HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Choi, S. J. Articles by Roodman, G. D. Search for Related Content PubMed PubMed Citation Articles by Choi, S. J. Articles by Roodman, G. D. Social Bookmarking                      What's this?

J Biol Chem, Vol. 274, Issue 39, 27747-27753, September 24, 1999

Identification of Human Asparaginyl Endopeptidase (Legumain) as an Inhibitor of Osteoclast Formation and Bone Resorption^*

Sun Jin Choi, Sakamuri V. Reddy, Rowena D. Devlin, Cheikh Menaa, Hoyeon Chung, Brendan F. Boyce^§, and G. David Roodman^¶

From the Departments of  Medicine/Hematology and ^§ Pathology, University of Texas Health Science Center, San Antonio, Texas 78284 and the ^¶ Audie Murphy Veterans Administration Hospital, San Antonio, Texas 78284

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

We screened a human osteoclast (OCL) cDNA expression library for OCL inhibitory factors and identified a clone that blocked both human and murine OCL formation and bone resorption by more than 60%. This clone was identical to human legumain, a cysteine endopeptidase. Legumain significantly inhibited OCL-like multinucleated cell formation induced by 1,25-dihydroxyvitamin D₃ (1,25-(OH)₂D₃) and parathyroid hormone-related protein (PTHrP) in mouse and human bone marrow cultures, and bone resorption in the fetal rat long bone assay in a dose-dependent manner. Legumain was detected in freshly isolated marrow plasma from normal donors and conditioned media from human marrow cultures. Furthermore, treatment of human marrow cultures with an antibody to legumain induced OCL formation to levels that were as high as those induced by 1,25-(OH)₂D₃. Implantation in nude mice of 293 cells transfected with the legumain cDNA and constitutively expressing high levels of the protein significantly reduced hypercalcemia induced by PTHrP by about 50%, and significantly inhibited the increase in OCL surface and in OCL number expressed per mm² bone area and per mm bone surface induced by PTHrP. These results suggest that legumain may be a physiologic local regulator of OCL activity that can negatively modulate OCL formation and activity.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The osteoclast (OCL),¹ the primary bone-resorbing cell, is derived from cells in the monocyte-macrophage lineage (1). Osteoclast activity and formation are regulated by both systemic hormones and locally produced factors. We have used an expression cloning approach with a cDNA expression library prepared from highly purified human OCL-like cells formed in vitro to identify local factors that modulate OCL formation. We have recently identified two novel stimulators of OCL formation, annexin II (2) and osteoclast stimulatory factor (3), and an inhibitor of OCL formation, human ScaI (4). We have screened this library for additional factors that enhance or inhibit OCL formation, and report here the identification and characterization of a previously unknown inhibitor of the OCL, legumain. Legumain inhibited OCL formation in both long term human and murine marrow cultures and blocked ⁴⁵calcium (⁴⁵Ca) release from fetal rat long bones stimulated by 1,25-dihydroxyvitamin D₃ (1,25-(OH)₂D₃). Interestingly, legumain was detected in normal human marrow plasma, and addition of an antibody to legumain induced high levels of OCL formation in human marrow cultures in the absence of any added stimulator of OCL formation. In addition, legumain was active in vivo and inhibited PTHrP-stimulated OCL bone resorption in nude mice implanted with human kidney fibroblast 293 cells genetically engineered to constitutively express human legumain.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Materials-- DNA transfection kits were obtained from Stratagene (La Jolla, CA), and cDNA synthesis kits were purchased from Amersham Pharmacia Biotech. PCR kits were obtained from Perkin-Elmer. All restriction enzymes used were from New England Biolabs (Beverly, MA) and Life Technologies, Inc., and chemicals were from Sigma. The 23c6 monoclonal antibody was generously provided by Dr. Michael Horton (St. Bartholomew's Hospital, London, United Kingdom).

Construction and Screening of a Mammalian cDNA Expression Library from Highly Purified Human Osteoclast-like Multinucleated Cells-- The construction and screening of the human cDNA expression library prepared from OCL-like multinucleated cells formed in vitro were done as described previously (2). Pools, whose conditioned media significantly and reproducibly inhibited tartrate-resistant acid phosphatase (TRAP)-positive multinucleated cell formation in murine marrow cultures and OCL-like cell formation in human bone marrow cultures, were progressively subfractionated until the subpool contained only one cDNA that inhibited OCL-like cell formation in both murine and human systems. The cDNA, initially named "osteoclast inhibitory peptide-2," was then sequenced by standard techniques, compared with cDNA sequences in the GenBank data base and was found to be identical to human legumain (GenBank accession no. Y09862).

