Isolation and characterization of a cDNA clone encoding a novel

JOURNAL OF CELLULAR PHYSIOLOGY 177:636 – 645 (1998)

Isolation and Characterization of a cDNA Clone Encoding a Novel Peptide (OSF) That Enhances Osteoclast Formation and Bone Resorption SAKAMURI REDDY,1 ROWENA DEVLIN,1 CHEIKH MENAA,1 RIKO NISHIMURA,2 SUN JIN CHOI,1 MARK DALLAS,1 TOSHI YONEDA,1 AND G. DAVID ROODMAN1* 1 Department of Medicine/Hematology/Endocrinology, the Veterans Administration Medical Center and The University of Texas Health Science Center, San Antonio, Texas 2 Department of Biochemistry, Osaka University, Osaka, Japan Using an expression cloning approach, we identified and cloned a novel intracellular protein produced by osteoclasts that indirectly induces osteoclast formation and bone resorption, termed OSF. Conditioned media from 293 cells transiently transfected with the 0.9 kb OSF cDNA clone stimulated osteoclast-like cell formation in both human and murine marrow cultures in the presence or absence 10-9 M 1,25-dihydroxyvitamin D3. In addition, conditioned media from 293 cells transfected with the OSF cDNA clone enhanced the stimulatory effects of 1,25(OH)2D3 on bone resorption in the fetal rat long bone assay. In situ hybridization studies using antisense oligomers showed expression of OSF mRNA in highly purified osteoclast-like cells from human giant cell tumors of the bone. Northern blot analysis demonstrated ubiquitous expression of a 1.3 kb mRNA that encodes OSF in multiple human tissues. Sequence analysis showed the OSF cDNA encoded a 28 kD peptide that contains a c-Src homology 3 domain (SH3) and ankyrin repeats, suggesting that it was not a secreted protein, but that it was potentially involved in cell signaling. Consistent with these data, immunoblot analysis using rabbit antisera against recombinant OSF demonstrated OSF expression in cell lysates but not in the culture media. Furthermore, recombinant OSF had a high affinity for c-Src, an important regulator of osteoclast activity. Taken together, these data suggest that OSF is a novel intracellular protein that indirectly enhances osteoclast formation and osteoclastic bone resorption through the cellular signal transduction cascade, possibly through its interactions with c-Src or other Src-related proteins. J Cell Physiol 177:636 – 645, 1998. © 1998 Wiley-Liss, Inc.

Recent evidence suggests that factors produced by osteoclasts (OCLs) play an important role in regulating OCL formation and activity. We have shown that interleukin (IL)-6 is produced by OCLs and can stimulate OCL-like multinucleated cell (MNC) formation in longterm marrow cultures (Kurihara et al., 1990). IL-6 plays an important role in the bone resorption process (Reddy et al., 1994) and may be involved in the increased bone remodeling in Paget’s disease. Similarly, transforming growth factor beta (TGFb; Oursler, 1994) is also produced by OCLs and inhibits OCL activity. Thus, it is likely that a multiplicity of factors produced by OCLs play a critical role in normal bone remodeling. To identify other such factors that enhance or suppress OCL formation and bone resorption, we constructed a cDNA mammalian expression library prepared from highly purified OCL-like cells formed in long-term human marrow cultures. Using this approach, we recently identified Annexin II as an autocrine factor secreted by OCL that can stimulate OCL formation and © 1998 WILEY-LISS, INC.

activity (Takahashi et al., 1994). This article describes the identification, cloning, and characterization of a cDNA clone encoding a novel intracellular 28 kD peptide, termed OSF, that is not secreted and indirectly enhances OCL formation and bone resorption. This peptide contains an Src homology (SH3) domain, ankyrin repeats, and most likely is involved in the Contract grant sponsor: The Veterans Administration; Contract grant sponsor: National Institutes of Health; Contract grant spoonsor: NIAMS; Contract grant numbers: AG 39529 and AR 41336; Contract grant sponsor: NIADDK; Contract grant number: AM 35188; Contract grant sponsor: NCI; Contract grant number: CA 40035; Contract grant sponsor: San Antonio Cancer Institute; Contract grant number: 5P30CA54174. *Correspondence to: Dr. G. David Roodman, Research Service (151), Audie Murphy Veterans Administration Hospital, 7400 Merton Minter Boulevard, San Antonio, TX 78284. E-mail: [email protected] Received 27 January 1998; Accepted 21 July 1998

