| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLES |
Department of Bone Biology and Osteoporosis Research, Merck Research Laboratories, West Point, Pennsylvania 19486
Address all correspondence and requests for reprints to: Le T. Duong, Ph.D., Department of Bone Biology and Osteoporosis Research, Merck Research Laboratories, West Point, Pennsylvania 19486.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
vß3 is highly
expressed in osteoclasts and was shown to play a critical role in
osteoclast function in vivo. The objective of this study
was to examine the role of vß3 integrin in
osteoclast formation in vitro using the inhibitory
disintegrin echistatin, an RGD-containing snake venom. We documented by
immunocytochemistry and Northern blot analysis that during murine
osteoclast-like cell (OCL) formation in a coculture of mouse
osteoblastic MB1.8 cells and bone marrow cells there is increased
expression of the v and ß3 integrin
subunits. Echistatin binds preferentially to the membrane fraction of
isolated enriched OCLs (IC50 = 0.6 nM), and
this binding is inhibited by vitronectin receptor-blocking polyclonal
antibodies. Additionally, cross-linking of radiolabeled echistatin to
OCLs, followed by immunoprecipitation with antibodies to vitronectin or
fibronectin receptors, shows that vß3
integrin is the predominant receptor for echistatin in this system. In
this coculture, echistatin completely inhibits the formation of
multinucleated OCLs, but not that of mononuclear prefusion OCLs (pOCs).
This inhibition is RGD and dose dependent (IC50 = 0.7
nM). We tested the hypothesis that inhibition of OCL
formation may be due to interference with pOC migration and found that
echistatin inhibited macrophage colony-stimulating factor-induced
migration and fusion of pOCs (IC50 = 1 and 0.6
nM, respectively). Echistatin inhibition of pOCs migration
and fusion is also RGD dependent. These results suggest that the
integrin vß3 plays a role in pOC
migration, which can explain the inhibitory effect of echistatin on
multinucleated osteoclast formation in vitro. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Integrins are transmembrane heterodimeric glycoproteins that mediate
cell-cell and cell-matrix interactions (8). Several studies have shown
that both mature osteoclasts and osteoclast-like cells (OCLs) generated
in tissue culture highly express the vitronectin receptor,
vß3 integrin, and the collagen/laminin
receptor, 2ß1 integrin (9, 10, 11). There are
differing reports on the localization of
vß3 integrin in osteoclasts; one study
localized the receptor to the basolateral and ruffled border membranes
(12, 13), whereas other reports also described its presence in the
sealing zone of resorbing osteoclasts (10, 14). Its role in osteoclast
function is thought to relate to osteoclast attachment to bone via
recognition of RGD-containing bone matrix proteins (15, 16). Monoclonal
antibodies against the vß3 integrin
inhibit osteoclast attachment to bone and bone resorption in
vitro (17, 18). Furthermore, osteoclasts adhere to plastic dishes
coated with ligands for the vß3 integrin
(vitronectin, fibrinogen, von Willebrand factor, osteopontin, and bone
sialoprotein) in an RGD-dependent manner via the ß3
integrin (16, 19, 20). Moreover, the RGD-containing disintegrin
echistatin, a snake venom peptide of 49 amino acids isolated from the
saw-scaled viper Echis carinatus, was previously reported to
inhibit integrin IIbß3-mediated platelet
aggregation (21) and bone resorption in vitro (15) and
in vivo (22); moreover, echistatin has been recently
reported to inhibit bone resorption in vivo in
estrogen-deficient mice and rats (23) and in mice with secondary
hyperparathyroidism (24).
In this study, echistatin was used to examine the role of
vß3 integrin in osteoclast differentiation
in vitro. vß3 integrin was
suggested to be the primary receptor for echistatin in mature
osteoclasts (15, 25). We used the murine coculture system of mouse
osteoblastic cells and bone marrow cells, which was developed to study
osteoclast development in vitro (26). In this system,
osteoclast-like multinucleated cells are formed in the presence of
osteotropic hormones, such as 1,25-dihydroxyvitamin D3
[1,25-(OH)2D3] or PTH. We found herein
that echistatin binds to the mononuclear osteoclasts and blocks the
migration of these cells, leading to the inhibition of osteoclast
multinucleation. Our data support the hypothesis that
vß3 integrin is the predominant receptor
for echistatin, and this integrin plays an important role in the
formation of multinucleated osteoclasts, probably through its function
in cell migration.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-32P]Deoxy (d)-CTP, [-32P]dATP,
Bolton and Hunter reagent, and radiolabeled protein standards were
obtained from Amersham (Arlington Heights, IL). Tissue culture media
were purchased from Life Technologies (Grand Island, NY),
FBS was obtained from JRH Bioscience (Lenexa, KS), collagenase was
obtained from Wako Chemicals (Richmond, VA), and dispase was obtained
from Boehringer Mannheim (Mannheim, Germany). Antisera to human
vitronectin receptor and human fibronectin receptor were purchased from
Life Technologies. Antibodies to the cytoplasmic domains
of v, 5, and ß5 integrin
subunits were purchased from Chemicon (Temecula, CA). Rabbit antiserum
to the ß3 cytoplasmic domain was a gift from Dr. M.