Murine and Human Bone Marrow Cultures-- The marrow cells (1 × 10⁶ cells/culture) from C57Bl mice were isolated and cultured for OCL-like multinucleated cell formation in the presence or absence of 10⁹ M 1,25-(OH)₂D₃ and varying concentrations of conditioned media from 293 cells transfected with the legumain cDNA, as described by Takahashi and co-workers (5). The cultures were then stained for TRAP using an acid phosphatase staining kit (Sigma). The TRAP-positive multinucleated cells containing three or more nuclei were counted with an inverted microscope. In selected experiments, conditioned media from 293 cells transfected with the legumain cDNA were added at various times after the initiation of murine marrow cultures stimulated with 1,25-(OH)₂D₃ (10⁹ M) (day 0, 2, 4, or 6 of culture). The cultures were continued for a total of 7 days, and the number of TRAP-positive multinucleated cells determined. In experiments to assess the effects of legumain on OCL apoptosis, conditioned media from 293 cells transfected with the legumain cDNA (5% v/v) were added to murine marrow cultures on day 4 and day 6 or only day 6 of culture. The cultures were continued for 24 h, and the percentage of apoptotic OCL-like cells was determined as described previously (6).

In experiments to assess the effects of osteotropic factors on expression of legumain, human bone marrow was cultured with IL-3 (3 ng/ml), stem cell factor (15 ng/ml), and granulocyte/macrophage colony-stimulating factor (200 pg/ml) for 10 days to induce CFU-GM colony formation, and the CFU-GM-derived colonies were isolated with a pipette. One hundred thousand cells/ml were then cultured in -minimum essential medium with 15% fetal calf serum. Salmon calcitonin (50 ng/ml), 1,25-(OH)₂D₃ (10⁹ M), hIL-1 (10 ng/ml), or media were then added to the cultures, and the cultures continued for 48 h. At the end of this culture period, the cells were harvested, RNA extracted, and the relative amounts of legumain mRNA and GAPDH mRNA compared by cycle-dependent reverse transcriptase-PCR. RNA was reverse-transcribed using murine Maloney virus reverse transcriptase, and then the PCR reaction was run at 94 °C for 1 min, then 94 °C for 30 s, then 59 °C for 30 s, followed by 72 °C for 1 min for 18-30 cycles. The primers for GAPDH were purchased from CLONTECH (Palo Alto, CA), and the primers for legumain were 5'-CAG TGC GTA AGA TCG TCT CCT-3' (sense) and 5'-CTC TGA TCA GCA CAC AGT CGG-3' (antisense).

Human long term marrow cultures were performed as described previously (6). Collection of bone marrow from normal donors was approved by the Institutional Review Board of the University of Texas Health Science Center, San Antonio, TX. In brief, human bone marrow nonadherent mononuclear cells from normal volunteers were cultured at 10⁶ cells/ml in -minimum essential medium, 20% horse serum with or without 10⁹ M 1,25-(OH)₂D₃, and varying concentrations of legumain conditioned media. Half of the medium was changed weekly. After 3 weeks, the cultures were harvested and stained with the 23c6 monoclonal antibody that identifies the OCL vitronectin receptor, and the 23c6-positive multinucleated cells scored. Multinucleated cells that cross-react with the 23c6 monoclonal antibody express calcitonin receptors and resorb bone (8).

Effects of Legumain Conditioned Media on Bone Resorption in Fetal Rat Organ Cultures in Vitro-- Timed pregnant rats were injected with 250 µCi of ⁴⁵Ca at day 18 of gestation, and 1 day later the rats were sacrificed by cervical dislocation, and the embryos removed. The explanted radii and ulnae were cultured on circles of membrane filter (mixed ester, 0.45 µm; Whatman, Hillsboro, OR) on stainless steel grids in 0.5 ml of chemically defined medium (Sigma) supplemented with 1 mg/ml bovine serum albumin (Sigma) and penicillin-streptomycin (50 units/ml and 50 mg/ml), in a controlled atmosphere of 5% CO₂ in air at 37 °C, as modified by Raisz and Niemann (9). The radii and ulnae were incubated for 24 h in control medium to allow for the removal of the exchangeable ⁴⁵Ca before transferring to equilibrated control or experimental medium. Bone resorption was stimulated with 10⁹ M 1,25-(OH)₂D₃ or 20 ng/ml PTHrP with or without varying concentrations of legumain conditioned media. Control and experimental media were then changed after 72 h, with bone explants incubated for a total of 5 days. Bone-resorbing activity was measured as percentage of total ⁴⁵Ca release from the bone into the media over the 5 days of incubation.

Tissue Distribution of Legumain-- The legumain cDNA probe was purified from the plasmid by restriction endonuclease digestion with EcoRI, separated by agarose gel electrophoresis, and recovered from the gel using a DEAE membrane (Schleicher & Schuell) according to manufacturer's protocol. Specific activity of greater than 2 × 10⁹ cpm/µg DNA was attained for the probe using a random priming kit (Life Technologies, Inc.). The filter was prehybridized for 3 h and then hybridized for 20 h with 10⁷ cpm/ml aqueous hybridization fluid. The filters were washed with 0.2× SSC and 0.1% SDS, exposed in photographic cassettes, and visualized by autoradiography. Northern blot analysis was done with human multiple tissue Northern blot (CLONTECH), as well as RNA from human bone marrow and MG63 cells.

