ofosteoclasts promoted by interleukin-1. J Biol Chem 273:8799–8805.

40. Shim J, Karin M 2002 The control of mRNA stability in re-sponse to extracellular stimuli. Mol Cells 14:323–331.

41. Waskiewicz AJ, Flynn A, Proud CG, Cooper JA 1997 Mito-gen-activated protein kinases activate the serine/threonine ki-nases Mnk1 and Mnk2. EMBO J 16:1909–1920.

42. Franchimont N, Reenaers C, Lambert C, Belaiche J, Bours V,Malaise M, Delvenne P, Louis E 2004 Increased expression ofreceptor activator of NF-kappaB ligand (RANKL), its recep-tor RANK and its decoy receptor osteoprotegerin in the colonof Crohn’s disease patients. Clin Exp Immunol 138:491–498.

43. Rogers A, Eastell R 2005 Circulating osteoprotegerin and re-ceptor activator for nuclear factor kappaB ligand: Clinical util-ity in metabolic bone disease assessment. J Clin EndocrinolMetab 90:6323–6331.

44. Holen I, Croucher PI, Hamdy FC, Eaton CL 2002 Osteopro-tegerin (OPG) is a survival factor for human prostate cancercells. Cancer Res 62:1619–1623.

45. Zhang J, Dai J, Qi Y, Lin DL, Smith P, Strayhorn C, MizokamiA, Fu Z, Westman J, Keller ET 2001 Osteoprotegerin inhibits

prostate cancer-induced osteoclastogenesis and prevents pros-tate tumor growth in the bone. J Clin Invest 107:1235–1244.

46. Croucher PI, Shipman CM, Lippitt J, Perry M, Asosingh K,Hijzen A, Brabbs AC, van Beek EJ, Holen I, Skerry TM,Dunstan CR, Russell GR, Van Camp B, Vanderkerken K 2001Osteoprotegerin inhibits the development of osteolytic bonedisease in multiple myeloma. Blood 98:3534–3540.

47. Kostenuik PJ 2005 Osteoprotegerin and RANKL regulatebone resorption, density, geometry and strength. Curr OpinPharmacol 5:618–625.

Address reprint requests to:N Franchimont, MD, PhD

Department of RheumatologyCHU Sart-Tilman B35B-4000 Liège, Belgium

E-mail: nathalie.franchimont@amgen.com

Received in original form August 21, 2006; revised form March 22,2007; accepted May 4, 2007.

IL-1� STIMULATES OPG 1361