the lim protein, limd1, regulates ap-1 activation

21
1 THE LIM PROTEIN, LIMD1, REGULATES AP-1 ACTIVATION THROUGH AN INTERACTION WITH TRAF6 TO INFLUENCE OSTEOCLAST DEVELOPMENT Yungfeng Feng 1,2,4 , Haibo Zhao 3,4 , Hilary F. Luderer 1,2 , Holly Epple 1,2 , F. Patrick Ross 2,3 , Steven L. Teitelbaum 3 , and Gregory D. Longmore 1,2 From the Departments of Medicine 1 , Cell Biology 2 , and Pathology 3 , Washington University, St. Louis, MO 63110 Running Title: LIM proteins and osteoclast development Address correspondence to: Gregory D. Longmore, Division of Hematology, Washington University School of Medicine, 660 South Euclid Avenue, St. Louis MO 63110, Tel. 314-362-8834; Fax. 314-362 8826; E-mail: [email protected] 4 These two authors contributed equally to this work. Increasingly a number of proteins important in the regulation of bone osteoclast development have been shown primarily influence osteoclastogenesis under conditions of physiologic or pathologic stress. Why basal osteoclastogenesis is normal and how these proteins regulate stress osteoclastogenic responses, as opposed to basal osteoclastogenesis, is unclear. LIM proteins of the Ajuba/Zyxin family localize to cellular sites of cell adhesion where they contribute to the regulation of cell adhesion and migration, translocate into the nucleus where they can affect cell fate, but are also found in the cytoplasm where their function is largely unknown. We show that one member of this LIM protein family, LIMD1, is uniquely upregulated during osteoclast differentiation and interacts with TRAF6, a critical cytosolic regulator of RANK-L regulated osteoclast development. LIMD1 positively affects the capacity of TRAF6 to activate AP-1, and LIMD1 -/- osteoclast precursor cells are defective in the activation of AP-1 and thus induction of NFAT2. LIMD1 -/- mice, although having normal basal bone osteoclast numbers and bone density, are resistant to physiological and pathologic osteoclastogenic stimuli. These results implicate LIMD1 as a potentially important regulator of osteoclast development under conditions of stress. Vertebrate bone development and homeostasis requires the exquisite balance between the bone resorptive capacity of osteoclasts and bone forming capacity of osteoblasts. Perturbation of this equilibrium can have pathologic consequences. For example, pathologic bone loss, as occurs in osteoporosis, rheumatoid arthritis, Paget’s disease of the bone (PDB), and tumor metastasis to bone represents enhanced net osteoclastic activity. Alternatively, defective osteoclastogenesis can lead to increased bone mass or osteopetrosis. These pathologies can result from enhanced or inhibited osteoclast development, altered osteoclast function without change in number of osteoclasts, or both. The extent of bone resorption is directly related to the control of osteoclast differentiation. Osteoclasts derive from bone marrow derived macrophages (BMDM) under the influence of macrophage colony stimulating factor (M-CSF), receptor activator of NF-κB ligand (RANK-L), and incompletely understood co-stimulatory factors acting through ITAM-containing receptors to give rise to large, motile, multinucleated, terminally differentiated osteoclasts (1) (2) (3). Other RANK-L related inflammatory cytokines, such as TNFα and IL-1β, also influence osteoclastogenesis and function, either independently or in synergy with RANK-L (4). While M-CSF is thought to largely provide a survival/proliferative signal to macrophage precursor cells, RANK-L signals are critical for osteoclast differentiation (1). The cellular receptor for RANK-L, RANK, is a member of the TNFR superfamily that includes the IL-1 receptor and Toll-like receptors (5). Like other TNFRs, RANK recruits adapter proteins after ligand-induced multimerization. A central family of such adapters is the TNF receptor-associated factors or TRAFs. RANK binds multiple TRAFs but only TRAF6 has been shown to be critical for osteoclast development and function (6,7). In the absence of TRAF6 or in the presence of inhibitory peptides osteoclast differentiation is blocked (7,8). TRAFs share a common C-terminal TRAF domain that serves to localize TRAFs to their target proteins and http://www.jbc.org/cgi/doi/10.1074/jbc.M607399200 The latest version is at JBC Papers in Press. Published on November 8, 2006 as Manuscript M607399200 Copyright 2006 by The American Society for Biochemistry and Molecular Biology, Inc. by guest on April 6, 2018 http://www.jbc.org/ Downloaded from

Upload: trinhtram

Post on 07-Feb-2017

222 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: THE LIM PROTEIN, LIMD1, REGULATES AP-1 ACTIVATION

1

THE LIM PROTEIN, LIMD1, REGULATES AP-1 ACTIVATION THROUGH AN INTERACTION WITH TRAF6 TO INFLUENCE OSTEOCLAST DEVELOPMENT

Yungfeng Feng1,2,4, Haibo Zhao3,4, Hilary F. Luderer1,2, Holly Epple1,2, F. Patrick Ross2,3, Steven L.

Teitelbaum3, and Gregory D. Longmore1,2 From the Departments of Medicine1, Cell Biology2, and Pathology3, Washington University, St.

Louis, MO 63110 Running Title: LIM proteins and osteoclast development

Address correspondence to: Gregory D. Longmore, Division of Hematology, Washington University School of Medicine, 660 South Euclid Avenue, St. Louis MO 63110, Tel. 314-362-8834; Fax. 314-362 8826; E-mail: [email protected] 4These two authors contributed equally to this work.

Increasingly a number of proteins important in the regulation of bone osteoclast development have been shown primarily influence osteoclastogenesis under conditions of physiologic or pathologic stress. Why basal osteoclastogenesis is normal and how these proteins regulate stress osteoclastogenic responses, as opposed to basal osteoclastogenesis, is unclear. LIM proteins of the Ajuba/Zyxin family localize to cellular sites of cell adhesion where they contribute to the regulation of cell adhesion and migration, translocate into the nucleus where they can affect cell fate, but are also found in the cytoplasm where their function is largely unknown. We show that one member of this LIM protein family, LIMD1, is uniquely upregulated during osteoclast differentiation and interacts with TRAF6, a critical cytosolic regulator of RANK-L regulated osteoclast development. LIMD1 positively affects the capacity of TRAF6 to activate AP-1, and LIMD1 -/- osteoclast precursor cells are defective in the activation of AP-1 and thus induction of NFAT2. LIMD1 -/- mice, although having normal basal bone osteoclast numbers and bone density, are resistant to physiological and pathologic osteoclastogenic stimuli. These results implicate LIMD1 as a potentially important regulator of osteoclast development under conditions of stress. Vertebrate bone development and homeostasis requires the exquisite balance between the bone resorptive capacity of osteoclasts and bone forming capacity of osteoblasts. Perturbation of this equilibrium can have pathologic consequences. For example, pathologic bone loss, as occurs in osteoporosis, rheumatoid arthritis, Paget’s disease of

the bone (PDB), and tumor metastasis to bone represents enhanced net osteoclastic activity. Alternatively, defective osteoclastogenesis can lead to increased bone mass or osteopetrosis. These pathologies can result from enhanced or inhibited osteoclast development, altered osteoclast function without change in number of osteoclasts, or both. The extent of bone resorption is directly related to the control of osteoclast differentiation. Osteoclasts derive from bone marrow derived macrophages (BMDM) under the influence of macrophage colony stimulating factor (M-CSF), receptor activator of NF-κB ligand (RANK-L), and incompletely understood co-stimulatory factors acting through ITAM-containing receptors to give rise to large, motile, multinucleated, terminally differentiated osteoclasts (1) (2) (3). Other RANK-L related inflammatory cytokines, such as TNFα and IL-1β, also influence osteoclastogenesis and function, either independently or in synergy with RANK-L (4). While M-CSF is thought to largely provide a survival/proliferative signal to macrophage precursor cells, RANK-L signals are critical for osteoclast differentiation (1). The cellular receptor for RANK-L, RANK, is a member of the TNFR superfamily that includes the IL-1 receptor and Toll-like receptors (5). Like other TNFRs, RANK recruits adapter proteins after ligand-induced multimerization. A central family of such adapters is the TNF receptor-associated factors or TRAFs. RANK binds multiple TRAFs but only TRAF6 has been shown to be critical for osteoclast development and function (6,7). In the absence of TRAF6 or in the presence of inhibitory peptides osteoclast differentiation is blocked (7,8). TRAFs share a common C-terminal TRAF domain that serves to localize TRAFs to their target proteins and

http://www.jbc.org/cgi/doi/10.1074/jbc.M607399200The latest version is at JBC Papers in Press. Published on November 8, 2006 as Manuscript M607399200

Copyright 2006 by The American Society for Biochemistry and Molecular Biology, Inc.

