A RANK/TRAF6-dependent Signal Transduction Pathway Is Essential for Osteoclast Cytoskeletal Organization and Resorptive Function
2002; Elsevier BV; Volume: 277; Issue: 46 Linguagem: Inglês
10.1074/jbc.m202009200
ISSN1083-351X
AutoresAllison Armstrong, Mark Tometsko, Moira Glaccum, Claire L. Sutherland, David Cosman, William C. Dougall,
Tópico(s)NF-κB Signaling Pathways
ResumoSignaling through receptor activator of nuclear factor-κB (RANK) is essential for the differentiation and activation of osteoclasts, the cell principally responsible for bone resorption. Animals genetically deficient in RANK or the cognate RANK ligand are profoundly osteopetrotic because of the lack of bone resorption and remodeling. RANK provokes biochemical signaling via the recruitment of intracellular tumor necrosis factor receptor-associated factors (TRAFs) after ligand binding and receptor oligomerization. To understand the RANK-mediated signal transduction mechanism in osteoclastogenesis, we have designed a system to recapitulate osteoclast differentiation and activation in vitro by transfer of the RANK cDNA into hematopoietic precursors genetically deficient in RANK. Gene transfer of RANK constructs that are selectively incapable of binding different TRAF proteins revealed that TRAF pathways downstream of RANK that affect osteoclast differentiation are functionally redundant. In contrast, the interaction of RANK with TRAF6 is absolutely required for the proper formation of cytoskeletal structures and functional resorptive activity of osteoclasts. Moreover, signaling via the interleukin-1 receptor, which also utilizes TRAF6, rescues the osteoclast activation defects observed in the absence of RANK/TRAF6 interactions. These studies are the first to define the functional domains of the RANK cytoplasmic tail that control specific differentiation and activation pathways in osteoclasts. Signaling through receptor activator of nuclear factor-κB (RANK) is essential for the differentiation and activation of osteoclasts, the cell principally responsible for bone resorption. Animals genetically deficient in RANK or the cognate RANK ligand are profoundly osteopetrotic because of the lack of bone resorption and remodeling. RANK provokes biochemical signaling via the recruitment of intracellular tumor necrosis factor receptor-associated factors (TRAFs) after ligand binding and receptor oligomerization. To understand the RANK-mediated signal transduction mechanism in osteoclastogenesis, we have designed a system to recapitulate osteoclast differentiation and activation in vitro by transfer of the RANK cDNA into hematopoietic precursors genetically deficient in RANK. Gene transfer of RANK constructs that are selectively incapable of binding different TRAF proteins revealed that TRAF pathways downstream of RANK that affect osteoclast differentiation are functionally redundant. In contrast, the interaction of RANK with TRAF6 is absolutely required for the proper formation of cytoskeletal structures and functional resorptive activity of osteoclasts. Moreover, signaling via the interleukin-1 receptor, which also utilizes TRAF6, rescues the osteoclast activation defects observed in the absence of RANK/TRAF6 interactions. These studies are the first to define the functional domains of the RANK cytoplasmic tail that control specific differentiation and activation pathways in osteoclasts. The structural and metabolic integrity of bone is maintained through the dynamic process of bone remodeling that results from the coordinate action of bone resorption by osteoclasts and the formation of new bone by osteoblasts. Osteoclasts are large, multinuclear cells that develop from a hematopoietic progenitor and are highly specialized for the resorptive process (1Teitelbaum S.L. Science. 