OCL STIMULATING FACTOR

cellular signal transduction cascade. The end result of this process is the expression of a potentially novel factor that is secreted by 293 cells that enhances OCL formation and bone resorption. MATERIALS AND METHODS Materials All restriction enzymes used were obtained from New England Biolabs (Beverly, MA). Chemicals were purchased either from Sigma Chemical Co. (St. Louis, MO) or from Promega (Madison, WI). The 23c6 monoclonal antibody was generously provided by Dr. Michael Horton (St. Bartholomew’s Hospital, London, UK). Screening of the mammalian expression library A mammalian cDNA expression library from highly purified human OCL-like MNCs was constructed using a CMV promoter driven eukaryotic expression vector, pcDNAI (Invitrogen, Carlsbad, CA) and screened for OCL stimulatory factors as described (Takahashi et al., 1994). Plasmid DNA isolated from the OCL mammalian expression library pools was transiently transfected into 293 cells (4 3 105) grown in individual 35 mm wells using the calcium phosphate method with a kit from Stratagene (La Jolla, CA) following the manufacturer’s protocol. Conditioned media obtained from each pool were collected after 48 h and tested at different concentrations (0.1–10% v/v) for their capacity to enhance MNC formation in human as well as in murine marrow cultures. At the end of the culture period, the number of MNCs formed in human and murine marrow cultures was scored as previously described (Takahashi et al., 1994) or tartrate-resistant acid phosphatase (TRAP) activity was measured in the culture lysates (Oreffo et al., 1988). Pools that contained MNC stimulatory activity were progressively subfractionated until they only contained one cDNA clone that enhanced MNC formation. This cDNA was then sequenced by standard techniques, and the sequence compared to those present in GenBank. Bone marrow cultures and bone resorption assays Human as well as murine bone marrow cells were cultured as described (Takahashi et al., 1994) to form OCL-like MNC in vitro. Human bone marrow nonadherent mononuclear cells were cultured with or without 10-9 M 1,25-(OH)2D3 or conditioned media from 293 cells transiently transfected with the OSF cDNA clone, hereafter termed “OSF conditioned media.” In selected experiments, neutralizing antibodies to human IL-1, tumor necrosis factor alpha (TNFa), or IL-6 at 500 mg/ml were added to cultures stimulated with OSF conditioned media. After 3 weeks, the human marrow cultures were fixed with 2% formaldehyde in phosphate-buffered saline (PBS) and stained with the 23c6 monoclonal antibody. Murine marrow cultures were overlaid on sterile dentin. At the end of the culture, they were removed and fixed in 2% glutaraldehyde and stained for TRAP activity. The number of TRAP-positive cells on each dentin slice was scored. The number of resorption pits and the area of the dentin resorbed were determined with an inverted microscope and Java image analysis software (Jandel Scientific, Corte Madona, CA).

637

Fetal rat bone organ cultures Fetal rat long bones were 45Ca radiolabeled in utero as described (Raisz and Neimann, 1969). Fetal bones (radii and ulnae) were dissected free of surrounding tissue and then cultured in 0.5 ml of chemically defined medium (Sigma) supplemented with 1 mg/ml bovine serum albumin and penicillin-streptomycin (50 units/ ml) at 37°C in an atmosphere of 5% CO2/air. The radii and ulnae were incubated for 24 h in control medium to allow removal of the exchangeable 45Ca before transferring to control or OSF conditioned media in the presence or absence of 10-9 M 1,25-(OH)2D3. Control or experimental media were changed after 72 h, and the bone explants incubated for a total of 5 days. Boneresorbing activity was measured as the percentage of total 45Ca released from the bone into the medium over 5 days of incubation. Amplification of the 5´ end of OSF mRNA We identified the 5´ end of the human OSF mRNA using a 5´ Amplifinder RACE kit (Clontech, Palo Alto, CA) following the manufacturer’s protocol. Total RNA was isolated from OCL-like giant cells that were purified from giant cell tumors (Chomczynski and Sacchi, 1987). cDNA was synthesized with the antisense OSF mRNAspecific primer (5´-TCTGCCTGCTCAGCCACATAG-3´) at positions 180 –200 bp relative to 11 ATG initiation codon. A single-stranded anchor oligonucleotide (3´-NH3GGAGACTTCCAAGGTCTTAGCTATCACTTAAGCAC5´) containing an EcoR1 restriction enzyme site, as underlined, was ligated to the 3 end of the cDNA. A portion of the cDNA was then used as a template for polymerase chain reaction (PCR) amplification, by using a primer complementary to the anchor and a nested OSF genespecific primer (5´-TAATTCATCTGGAGTTCTGGG-3´) position 72–92 bp. The PCR reaction mixture was incubated at 94°C for 1 min and then 35 cycles of PCR were conducted at 94°C for 45 sec, 60°C for 45 sec, and 72°C for 2 min. The PCR product (312 bp) was subcloned into the TA cloning vector (Invitrogen, San Diego, CA) and sequenced with T7 and SP6 promoter-specific complementary primers. Northern blot analysis Approximately 2 Mg of poly(A)1 RNA isolated from multiple human tissues was separated on a denaturing formaldehyde 1.2% agarose gel and blot transferred to a charged nylon membrane (Clontech) and fixed by ultraviolet (UV) irradiation. The filters were hybridized with a [32P]-labeled OSF cDNA probe in a solution containing 25 mM KPO4, pH 7.4, 53 SSC, 53 Denhardt’s solution, 50 mg/ml salmon sperm DNA, and 50% formamide. After washing in 0.23 SSC, 0.1% sodium dodecyl sulfate (SDS) solution at 55°C, the filters were exposed to Kodak XAR film at -70°C. In situ hybridization OCL-like giant cells were isolated from human giant cell tumors of bone as described (Reddy et al., 1994). The cells were plated in LabTek chamber slides and allowed to adhere. The cells were fixed in 4% paraformaldehyde and permeabilized by treatment with 0.2 M HCl for 5 min at room temperature. Nonspecific background was reduced by treatment with 0.25% (v/v)