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

Page 2: THE LIM PROTEIN, LIMD1, REGULATES AP-1 ACTIVATION

2

results in oligomerization of the N-terminal effector domain leading to the activation of IKK, NF-κB, the MAPKs (particularly JNK and p38), and AP-1 (9-11). Recently, TRAF6 was also found to form a complex with the atypical PKC-interacting adapter protein p62 in osteoclasts (aPKC/p62/TRAF6 complex), and this complex was shown to be important for the activation of NF-κB by RANK-L (12). The importance of p62 to bone physiology is evident as mutations in p62 have been identified in a group of patients with 5q35-linked Paget’s disease (13), and deletion of the p62 gene in mice results in normal basal bone structure but inhibited osteoclastogenic response to PTH challenge (12).

The NF-κB and AP-1 transcriptional complexes are particularly critical for osteoclast development (14) (15,16). AP-1 synergizes with the nuclear factor of activated T cells, NFAT1 (or NFATc2), transcription factor to induce transcription of NFAT2 (or NFATc1) (17) (18). NFAT2 expression is a critical cell fate determinant for osteoclast development (19). While there is abundant genetic evidence demonstrating the importance of AP-1 and NF-kB component proteins in the regulation of osteoclastogenesis (for review see (14)), it is still not completely understood how RANK-L induced TRAF6 activation leads to the activation of these transcriptional complexes. LIM domains are unique protein-protein interacting module found in multiple proteins throughout all cellular compartments. The Ajuba/Zyxin family, are cytosolic, complex LIM proteins that associate with cellular cytoskeletal components particularly at sites of cell adhesion where they can regulate cell-cell adhesion and migration (20-22), shuttle from sites of cell adhesion into the nucleus where they have the potential to affect cell fate (21,23,24), and are also found in the cytoplasm, free of cytoskeletal association, where their function is incompletely understood.

Recently one member, Ajuba, was found to interact with the p62 adapter protein and affect IL-1-induced NF-κB activation in epithelia (25). Within this family Ajuba is most closely related to LIMD1. LIMD1 was originally identified as a gene present at chromosome locus 3p21 in humans (26), which in human twin studies has been a locus implicated as containing gene(s) important for the regulation of bone mineral density (27). We found that the expression of LIMD1 is uniquely regulated during

osteoclast differentiation, interacts with TRAF6, and affects TRAF6’s ability to activate AP-1. LIMD1 null mice, although having normal basal bone osteoclast numbers and bone density, are resistant to physiological and pathologic osteoclastogenic stimuli. We discuss LIMD1’s potential role as a positive regulator of TRAF6 activity, only during states of osteoclastogenic stress.

Experimental Procedures

Antibodies, and Protein Purification - Mouse LIMD1 cDNA was PCR amplified from a mouse kidney cDNA library, sequenced, and subcloned. The full length protein and the N-terminal PreLIM region were cloned into pBacPAK9 containing a 6XHis-Flag tag (HF-PreLIM), baculovirus generated, and proteins purified from infected Sf9 cells using Talon metal affinity resin (Clontech). Polyclonal rabbit antiserum was raised against the purified PreLIM peptide and partially purified by mixing with purified HF-PreLIM peptide charged to PVDF membrane. Ajuba antiserum has previously been described (28) (20). Mouse monoclonal antibodies against p62 were from BD Transduction Laboratories (San Diego, CA). TRAF6 mouse monoclonal, and NFAT2, c-Jun, c-Fos, and TRAF6 rabbit polyclonal antibodies were from Santa Cruz (Santa Cruz, CA). M2AG (mouse monoclonal anti-Flag antibody immobilized on Agarose), HRP-conjugated monoclonal Anti-Flag, HRP-conjugated monoclonal Anti-Myc, HRP-conjugated monoclonal Anti-HA antibodies were all from Sigma (St. Louis, MO). Phospho-ERK, ERK1, phosphor-JNK, JNK1, phosphor-p38, p38, and phospho-c-Jun rabbit polyclonal antibodies were all from Cell Signaling (Beverly, MA). Nucleophosmin antibodies were from J. Weber (Washington University). P65 RelA antiserum was from Oncogene Research Products (San Diego, CA). In Vivo Bone Analyses - Bone mineral density (BMD) was determined in male mice, 6-7 weeks of age, by dual-energy x-ray absorptiometry (DEXA) using a PIXCImus2 scanner (Lunar Corporation, Madison, WI). Seven-week-old female mice were injected subcutaneously every 6 hours for 3 days with 10 µg hPTH peptide (1-34) (Bachem). Animals were sacrificed 24 hours after the last PTH injection. GST and GST-RANK-L were prepared, as described (29). Eight to ten

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

Page 3: THE LIM PROTEIN, LIMD1, REGULATES AP-1 ACTIVATION

3

week old mice were subcutaneously injected daily for 7 days at the base of the skull with 100 µg GST or GST-RANK-L, then sacrificed. In both instances following sacrifice calvariae were isolated, fixed overnight in 10% neutral buffered formalin, and decalcified in 14% EDTA for 4-5 days. Following sample dehydration, paraffin embedded sections were prepared and stained for TRAP with a hematoxylin counterstain. Histomorphometric analysis of Osteoclast number per millimeter of trabecular surface and percent surface covered by osteoclasts were measured and analyzed using Osteomeasure (OsteoMetrics, Atlanta, GA) in a blinded fashion. Three to four calvariae slices per mouse were analyzed and statistics performed with the unpaired Student’s t test. K/BxN serum (200 ml, IP) was injected into 8 week old mice on days 1 and 4. On day 7 clinical evidence for ankle inflammation was assessed and ankle swelling measured. Mice were sacrificed and ankles and foot prepared for histological analysis. Serum was also obtained on day 7 and levels of type I collagen C-terminal fragments determined, as described by manufacturer (Nordic Biosciences, Denmark). Bone Marrow Macrophage Isolation and Osteoclast Differentiation - Primary bone marrow derived macrophages (BMDM) were extracted from femora and tibia of 6- to 8-week-old mice and cultured overnight in α-10 medium (LPS-free α-MEM containing 10% inactivated fetal bovine serum). Nonadherent cells were collected by centrifugation and re-plated in fresh α-10 medium containing 1/10 vol. of CMG 14-12 culture supernatant (which was equivalent to 130 ng/ml of recombinant M-CSF) for 4 days. Fresh media and M-CSF were supplemented every other day. For osteoclast differentiation cells were washed in PBS, lifted and reseeded at 1.5 x 106 cells per 10 cm dish, in osteoclast differentiation medium (α-10 medium containing 10 ng/ml M-CSF and 100 ng/ml of recombinant RANK-L). Media was changed every other day. TRAP+ mononuclear prefusion osteoclast precursors (preOCs) are present after 3 days of differentiation while TRAP+ mature, terminally differentiated, multinucleated Osteoclast cells (OCs) are produced after 5 or 6 days in culture. TRAP staining was performed as described by manufacturer (Sigma).