2000; 2889: 1504-1508Crossref Scopus (2967) Google Scholar). Regulation of bone remodeling occurs through multiple mechanisms that ultimately converge on the interaction of osteoclasts or their precursors with osteoblasts and bone marrow stromal cells. Two key factors supplied by the stromal environment are CSF-1 1The abbreviations used are: CSF, colony-stimulating factor; IL, interleukin; JNK, c-Jun kinase; NF-κB, nuclear factor-κB; RANK, receptor activator nuclear factor-κB; RANKL, receptor activator nuclear factor-κB ligand; TNF, tumor necrosis factor; TNFR, tumor necrosis factor receptor; TRAF, tumor necrosis factor receptor-associated factor; TRAP, tartrate-resistant acid phosphatase; PBS, phosphatebuffered saline; RT, reverse transcription; HPRT, hypoxanthine-guanine phosphoribosyl-transferase 1The abbreviations used are: CSF, colony-stimulating factor; IL, interleukin; JNK, c-Jun kinase; NF-κB, nuclear factor-κB; RANK, receptor activator nuclear factor-κB; RANKL, receptor activator nuclear factor-κB ligand; TNF, tumor necrosis factor; TNFR, tumor necrosis factor receptor; TRAF, tumor necrosis factor receptor-associated factor; TRAP, tartrate-resistant acid phosphatase; PBS, phosphatebuffered saline; RT, reverse transcription; HPRT, hypoxanthine-guanine phosphoribosyl-transferaseand the TNF family member, RANKL (also called TRANCE, ODF, OPGL) (2Riggs B.L. Baron R. Boyle W.J. Drezner M. Manolagas S. Martin T.J. Stewart A.F. Suda T. Yasuda H. Aubin J. Goltzman D. J. Bone Miner. Res. 2000; 15: 2293-2296PubMed Google Scholar), as confirmed by the osteopetrotic phenotypes of the op/op mice that are mutated in the CSF-1 gene (3Yoshida H. Hayashi S. Kunisada T. Ogawa M. Nishikawa S. Okamura H. Sudo T. Shultz L.D. Nishikawa S. Nature. 1990; 345: 442-444Crossref PubMed Scopus (1492) Google Scholar) and the RANKL knockout mice (4Kong Y.Y. Yoshida H. Sarosi I. Tan H.L. Timms E. Capparelli C. Morony S. Oliveira-dos Santos A. Van G. Itie A. Khoo W. Wakeham A. Dunstan C.R. Lacey D.L. Mak T.W. Boyle W.J. Penninger J.M. Nature. 1999; 397: 315-323Crossref PubMed Scopus (2799) Google Scholar). It is now widely accepted that most osteotropic agents including IL-1, IL-6, IL-11, IL-17, TNF-α, prostaglandin E2, parathyroid hormone, and 1,25-dihydroxyvitamin D3 (5Hofbauer L.C. Dunstan C.R. Lacey D.L. Boyle W.J. Riggs B.L. J. Bone Miner. Res. 2000; 15: 2-12Crossref PubMed Scopus (1011) Google Scholar) affect bone resorption primarily by enhancing stromal cell production of RANKL. RANKL affects bone resorption and bone density by influencing the osteoclast population at multiple stages. Not only does RANKL drive the differentiation of osteoclasts from multipotential progenitors, thereby expanding the pool of osteoclasts available for bone resorption (6Yasuda H. Shima N. Nakagawa N. Yamaguchi K. Kinosaki M. Mochizuki S. Tomoyasu A. Yano K. Goto M. Murakami A. Tsuda E. Morinaga T. Higashio K. Udagawa N. Takahashi N. Suda T. Proc. Natl. Acad. Sci. U. S. A. 1998; 95: 3597-3602Crossref PubMed Scopus (3500) Google Scholar), RANKL also activates resorption and enhances survival of existing mature osteoclasts in vitro (7Burgess T.L. Kaufman S. Ring B.D. Van G. Capparelli C. Kelley M. Hsu H. Boyle W.J., Hu, S. Lacey D.L. J. Cell Biol. 1999; 145: 527-538Crossref PubMed Scopus (601) Google Scholar, 8Lacey D.L. Timms E. Tan H.L. Kelley M.J. Dunstan C.R. Burgess T. Elliott R. Colombero A. Elliott G. Scully S. Hsu H. Sullivan J. Hawkins N. Davy E. Capparelli C. Eli A. Qian Y.X. Kaufman S. Sarosi I. Shalhoub V. Senaldi G. Guo J. Delaney J. Boyle W.J. Cell. 1998; 93: 165-176Abstract Full Text Full Text PDF PubMed Scopus (4546) Google Scholar) and in vivo(9Lacey D.L., Lu, J. Kaufman S. Van G. Qiu W. Rattan A. Scully S. Fletcher J.T. Kelley M. Burgess T.L. Boyle W.J. Polverino A.J. Am. J. Pathol. 