Ginsburg (Scripps Research Institute, La Jolla, CA).
1,25-(OH)2D3 was a gift from Dr. Milan R.
Uskokovic (Hoffman-La Roche, Inc., Nutley, NJ). BALB/c
mice were obtained from Taconic Farms, Inc. (Germantown,
NY). All animals were cared for and housed under conditions stated in
the Institutional Animal Care and Use Committee (IACUC) Guide for the
Care and Use of Laboratory Animals, and the studies were reviewed and
approved by the Merck Research Laboratories Institutional Animal Care
and Use Committee.
Coculture system and enrichment of OCLs
Murine osteoclast-like multinucleated and mononuclear cells
(OCLs) were prepared from BALB/c mice in a coculture system, as
reported previously with slight modifications (26). In short, a mouse
osteoblastic MB1.8 cell line was used instead of primary osteoblastic
cells (27). MB1.8 cells were plated at 1 x 104
cells/cm2. After 24 h, bone marrow cells isolated from
tibiae of 6- to 8-week-old male mice were added (2.5 x
104 cells/cm2) to the monolayer of MB1.8 cells.
The coculture was maintained in MEM containing 10% FBS and 10
nM 1,25-(OH)2D3. OCLs were
formed within 7 days and identified by the staining for
tartrate-resistant acid phosphatase (TRAP), which is a marker enzyme
for osteoclasts (26). TRAP-positive cells with three or more nuclei
were counted as multinucleated cells. These cells were shown to possess
calcitonin receptors and form resorption pits on bovine cortical bone
slices (28). The formation of OCLs in this coculture system was
quantitated as the number of TRAP-positve multinucleated cells per well
of a 24-well tissue culture plate.
The cocultures were maintained in 150-mm tissue culture dishes for 7
days, then washed twice with PBS and treated with 0.1% collagenase and
0.1% dispase in PBS for 20 min at 37 C. The dishes were vigorously
washed with PBS (three times). The nonadhering cells were collected by
this treatment and washed with PBS (three times) before membrane
preparation. The adhering cells were immediately stained for TRAP
activity, and the purity of the enriched OCLs was determined by the
number of all TRAP-positive cells over total cell number. Membranes
were prepared by scraping the cells into 10 mM HEPES, pH
6.5, and 1 mM EDTA containing leupeptin (20 µg/ml),
soybean trypsin inhibitor (20 µg/ml), and
phenylmethylsulfonylfluoride (1 mM). The crude membrane
fraction was then collected by two-step centrifugation as described
previously (29). As a control, the calvarial osteoblastic cells were
also cultured without bone marrow cells and with
1,25-(OH)2D3 treatment for 7 days, from
which membranes were prepared as described above.
The mononuclear prefusion osteoclasts (pOCs) were prepared as described
previously (27) with the following modifications. The murine coculture
of bone marrow cells and MB1.8 cells was maintained in 150-mm tissue
culture dishes for 6 days. pOCs were then detached using 10
mM EDTA solution after removing MB1.8 cells with
collagenase-dispase, followed by washing with 0.1% BSA in MEM
(three times). Preparations of pOCs (
95% pure as determined by TRAP
staining) were used for cell fusion assays and cell migration
assays.
RNA isolation and Northern blot analysis
Total cellular RNA was isolated by guanidine isothiocyanate and
phenol extraction (30). Twenty micrograms of total RNA were separated
using formaldehyde-agarose gel electrophoresis, followed by transfer
onto nylon filters (Hybond-N, Amersham, Arlington Heights,
IL). Human v and ß3 complementary DNA
(cDNA) probes were cloned from a human umbilical vein endothelial cell
cDNA library by PCR using the published human sequences (31, 32). Mouse
5 and human ß1 and ß5 cDNA
probes were also cloned by PCR from a mouse OCL cDNA library and a
human placenta cDNA library, respectively (33, 34, 35). Human
glyceraldehyde 3-phosphate dehydrogenase cDNA was a gift from Dr.
Robert W. Allen (American Red Cross, St. Louis, MO). cDNAs were
labeled with [-32P]dCTP using a random primer a DNA
labeling kit (Pharmacia Biotech, Piscataway, NJ). A 40-mer
oligo DNA (AGCCGTTGTCGACGAC CAGCGCAGCGATATCGTCATCAT) for mouse
ß-actin was synthesized and labeled with [-32P]dATP
using a DNA tailing kit (Boehringer Mannheim). Hybridizations were
performed in 40% formamide, 5 x SSC (standard saline citrate),
0.1% SDS, 0.1% Ficoll, 0.1% polyvinylpyrolidone, 0.1% BSA, and 200
µg/ml sonicated salmon sperm DNA at 42 C overnight and washed twice
(30 min each time) at 55 C in 0.1 x SSC and 0.1% SDS. The
filters were dried and exposed to XAR-2 films (Eastman Kodak Co., Rochester, NY).