Osteoclast bone matrix resorption assay - BMDM were plated on synthetic calcium matrices (BD Biosciences), and allowed to differentiate in the presence of 100 ng/ml RANK-L and 10 ng/ml MM-CSF for 10 days. Cells were removed with a bleach solution and calcium matrix washed 3 times with water. Resorption pits were analyzed by phase contrast microscopy and the total area of bone resorption was determined using Metamorph program (Molecular Devices). Retroviral production and macrophage transduction - Full-length mouse LIMD1 was cloned into the pMX-IRES-Bsr retrovirus vector (30), and transiently transfected into Plat-E packaging cells (31) using FuGENE 6 Transfection Reagent (Roche). Virus was collected 48 h after transfection. BMDMs were infected with virus for 24 h in the presence of M-CSF and 4 µg/ml polybrene (Sigma). Cells were then selected in the presence of M-CSF and 1 µg/ml Blasticidin (Calbiochem) for 3 days. Subcellular Fractionation - Subconfluent cultures of BMDMs, preOCs or OCs were starved of serum, M-CSF, and RANK-L in α-MEM for 6 hours, then stimulated with RANK-L (100 ng/ml). Cell fractionation (nuclear, cytosolic, and total cell lysates) was performed as described (25). Protein concentration was determined with Bio-Rad protein assay kit. For Western blots, 5 µg of nuclear extract or 25 µg of cytosolic or total cell lysate were loaded in each lane. Immunoprecipitation and Western blots - For immunoprecipitation cells were harvested, washed with cold PBS, and lysed with IP buffer (20 mM HEPES, pH 7.5, 120 mM NaCl, 5 mM NaF, 1 mM sodium orthovanidate, 0.5 mM EDTA, 1 mM DTT, 5% glycerol, 0.1% NP-40, and protease inhibitor cocktail from Sigma). Extracts were clarified by centrifugation at 15,000g for 15 min. For each IP, cell extract proteins were mixed with primary antibody or pre-immune serum on ice for 2 hours, and then incubated with 25 µl protein AG/slurry (1:1 v/v) overnight with gentle rotation at 4ºC. The immunoprecipitates were washed five times with IP buffer, and boiled in SDS-loading buffer. After SDS-PAGE, under reducing conditions, products were transferred to nitrocellulose ECL membrane

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

Page 4: THE LIM PROTEIN, LIMD1, REGULATES AP-1 ACTIVATION

4

and subjected to Western blot analysis with ECL detection reagent (Amersham). Electrophoretic Mobility Shift Assay (EMSA) - DNA oligos were labeled with biotin at the 5’ ends during synthesis, annealed and purified as described (25). 3 ug of nuclear extract were mixed with EMSA binding buffer containing 10 mM HEPES, pH 7.5, 1.5 mM MgCl2, 50 mM KC1, 2.5% glycerol, 1 mM DTT, 0.5 mM EDTA, 0.2 ug of poly (dI:dC), 0.2 nM NF-κB or AP-1 binding site to a final volume of 20 µl and kept at room temperature for 30 min. The mixture was then subjected to 6% polyacrylamide gel electrophoresis. Biotin-complex signals were developed with Pierce kit following the manufacturer’s instruction. Oligos used: NF-κB: 5’-Biotin-AAGTTGAGGGGACTTTCCCAGGCT-3’ and 5’-Biotin-AGCCTGGGAAAGTCCCCTCAACTT-3’. AP-1: 5’-Biotin-ACGCTTGATGACTCAGCCGGAAT-3’ and 5’-Biotin-ATTCCGGCTGAGTCATCAAGC-3’. Luciferase Assay – HEK293T cells (6 x 104 cells/well) were transfected with pAP-1-luciferase (0.025 µg/well, Stratagene), pTK-renilla luciferase (0.025 µg/well, Promega), pcDNA3-Flag-mLIMD1 isoforms (0.1 µg/well), pcDNA3-Flag-mTRAF6 (0.1 µg/well). Total plasmid amount was balanced to 0.25 µg/well with pcDNA3-Stop-mLIMD1 (stop codon after the ATG start site) or pcDNA3, as needed. Forty-eight hours post transfection cells were lysed in 100 µl lysis buffer (Promega) and firefly and renilla luciferase activity determined using substrates from Promega and a luminometer.

RESULTS LIMD1 associates with TRAF6

LIMD1 was found to interact with the atypical PKC interacting adapter protein p62/sequestosome in a yeast 2-hybrid protein-protein interactive screen (25). Since the LIMD1 gene localizes to a chromosomal locus, 3p21 (26), a region implicated in the regulation of bone density in humans (27) and p62 appears to be important for bone osteoclast development (12), we asked whether LIMD1 might contribute to osteoclast development.

The expression pattern of LIMD1, and related LIM proteins, during osteoclast

differentiation was determined. Bone marrow derived macrophages (BMDM) were isolated from wild type adult mice, expanded in M-CSF, and then induced to differentiate into osteoclasts by adding RANK-L. At each of six days in cultures containing M-CSF and RANK-L, total cell protein extracts were prepared and the level of LIMD1 protein determined by quantitative Western blot. BMDM were found to express low levels of LIMD1 protein (Fig. 1), but during RANK-L-induced osteoclast differentiation LIMD1 protein levels significantly increased (Fig. 1). This occurred, in part, at the level of transcriptional regulation as RT-PCR analysis revealed an increase in LIMD1 mRNA during osteoclast differentiation (data not shown). The induced expression of LIMD1 protein during osteoclast differentiation was similar to previously reported induced expression of TRAF6 (12), an important cytosolic regulator of osteoclast development (Fig. 1). Ajuba, a closely related LIM protein abundant in epithelia (20) that also interacts with p62 (25), was present in BMDM, at a low level, but in contrast to LIMD1 its level did not change during osteoclast differentiation (Fig. 1). Likewise, the level of other related LIM proteins: WTIP, Zyxin and LPP, did not fluctuate significantly during osteoclast differentiation (data not shown). These results indicated that of the Ajuba/Zyxin LIM protein family LIMD1 protein expression was uniquely upregulated during osteoclast differentiation, and thus, might contribute to osteoclast development and bone homeostasis.

When bone osteoclast progenitors are stimulated with RANK-L, p62 assembles into a multiprotein complex containing TRAF6 and aPKC (12). Since TRAF6 has been shown to be a critical regulator of RANK-L induced osteoclast development and function (6,7), we asked whether LIMD1 associates with TRAF6. HEK293T cells were co-transfected with epitope-tagged plasmids expressing LIMD1 and TRAF6, LIMD1 was immunoprecipitated, and bound products Western blotted for the presence of TRAF6. LIMD1 was found to associate with TRAF6 (Fig. 2A). Mapping studies identified the C-terminal LIM region as directing LIMD1’s interaction with TRAF6 (Fig. 2A). The N-terminal PreLIM region of LIMD1 did not interact with TRAF6 (Fig. 2A). The association of LIMD1 with TRAF6 was specific, as related LIM proteins LPP and Zyxin did not co-

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

Page 5: THE LIM PROTEIN, LIMD1, REGULATES AP-1 ACTIVATION

5

immunoprecipitate with TRAF6 (Fig. 2B), nor did TRAF2 interact with LIMD1 (data not shown).

To establish that LIMD1 and TRAF6 interact in physiologically relevant primary cells expressing endogenous levels of each protein, LIMD1 was immunoprecipitated from primary prefusion osteoclast precursors (preOCs) and bound products Western blotted for the presence of TRAF6. LIMD1 immunoprecipitation from wild type cells contained TRAF6 when compared to LIMD1 immunoprecipitates from LIMD1 -/- cells (see later for description and characterization of LIMD1 -/- mice), but only a small fraction of total cell TRAF6 interacted with LIMD1 (Fig. 2C). In addition LIMD1 immunoprecipitates also contained p62 and the aPKC, PKCζ (Fig. 2C). These results indicated that LIMD1 associated with TRAF6 in primary osteoclast precursors, in a multiprotein complex.