2000; 157: 435-448Abstract Full Text Full Text PDF PubMed Scopus (339) Google Scholar). An essential role for the RANKL receptor, RANK, in osteoclast differentiation in vivo is demonstrated by the lack of osteoclasts and resulting osteopetrosis in RANK−/− animals (10Dougall W.C. Glaccum M. Charrier K. Rohrbach K. Brasel K., De Smedt T. Daro E. Smith J. Tometsko M.E. Maliszewski C.R. Armstrong A. Shen V. Bain S. Cosman D. Anderson D. Morrissey P.J. Peschon J.J. Schuh J. Genes Dev. 1999; 13: 2412-2424Crossref PubMed Scopus (1171) Google Scholar, 11Li J, S.I. Yan X.Q. Morony S. Capparelli C. Tan H.L. McCabe S. Elliott R. Scully S. Kaufman S. Juan S.C. Sun Y. Tarpley J. Martin L. Christensen K. McCabe J.K.P. Hsu H. Fletcher F. Dunstan C.R. Lacey D.L. Boyle W.J. Proc. Natl. Acad. Sci. U. S. A. 2000; 97: 1566-1571Crossref PubMed Scopus (926) Google Scholar). Thus, RANK plays a key role in bone metabolism as the critical signaling receptor responsible for osteoclast differentiation, activation, and survival. RANK, like many of the TNFR family proteins, transmits biochemical signals after recruitment of intracellular adaptor TNF receptor-associated factor (TRAF) proteins. Activation of NF-κB, JNK, and mitogen-activated protein kinase pathways by RANK occurs via TRAFs 2, 5, and 6 (reviewed in Ref. 12Inoue J. Ishida T. Tsukamoto N. Kobayashi N. Naito A. Azuma S. Yamamoto T. Exp. Cell Res. 2000; 254: 14-24Crossref PubMed Scopus (360) Google Scholar). The relevance of each of these biochemical signals in osteoclast biology is underscored by the osteopetrotic phenotypes of the p50/p52 NF-κB double knockout mouse (13Franzoso G. Carlson L. Xing L. Poljak L. Shores E.W. Brown K.D. Leonardi A. Tran T. Boyce B.F. Siebenlist U. Genes Dev. 1997; 11: 3482-3496Crossref PubMed Scopus (853) Google Scholar, 14Iotsova V. Caamano J. Loy J. Yang Y. Lewin A. Bravo R. Nat. Med. 1997; 3: 1285-1289Crossref PubMed Scopus (862) Google Scholar) and the fos−/− mouse (15Grigoriadis A.E. Wang Z.-Q. Cecchini M.G. Hofstetter W. Felix R. Fleisch H.A. Wagner E.F. Science. 1994; 266: 443-448Crossref PubMed Scopus (1056) Google Scholar), as well as the inhibition ofin vitro osteoclast differentiation after blockade of p38 mitogen-activated protein kinase activity (16Matsumoto M. Sudo T. Saito T. Osada H. Tsujimoto M. J. Biol. Chem. 2000; 275: 31155-31161Abstract Full Text Full Text PDF PubMed Scopus (468) Google Scholar). TRAF6, which binds to a RANK structural motif distinct from the region that binds TRAFs 1, 2, 3, and 5, plays an essential role in RANKL-mediated NF-κB activation in transfected cell systems (17Darnay B.G., Ni, J. Moore P.A. Aggarwal B.B. J. Biol. Chem. 1999; 274: 7724-7731Abstract Full Text Full Text PDF PubMed Scopus (353) Google Scholar, 18Galibert L. Tometsko M.E. Anderson D.M. Cosman D. Dougall W.C. J. Biol. Chem. 1998; 273: 34120-34127Abstract Full Text Full Text PDF PubMed Scopus (276) Google Scholar). TRAF6 also provides a functional connection between RANK signaling and the activation of c-Src kinase, Akt/protein kinase B, and phosphatidylinositol 3-kinase (19Wong B.R. Kim N. Arron J.R. Vologodskaia M. Hanafusa H. Choi Y. Mol. Cell. 1999; 4: 1041-1049Abstract Full Text Full Text PDF PubMed Scopus (509) Google Scholar). Knockouts of c-Src (20Soriano P. Montgomery C. Geske R. Bradley A. Cell. 1991; 64: 693-702Abstract Full Text PDF PubMed Scopus (1786) Google Scholar) and TRAF6 (21Lomaga M.A. Yeh W.