638

REDDY ET AL.

acetic anhydride in 0.1 M triethanolamine-HCl, pH 8.0, for 5 min. The slides were washed in RNAse-free water and prehybridized for 1 h at 42°C in a solution containing 50% formamide, 53 Denhardt’s solution, 53 SSC, 50 mM sodium phosphate, pH 6.5, 250 mg/ml sheared salmon sperm DNA, 250 mg/ml yeast tRNA, and 0.1% SDS. Hybridization was carried out in 50 ml of the same buffer containing [35S]-labeled antisense or sense oligomers to OSF mRNA (106 cpm/ml) at 42°C for 24 h. The slides were washed at 37°C in 23 SSC to remove unbound probe. The slides were then washed once for 30 min in 50% formamide/23 SSC/2-ME at 42°C, twice for 10 min in 23 SSC-0.1% SDS at room temperature, and once for 10 min in 0.13 SSC-0.1% SDS at room temperature. The slides were air dried and autoradiography was performed using Kodak nuclear tracking emulsion (NTB2) and developing reagents as previously described (Reddy et al., 1996). The slides were counterstained with 0.5% (v/v) methyl green. Cells hybridized with a sense oligomer served as controls for these experiments. Synthesis of recombinant OSF Recombinant OSF was synthesized in BL21 E. coli strain using the pET14b vector system (Novagen, Inc., Madison, WI) as described by the manufacturer’s protocol. The nucleotide sequence encoding the OSF peptide was PCR amplified as described using sense (5´-CCAGGAAGCATATGTCGAAGCCGCCACCC-3´) and antisense (5´-AAGGATCCTATTAATCTGAGTCTTCATC-3´) primers. Sequences underlined represent NdeI and BamH1 restriction endonuclease sites added to the sense and antisense primer sequences, respectively, for cloning purposes. The PCR product (approximately 650 bp) amplified using the OSF cDNA template was digested with NdeI and BamH1 restriction endonucleases and then cloned into the pET14b vector in frame with the 6x His tag followed by a thrombin cleavage site. The plasmid construct, 5FpET2, was transformed into the BL21 (DE3) pLysS E. coli host, and the synthesis of recombinant OSF was induced by 0.4 mM IPTG. Recombinant OSF (28 kD) was purified to homogeneity by binding to a nickel affinity column followed by elution with 13 elution buffer containing 0.5 M imidazole, 0.5 M NaCl, and 20 mM Tris base, pH 7.9. 6x His tag was then removed by cleavage with biotinylated thrombin in 20 mM Tris-HCl, pH 8.4, 150 mM NaCl and 2.5 mM CaCl2 containing cleavage buffer. Thrombin-free recombinant OSF was then obtained by incubating with streptavidin- coupled magnetic beads (Dynal, Inc., Great Neck, NY) followed by separation using a magnet. Immunoblot and immunocytochemical analysis Polyclonal antisera against E. coli recombinant OSF were raised in rabbits (Biodesign, Kennebunk, ME) and used to determine OSF expression by immunoblot analysis. Total 293 cell lysates were prepared using lysis buffer (20 mM Tris, pH 7.4, NaCl 150 mM, 1% Triton X100, 10% glycerol, 1.5 mM MgCl2, 1 mM EGTA, 200 mM sodium vanadate, 1 mM PMSF, and aprotinin [1 mg/ml]). Samples were then subjected to SDS-polyacrylamide gel electrophoresis (PAGE), using 12% gels (BioRad, Hercules, CA). For immunoblot analysis, proteins were transferred from

Fig. 1. The effect of OSF conditioned media on human bone marrow MNC formation. Various concentrations of conditioned media from 293 cells transfected with the OSF cDNA clone were added to normal human bone marrow long-term cultures in the absence (A) or in the presence (B) 10-9 M 1,25-(OH)2D3. However, due to the potent synergism between OSF conditioned media and 1,25-(OH)2D3, vastly different percentages of OSF conditioned media are used in panels A and B. After 3 weeks, the cultures were fixed and stained with the 23c6 monoclonal antibody. Results represent the mean 1 SEM for four determinations for a typical experiment. Similar results were seen in four independent experiments. *P , 0.05 compared to cultures without conditioned media.