Since LIMD1 interacts with both p62 and TRAF6, and TRAF6 also associates with p62 (32), it is possible that the LIMD1-TRAF6 interaction occurs indirectly through an interaction of each protein with p62. To determine whether LIMD1 could interact directly with TRAF6, in the absence of p62, we purified His-Flag-tagged LIMD1 protein (HF-LIMD1) from baculovirus-infected Sf9 insect cells and GST-TRAF6 protein from bacteria (Fig. 2D, lower panel). When mixed, in vitro, LIMD1 and TRAF6 proteins readily interacted, at low concentration, while LIMD1 did not interact with control GST (Fig. 2D) indicating that LIMD1 could interact directly with TRAF6, in the absence of p62.

To determine whether the presence of p62 influenced the interaction between LIMD1 and TRAF6 HEK293T cells were cotransfected with fixed amounts of LIMD1 and TRAF6 and increasing amounts of p62. LIMD1 was immunoprecipitated and bound products Western blotted for the presence of TRAF6 and p62. Although p62 was detected in LIMD1 immunoprecipitates it did not influence the amount of TRAF6 associated with LIMD1 (Fig. 2E). In summary LIMD1 associated with TRAF6 in cells, including primary osteoclast precursors. Although LIMD1, p62, and TRAF6 were present in a multiprotein complex, p62 did not affect the amount of TRAF6 associated with LIMD1. Finally LIMD1 could interact directly with TRAF6, in vitro, in the absence of p62.

LIMD1 positively influences TRAF6’s ability to activate AP-1 TRAF6 regulates osteoclast differentiation by mediating the activation of AP-1 and NF-κB; two important transcriptional complexes required for osteoclast differentiation (14-16), in response to RANK-L stimulation. To determine if the interaction between LIMD1 and TRAF6 was functionally relevant HEK293T cells were co-transfected with an AP-1 regulated luciferase reporter plasmid and LIMD1 expressing plasmids (Fig. 3A), in the absence or presence of TRAF6. Luciferase activity was determined and values normalized for transfection efficiency. As expected (10), TRAF6 alone (no added LIMD1) induced a 2-fold increase in AP-1 activity (Fig. 3B, column set 1). In the presence of exogenous LIMD1, TRAF6-mediated AP-1 activity was further enhanced another 2-3 fold (Fig. 3B, column set 2). In control experiments, expression of LIMD1 alone in the absence of TRAF6 had no effect upon AP-1 activity (Fig 3B, column set 2, white column). When excess p62 was added LIMD1 mediated potentiation of TRAF6 AP-1 activation was inhibited (Fig. 3C, column 4). Indeed p62 alone inhibited TRAF6 induced AP-1 activation (Fig. 3C, column 5). These results, coupled with results in Fig. 2E, suggested the possibility of an AP-1 activating LIMD1-TRAF6 complex, and a p62-TRAF6 or LIMD1-p62-TRAF6 inhibitory complexes. Nonetheless, under these over-expression conditions, LIMD1 enhanced TRAF6 mediated activation of AP-1.

Since the LIM region (3 LIM domains) of LIMD1 mediated its interaction with TRAF6 (Fig. 2A) we asked whether any specific LIM domain preferentially interacted with TRAF6 to effect AP-1 activation, and whether a physical interaction between LIMD1 and TRAF6 was required for LIMD1 to enhance TRAF6 activation of AP-1. HEK293T cells were co-transfected with myc-tagged TRAF6 and various Flag-tagged LIMD1 mutants, as depicted in Figure 3A. The PreLIM region alone did not interact with TRAF6 (Figs. 2A and 3D) or cooperate with TRAF6 to activate AP-1 (Fig. 3B, column set 3). When either LIM1 or LIM2 domains, but not the LIM3 domain, were added to the PreLIM region of LIMD1, these LIMD1 isoforms now associated with TRAF6 (Fig. 3D). The LIM1 interaction appeared stronger than LIM2, in this assay, as more LIM1 containing protein co-immunoprecipitated with TRAF6 (Fig. 3D). Next

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

Page 6: THE LIM PROTEIN, LIMD1, REGULATES AP-1 ACTIVATION

6

we asked whether physical mapping correlated with the ability of specific LIM domains to augment TRAF6 mediated AP-1 activation. Only LIM1, not LIM2 or LIM3, cooperated with TRAF6 to activate AP-1 (Fig. 3B, column set 4 versus 5 and 6, respectively).

Mapping studies of the domain in TRAF6 that mediates its interaction with LIMD1 revealed multiple points of contact between TRAF6 and LIMD1. The first LIM domain of LIMD1 interacted with the ring domain (1-143) of TRAF6 while the second LIM domain of LIMD1 interacted with the C-terminal TRAF domain (Suppl. Fig. 1). In summary these experiments indicated that LIMD1, through an interaction with TRAF6, enhanced the capacity of TRAF6 to activate AP-1. A role for LIMD1 in bone osteoclast development

To determine the biological significance of the LIMD1-TRAF6 interaction LIMD1 null mice were generated by homologous recombination. Two different ES clones were used to generate chimeric SV129J-C57BL/6 mice and mice heterozygous for the targeted allele were recovered. Interbreeding between heterozygous mice resulted in the birth of mice of all three genotypes in the expected Mendelian ratios. PCR and Southern blot analyses of genomic DNA (Suppl. Fig. 2, and data not shown) from adult tails revealed that LIMD1 -/- mice were viable and reached adulthood without any obvious phenotypes. LIMD1 -/- mice were fertile. Western blot analyses of primary fibroblasts (MEFs) from LIMD1 null (-/-), heterozygote (+/-), or littermate wild type (+/+) control mice demonstrated that there was no detectable LIMD1 protein in LIMD1 -/- cells, and that the expression of related Ajuba/Zyxin family members: Ajuba, Zyxin, LPP, and Trip6 were not significantly altered by deletion of the LIMD1 gene (Suppl. Fig. 2, and data not shown).

There was no gross phenotypic difference between adult (8-16 week old) LIMD1 -/- and control wt mice. Dual-energy X-ray absorptiometry (DEXA) analysis of long bones and the axial skeleton of mice revealed no difference in bone density (data not shown), and histological analysis of long bones and calvarial bones revealed no difference in the number of osteoclast present in the basal state of LIMD1 -/- mice (Fig. 4C and D, white columns – no experimental manipulation). To determine if osteoclast precursor numbers, at the

basal state, was altered in LIMD1 -/- mice, equal numbers of BMDM isolated from LIMD1 +/+ and -/- mice were placed in ex vivo cultures containing M-CSF and RANK-L, and the number of TRAP+, multinucleated, mature OCs formed scored. There was no significant difference in the number of OCs formed between genotypes (Fig. 4A). When equal numbers of BMDM were differentiated on a bone matrix and resorption measured (pit formation) there was again no significant functional difference noted between genotypes (Fig. 4B, and suppl. Fig. 4).

Increasingly analysis of mice deficient in genes important for bone osteoclast biology have revealed basal osteoclast numbers and function that are normal, while there is a clear defect in the osteoclastogenic response to physiologic or pathologic stressors (12,33-36). Therefore we sought to determine the in vivo response of LIMD1 -/- mice to three different osteoclastogenic stimuli: short course parathyroid hormone (PTH) (33) (12), RANK-L (33) (36), and a mouse model of serum-induced inflammatory arthritis (37) (35).

LIMD1 -/- and control wt mice were injected with either: (1) hPTH(1-34) peptide four times daily for 3 days, (2) recombinant GST-RANK-L, or control GST, for 7 consecutive days, or (3) serum from K/BxN mice, a strain that develops spontaneous arthritis (37), on day 1 and 4. The osteoclastogenic response was determined histologically, following TRAP staining (to identify osteoclasts) of calvaria (PTH and RANK-L) or ankle/foot joints and bones (K/BxN serum). In all cases basal numbers of osteoclasts (OC number per mm bone surface or OC area of the bone surface) were not significantly different in control uninjected wt versus LIMD1 -/- mice (Fig. 4). While control wt mice exhibited the expected increase in OC number following each maneuver, the osteoclastogenic response in LIMD1 -/- mice was significantly blunted to all three stimuli (Fig. 4C-E). Histological data for graphs C, D, and E in figure 4 are shown in supplemental figure 3.