-C. Sarosi I. Duncan G.S. Furlonger C., Ho, A. Morony S. Capparelli C. Van G. Kaufman S. van der Heiden A. Itie A. Wakeham A. Khoo W. Sasaki T. Cao Z. Penninger J.M. Paige C.J. Lacey D.L. Dunstan C.R. Boyle W.J. Goeddel D.V. Mak T.W. Genes Dev. 1999; 13: 1015-1024Crossref PubMed Scopus (1060) Google Scholar) are also osteopetrotic, although the bone defects in these animals are not the result of lack of osteoclasts as seen in the p50/p52 NF-κB double knockout, RANK−/−, and the fos−/− mice, but rather are caused by defective osteoclast activity. RANK signaling is essential for the initiation of osteoclast differentiation. However, transfection studies in heterologous cells have demonstrated that RANK signal transduction is mediated via multiple distinct and selective TRAF-binding motifs (reviewed in Ref.12Inoue J. Ishida T. Tsukamoto N. Kobayashi N. Naito A. Azuma S. Yamamoto T. Exp. Cell Res. 2000; 254: 14-24Crossref PubMed Scopus (360) Google Scholar). To define the relevance of these distinct RANK cytoplasmic motifs and downstream biochemical pathways in the context of each stage of osteoclast differentiation and function, we have established an experimental system in which transfer of the RANK gene into RANK−/− hematopoietic precursors will restore the formation and activation of osteoclasts. Using a panel of RANK mutants selectively incapable of binding distinct TRAF proteins, we have established that there are redundant RANK signaling paths necessary for the differentiation of hematopoietic precursors into multinuclear osteoclasts and for the up-regulation of genes which function during osteoclastogenesis. In contrast to RANK-dependent differentiation signals, the sequence elements in the RANK cytoplasmic domain necessary for TRAF6 binding are nonredundant and critical for the proper cytoskeletal organization and resorptive activities of osteoclasts. Each of the human RANK cDNA constructs has been previously described (18Galibert L. Tometsko M.E. Anderson D.M. Cosman D. Dougall W.C. J. Biol. Chem. 1998; 273: 34120-34127Abstract Full Text Full Text PDF PubMed Scopus (276) Google Scholar) except for RANK Δ340–421/545–616. Full-length and mutated human RANK constructs were subcloned into the LZRS-pBMNZ vector (22Kinsella T.M. N. G.P. Hum. Gene Ther. 1996; 7: 1405-1413Crossref PubMed Scopus (668) Google Scholar), and production of infectious retroviral vector particles was performed in 293-E Phoenix packaging cells as described. Normalized expression of each of the RANK constructs was confirmed by analysis of RANK surface expression using an antibody against human RANK (monoclonal antibody M330) (23Anderson D.M. Maraskovsky E. Billingsley W.L. Dougall W.C. Tometsko M.E. Roux E.R. Teepe M.C. DuBose R.F. Cosman D. Galibert L. Nature. 1997; 390: 175-179Crossref PubMed Scopus (1905) Google Scholar) on infected cells. NIH3T3 cells were infected with varying amounts of each RANK retroviral construct in the presence of 5 μg/ml Polybrene. Forty-eight hours after infection, cells were lifted with Versene and stained with an antibody to hRANK (2.5 μg/ml), and fluorescein isothiocyanate-conjugated anti-mouse IgG1 secondary (1:500) and positive cells were analyzed by flow cytometry. The generation of the RANK−/− mice has been described previously (10Dougall W.C. Glaccum M. Charrier K. Rohrbach K. Brasel K., De Smedt T. Daro E. Smith J. Tometsko M.E. Maliszewski C.R. Armstrong A. Shen V. Bain S. Cosman D. Anderson D. Morrissey P.J. Peschon J.J. Schuh J. Genes Dev. 1999; 13: 2412-2424Crossref PubMed Scopus (1171) Google Scholar). Spleen cells were isolated from 3–6-week-old RANK−/− mice, and T cells, erythroid cells, and neutrophils were depleted using rat antibodies against CD3, Ter-119, and GR-1 (BD PharMingen, San Diego, CA) and sheep anti-rat conjugated magnetic beads (Dynabeads M-450; Dynal, Oslo, Norway). Cells were incubated for 48 h in 40 ng/ml hCSF-1 (R&D Systems, Minneapolis, MN) and infected using retroviral supernatants (multiplicity of infection of 5) in the presence of recombinant fibronectin fragments (Retronectin, PanVera Corp., Madison, WI) for 48 h in 40 ng/ml CSF-1. Following infection, spleen cells were harvested and plated in α-minimal essential medium plus 10% fetal bovine serum containing 40 ng/ml CSF-1 and the presence or absence of 200 ng/ml mRANKL as indicated. Recombinant murine RANKL is an N-terminal fusion of a leucine zipper trimerization domain (24Fanslow W.C. Srinivasan S. Paxton R. Gibson M.G. Spriggs M.K. Armitage R.J. Semin. Immunol. 