SDS gels onto a NitroBind nitrocellulose membrane (Micron Separations Inc., Westboro, MA). After blocking with 5% non-fat dry milk in 150 mM NaCl, 50 mM Tris, pH 7.2, 0.05% Tween 20 (TBST) buffer, the membrane was incubated for 1 h with the polyclonal antibody to OSF diluted 1:4,000 in 5% non-fat dry milk-TBST. The blots then were incubated for 1 h with horseradish peroxidase-conjugated goat antirabbit IgG (Sigma) diluted 1:10,000 in 5% non-fat dry milk-TBST and developed using an ECL system (Amersham, Arlington Heights, IL). For immunocytochemical studies, OCL-like cells were purified from human giant cell tumors of bone and cultured in LabTech slides for 24 h, as described previously (Reddy et al., 1994). The cells were then fixed with 2% formaldehyde for 20 min and washed with PBS. In order to reduce the nonspecific binding of antibodies, the cells were treated with 1% bovine serum albumin in PBS for 30 min. The cells were then incubated with a rabbit antiserum raised against recombinant OSF or preimmune serum at a 1:200 dilution in PBS for 1 h and washed three times with PBS. The cells were then incubated with biotinylated antirabbit IgG and stained using an ABC kit from Vector Laboratories (Burlingame, CA) following the manufacturer’s protocol. Synthesis of glutathione-S-transferase (GST)-cSrc/SH3-SH2 domain fusion protein c-Src cDNA, kindly provided by Dr. Hanafusa (Rockefeller University, New York, NY), was used as the template to PCR amplify the SH3-SH2 domains using a sense primer 5´-CGCGGATCCGGGAGCAGCAAGAGCAAGCCC-3´ and an antisense primer, 5´-CGCGAATTCCCTATAGGTTCTCTCCAGGCTG-3´. The sequences underlined in the sense and antisense primers


639

Fig. 2. The effect of OSF conditioned media on mouse bone marrow MNC formation and dentin resorption. Various concentrations of conditioned media from 293 cells transfected with OSF cDNA clone were added to normal mouse bone marrow cultures in the absence (A) or in the presence (B) 10-9 M 1,25-(OH)2D3 to normal mouse bone marrow cultures on dentin slices. However, different percentages of OSF conditioned media were used in panels A and B due to the synergistic effect of OSF conditioned media and 1,25-(OH)2D3. After 6 days, the dentin slices were fixed and stained for TRAP. The number of TRAPpositive MNC was counted. The number and areas of resorption lacunae were determined using an image analysis system (C). The number of pits per slice (D), the area of resorption lacunae formed per OCL (E), and the number of pits per OCL (F) were also determined. Results represent the mean 1 SEM for four cultures. Similar results were seen in three independent experiments. *P , 0.05 compared to control cultures containing conditioned media from mock-transfected 293 cells plus 10-9 M 1,25-(OH)2D3.

represent BamH1 and EcoR1 restriction enzyme sites, respectively, added in order to clone the PCR amplified product in frame with GST in the pGS-Tag vector. The purified E. coli recombinant GST-c-Src/SH3-SH2 domain fusion protein was then coupled to glutathioneagarose beads (Ron and Dressler, 1992). Immunoprecipitation of OSF with anti-c-Src The OSF cDNA clone was transiently expressed in 293 cells. The cell lysates were prepared using cell lysis

buffer. Five hundred micrograms of total cellular extract was incubated with 1 mg of anti-c-Src antibody (clone 317) for 1 h, and the immune complex was separated using a protein G agarose conjugate. A control using mouse IgG was run in parallel. The immune complex was washed four times with buffer containing 1% Triton X100, 20 mM Tris, pH 7.4, 150 mM NaCl, and 1 mM EGTA and subjected to SDS-PAGE. Western blot analysis was performed using OSF antisera as described above.

640

REDDY ET AL.

Fig. 3. The effect of OSF conditioned media on fetal rat long bone resorption. The conditioned media (CM) from 293 cells transfected with the OSF cDNA clone (70%, v/v) were added to fetal rat long bones with 10-9 M 1,25-(OH)2D3. The bones were treated and processed as described under Materials and Methods. Results represent the mean 1 SEM for four determinations. Similar results were seen in three independent experiments. OSF conditioned media with 1,25-(OH)2D3 (10-9 M) was significantly different from 1,25-(OH)2D3 with conditioned media from mock-transfected 293 cells. *P , 0.01 compared to control.

Statistical analysis Results are presented as the means 1 SEM for quadruplicate determinations. Differences between means were assessed by a two-way analysis of variance for repeated measures. Results were considered significantly different for P , 0.05. RESULTS Screening of the cDNA expression library The cDNA expression library was fractionated into 200 pools, each containing 2,000 clones. Pools were initially screened by transiently transfecting 293 cells with cDNAs from each pool and testing conditioned media from the transfected 293 cells for their capacity to enhance OCL-like MNC formation in human marrow cultures. Conditioned media (0.1–5%) obtained from each pool were added to normal human bone marrow long-term cultures in the presence or absence of 10-9 M 1,25-(OH)2D3. The number of OCL-like MNCs that reacted with the 23c6 monoclonal antibody was counted and compared with cultures not containing conditioned media or cultures treated with conditioned media from 293 cells that were mock transfected. One of the six positive pools (number 24) that did not contain known OCL-stimulating factors, as assessed by PCR, was further subdivided into ten subpools, each containing 100 –200 clones. cDNAs from the individual subpools were tested for their ability to stimulate OCLlike MNC formation as described (Takahashi et al., 1994). A single positive subpool (number 24.5) was identified and further fractionated into 96-well plates containing one clone per well. Conditioned medium obtained from each of these clones was tested for its capacity to stimulate OCL-like MNC formation. One positive clone was detected and termed OSF. Conditioned media from 293 cells transfected with the isolated cDNA clone stimulated 23c6-positive cell formation in the human marrow culture system approxi-