For K/BxN serum-injected mice both wt and LIMD1 -/- mice developed the expected inflammatory response (Suppl. Fig. 3C). In agreement with a diminished osteoclastogenic response of LIMD1 -/- mice to K/BxN serum there was also evidence for less bone matrix derived collagen type I degradation, in serum assays (Fig. 4E). Therefore, although LIMD1 -/- mice have normal basal bone structure they are relatively

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

Page 7: THE LIM PROTEIN, LIMD1, REGULATES AP-1 ACTIVATION

7

resistant to PTH and RANK-L stimulated osteoclastogenesis. Furthermore, they are also resistant to the pathologic osteoclastogenic response induced by induction of arthritis following injection of K/BxN mouse serum.

LIMD1 influences activation of AP-1, and subsequent NFAT2 expression, in primary osteoclast precursor cells The transcription factor complex AP-1 (14) and NFAT proteins (19) are important regulators of osteoclastogenesis. These transcription factors cooperate to effect the transcription of genes critical for osteoclast cell fate determination (17,38). In response to RANK-L, NFAT2 (NFATc1) transcription is induced by the cooperative action of an AP-1/NFAT1 (NFATc2) transcription complex. TRAF6 activity is central to the activation of AP-1 by RANK-L in developing OCs. Since LIMD1 interacted with TRAF6 (Fig. 2), affected TRAF6 mediated AP-1 activation in transient co-transfection experiments (Fig. 3), and influenced RANK-L induced osteoclast development in mice in vivo (Fig. 4) we asked whether, in the absence of LIMD1, RANK-L-induced activation of AP-1, NFAT2, and NF-κB is altered in developing primary osteoclasts. BMDM were isolated from LIMD1 -/- and wt adult mice, and expanded in M-CSF for 4 days. RANK-L (100 ng/ml) was then added for 3 days to generate immature prefusion osteoclasts (preOCs) or for 6 days to generate mature fully differentiated, multinucleated osteoclasts (OCs). At each stage of differentiation cells were starved of serum, M-CSF, and RANK-L for 6 hours (0 time point) and then stimulated with only RANK-L (100 ng/ml) for the times indicated. Nuclear extracts and cytosolic extracts were prepared from each time point for subsequent analyses. Nucleophosmin (nucleolar protein) and Actin Western blots were used to confirm equal loading between samples for nuclear and cytosolic extracts, respectively. In BMDM only slight changes in RANK-L-induced AP-1 and NF-κB activity were apparent in LIMD1 -/- compared to wt control cells (Fig. 5A). However, in preOC cells RANK-L-induced AP-1 activation was dramatically inhibited in LIMD1 -/- cells (Fig. 5B, left panel). By both EMSA assay and quantitative Western blot for nuclear c-Fos, c-Jun, and phospho-c-Jun protein levels AP-1 activity was diminished in the absence of LIMD1. NF-κB activity, as assessed by EMSA, was also inhibited but not to the same degree as AP-

1 (Fig. 5B, left panel). We also determined the level of c-Jun and c-Fos proteins in cytosolic extracts of LIMD1 -/- relative to control wt cells. Both c-Jun and c-Fos protein levels were decreased (Fig. 5B, right panels). In terminally differentiated, mature OC cells AP-1 activity was still diminished in LIMD1 -/- cells versus wt controls, while NF-κB activity was minimally altered (Fig. 5C). The minimal effect of LIMD1 deletion on RANK-L signaling in BMDM with more dramatic effects seen in day 3 preOCs and day 6 OCs may reflect the induction of LIMD1 and TRAF6 protein expression during osteoclast differentiation (see Fig. 1). Importantly, reintroduction of LIMD1 into LIMD1 -/- BMDM prior to osteoclast differentiation completely rescued all RANK-L-induced signaling abnormalities of LIMD1 -/- preOCs (Fig. 6). NF-κB activity, as determined by nuclear EMSA and p65 RelA protein nuclear translocation (Fig. 6A); AP-1 activity, as determined by nuclear EMSA, nuclear c-Jun and c-Fos protein levels (Fig. 6A), and total cellular c-Jun and c-Fos levels (Fig. 6B) all normalized following reintroduction of LIMD1 into LIMD1 -/- cells. The absence of LIMD1 also affected the induction of NFAT2 expression during RANK-L induced OC differentiation. BMDM were expanded in M-CSF and then induced to form preOCs by adding RANK-L to cultures for 3 days. Nuclear and cytosolic extracts were prepared and Western blotted for NFAT2. In both nuclear and cytosolic extracts the level of NFAT2 expression was dramatically less in LIMD1 -/- cells (Fig. 6C). Importantly reintroduction of LIMD1 into LIMD1 -/- cells normalized NFAT2 levels (Fig. 6C). Taken together these results indicated that LIMD1 affects RANK-L activation of AP-1 during osteoclast differentiation, in a cell-intrinsic manner. Due to the diminished AP-1 activity little NFAT2 was produced, possibly explaining why LIMD1 -/- mice were resistant to RANK-L challenge. Diminished NF-κB activity in LIMD1 -/- cells may also contribute to the blunted osteoclastogenic response of LIMD1 -/- mice. LIMD1 affects RANK-L induced JNK and p38 activation RANK-L stimulated TRAF6 regulates AP-1 activity through activation of MAPKs. JNK1 and p38 activity, in particular, are important for osteoclastogenesis (17,39,40). In RANK-L stimulated LIMD1 -/- day 3 preOCs c-Jun

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

Page 8: THE LIM PROTEIN, LIMD1, REGULATES AP-1 ACTIVATION

8

phosphorylation was found to be dramatically decreased (Fig. 5B), suggesting that JNK activity, in particular, was inhibited in the absence of LIMD1. Thus we determined the activation profile of the MAPKs ERK, JNK, and p38 during RANK-L-induced osteoclast differentiation. Day 3 prefusion osteoclast progenitors (preOCs) were generated from wt and LIMD1 -/- mice. Total cell lysates were then analyzed for ERK, JNK, and p38 activity following RANK-L stimulation. The most notable differences were diminished JNK and p38 activation in LIMD1 -/- preOCs (Fig. 7). There were minimal or no change in RANK-L-induced ERK activity (Fig. 7). Reintroduction of LIMD1 into LIMD1 -/- cells corrected the defect in JNK activation by RANK-L (data not shown). These results indicated that during osteoclast differentiation LIMD1 affected RANK-L-induced JNK and p38 activation, particularly at the preOC stage.

DISCUSSION

We have identified a new TRAF6 interacting protein, the cytosolic LIM protein LIMD1, which positively affects TRAF6 activity. LIMD1 interacts with TRAF6 to enhance TRAF6 induced activation of AP-1. This interaction is biologically relevant. Although LIMD1 -/- mice have normal basal bone structure they exhibit resistance to osteoclastogenic challenges, both physiologic (PTH and RANK-L) and pathologic (serum-induced arthritis). Mechanistically LIMD1 associates with TRAF6 in primary osteoclasts, and osteoclast precursors from LIMD1 -/- mice are defective in the activation of AP-1 activation, and thus NFAT2 induction, in response to RANK-L. These results indicate that LIMD1 positively regulates TRAF6 signals, and thus, OC development but only in times of physiologic or pathologic stress. This is not a unique, or novel phenotype. Increasingly mice genetically deficient in a number of proteins important for the regulation of osteoclast development exhibit normal basal osteoclastogenesis yet are resistant to osteoclastogenic challenges. These include p62/sequestosome (12), NIK (NF-κB inducing kinase) (33) (35), the LIM protein FHL2 (four and a half LIM proteins) (36), and IKKα (inhibitor of κB kinase α) (34). How then could LIMD1 influence osteoclast development only in times of stress?