1994; 6: 267-278Crossref PubMed Scopus (149) Google Scholar) with residues 134–316 of murine RANKL (23Anderson D.M. Maraskovsky E. Billingsley W.L. Dougall W.C. Tometsko M.E. Roux E.R. Teepe M.C. DuBose R.F. Cosman D. Galibert L. Nature. 1997; 390: 175-179Crossref PubMed Scopus (1905) Google Scholar). Tartrate-resistant acid phosphatase (TRAP) staining was done at 37 °C using the leukocyte acid phosphatase kit (Sigma kit 387-A) in the presence of tartrate as described (10Dougall W.C. Glaccum M. Charrier K. Rohrbach K. Brasel K., De Smedt T. Daro E. Smith J. Tometsko M.E. Maliszewski C.R. Armstrong A. Shen V. Bain S. Cosman D. Anderson D. Morrissey P.J. Peschon J.J. Schuh J. Genes Dev. 1999; 13: 2412-2424Crossref PubMed Scopus (1171) Google Scholar). Calcitonin binding was performed by blocking cells in 5% nonfat dry milk followed by incubation with 1 μCi/ml 125I-labeled salmon calcitonin (Amersham Biosciences; specific activity = 2000 Ci/mmol). After cells were washed and solubilized in 0.5 mNaOH, total bound counts measured using a γ counter. The specificity of the calcitonin radioligand binding assay was determined by competition of the signal using 100× molar excess of unlabeled salmon calcitonin (Calbiochem, San Diego, CA). Resorption of the CaPO4 matrix (BioCoatTMOsteologicTM discs; BD Biosciences, Bedford, MA) was measured after treatment of slides in bleach and staining with 0.5% alizarin red. Quantitation of the resorbed area was performed by measuring the surface area of cleared CaPO4 resorption pits using Image Pro Plus 4.1 software from Media Cybernetics (Silver Spring, MD). Values are expressed as relative resorbed area in percentages compared with that measured in cells transduced with full-length hRANK and treated with RANKL and were measured from wells in triplicate for each experiment and expressed as mean ± standard deviation. For the pit formation/resorption lacunae assay, transduced cells were cultured in α-minimal essential medium plus 10% fetal bovine serum containing 40 ng/ml CSF-1 and with either RANKL (200 ng/ml), IL-1β (20 ng/ml), or a combination of RANKL and IL-1β. After culture for a period of 7 days, cells were stained for TRAP and removed from the dentin slices, and the resorption pits were visualized by staining with 1% toluidine blue. RNA was isolated from infected splenic cultures 24, 48, 72, and 120 h after retroviral infection. Treatment of cultures with RANKL was begun at the 48-h time point after the initiation of retroviral infection. cDNA corresponding to 10 ng of total RNA was used per reaction in quantitative PCR performed on a GeneAmp 9600 thermocycler in the presence of the dye SYBR Green, and analyzed by the GeneAmp 5700 sequence detection system (Applied Biosystems, Foster City, CA). We used primers corresponding to cathepsin K, TRAP, OC-116kDa, c-Src, MMP-9, HPRT, carbonic anhydrase II, calcitonin receptor, hRANK, and CD61; the correct sequence of the reaction products was confirmed by DNA sequencing. All samples were normalized for equivalent cDNA amounts using the HPRT signal. Cells were cultured on coverslip-thick microscope slides. After incubation for 5 days, the medium was aspirated and cells were washed twice in PBS and then fixed in a solution of 3% paraformaldehyde for 10 min followed by quenching in 50 mm