Fig. 4. Nucleotide and deduced amino acid sequence of OSF mRNA. The conserved SH3 domain analogous to GRB2 and Fyn proto-oncogene product is boxed. A potential N-glycosylation site is hatch-boxed. The proline-rich sequence is underlined and the ankyrin repeat containing regions are shaded. The stop codon is denoted by an asterisk.

mately 200-fold compared to the initial activity detected with pool 24. Effects of OSF conditioned media on MNC formation and bone-resorbing capacity We tested the effects of conditioned media from 293 cells transiently transfected with the OSF cDNA clone on OCL-like MNC formation in mouse and human bone marrow cultures. In the human bone marrow cultures, the conditioned media increased 23c6 antibody-positive MNC formation at a concentration of 10% v/v (Fig. 1A). In mouse bone marrow cultures, TRAP-positive MNC formation was increased at a concentration of 30% v/v in the absence of 1,25-(OH)2D3 (Fig. 2A). In the presence of 10-9 M 1,25-(OH)2D3, the OSF conditioned media significantly increased MNC formation at concentrations as low as 0.025% v/v in the human and murine bone marrow cultures (Figs. 1B, 2B). When dentin slices were placed in the murine marrow cultures stimulated with conditioned media from OSF cDNA-transfected 293 cells, the area of dentin resorbed and the number of pits per dentin slice were significantly increased, as were the number of TRAP-positive cells per dentin slice in the presence of 10-9 M 1,25-(OH)2D3 (Fig. 2C,D). Treatment of murine marrow cultures with OSF conditioned media significantly increased the pit area and the number of pits per OCL formed compared to that seen with 1,25-(OH)2D3 (10-9 M) treatment alone (Fig. 2E,F). The number of pits per OCL in cultures treated with OSF was approximately one. These data suggest that OSF not only recruits OCLs, but also


641

Fig. 6. Northern blot analysis of OSF mRNA expression in human tissues. Poly(A1) RNA (2 mg) isolated from multiple human tissues was separated by formaldehyde agarose gel electrophoresis and blot transferred onto a nylon membrane. The filter was hybridized with [32P]-radiolabeled OSF cDNA probe as described.

Fig. 5. 5´ RACE analysis of OSF mRNA derived from OCL-like cells. The PCR product (312 bp) is shown by an arrow (lane 1). M, DNA size marker.

activates them to resorb bone and form larger pits per OCL. Similarly, in the fetal rat bone resorption assay, the cultures treated with OSF conditioned media (70% v/v) from 293 cells did not significantly increase 45Ca release over the mock-transfected 293 cell conditioned media, but significantly enhanced the effect of 1,25(OH)2D3 (Fig. 3). Sequence analysis of OSF mRNA The cDNA clone isolated contained a 900 bp sequence. A GenBank homology search against the OSF mRNA nucleotide sequence (accession number U63717) revealed a high degree of homology (.80%) with partial sequence of human pancreatic EST cDNA (Takeda et al., 1993). Recently, a murine homologue, SH3P2 protein, containing an SH3 domain and three ankyrin repeat regions, has been isolated (Sparks et

al., 1996a). OSF and SH3P2 exhibit 67% identity of nucleotide sequence and 95% identity with respect to deduced amino acid sequence. However, there is a limited identity (44%) for the 5´ untranslated sequence. Sequence analysis further showed one large open reading frame spanning 642 bp and lacked a poly(A1) signal sequence. The OSF mRNA sequence and the deduced 214 amino acid sequence encoding a 28 kD peptide is shown in Figure 4. The OSF peptide contained a proline-rich region located at the N-terminal end, a potential N-glycosylation site, and lacked an N-terminal signal sequence or hydrophobic sequence that could serve as a transmembrane domain, suggesting that OSF was not secreted. Protein motif database searches against the OSF peptide sequence further revealed the presence of the Src homology 3 domain analogous to the GRB2 (human growth factor receptor bound protein)/Fyn proto-oncogene tyrosine kinase with 62% and 63% homology, respectively, and ankyrin repeats, further supporting OSF being an intracellular protein. Analysis of OSF gene expression In order to determine if the OSF cDNA clone was a full-length clone that contained the entire 5´ end of the OSF mRNA, 5´ RACE was performed as described us-

642

REDDY ET AL.