Expression of Recombinant Legumain-- From the pcDNA1 mammalian expression clone, the coding region of the legumain cDNA was generated by PCR amplification with two primers, GCC ATA TGG TTT GGA AAG TAG CTG T and GAG AAT TCT TCT CAG AAT AAA GAC TCC T. The DNA fragment generated by PCR was digested with NdeI (recognition sequence underlined) and BamHI (to cleave the cloning site in the vector) and ligated into the PET 14b vector. Twelve transformants were isolated and confirmed by sequence analysis. The Escherichia coli recombinant protein was produced and purified on His-Bind resin (Novagen Co., Madison, WI) as described previously (4). The recombinant legumain fusion protein was dissolved in 10 mM NaOH. This purified recombinant legumain was used for the generation of a rabbit polyclonal antibody, and for testing in mouse and human bone marrow cultures and fetal rat organ culture assays.

Generation of Polyclonal Anti-legumain Antisera-- The recombinant legumain protein was emulsified in PBS containing complete Freund's adjuvant, and a rabbit was immunized with 200 µg of emulsified protein in a total volume of 0.2 ml. After 14 days, the animal was hyperimmunized by injection of 200 µg/0.2 ml of the recombinant legumain preparation. At least three repeat challenges were performed once every 14 days. The specific reactivity of the legumain antiserum was determined by testing its capacity to detect legumain in Western blots.

SDS-PAGE and Western Immunoblotting-- Crude legumain from an E. coli lysate induced with isopropyl-1-thio--D-galactopyranoside, purified legumain recombinant protein, or concentrated conditioned medium from 293 cells transfected with the legumain cDNA were boiled in SDS sample buffer and loaded onto an 8% PAGE slab gel. To detect the protein profile, the PAGE gel was stained with Coomassie Blue. Electrophoretic transfer of proteins from polyacrylamide to nitrocellulose (Schleicher & Schuell) was performed using a semi-dry blotting unit (Fisher) at 20 V for 45 min. After transfer, the nitrocellulose membrane was blocked with 5% skim milk and then blotted with the legumain polyclonal antibody. The nitrocellulose membrane was then washed and reacted with horseradish peroxidase-labeled anti-rabbit IgG and visualized with the ECL system (Amersham Pharmacia Biotech) on the Kodak X-AR5, according to manufacturer's protocol.

Generation of Stably Transfected 293 Cell Lines That Constitutively Express Legumain (293-Legumain Cells)-- 293 cells (2 × 10⁶) were stably transfected with the legumain cDNA in the pIRES/neo vector (293-legumain cells) or the empty vector (293 cells) (CLONTECH) and then cultured in media containing 400 µg of G418. Every 4 days, the selection media containing G418 was changed, and after 2 weeks of culture in selection media, single colonies were isolated using cloning rings. Each colony was tested for legumain expression by Western blot analysis. The biological activity of legumain produced by these colonies was determined by treating fetal rat long bone organ cultures with serum-free media lacking G418 that has been conditioned by the cells for 48 h, and determining ⁴⁵Ca release as described above.

Effect of Human Legumain on Bone Resorption in Vivo Using the 293-Legumain Cell Line-- Four-week-old nude mice (4 mice/group) were injected in the thigh muscles with 5 × 10⁶ 293 cells or 293-legumain cells in 100 µl of PBS 3 weeks before treating the mice with PTHrP. Mice with palpable tumors were injected with 2 µg of PTHrP in 100 µl of phosphate-buffered saline four times per day subcutaneously for 5 days using a protocol that we have shown previously causes hypercalcemia and increased bone resorption (10). Blood from the retro-orbital sinus was sampled at base line and 2 h after the second injection each day, and ionized blood calcium levels were measured using a 634 ISE Ca²⁺ pH analyzer (Ciba-Corning Diagnostics Corp., Medfield, MA). The effects of legumain and PTHrP on bone histomorphometry were assessed in the calvariae, which were removed intact from the mice at the end of the experiment. The fixation and staining procedures were as described by Boyce et al. (11).

The following variables were measured in two representative sections from each calvaria using a digitizing table and the Bioquant computerized image analysis system (R&M Biometrics, Inc., Nashville, TN) and standard histomorphometric techniques (12): 1) total bone area; 2) the number of OCLs within the bone marrow spaces, expressed per mm² total bone area and per mm bone/bone marrow interface; and 3) the extent of OCLs along the bone/bone marrow interface, expressed as a percentage of the total interface length. Seven consecutive fields of equal length were counted in each slide. Serum samples from these mice were assayed for the relative amounts of legumain detectable by serially diluting the samples in PBS and performing Western blot analysis as described above.