NH4Cl. Cells were blocked using 5% normal goat sera in PBS. Cells were then incubated with antibodies against human RANK (M330), mouse CD61/β3 integrin (clone 2C9.G2; BD PharMingen, San Diego, CA), and the appropriate fluorescent secondary antibodies. Alternatively, cells were permeabilized in 0.1% Triton X-100, 5% normal goat sera in PBS and incubated with antibodies against hRANK, c-Src (M327v-Src, Ab-1; Calbiochem), c-Cbl (clone C-15; Santa Cruz Biotechnology), and the appropriate fluorescent secondary antibodies. For visualization of the F-actin, cells were permeabilized as above and then incubated with fluorescently conjugated phalloidin (Molecular Probes, Eugene, OR) at 1:100. To visualize nuclei, cells were permeabilized as described in the presence of RNase A at 100 μg/ml and incubated with Sytox (Molecular Probes) at 1:20,000 in 5% normal goat serum and 0.1% Triton X-100. The complete absence of osteoclasts and resulting osteopetrotic phenotype of the RANK knockout mice revealed an obligate role for RANK in normal bone remodeling (10Dougall W.C. Glaccum M. Charrier K. Rohrbach K. Brasel K., De Smedt T. Daro E. Smith J. Tometsko M.E. Maliszewski C.R. Armstrong A. Shen V. Bain S. Cosman D. Anderson D. Morrissey P.J. Peschon J.J. Schuh J. Genes Dev. 1999; 13: 2412-2424Crossref PubMed Scopus (1171) Google Scholar, 11Li J, S.I. Yan X.Q. Morony S. Capparelli C. Tan H.L. McCabe S. Elliott R. Scully S. Kaufman S. Juan S.C. Sun Y. Tarpley J. Martin L. Christensen K. McCabe J.K.P. Hsu H. Fletcher F. Dunstan C.R. Lacey D.L. Boyle W.J. Proc. Natl. Acad. Sci. U. S. A. 2000; 97: 1566-1571Crossref PubMed Scopus (926) Google Scholar). However, it is not clear how the distinct TRAF-binding domains of the RANK cytoplasmic domain contribute to osteoclast formation. The C-terminal 72 amino acids of the human RANK cytoplasmic domain are capable of binding to multiple TRAFs (TRAFs 1, 2, 3, and 5) and transmit signaling through NF-κB- and JNK-dependent pathways (18Galibert L. Tometsko M.E. Anderson D.M. Cosman D. Dougall W.C. J. Biol. Chem. 1998; 273: 34120-34127Abstract Full Text Full Text PDF PubMed Scopus (276) Google Scholar, 25Darnay B.G. Haridas V., Ni, J. Moore P.A. Aggarwal B.B. J. Biol. Chem. 1998; 273: 20551-20555Abstract Full Text Full Text PDF PubMed Scopus (361) Google Scholar). A separate region of the human RANK sequence (amino acids 340–421) mediates direct binding to TRAF6 only and is both necessary and sufficient for NF-κB activation in transfected 293 cells (17Darnay B.G., Ni, J. Moore P.A. Aggarwal B.B. J. Biol. Chem. 1999; 274: 7724-7731Abstract Full Text Full Text PDF PubMed Scopus (353) Google Scholar, 18Galibert L. Tometsko M.E. Anderson D.M. Cosman D. Dougall W.C. J. Biol. Chem. 1998; 273: 34120-34127Abstract Full Text Full Text PDF PubMed Scopus (276) Google Scholar). To define the functional role of these RANK cytoplasmic domains within the physiological context of the osteoclast, we reconstituted osteoclast differentiation and activation of RANK−/− hematopoietic progenitors using retroviral gene transfer of various human RANK cDNA constructs that selectively lack the ability to bind specific TRAF proteins. We confirmed that infection of fibroblasts with each of the different RANK retroviral supernatants leads to equivalent levels of surface expression (Fig. 1 A). Splenic cells purified from RANK−/− animals and enriched for hematopoietic progenitors cannot form osteoclasts in vitroafter addition of RANKL and CSF-1 (10Dougall W.C. Glaccum M. Charrier K. Rohrbach K. Brasel K., De Smedt T. Daro E. Smith J. Tometsko M.E. Maliszewski C.R. Armstrong A. Shen V. Bain S. Cosman D. Anderson D. Morrissey P.J. Peschon J.J. Schuh J. Genes Dev. 1999; 13: 2412-2424Crossref PubMed Scopus (1171) Google Scholar) and therefore