Fig. 7. Localization of OSF expression in highly purified giant cells from human osteoclastoma of bone. Giant cells were hybridized with [35S]-labeled control sense oligomer (A) and antisense oligomer to OSF mRNA (B) as described in Materials and Methods. Immunocy-

tochemical staining of giant cells with rabbit preimmune serum (C) and OSF antiserum (D). Magnifications for A and B were 3100. Magnifications for C and D were 350.

ing total RNA derived from OCL-like cells isolated from human osteoclastomas as the template. As shown in Figure 5, a single PCR product was obtained which, upon sequence analysis, revealed that the 5´ end of the OSF cDNA was shorter by 36 bp compared to the 5´ end of OSF mRNA. Northern blot analysis of 2 mg of poly(A1) RNA isolated from different normal human tissues showed ubiquitous expression of transcripts (1.3 kb) encoding the OSF peptide (Fig. 6). This would also suggest that the isolated cDNA clone is truncated with approximately 0.4 kb of the 3´ untranslated sequence. In situ hybridization studies were performed to confirm the expression of OSF mRNA in OCLs. As shown in Figure 7, giant cells isolated from human osteoclastomas demonstrated strong hybridization with antisense oligomers to OSF mRNA (Fig. 7B) in contrast to sense oligomers (Fig. 7A), further demonstrating OSF mRNA expression in OCLs. OCL-like cells isolated from human osteoclastomas, as well as OCL-like cells formed in long-term human marrow cul-

tures (MacDonald et al., 1986), also demonstrated positive immunocytochemical staining using rabbit antisera (1:100 dilution) raised against recombinant OSF (Fig. 7D). In contrast, a control IgG derived from preimmunized rabbit serum did not demonstrate positive staining (Fig. 7C). Immunoblotting of OSF and affinity binding with c-Src Total cellular extracts obtained from human marrow cultures, 293 cells transiently transfected with the OSF cDNA expression construct, and mock-transfected control cell lysate samples were analyzed on SDSPAGE. Western blot analysis using rabbit antisera against OSF detected a single band identical with the predicted size. However, OSF or human bone marrow culture conditioned media that were not concentrated or were 20-fold concentrated using Centriprep-10 filters did not demonstrate OSF expression by immuno-


643

Fig. 8. Immunoblot analysis of OSF. Western blot analysis of OSF expression in 293 cells transiently transfected with the OSF cDNA clone in parallel with mock-transfected control cell lysates. OSF conditioned media and human bone marrow cultures were concentrated 20-fold by ultrafiltration. E. coli produced recombinant (r-OSF) served as a positive control. The total cell lysates were prepared using cell lysis buffer. About 50 mg protein of each sample was subjected to SDS-PAGE and blot transferred onto a NitroBind membrane. Immune detection was performed using rabbit antisera raised against recombinant OSF and ECL system. E. coli recombinant OSF (r-OSF) containing 6x His tag has slightly less mobility compared to the native form and is shown as a positive control.

blot analysis using OSF antisera (Fig. 8), suggesting that OSF was not secreted into the culture media. Since the protein structure of OSF contains a proline-rich sequence at the N-terminal end, we then tested the in vitro affinity binding of recombinant OSF with the GST-c-Src/SH3-SH2 fusion protein. As evident from Figure 9A, E. coli-produced recombinant OSF demonstrated affinity binding with the c-Src/SH3SH2 domain. However, the GST alone did not bind OSF in vitro. Furthermore, OSF was immunoprecipitated from the 293 cells transiently transfected with the OSF cDNA clone with an antibody to c-Src (Fig. 9B). Our observation of in vitro stimulation of OCL formation and bone resorptive activity by conditioned media from 293 cells transfected with the OSF cDNA clone suggests that transient expression of OSF cDNA in 293 cells at high levels results in the release of other cellular factors/products into the culture media. Preliminary studies of treatment of OSF conditioned media with neutralizing antibodies to known OCL-inducing factors such as IL-1, IL-6, TNF}, and rabbit antisera against recombinant OSF which did not immunoprecipitate OSF, also failed to block the stimulation of MNC formation by OSF conditioned media (Fig. 10). Assay of OSF conditioned media for granulocyte macrophage-colony-stimulating factor (GM-CSF) by enzyme-linked immunosorbent assay (ELISA) did not detect any GM-CSF. DISCUSSION Formation of OCL-like MNCs in long-term bone marrow cultures is enhanced by osteotropic factors such as 1,25-(OH)2D3, parathyroid hormone (PTH), IL-1, MCSF, and by IL-6 (MacDonald et al., 1986, 1987; Takahashi et al., 1986). We have shown that OCLs secrete factors such as IL-6 that stimulate its own formation and bone-resorbing capacity. To identify other cellular factors that stimulate OCL formation and activity, we further screened a mammalian cDNA expression li-