One possibility is that bone OC LIMD1 expression is regulated, in vivo. During basal osteoclastogenesis LIMD1 levels are low, while in response to stress LIMD1 levels increase and contribute to osteoclastogenesis, in these settings; as seen for a related LIM protein FHL2 (36). Some consider ex vivo differentiation of BMDM into OC to represent a “stress” situation, and not representative of basal osteoclastogenesis. During ex vivo OC differentiation LIMD1 levels are low in BMDM and increase during OC differentiation. Despite multiple approaches, we have not been able to correlate this with an in vivo response as our antisera (2 different ones) to LIMD1 do not specifically detect LIMD1 in bone immunohistochemical analyses.

In response to RANK-L, TRAF6 is recruited to the cytoplasmic tail of the RANK receptor. RANK has multiple TRAF6 binding sites and it appears that the number of TRAF6 binding sites is important to effect osteoclastogenesis. Decreasing the number of TRAF6 binding sites in RANK reduced its capacity to support osteoclastogenesis while over-expressing CD40 or increasing the number of TRAF6 binding sites in CD40 allowed CD40 activation to now support osteoclastogenesis (41,42). These data suggest that the strength of TRAF6 signal (i.e., quantitative) is a critical factor in its osteoclastogenic regulation. Therefore, LIMD1 may function, in effect, to increase the amount of TRAF6 recruited to RANK receptors following RANK-L stimulation. However, over-expression of LIMD1 did not increase the amount of TRAF6 co-immunoprecipitated with the RANK receptor (data not shown), making the possibility that LIMD1 affects TRAF6 recruitment to RANK receptor less likely. In response to RANK-L, TRAF6 interacts with p62 that also associates with the atypical Protein Kinase C (aPKC) to form a TRAF6/p62/aPKC multiprotein complex (12). Where this complex localizes in cells, whether it influences the amount of TRAF6 associated with the RANK receptor, and what effect it has upon total cellular TRAF6 activity is not known, but stress osteoclastogenesis in p62 null mice is inhibited (basal osteoclastogenesis is normal in p62 null mice) (12). LIMD1 interacts with p62, TRAF6, and aPKC and LIMD1 immunoprecipitates from primary osteoclasts contain, in addition to TRAF6, p62 and aPKC. Therefore, LIMD1 may affect the assembly

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

Page 9: THE LIM PROTEIN, LIMD1, REGULATES AP-1 ACTIVATION

9

and function of the TRAF6/p62/aPKC multi-protein complex. How this complex affects TRAF6 activity, and whether this identifies an active pool of TRAF6 distinct from TRAF6 recruited to the RANK receptor requires further investigation. The LIMD1/TRAF6/p62/aPKC complex may represent a sequential cytosolic signaling complex formed following RANK receptor recruitment of TRAF6, as has been shown for signaling by the RANK-related TNF receptor I (43). Finally although LIMD1 and p62 interact with TRAF6 and LIMD1 positively affects TRAF6 activity, the bone phenotype and osteoclastogenic response of LIMD1 and p62 null mice are less severe than that observed in TRAF6 null mice (6,7). This may reflect the small proportion of cellular TRAF6 detected in LIMD1 immunoprecipitates (Fig. 2C). TRAF6 clearly also functions in a LIMD1-independent manner. Another LIM protein, four and a half LIM domains 2 (FHL2), was recently found to affect osteoclastogenesis (36) and osteoblast development (44). FHL2 interacts with TRAF6 but in contrast to LIMD1, this interaction negatively influences TRAF6 activity by displacing TRAF6 from RANK. FHL2 is distinct from the LIMD1 family of LIM proteins as it is a LIM domain only protein, not a complex type LIM protein like LIMD1 (21). FHL2 also functions prominently in the nucleus as a transcriptional coactivator of androgen receptors (45), AP-1 (46), and Runx2 (44). Whether the activity of FHL2 as an AP-1 transcriptional coactivator contributes to its role in osteoclastogenic

regulation has not been addressed. The LIMD1 family of LIM proteins also translocate into the nucleus (47), and recently LIMD1 was found to interact with the retinoblastoma gene (Rb) to affect Rb transcriptional regulation of E2F target genes (47). Interestingly Rb also functions as a direct transcriptional coactivator of Runx2 to promote osteoblast differentiation (48). Considering that FHL2 affects both osteoclast and osteoblast differentiation, it is compelling to consider whether LIMD1 might also affect osteoblast development and function.

Finally, mutations in the p62 gene have been identified in a subset of patients suffering from Paget’s disease (13), but precisely how these p62 mutations contribute to the bone pathology of Paget’s disease is not clear. Since LIMD1 interacts with p62, and LIMD1 null and p62 null mice have overlapping phenotypes, this raises the possibility that LIMD1 might influence the manifestations of p62 mutations present in Paget’s disease of the bone.

ACKNOWLEDGEMENTS

This work was supported by grants CA75315 and the Washington University/Pfizer biomedical research program to GDL; AR032788, AR046523, and AR048853 to SLT; and AR046852 and AR048812 to FPR. We thank Dr. Roberta Faccio for helpful suggestions and comments.

REFERENCES

1. Boyle, W. J., Simonet, W. S., and Lacey, D. L. (2003) Nature 423(6937), 337-342 2. Koga, T., Inui, M., Inoue, K., Kim, S., Suematsu, A., Kobayashi, E., Iwata, T., Ohnishi,

H., Matozaki, T., Kodama, T., Taniguchi, T., Takayanagi, H., and Takai, T. (2004) Nature 428(6984), 758-763

3. Reddy, S. V. (2004) Crit Rev Eukaryot Gene Expr 14(4), 255-270 4. Kwan Tat, S., Padrines, M., Theoleyre, S., Heymann, D., and Fortun, Y. (2004) Cytokine

Growth Factor Rev 15(1), 49-60 5. Dunne, A., and O'Neill, L. A. (2003) Sci STKE 2003(171), re3 6. Lomaga, M. A., Yeh, W., 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., Paige, C., Lacey, D., Dunstan, C., Boyle, W., Goeddel, D. V., and Mak, T. W. (1999) Genes & Dev. 13, 1015-1024

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

Page 10: THE LIM PROTEIN, LIMD1, REGULATES AP-1 ACTIVATION

10

7. Naito, A., Azuma, S., Tanaka, S., Miyazaki, T., Takaki, S., Takatsu, K., Nakao, K., Nakamura, K., Katsuki, M., Yamamoto, T., and Inoue, J. (1999) Genes Cells 4(6), 353-362

8. Ye, H., Arron, J. R., Lamothe, B., Cirilli, M., Kobayashi, T., Shevde, N. K., Segal, D., Dzivenu, O. K., Vologodskaia, M., Yim, M., Du, K., Singh, S., Pike, J. W., Darnay, B. G., Choi, Y., and Wu, H. (2002) Nature 418(6896), 443-447

9. Arch, R. H., Gedrich, R. W., and Thompson, C. B. (1998) Genes Dev 12(18), 2821-2830 10. Baud, V., Liu, Z. G., Bennett, B., Suzuki, N., Xia, Y., and Karin, M. (1999) Genes Dev

13(10), 1297-1308 11. Bradley, J. R., and Pober, J. S. (2001) Oncogene 20(44), 6482-6491 12. Duran, A., Serrano, M., Leiyges, M., Flores, J. M., Picard, S., Brown, J. P., and Diaz-

Meco, M. T. (2004) Dev. Cell 6, 303-309 13. Laurin, N., Brown, J. P., Morissette, J., and Raymond, V. (2002) Am J Hum Genet 70(6),