Statistical Analysis-- In vitro results are reported as the mean ± S.E. for five replicate samples and were compared by Student's t test. Results were considered significantly different for p < 0.05. The in vivo results were compared using Student's Neuman-Keuls test and are representative of the findings in three to five separate experiments.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Screening of the Osteoclast cDNA Expression Library-- The cDNA expression library was screened by testing the effects of conditioned media from 293 cells transfected with cDNA from each pool for their capacity to inhibit OCL-like multinucleated cell formation in mouse bone marrow cultures. Conditioned media (1-10%) were added to mouse bone marrow cultures in the presence of 10⁹ M 1,25-(OH)₂D₃. The number of TRAP-positive multinucleated cells was counted and compared with the control cultures treated with conditioned media from 293 cells transfected with the empty pcDNA1 vector. A total of five pools that reproducibly inhibited multinucleated cell formation in murine marrow cultures treated with 10⁹ M 1,25-(OH)₂D₃ were detected from the original 200 pools. The positive pools were then screened by PCR for the presence of transforming growth factor  or -interferon mRNA, factors known to inhibit OCL formation. -Interferon cDNA was detected in one of the pools (data not shown), which was not screened further.

The remaining four positive cDNA pools were separated into 12 subpools each, containing 100-200 clones/pool. The cDNAs from the individual pools were transfected into 293 cells, and the conditioned media tested for their effects on TRAP-positive multinucleated cell formation in mouse bone marrow cultures. One of the positive subpools containing 200 clones was fractionated into two 96-well plates, with one clone per well. Conditioned medium from each row and column pool of these matrices was tested for its capacity to inhibit multinucleated cell formation in mouse marrow cultures.

Identification of Inhibitory Clone as the cDNA for Legumain-- The inhibitory cDNA was approximately 1.9 kilobase pairs in length. The cDNA sequence of the clone was determined, and it encoded a 1299-base pair open reading frame and had poly(A) adenylation signal and a poly(A) tail. Comparison of the coding sequence to sequences in the GenBank data base revealed that it was identical to the recently reported human legumain (human asparaginyl endopeptidase) (13, 14). The cDNA sequence and the deduced amino acid sequence also had about 30% amino acid and 50% DNA sequence homology with plant legumain (asparaginyl endopeptidase) (15) and blood fluke hemoglobinase. The deduced amino acid sequence of legumain suggested that it was a protein that contained a 17-amino acid N-terminal cleavable signal peptide. Human legumain also had four possible glycosylation sites and one RGD sequence.

Tissue Distribution of Legumain-- Legumain mRNA was expressed in all tissues, and very low levels were detectable in brain compared with the other tissues, similar to results reported by Chen et al. (13, 14). Significant expression levels of legumain mRNA were also present in bone marrow, but MG63 osteosarcoma cells did not express legumain (data not shown). Immunocytochemical analysis using our polyclonal antibody to legumain of cells in human marrow cultures confirmed that OCL-like multinucleated cells and mononuclear cells expressed legumain (data not shown).

Effects of Legumain Conditioned Media on OCL-like Multinucleated Cell Formation-- We then tested the effects of conditioned media from 293 cells transfected with the legumain cDNA on OCL-like multinucleated cell formation in mouse and human bone marrow cultures. In the mouse bone marrow cultures, legumain conditioned medium inhibited TRAP-positive multinucleated cell formation in the presence of 1,25-(OH)₂D₃ in a dose-dependent fashion at concentrations from 1% to 10% (v/v) (Fig. 1A) and maximally decreased TRAP-positive multinucleated cell formation approximately 60% at a concentration of 10% (v/v). Addition of legumain conditioned media for only the first 2 days of culture did not significantly inhibit TRAP-positive multinucleated cell formation in the cultures. Addition of legumain between days 2 and 6 of cultures markedly inhibited multinucleated cell formation (Fig. 1B). The level of inhibition of OCL formation when legumain was added during the second phase of the cultures was similar to that when legumain was present for the entire culture period. Legumain conditioned media also decreased 23c6-positive multinucleated cell formation in human marrow cultures induced by 10⁹ M 1,25-(OH)₂D₃ approximately 80% at a concentration of 5% (v/v) compared with the control cultures (Fig. 1A). Legumain conditioned media also inhibited multinucleated cell formation in murine and human marrow cultures treated with PTHrP (data not shown). However, although legumain conditioned media inhibited OCL formation (Fig. 1B), it did not significantly increase the percentage of apoptotic OCLs in these cultures (control: 13 ± 1% versus legumain-treated: 8.6 ± 0.6%) when legumain was added for the last 48 h of the cultures.

View larger version (33K): [in this window] [in a new window]   Fig. 1.   Effects of legumain on osteoclast formation in murine and human marrow cultures. A, human and murine marrow cultures were treated with conditioned media from 293 cells transfected with human legumain cDNA as described under "Experimental Procedures." Legumain conditioned media inhibited mouse TRAP(+) multinucleated cell formation and human 23c6(+) multinucleated cell formation stimulated with 109 M 1,25-(OH)2D3 in a dose-dependent manner, compared with media from 293 cells transfected with the empty vector. B, murine marrow cultures were treated with legumain conditioned media (10% v/v) during days 2-4 or days 4-6 of the culture as described under "Experimental Procedures." Legumain conditioned media significantly inhibited mouse osteoclast formation compared with cultures treated for 6 days with the control media from 293 cells transfected with the empty vector. Results in A and B represent the mean ± S.E. for four determinations in a typical experiment. A similar pattern of results was seen in three independent experiments.