Fig. 9. Binding of OSF to c-Src. A: In vitro affinity binding of OSF with c-Src/SH3-SH2 domain. Agarose-glutathione beads were coupled with the GST-c-Src/SH3-SH2 domain fusion protein. The conjugate was incubated with 100 ng of recombinant OSF at room temperature for 1 h in binding buffer containing 1% Triton X100, 20 mM Tris, pH 7.4, 10% glycerol, 100 mM NaCl, 1 mM EDTA, and 0.1% bovine serum albumin. The sample was spun down briefly and washed four times with the same buffer and separated on SDS-PAGE. Western blot analysis was then performed using rabbit antisera raised against recombinant OSF. B: Immunoprecipitation of OSF with anti-c-Src followed by Western blot analysis with anti-OSF. OSF was immunoprecipitated with anti-c-Src or control IgG from 293 cell lysates. The immunoprecipitates were subjected to SDS-PAGE and Western blot was done with anti-OSF as described in Materials and Methods.

brary which we have recently constructed from highly purified human OCL-like MNCs. In this article, we found that OSF is a novel intracellular protein that indirectly enhances OCL-like cell formation and bone resorption. We cannot state with certainty which cell type in our marrow cultures expresses OSF, because our highly enriched MNCs still contain 10 –20% mononuclear cells. Furthermore, the

644

REDDY ET AL.

Fig. 10. Effects of anti-IL-1, anti-IL-6, anti-TNF, and anti-OSF on the stimulatory effects of conditioned media from 293 cells transfected with the OSF cDNA clone on MNC formation. Human marrow cells were cultured with 10% OSF conditioned media in the presence or absence of 500 ng/ml of neutralizing antibodies to IL-1, IL-6, TNFa, or OSF (1:100 dilution of antisera). These concentrations of the antibodies are sufficient to neutralize 1 ng/ml of the appropriate cytokine. Results represent the mean 1 SEM for a typical experiment.

expression of OSF mRNA observed in highly purified giant cells from osteoclastomas by in situ hybridization, as well as our immunocytochemical studies, suggests that the OSF is expressed in OCLs. However, using semiquantitative PCR analysis, we could not detect significant changes in OSF mRNA expression in OCLs induced with osteotropic factors such as PTH (50 ng/ml), 1,25-(OH)2D3 (10-8 M), and dexamethasone (10-9 M) (data not shown). Expression of the OSF mRNA transcript in different human tissues, as evident from Northern blot analysis, suggests the ubiquitous expression of OSF, analogous to other regulators of OCL activity such as c-Src or TGFb. The basis for differences in responsivity of murine and human marrow cultures in vitro to form OCL-like cells to OSF conditioned media is unclear. Differences in response to osteotropic factors have been reported in human and murine systems. IL-6, for example, enhances only OCL precursor proliferation in murine systems and requires other factors such as 1,25-(OH)2D3 or PTHrP to induce murine OCL formation and bone resorption (De La Mata et al., 1995). In contrast, in human systems, IL-6 by itself can induce human OCL formation and bone resorption (Kurihara et al., 1990). Similarly, GM-CSF does not induce OCL formation in murine systems in the absence of 1,25-(OH)2D3, but can induce OCL-like cell formation in human systems by itself, but its effects are enhanced by 1,25-(OH)2D3 (MacDonald et al., 1986). At present, the mechanism responsible for the effects of OSF on MNC formation is unknown, but most likely it works through the cellular signaling cascade to induce secretion of a factor that can enhance OCL activity. This seems likely because OSF is expressed by OCLs, but is not secreted. Lack of a hydrophobic sequence that could serve as a transmembrane domain, or an N-terminal hydrophobic sequence for secretion or an endoplasmic reticulum translocation, further sup-