provide a null background for the osteoclast rescue experiments. These cells were infected with retroviruses encoding the different RANK constructs or a control lacZ virus, and cultured using osteoclastogenic conditions (RANKL + CSF-1, see "Materials and Methods"). Specific expression of the transgenic mRNAs (Fig. 1 B) and total or surface protein level (Fig. 1, C and D) could be detected for each of the RANK constructs only in infected cells. The expression of the full-length RANK transgene mRNA and protein was consistently lower by a factor of 2–2.5-fold relative to the other RANK construct transgenes (Fig. 1, B and C). Thus, any deficiency in osteoclast differentiation or function was not caused by lack of expression of the mutated RANK constructs. The formation of differentiated osteoclasts was monitored in situ by expression of the enzyme marker TRAP in multinuclear cells. Osteoclasts failed to form from RANK−/− progenitors that were infected with a control virus encoding lacZ and then cultured with RANKL and CSF-1 (Fig.2). These cells expressed CD11b+, Mac-2+, and F4/80+ (data not shown) consistent with a macrophage phenotype. Retroviral-mediated expression of the full-length human RANK gene efficiently rescued multinuclear TRAP+ osteoclast differentiation even in the absence of exogenously added RANKL (Fig. 2, A andB). Addition of RANKL to the culture clearly increased the size of TRAP+ osteoclasts. RANK constructs lacking the C-terminal TRAF binding domain (RANK Δ545–616 or RANK Δ421–616) also promoted the differentiation of large TRAP+ osteoclasts in the presence of RANKL (Fig. 2,A and B), although to a lesser extent than full-length RANK. The absence of significant numbers of TRAP+ cells without addition of RANKL and the reduced number of multinuclear TRAP+ cells overall suggest that the differentiation signal from these RANK constructs is less potent. Cells transduced with a RANK construct lacking the entire cytoplasmic domain (RANK Δ340–616) were completely incapable of osteoclast differentiation (Fig. 2, A and B). A RANK construct lacking all TRAF binding sites (RANK Δ340–421/Δ545–616) was also completely incapable of forming TRAP+ osteoclasts in the presence of RANKL and CSF-1 (Fig. 2) and was indistinguishable from the C-terminal deletion RANK Δ340–616. To specifically address the role of RANK/TRAF6 signaling in osteoclast differentiation, RANK−/− cells were infected with the RANK Δ340–421 retroviral construct and cultured under the same conditions as described above. TRAP+ osteoclast formation was observed at efficiencies equal to that observed with the C-terminal RANK deletions (Fig. 2 B), although the cells were clearly more compact than, and morphologically distinct from, cells derived from the transduction of wild-type RANK or C-terminal RANK deletions. Higher magnification of these cells revealed that they contained more dendritic or lamellipodia-like processes and were less spread out than TRAP+ osteoclasts derived from the full-length RANK (Figs.5 and 6). Importantly, cells derived from the C-terminal deletion RANK constructs or the internal deletion of the TRAF6 binding site (RANK Δ340–421) were frequently multinuclear (Fig. 2 B). Quantitation of the nuclei in TRAP + osteoclasts transduced and differentiated with the various RANK constructs indicated that there was no significant difference in the mean nuclei number in each case. The reduced efficiencies of TRAP+ osteoclast formation mediated by the different RANK/TRAF-binding mutants indicate that both the RANK C-terminal (TRAFs 1, 2, 3, and 5) binding element and the membrane-proximal TRAF6 binding element contribute to osteoclast differentiation, but in the absence of either individual TRAF binding site, osteoclast differentiation still proceeds. This