Effects of Legumain Conditioned Media on Bone Resorption in the Fetal Rat Long Bone Organ Cultures-- ⁴⁵Ca-labeled fetal rat long bones were stimulated with 10⁹ M 1,25-(OH)₂D₃ and treated with varying concentrations of legumain conditioned media. Legumain inhibited ⁴⁵Ca release about 70% at 32% (v/v) compared with conditioned media from 293 cells transfected with the empty pcDNA1 vector (Fig. 2A). Legumain conditioned media also inhibited PTHrP-stimulated bone resorption in a dose-dependent manner and inhibited ⁴⁵Ca release by about 70% at a concentration of 32% (v/v) compared with conditioned media from 293 cells transfected with the empty pcDNA1 vector (Fig. 2B).

View larger version (33K): [in this window] [in a new window]   Fig. 2.   Effects of legumain conditioned media on bone resorption in fetal rat long bone organ cultures. Fetal rat organ cultures were stimulated with 109 M 1,25-(OH)2D3 (A) or 20 ng/ml PTHrP (B) and treated with conditioned media from 293 cells transfected with the human legumain cDNA as described under "Experimental Procedures." Legumain conditioned media inhibited 45Ca release in a dose-dependent manner compared with control conditioned media. Results represent the mean ± S.E. for six determinations. A similar pattern of results was seen in three independent experiments.

Effects of Osteotropic Factors on Expression of Legumain mRNA in Human OCL Precursors-- Treatment of highly purified human OCL precursors (CFU-GM-derived cells) with 1,25-(OH)₂D₃ or IL-1, factors that stimulate OCL formation, decreased the relative levels of legumain mRNA compared with control cultures. Comparison of the relative amounts of legumain mRNA to that of GAPDH (a housekeeping gene used to control for mRNA loading) demonstrated that the ratio of legumain mRNA to GAPDH mRNA for IL-1- and 1,25-(OH)₂D₃-treated cultures was 0.4 or 1.04, respectively, compared with 3.05 and 4.0 for control or calcitonin-treated cultures (Fig. 3).

View larger version (74K): [in this window] [in a new window]   Fig. 3.   Effects of osteotropic factors on expression of legumain mRNA in human OCL precursors. Human OCL precursors (CFU-GM-derived cells) were treated for 48 h with 1,25-(OH)2D3 (109 M), IL-1 (10 ng/ml), calcitonin (50 ng/ml), or media alone. At the end of the culture period, RNA was extracted, and reverse transcriptase-PCR analysis of the mRNA present in these cultures was done for both legumain and for the housekeeping gene, GAPDH, as described under "Experimental Procedures." Relative levels of legumain mRNA to GAPDH mRNA in the cultures were determined. The ratio of legumain to GAPDH mRNA levels in control cultures was 3.05, for calcitonin-treated cultures was 4.0, and these values were not significantly different. In contrast, the ratio of legumain mRNA to GAPDH mRNA in IL-1 and 1,25-(OH)2D3 cultures was 0.4 and 1.04, respectively. A similar pattern of results was seen in two independent experiments.

Isolation of Recombinant Legumain-- Recombinant legumain was expressed in E. coli BL 21 by induction with 1 mM isopropyl-1-thio--D-galactopyranoside using the PET14b system. The cell pellet was solubilized with 6 M GnHCl and purified with a His-Bind resin column as described above. Recombinant legumain was eluted with a 50-80 mM imidazole gradient in 6 M GnHCl and precipitated by double-distilled H₂O dialysis. After re-chromatography in the presence of 6 M GnHCl on the His-Bind column, about 1 mg of recombinant legumain was obtained from 1 liter of E. coli culture with more than 95% purity.

Detection of Legumain in Conditioned Media of 293 Cells and Human Bone Marrow Supernatants and Cultures-- Legumain cDNA was transfected into 293 cells, and the conditioned media were harvested and concentrated on an Amicon 10-kDa membrane. The concentrated fraction was analyzed by PAGE and immunoblotted with the anti-legumain polyclonal antibody. A slightly larger 50-kDa band was detected in 293 cell conditioned media transfected with legumain cDNA compared with the control E. coli-derived recombinant protein (20 ng), which ran as a 46-kDa band (Fig. 4A), most likely because it was not glycosylated. In contrast, no band was detected in concentrated media from 293 cells transfected with the empty vector. Analysis of freshly isolated normal human bone marrow plasma that was assayed immediately after collection and conditioned media from human bone marrow cultures showed two bands, one at 50 kDa and a second at 30 kDa. Compared with control recombinant legumain (Fig. 4B), human bone marrow plasma had at least 50 ng/ml legumain.