ports the intracellular localization of OSF. The OSF protein structure contains potential motifs such as proline-rich sequences, an SH3 domain, and ankyrin repeats which could be responsible for OSF interactions with other cellular proteins. Ankyrin repeats are present in proteins involved in cell cycle and tissue differentiation. These ankyrin repeats have also been shown to be confined to specific membrane domains and bind integral membrane proteins, such as the (Na1 1 K1)-ATPase and the voltage-dependent Na1 channel in hematopoietic cells (Lux et al., 1990). The presence in OSF of an SH3 domain homologous to the GRB2/Fyn proto-oncogene tyrosine kinase further suggests that OSF may mediate its effects through its interactions with nonreceptor tyrosine kinases such as c-Src, which plays an important role in osteoclastic bone resorption (Soriano et al., 1991) and is highly expressed in OCLs. The high affinity binding of OSF with c-Src in vitro suggests a possible role for OSF in the bone resorption process through protein interactions with the cellular signal transduction machinery. Distinct ligand preferences for c-Src/SH3 domains mediate protein-protein interactions through binding of short proline-rich regions in the ligand proteins (Sparks et al., 1996b). This interaction of OSF with c-Src results in secretion of a soluble stimulator of OCL formation. The identity of this secondary factor(s) that stimulates OCL formation in the conditioned media from cells transfected with the OSF cDNA is unknown. The potential of this factor(s) to stimulate OCL formation and bone resorptive activity appears to be significant when fed to the marrow cultures. It is not IL-1, IL-6, GM-CSF, or other cytokines known to stimulate OCL formation. The availability of recombinant OSF should help to identify these factors involved in OCL formation. ACKNOWLEDGMENTS We thank Bibi Cates for preparation of this manuscript. A Research Award was given to Dr. Reddy from the Paget’s Foundation, New York. LITERATURE CITED Chomczynski, P., and Sacchi, N. (1987) Single-step method of RNA isolation by acid guanidium thiocyanate-phenol-chloroform extraction. Anal. Biochem., 162:156 –159. De La Mata, J., Uy, H.L., Guise, T.A., Story, B., Boyce, B.F., Mundy, G.F., and Roodman, G.D. (1995) IL-6 enhances hypercalcemia and bone resorption mediated by PTH-rP in vivo. J. Clin. Invest., 95: 2846 –2852. Kurihara, N., Bertolini, D., Suda, T., Akiyama, Y., and Roodman, G.D. (1990) IL-6 stimulates osteoclast-like multinucleated cell formation in long-term human marrow cultures by inducing IL-1 release. J. Immunol., 144:4226 – 4230. Lux, S.E., John, K.M., and Bennett, V. (1990) Analysis of cDNA for human erythrocyte ankyrin indicates a repeated structure with homology to tissue-differentiation and cell cycle control proteins. Nature, 344:36 – 42. MacDonald, B.R., Mundy, G.R., Clark, S., Wang, E.A., Kuehl, T.J., Stanley, E.R., and Roodman, G.D. (1986) Effects of human recombinant CSF-GM and highly purified CSF-1 on the formation of multinucleated cells with osteoclast characteristics in long-term bone marrow cultures. J. Bone Miner. Res., 1:227–233. MacDonald, B.R., Takahashi, N., McManus, L.M., Holahan, J., Mundy, G.R., and Roodman, G.D. (1987) Formation of multinucleated cells that respond to osteotropic hormones in long-term human bone marrow cultures. Endocrinology, 120:2326 –2333. Oreffo, R.O.C., Teti, A., Triffitt, J.T., Francis, M.J.O., Carano, A., and Zallone, A.Z. (1988) Effect of vitamin A on bone resorption: Evi-


dence for direct stimulation of isolated chicken osteoclasts by retinol and retinoic acid. J. Bone Miner. Res., 3:203–210. Oursler, M.T. (1994) Osteoclast synthesis, secretion and activation of latent transforming growth factor beta. J. Bone Miner. Res., 9:443– 452. Raisz, L.G., and Neimann, I. (1969) Effect of phosphate, calcium and magnesium on bone resorption and hormonal response in tissue culture. Endocrinology, 85:446 – 452. Reddy, S.V., Takahashi, S., Dallas, M., Williams, R.E., Neckers, L., and Roodman, G.D. (1994) IL-6 antisense deoxyoligonucleotides inhibit bone resorption by giant cells from human giant cell tumors of bone. J. Bone Miner. Res., 9:753–757. Reddy, S.V., Singer, F.R., Mallette, L., and Roodman, G.D. (1996) Detection of measles virus nucleocapsid transcripts in circulating blood cells and hematopoietic progenitors from patients with Paget’s disease. J. Bone Miner. Res., 11:1602–1607. Ron, D., and Dressler, H. (1992) pGSTag - A versatile bacterial expression plasmid for enzymatic labeling of recombinant proteins. Biotechniques, 13:866 – 869. Soriano, P., Montgomery, C., Gieske, R., and Bradley, A. (1991) Targeted disruption of the c-src proto-oncogene leads to osteopetrosis in mice. Cell, 64:693–702.

645

Sparks, A.B., Hoffman, N.G., McConnell, S.J., Fowlkes, D.M., and Kay, B.K. (1996a) Cloning of ligand targets: Systemic isolation of SH3 domain-containing proteins. Nature Biotech., 14:741– 744. Sparks, A.B., Rider, J.E., Hoffman, N.G., Fowlkes, D.M., Quilliam, L.A., and Kay, B.K. (1996b) Distinct ligand preferences of SH3 domains from Src, Yes, AG1, Cortactin, p53BP2, PLC8, Crk and GrG2. Proc. Natl. Acad. Sci. USA, 93:1540 –1544. Takahashi, N., MacDonald, B.R., Hon, J., Winkler, M.E., Derynck, R., Mundy, G.R., and Roodman, G.D. (1986) Recombinant human transforming growth factor-a stimulates the formation of osteoclast-like cells in long-term human marrow cultures. J. Clin. Invest., 78:894 – 898. Takahashi, S., Reddy, S.V., Chirgwin, J.M., Devlin, R., Haipek, C., Anderson, J., and Roodman, G.D. (1994) Cloning and identification of Annexin II as an autocrine/paracrine factor that increases osteoclast formation and bone resorption. J. Biol. Chem., 269:28696 – 28701. Takeda, J., Yano, H., Eng, S., Zeng, Y., and Bell, G.I. (1993) A molecular inventory of human pancreatic islets: Sequence analysis of 1000 cDNA clones. Hum. Mol. Genet., 2:1793–1798.