compensatory signaling toward osteoclast differentiation is completely abrogated upon removal of all TRAF binding elements. The relative differences in osteoclast differentiation did not vary substantially over several independent viral transductions with the RANK constructs.Figure 6Actin cytoskeleton of osteoclasts after transduction with RANK constructs. RANK−/− cells were infected with each of the RANK retrovirus constructs or a retrovirus encodinglacZ and differentiated in the presence of RANKL and CSF-1 for 5 days. Cells were fixed with paraformaldehyde and processed for F-actin fluorescence using Alexa 568-conjugated phalloidin (red) and nuclei were visualized with Sytox (yellow). A, cells differentiated after transduction with the full-length/wild-type (WT) RANK, or either of the C-terminal RANK deletions (RANK Δ545–616 or RANK Δ421–616) exhibited distinct podosome assembly (F-actin rings) in the periphery of most cells. Cells differentiated after transduction with the RANK/TRAF6 binding mutant (RANK Δ340–421) showed a redistribution of the actin to the ends of lamellipodia-like structures. Images were taken with a 20× objective. B, higher magnification (images were taken with a 60× objective) of the F-actin fluorescence in cultures transduced with the controllacZ virus, wild-type RANK, and the RANK/TRAF6 binding mutant (RANK Δ340–421).View Large Image Figure ViewerDownload (PPT) In addition to multinuclearity and TRAP staining, we have monitored the expression of multiple mRNAs and proteins expressed specifically during osteoclast differentiation to confirm that cells infected with RANK constructs lacking either the TRAF6 binding site or the separate binding site for TRAFs 1, 2, 3, and 5 still commit and differentiate into the osteoclast lineage. Expression of the osteoclast marker calcitonin receptor was detected using a radioligand binding assay. We found that expression of the full-length RANK transgene led to specific125I-calcitonin binding that was enhanced further by the addition of RANKL (Fig. 3). The expression of calcitonin receptor was also confirmed in cultures infected with either the C-terminal RANK deletions (RANK Δ545–616 or RANK Δ421–616) or the internal TRAF6 deletion (RANK Δ340–421). However, similar to the reduced efficiency of TRAP positive osteoclast differentiation, these cultures exhibited a reduction in overall125I-calcitonin binding as compared with osteoclasts differentiated via the full-length RANK protein. Quantitative RT-PCR analysis of several osteoclast differentiation marker mRNAs (MMP-9, cathepsin K, β3 integrin, carbonic anhydrase II, c-Src, calcitonin receptor, TRAP) demonstrated significant increases of each of these mRNAs in cells cultured in the presence of RANKL after infection with the full-length RANK (Fig.4). In addition, we confirmed protein expression of c-Src (Fig. 5) as well as surface expression of CD61/β3 integrin, a subunit of the vitronectin receptor (data not shown). Osteoclast marker mRNAs were not detected in cells infected with the control lacZ virus or cells infected with the RANK constructs that cannot bind TRAFs (RANK Δ340–421/Δ545–616 and deletion RANK Δ340–616). Similar to the TRAP staining and calcitonin receptor expression, RANK constructs unable to bind TRAF6 (RANK Δ340–421) or TRAFs 1, 2, 3, and 5 (RANK Δ545–616 or RANK Δ421–616) still promoted osteoclast marker expression but to a lesser degree than the full-length RANK (Fig. 4). These data are consistent with the TRAP/multinuclear osteoclast potential (Fig. 2) and suggest that TRAF/RANK interactions lead to a similar capacity for osteoclast differentiation. The c-Src kinase is highly expressed in osteoclasts and serves as a specific determinant of osteoclast differentiation fro
Referência(s)