View larger version (34K): [in this window] [in a new window]   Fig. 4.   Western blot analysis of human legumain in 293 cells, human bone marrow plasma, and human marrow cultures. A, concentrated conditioned medium (×20) was subjected to Western blot analysis using a polyclonal antibody generated against recombinant human legumain. Conditioned medium from 293 cells transfected with legumain cDNA showed a 50-kDa band that was larger than recombinant human legumain expressed in E. coli. Control medium from 293 cells transected with the empty vector (Vector CM) and untransfected 293 cell conditioned media also expressed a 50-kDa legumain band. B, human bone marrow plasma and conditioned media from human bone marrow cultures showed two bands for legumain, 50 and 30 kDa, respectively. Detection of both bands was competed by addition of recombinant legumain. The expression level of human legumain in human bone marrow plasma was more than 50 ng/ml.

Neutralizing Effect of Legumain Antisera on Human and Mouse Bone Marrow Culture-- Since bone marrow cultures contain large amounts of legumain, we determined if the legumain secreted by these cells was inhibiting OCL formation in these cultures. In human bone marrow cultures, addition of preimmune sera did not increase 23c6-positive multinucleated cell formation above control levels. In contrast, addition of the polyclonal antibody against legumain increased 23c6-positive multinucleated cell formation 4-fold at a dilution of 1:2500 (Fig. 5A). To confirm that the anti-legumain was neutralizing the effects of legumain in human marrow cultures, we tested the capacity of the anti-legumain to block the inhibitory effects of recombinant legumain in murine marrow cultures treated with PTHrP. Addition of the legumain antiserum (1:2000) to murine marrow cultures treated with recombinant legumain (5 ng/ml) blocked the inhibitory effects of legumain (Fig. 5B). The legumain antiserum did not block the effects of PTHrP in murine marrow cultures (multinucleated cells/culture: PTHrP-treated = 33 ± 8.5 (mean ± S.E.) versus PTHrP + anti-legumain = 21.5 ± 1.7; p = 0.233).

View larger version (22K): [in this window] [in a new window]   Fig. 5.   Effects of an antibody to human legumain on osteoclast formation. A, anti-human legumain was added to human marrow cultures that did not contain any exogenous stimulator of osteoclast formation. Anti-legumain enhanced basal 23c6(+) multinucleated cell formation in human bone marrow cultures at a dilution of 1:2500. Multinucleated cell formation in the presence of anti-legumain was almost to the same level as that seen in cultures stimulated with 109 M 1,25-OH)2D3. B, polyclonal antibody generated against human legumain was added to mouse bone marrow cultures treated with 5 ng/ml human legumain expressed in 293 cells as described under "Experimental Procedures." Human legumain inhibited TRAP(+) multinucleated cell formation stimulated by PTHrP, and this was reversed by the polyclonal anti-legumain antisera. Sera obtained from rabbits prior to immunization with legumain (pre-bleed sera) did not block the inhibitory effects of legumain on osteoclast formation. Results represent the mean ± S.E. for four determinations for a typical experiment. A similar pattern of results was seen in three independent experiments.

In Vivo Effects of Legumain on Bone Resorption Stimulated by PTHrP-- Conditioned media from 293 cell lines stably transfected with legumain were examined for expression of legumain by Western blot and for biological activity in fetal rat long bone culture. Stable clones 3-1 and 3-3 expressed high levels of legumain (approximately 50-100 ng/ml per 4 × 10⁵ cells plated for 48 h) in vitro. These conditioned media blocked bone resorption consistently in fetal rat organ cultures (data not shown). The 3-1 clone was used for the in vivo experiments described below because it produced the highest levels of legumain (approximately 100 ng/ml in vitro).

In nude mice bearing 293-legumain cells, hypercalcemic effects of PTHrP were attenuated compared with mice bearing control 293 cells. As shown in Fig. 6, administration of PTHrP for 5 days in these animals significantly increased whole blood ionized calcium to >1.4 mmol/liter. In contrast, whole blood ionized calcium levels were significantly reduced in mice bearing 293-legumain producing tumors and treated with PTHrP compared with mice bearing 293 control tumors and treated with PTHrP (Fig. 6). Whole blood ionized calcium levels did not differ significantly among mice injected with 293-legumain cells or control 293 cells and treated with vehicle, but were significantly less than those seen in animals receiving PTHrP. Histomorphometric analysis of calvarial sections from animals from these experiments showed that mice injected with 293-legumain cells and treated with PTHrP had significantly fewer OCLs and OCL surfaces and smaller bone marrow spaces than animals bearing control 293 tumors treated with PTHrP (Fig. 7, a and b). OCL numbers and OCL surfaces in the 293-legumain animals treated with PTHrP were nonetheless higher than those in control animals bearing 293 tumors, but not treated with PTHrP. Animals bearing legumain-producing tumors and treated with vehicle had similar numbers of OCLs and bone-resorbing surfaces as mice bearing control tumors. Gross dissection of animals bearing 293-legumain producing tumors and 293 cell tumors reveals that there were no differences detected in any organ system except in bone. Western blot analysis of serum samples from mice bearing control or legumain producing 293 cells confirmed that increased levels of legumain were present in sera from mice bearing 293-legumain tumors compared with controls (Fig. 7c).