1582-1588 14. Wagner, E. F., and Matsuo, K. (2003) Ann Rheum Dis 62 Suppl 2, ii83-85 15. Franzoso, G., Carlson, L., Xing, L., Poljak, L., Shores, E. W., Brown, K. D., Leonardi,

A., Tran, T., Boyce, B. F., and Siebenlist, U. (1997) Genes Dev 11(24), 3482-3496 16. Iotsova, V., Caamano, J., Loy, J., Yang, Y., Lewin, A., and Bravo, R. (1997) Nat Med

3(11), 1285-1289 17. Ikeda, F., Nishimura, R., Matsubara, T., Tanaka, S., Inoue, J., Reddy, S. V., Hata, K.,

Yamashita, K., Hiraga, T., Watanabe, T., Kukita, T., Yoshioka, K., Rao, A., and Yoneda, T. (2004) J Clin Invest 114(4), 475-484

18. Macian, F., Lopez-Rodriguez, C., and Rao, A. (2001) Oncogene 20(19), 2476-2489 19. Takayanagi, H., Kim, S., Koga, T., Nishina, H., Isshiki, M., Yoshida, H., Saiura, A.,

Isobe, M., Yokochi, T., Inoue, J., Wagner, E. F., Mak, T. W., Kodama, T., and Taniguchi, T. (2002) Dev Cell 3(6), 889-901

20. Marie, H., Pratt, S. J., Betson, M., Epple, H., Kittler, J. T., Meek, L., Moss, S. J., Troyanovsky, S., Attwell, D., Longmore, G. D., and Braga, V. M. (2003) J Biol Chem 278(2), 1220-1228

21. Kadrmas, J. L., and Beckerle, M. C. (2004) Nat Rev Mol Cell Biol 5(11), 920-931 22. Pratt, S. J., Epple, H., Ward, M., Feng, Y., Braga, V. M., and Longmore, G. D. (2005) J

Cell Biol 168(5), 813-824 23. Nix, D. A., and Beckerle, M. C. (1997) J Cell Biol 138(5), 1139-1147 24. Kanungo, J., Pratt, S. J., Marie, H., and Longmore, G. D. (2000) Mol Biol Cell 11(10),

3299-3313 25. Feng, Y., and Longmore, G. D. (2005) Mol Cell Biol 25(10), 4010-4022 26. Kiss, H., Kedra, D., Yang, Y., Kost-Alimova, M., Kiss, C., O'Brien, K. P., Fransson, I.,

Klein, G., Imreh, S., and Dumanski, J. P. (1999) Hum Genet 105(6), 552-559 27. Wilson, S. G., Reed, P. W., Bansal, A., Chiano, M., Lindersson, M., Langdown, M.,

Prince, R. L., Thompson, D., Thompson, E., Bailey, M., Kleyn, P. W., Sambrook, P., Shi, M. M., and Spector, T. D. (2003) Am J Hum Genet 72(1), 144-155

28. Goyal, R. K., Lin, P., Kanungo, J., Payne, A. S., Muslin, A. J., and Longmore, G. D. (1999) Mol Cell Biol 19(6), 4379-4389

29. McHugh, K. P., Hodivala-Dilke, K., Zheng, M. H., Namba, N., Lam, J., Novack, D., Feng, X., Ross, F. P., Hynes, R. O., and Teitelbaum, S. L. (2000) J Clin Invest 105(4), 433-440

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

Page 11: THE LIM PROTEIN, LIMD1, REGULATES AP-1 ACTIVATION

11

30. Onishi, M., Nosaka, T., Misawa, K., Mui, A. L., Gorman, D., McMahon, M., Miyajima, A., and Kitamura, T. (1998) Mol Cell Biol 18(7), 3871-3879

31. Morita, S., Kojima, T., and Kitamura, T. (2000) Gene Ther 7(12), 1063-1066 32. Sanz, L., Diaz-Meco, M. T., Nakano, H., and Moscat, J. (2000) EMBO J 19, 1576-1586 33. Novack, D. V., Yin, L., Hagen-Stapleton, A., Schreiber, R. D., Goeddel, D. V., Ross, F.

P., and Teitelbaum, S. L. (2003) J Exp Med 198(5), 771-781 34. Chaisson, M. L., Branstetter, D. G., Derry, J. M., Armstrong, A. P., Tometsko, M. E.,

Takeda, K., Akira, S., and Dougall, W. C. (2004) J Biol Chem 279(52), 54841-54848 35. Aya, K., Alhawagri, M., Hagen-Stapleton, A., Kitaura, H., Kanagawa, O., and Veis

Novack, D. (2005) J Clin Invest 115(7), 1848-1854 36. Bai, S., Kitaura, H., Zhao, H., Chen, J., Muller, J. M., Schule, R., Darnay, B., Novack, D.

V., Ross, F. P., and Teitelbaum, S. L. (2005) J Clin Invest 115(10), 2742-2751 37. Matsumoto, I., Maccioni, M., Lee, D. M., Maurice, M., Simmons, B., Brenner, M.,

Mathis, D., and Benoist, C. (2002) Nat Immunol 3(4), 360-365 38. Teitelbaum, S. L. (2004) J Clin Invest 114(4), 463-465 39. Matsumoto, M., Sudo, T., Saito, T., Osada, H., and Tsujimoto, M. (2000) J Biol Chem

275(40), 31155-31161 40. David, J. P., Sabapathy, K., Hoffmann, O., Idarraga, M. H., and Wagner, E. F. (2002) J

Cell Sci 115(Pt 22), 4317-4325 41. Kadono, Y., Okada, F., Perchonock, C., Jang, H. D., Lee, S. Y., Kim, N., and Choi, Y.

(2005) EMBO Rep 6(2), 171-176 42. Gohda, J., Akiyama, T., Koga, T., Takayanagi, H., Tanaka, S., and Inoue, J. (2005) Embo

J 24(4), 790-799 43. Micheau, O., and Tschopp, J. (2003) Cell 114(2), 181-190 44. Gunther, T., Poli, C., Muller, J. M., Catala-Lehnen, P., Schinke, T., Yin, N., Vomstein,

S., Amling, M., and Schule, R. (2005) Embo J 24(17), 3049-3056 45. Muller, J. M., Isele, U., Metzger, E., Rempel, A., Moser, M., Pscherer, A., Breyer, T.,

Holubarsch, C., Buettner, R., and Schule, R. (2000) Embo J 19(3), 359-369 46. Morlon, A., and Sassone-Corsi, P. (2003) Proc Natl Acad Sci U S A 100(7), 3977-3982 47. Sharp, T. V., Munoz, F., Bourboulia, D., Presneau, N., Darai, E., Wang, H. W., Cannon,

M., Butcher, D. N., Nicholson, A. G., Klein, G., Imreh, S., and Boshoff, C. (2004) Proc Natl Acad Sci U S A 101(47), 16531-16536

48. Thomas, D. M., Carty, S. A., Piscopo, D. M., Lee, J. S., Wang, W. F., Forrester, W. C., and Hinds, P. W. (2001) Mol Cell 8(2), 303-316