View larger version (20K): [in this window] [in a new window]   Fig. 6.   Effects of legumain on whole blood ionized calcium levels in mice treated with PTHrP. Mice bearing control or legumain-producing tumors were injected with PTHrP (2 µg four times per day for 5 days or vehicle (PBS) in a volume of 100 µl as described under "Experimental Procedures." Data represent the mean ± S.E. of whole blood ionized calcium levels for 4 mice at each time point. Significant hypercalcemia occurred by days 4 or 5 of PTH treatment in mice bearing control tumors and was inhibited by legumain (p < 0.05). The results represent the mean ± S.E. for six determinations for four independent experiments. **, p < 0.01 compared with the other three treatment groups. , p < 0.05 compared with animals bearing 293 control tumors and treated with PTHrP.

View larger version (53K): [in this window] [in a new window]   Fig. 7.   Effects of legumain on bone resorption in calvaria of mice treated for 5 days with PTHrP, as described under "Experimental Procedures." a, increased numbers of osteoclasts (arrow) are seen resorbing bone along the bone/bone marrow interface in mice treated with PTHrP. Mice bearing legumain producing 293 cell tumors (B) treated with PTHrP had lower numbers of osteoclasts and decreased marrow spaces compared with the mice bearing control 293 cell tumors treated with PTHrP (A). Few osteoclasts were observed in the control mice bearing 293 cells with (C) or without legumain (D), and bone marrow spaces were smaller than those seen in the PTHrP-treated mice. b, Histomorphometric analysis of the effects of legumain on osteoclast activity induced by PTHrP in vivo. Osteoclast number per mm2 bone area and per mm bone/bone marrow interface, and osteoclast surface percentage were significantly reduced in mice bearing legumain producing tumors and treated with PTHrP (p < 0.0001) compared with the control mice treated with PTHrP. However, no difference was observed between vehicle (PBS)-treated mice bearing control or legumain producing tumors. c, Western blot analysis of sera from mice bearing 293-legumain tumors and control 293 cell tumors. Sera were serially diluted and then subjected to Western blot analysis with anti-legumain as described under "Experimental Procedures."

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

We have cloned and characterized a previously unknown inhibitor of OCL activity, legumain, which inhibits OCL formation in both murine and human cultures in vitro and PTH-stimulated bone resorption in vivo. Legumain is a member of the mammalian cysteine protease family, the asparaginyl endopeptidases, which have been identified recently (13, 14). Legumain is present in significant amounts in normal human marrow and in human marrow cultures, and neutralization of legumain in human marrow cultures, in the absence of any added stimulator of OCL formation, induced OCL formation to levels similar to those induced by 1,25-(OH)₂D₃, one of the most potent inducers of OCL formation in vitro. Treatment of mice bearing legumain-producing tumors with PTHrP decreased the effects of PTHrP on bone remodeling. These data suggest that legumain may act as a physiologic regulator of OCL activity and serve to limit the number of OCLs normally present in bone. OCLs are relatively rare cells on normal bone surfaces (0.3/mm²) (16). Large numbers of OCLs would be undesirable in normal bone, leading to osteopenia. The significant amounts of legumain present in normal bone may explain why exposure of mice to exogenous legumain did not decrease basal OCL activity. However, our in vitro and in vivo data demonstrate that stimulators of OCL formation can overcome the inhibitory effects of legumain and increase the numbers of OCLs formed on the bone surface in response to appropriate stimuli, suggesting that legumain is a negative modulator of OCL activity.

Legumain appears to block the later stages of OCL formation. Legumain inhibited OCL formation from mononuclear precursors in murine cultures only when it was added in the second half of the culture period, suggesting that it was inhibiting the differentiation/fusion phase of osteoclastogenesis, rather than the proliferative phase. Biskobing and co-workers (17) have previously shown that osteoclastogenesis occurs in discrete stages in murine marrow cultures. Days 1 through 3 of murine marrow cultures is the phase in which the OCL precursor pool proliferates, and days 4 through 8 is the stage when the cells differentiate to post-mitotic cells and then fuse to form OCLs. Legumain also had no effect on OCL apoptosis. Interestingly, two factors that stimulate OCL formation (IL-1 or 1,25-(OH)₂D₃) decreased the relative levels of legumain mRNA in human OCL precursors. In contrast, calcitonin, an inhibitor of OCL formation, did not have a significant effect on levels of legumain mRNA. The data suggest that the mechanism of action of legumain is similar to that of another recently described inhibitor of the OCL, osteoprotegerin, a member of the tumor necrosis factor  receptor family (18) that also inhibits the differentiation/fusion stage of OCL formation. However, legumain is structurally unrelated to osteoprotegerin and has different effects in vivo (19). The observations that factors that stimulate OCL formation down-regulate legumain expression while inhibitors of OCL formation do not, also suggest that legumain may be an important regulator of OCL formation. These results further suggest that stimulators of OCL formation may in part increase OCL formation by decreasing production of legumain.