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

Page 12: THE LIM PROTEIN, LIMD1, REGULATES AP-1 ACTIVATION

12

FIGURE LEGENDS Fig. 1. LIMD1 expression is regulated during osteoclast differentiation. BMDM from wt mice, were expanded in M-CSF, and osteoclast differentiation initiated by adding RANK-L (100 ng/ml). Each day following RANK-L addition an aliquot of cells were lysed, and equal amounts of protein from each sample run on SDS-PAGE, and Western blotted for the indicated proteins. Fig. 2. LIMD1 interacts with TRAF6. A, A stick figure representation of the modular organization of LIMD1. The C-terminal LIM region is comprised of three LIM domains, while the PreLIM region is the N-terminal end of the protein. HEK293T cells were transiently co-transfected with myc-TRAF6, full length Flag-LIMD1, Flag-PreLIM region, or Flag-LIM region as indicated. LIMD1 isoforms were immunoprecipitated (anti-Flag) and bound products Western blotted for the presence of TRAF6 (anti-myc) and LIMD1 (anti-Flag). Input controls are shown on the right panels. B, HEK293 cells were transiently co-transfected with Flag-TRAF6, myc-tagged LIM proteins: LPP, Zyxin, or LIMD1. TRAF6 was immunoprecipitated (anti-Flag) and bound products Western blotted for the presence of bound LIM protein (anti-myc) and TRAF6 (anti-Flag). Input controls are shown in lanes 3 and 4 of each panel. C, LIMD1 +/+ or -/- BMDM were differentiated to form preOCs, cells lysed and endogenous LIMD1 immunoprecipitated. Bound products were Western blotted for the presence of TRAF6, p62, PKCζ, and LIMD1. Input controls are shown in lanes 3 and 4. D, Aliquots of purified His-Flag-tagged LIMD1 (HF-LIMD1), purified GST, and GST-TRAF6 were run on SDS-PAGE and the gel Coumassie stained (lower panel). Control GST (lane 1) or GST-TRAF6 (lane 2) proteins were mixed with HF-LIMD1. Glutathione-agarose beads added, pelleted, washed and bound products incubated with Prescission enzyme to cleave TRAF6 from GST. The final reaction was then run on SDS-PAGE and Western blotted for LIMD1 and TRAF6. Input controls are shown in lanes 3 and 4. In lane 2 both cleaved TRAF6 and uncleaved GST-TRAF6 are present. E, p62 does not affect the amount of TRAF6 associated with LIMD1. As in A and B HEK293T cells were transfected with the indicated plasmids, LIMD1 immunoprecipitated and bound products Western blotted for the presence and amount of TRAF6, p62 and LIMD1. Fig. 3. LIMD1 affects TRAF6 activation of AP-1. A, A stick figure representation of LIMD1 mutants used. NES, nuclear export signal. B and C, AP1 transcription assays. HEK293T cells were co-transfected with the following plasmids: AP1 regulated firefly-luciferase, pTK-renilla luciferase, and CMV promoter driven LIMD1, TRAF6, or p62 expressing plasmids, as indicated. In all tubes equal amount of DNA were transfected. Following transfection Firefly-luciferase activity was determined. All samples were normalized to renilla luciferase levels. White columns are absence of TRAF6, black columns presence of TRAF6. D, Mapping of LIM domain that interacts with TRAF6. HEK293T cells were cotransfected with myc-tagged TRAF6 and Flag-tagged LIMD1 mutants. LIMD1 isoforms were immunoprecipitated (anti-Flag) and bound products Western blotted for the presence of TRAF6 (anti-myc) and LIMD1 (anti-Flag). Input controls are shown in lanes 7-12. NS identifies a nonspecific band detected in Flag Western blots. Fig. 4. LIMD1 -/- mice are resistant to physiologic and pathologic stress osteoclastogenic stimuli. A, Osteoclast precursor numbers in LIMD1 +/+ and LIMD1 -/- mice. Equal numbers of BMDM were cultured in 10 ng/ml M-CSF and 100 ng/ml RANK-L. After 5 days cultures were stained with TRAP and multinucleated mature OCs counted and reported. B, Osteoclast resorption assay. Equal numbers of BMDM were plated of bone matrix and differentiated in M-CSF and RANK-L. Resorption pits were identified and quantified. C-E, LIMD1 +/+ or -/- mice were uninjected or injected with PTH (C), n=6 for each, GST or GST-RANK-L (D), n=4 for each, or uninjected or injected with serum from K/BxN mice (E) n=4 for each and histomorphometric analyses of TRAP stained calvariae (C,D) or ankles and foot (E) performed. The number of TRAP+ OC per mm bone surface and the % bone surface covered by OC are graphed. * Indicates a significant difference between +/+ and -/- samples (P<0.01, students t test A and B; P<0.03 for C).

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

Page 13: THE LIM PROTEIN, LIMD1, REGULATES AP-1 ACTIVATION

13

Fig. 5. LIMD1 -/- osteoclast precursors are defective in AP-1 and NF-κB activation in response to RANK-L stimulation. BMDM (A), day 3 preOCs (B), and day 6 mature OCs (C) were starved of serum, M-CSF, and RANK-L for 6 hours then stimulated with RANK-L (100 ng/ml) for the indicated times. Nuclear and cytosolic extracts were prepared from each time point. Nuclear extracts were analyzed (A, B, C) by EMSA to detect NF-κB and AP-1 activity, Western blotted for AP-1 components c-Fos, c-Jun, phospho-c-Jun (P-c-Jun), and as a loading control the nucleolar protein nucleophosmin. Cytosolic extracts (B only) were analyzed by Western blot for AP-1 component c-Fos and c-Jun, and as a loading control actin. Fig. 6. LIMD1 re-expression in LIMD1 -/- cells rescues RANK-L induced AP-1 and NF-κB activation, and induction of NFAT2. A and B, LIMD1 -/- and LIMD1 rescued LIMD1 -/- (-/-, +LIMD1) preOCs were starved of serum, M-CSF, and RANK-L for 6 hours then stimulated with RANK-L (100 ng/ml) for the indicated times. Nuclear and cytosolic extracts were prepared from each time point. Nuclear extracts (A) were analyzed by EMSA to detect NF-κB and AP-1 activity and Western blotted for AP-1 components c-Fos and c-Jun, nuclear translocation of NF-κB p65 RelA, and the nucleolar protein nucleophosmin served as a loading control. Cytosolic extracts (B) were analyzed by Western blot for AP-1 components c-Fos and c-Jun, p65 RelA, and Actin served as a loading control. C, Wild type (+/+), LIMD1 -/- (-/-), and LIMD1 -/- rescued with LIMD1 (-/-, +LD1) BMDM were differentiated to preOCs by adding M-CSF and RANK-L (100 ng/ml) for 3 days. Nuclear (left) and cytosolic (right) extracts were then Western blotted for the level of NFAT2. Nucleophosmin and Actin served as loading controls for nuclear and cytosolic extracts, respectively. Fig. 7. JNK and p38 activation are diminished in LIMD1 -/- developing osteoclasts. Day 3 preOCs were starved of serum, M-CSF, and RANK-L for 6 hours then stimulated with RANK-L (100 ng/ml) for the indicated times. Total cell extracts were prepared from each time point and Western blots for phosphorylated ERK (P-ERK), JNK (P-JNK), and p38 (P-p38) (activated enzymes) were performed. Equivalent loading per lane was demonstrated through Western blots for total ERK, JNK, or p38 protein.

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

Page 14: THE LIM PROTEIN, LIMD1, REGULATES AP-1 ACTIVATION

14

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

Page 15: THE LIM PROTEIN, LIMD1, REGULATES AP-1 ACTIVATION

15

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

Page 16: THE LIM PROTEIN, LIMD1, REGULATES AP-1 ACTIVATION

16

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

Page 17: THE LIM PROTEIN, LIMD1, REGULATES AP-1 ACTIVATION

17

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

Page 18: THE LIM PROTEIN, LIMD1, REGULATES AP-1 ACTIVATION

18

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

Page 19: THE LIM PROTEIN, LIMD1, REGULATES AP-1 ACTIVATION

19

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

Page 20: THE LIM PROTEIN, LIMD1, REGULATES AP-1 ACTIVATION

20

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

Page 21: THE LIM PROTEIN, LIMD1, REGULATES AP-1 ACTIVATION

Teitelbaum and Gregory D. LongmoreYungfeng Feng, Haibo Zhao, Hilary F. Luderer, Holly Epple, F. Patrick Ross, Steven L.

TRAF6 to influence osteoclast developmentThe LIM protein, LIMD1, regulates AP-1 activation through an interaction with

published online November 8, 2006J. Biol. Chem. 

  10.1074/jbc.M607399200Access the most updated version of this article at doi:

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

to choose from all of JBC's e-mail alertsClick here

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from