The molecular mechanism responsible for legumain's inhibition of OCL formation is unknown. A similar activity has been detected in Schistosoma mansoni and acts as a hemoglobinase to degrade host hemoglobin (20, 21). Legumain has an active site highly specific for substrates with asparaginyl bonds and cleaves only some of the asparaginyl bonds in polypeptides (22). However, it is unknown if this enzyme activity plays a role in the inhibitory effects of legumain on OCL formation/activity. It is unlikely that legumain inhibits OCL formation by cleaving PTHrP, since it also inhibits 1,25-(OH)₂D₃-stimulated OCL formation in murine and human cultures as well as OCL formation stimulated by RANK ligand, a recently described osteoclast differentiation factor (19). Alternatively, legumain also contains an RGD sequence that could participate in the inhibitory activity of legumain. RGD-containing peptides can inhibit OCL bone resorption (23).

Further studies are currently under way to address the potential physiologic role of legumain in bone remodeling and its potential role in pathologic states of bone remodeling.

	FOOTNOTES

^* This work was supported by research funds from the Veterans Administration and by National Institutes of Health NIAMS Grants AR44603 and AR41336 and NIDDK Grant AG 13625.The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

To whom correspondence should be addressed: Research Service (151), Audie Murphy Veterans Administration Hospital, 7400 Merton Minter Blvd., San Antonio, TX 78284. Tel.: 210-617-5319; Fax: 210-567-4705; E-mail: roodman@uthscsa.edu.

	ABBREVIATIONS

The abbreviations used are: OCL, osteoclast; PBS, phosphate-buffered saline; IL, interleukin; PCR, polymerase chain reaction; PAGE, polyacrylamide gel electrophoresis; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; CFU-GM, colony forming unit-granulocyte macrophage; TRAP, tartrate-resistant acid phosphatase; 1,25-(OH)₂D₃, 1,25-dihydroxyvitamin D₃; PTHrP, parathyroid hormone-related protein.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				R. V. Murthy, G. Arbman, J. Gao, G. D. Roodman, and X.-F. Sun Legumain Expression in Relation to Clinicopathologic and Biological Variables in Colorectal Cancer Clin. Cancer Res., March 15, 2005; 11(6): 2293 - 2299. [Abstract] [Full Text] [PDF]

				E. Gazzerro, R. C. Pereira, V. Jorgetti, S. Olson, A. N. Economides, and E. Canalis Skeletal Overexpression of Gremlin Impairs Bone Formation and Causes Osteopenia Endocrinology, February 1, 2005; 146(2): 655 - 665. [Abstract] [Full Text] [PDF]

				D. N. Li, S. P. Matthews, A. N. Antoniou, D. Mazzeo, and C. Watts Multistep Autoactivation of Asparaginyl Endopeptidase in Vitro and in Vivo J. Biol. Chem., October 3, 2003; 278(40): 38980 - 38990. [Abstract] [Full Text] [PDF]

				K. Tarte, F. Zhan, J. De Vos, B. Klein, and J. Shaughnessy Jr Gene expression profiling of plasma cells and plasmablasts: toward a better understanding of the late stages of B-cell differentiation Blood, July 15, 2003; 102(2): 592 - 600. [Abstract] [Full Text] [PDF]

				C. Liu, C. Sun, H. Huang, K. Janda, and T. Edgington Overexpression of Legumain in Tumors Is Significant for Invasion/Metastasis and a Candidate Enzymatic Target for Prodrug Therapy Cancer Res., June 1, 2003; 63(11): 2957 - 2964. [Abstract] [Full Text] [PDF]

				H. Zhou, V. Kartsogiannis, J. M. W. Quinn, C. Ly, C. Gange, J. Elliott, K. W. Ng, and M. T. Gillespie Osteoclast Inhibitory Lectin, a Family of New Osteoclast Inhibitors J. Biol. Chem., December 6, 2002; 277(50): 48808 - 48815. [Abstract] [Full Text] [PDF]

				D.P. Dickinson CYSTEINE PEPTIDASES OF MAMMALS: THEIR BIOLOGICAL ROLES AND POTENTIAL EFFECTS IN THE ORAL CAVITY AND OTHER TISSUES IN HEALTH AND DISEASE Crit. Rev. Oral. Biol. Med., May 1, 2002; 13(3): 238 - 275. [Abstract] [Full Text]

				H. Zhou, V. Kartsogiannis, Y. S. Hu, J. Elliott, J. M. W. Quinn, W. J. McKinstry, M. T. Gillespie, and K. W. Ng A Novel Osteoblast-derived C-type Lectin That Inhibits Osteoclast Formation J. Biol. Chem., April 27, 2001; 276(18): 14916 - 14923. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Choi, S. J. Articles by Roodman, G. D. Search for Related Content PubMed PubMed Citation