Immunocytochemistry
To examine localization of the vß3
integrin by immunofluorescence, cells were fixed in 4%
paraformaldehyde at room temperature for 15 min. Cells were blocked in
1% BSA in PBS and stained with a hamster antimouse ß3
integrin monoclonal antibody (PharMingen, San Diego, CA),
at a 1:200 dilution for 1 h at room temperature, followed by the
addition of fluorescein-conjugated goat antihamster IgG
(Organon Teknika, Durham, NC). All slides were mounted
with Fluorotec mounting medium (Accurate Chemical Co., Westbury, NY).
Fluorescent samples were viewed under a Microphot-FX microscope
(Nikon, Garden City, NY).
Binding and localization of Echistatin
Iodination of echistatin. Peptide (10 µg) dissolved in 0.1
M borate buffer (pH 8.5) was iodinated using Bolton and
Hunter reagent according to the method described by the manufacturer.
The reaction mixture was purified on a Sep-Pak C18
cartridge (Millipore Corp., Milford, MA). The cartridge
was washed with H2O and then eluted with 60%
acetonitrile-0.1% trifluoroacetic acid. Labeled echistatin was
immediately neutralized and stored in 0.1 mM HEPES (pH 7.4)
containing 1% BSA. The specific activity was 2200 µCi/mmol.
Autoradiography of [125I]echistatin
binding. Osteoblastic cells and bone marrow cells were cocultured
on glass coverslips (12 mm in diameter) in the presence of 10
nM 1,25-(OH)2D3
as described above. The cells were washed twice with prewarmed MEM
containing 0.5% BSA and incubated with [125I]echistatin
(20 nCi) in MEM containing 0.5% BSA in the presence or absence of 1
µM cold echistatin. After incubation for 10 min at room
temperature, cells were washed with cold MEM (four times) containing
0.5% BSA and fixed in 0.1% glutaraldehyde and 4% paraformaldehyde,
followed by TRAP activity staining as described above. The coverslips
were mounted on a glass slide, coated with NTB2 emulsion (Eastman Kodak Co.), incubated in the dark at 4 C, and developed in Kodak
D-19 developer. The pictures were taken in the same field under either
phase contrast or episcopic polarization illumination, using IGS filter
block (Nikon, Inc., Garden City, NY) with GIF prefilter (Nikon, Inc.)
on a Microphot-FX microscope.
Cross-linking of [125I]echistatin.
Enriched preparations of TRAP-positive OCLs were washed twice with PBS
and once with HEPES buffer (50 mM HEPES, 140
mM NaCl, 1 mM
MgCl2, and 1 mM CaCl2, pH 7.4)
and harvested in HEPES buffer containing 1 mM
phenylmethylsulfonylfluoride and leupeptin (20 µg/ml). The cells were
incubated with [125I]echistatin (20 nCi) in the presence
or absence of 1 µM echistatin for 30 min at room
temperature. Bis(sulfosuccinimidyl) suberate (Pierce,
Rockford, IL) was added at final concentration of 0.2 mM,
and cells were incubated for 30 min at room temperature to cross-link
[125I]echistatin. After the addition of Tris (pH 8.0) at
a final concentration of 50 mM to stop the reaction, cells
were washed three times with 10 mM Tris and 140
mM NaCl, pH 7.4, and solubilized in 10 mM Tris
and 140 mM NaCl, pH 7.4, containing 2% Nonidet P-40 for 30
min before centrifugation. The supernatant was incubated with 0.5% BSA
and 0.05% Tween-20 in the presence or absence of antihuman vitronectin
receptor antiserum for 2 h at 4 C, followed by protein A-agarose
(Pharmacia Biotech) for 1 h at room temperature. The
polyclonal antivitronectin receptor (vß3
integrin) antiserum was preabsorbed overnight (4 C) over a confluent
monolayer of a human embryonic kidney 293 cell line that was stably
transfected with human vß5 (36). As 293
cells were previously shown to lack vß3
expression (37), we found that this preabsorption step removed all
cross-reactivity of this antiserum to other v-associated
integrins, in particular to vß5. The
preabsorbed antiserum was then used in immunoprecipitation and blocking
studies. The immunoprecipitated products were then separated on a 6%
SDS-PAGE (38). The gels were fixed, dried, and exposed to XAR-2 films
with intensifying filters at -70 C.
Echistatin membrane binding. Membrane fractions prepared from OCLs were resuspended (40 µg/ml) in 10 mM HEPES, pH 7.5, and 150 mM NaCl containing 0.5% BSA. [125I]Echistatin (10 nCi/tube) was added to membrane suspension and incubated for 1 h at room temperature. Specific binding was determined by adding 1 µM unlabeled echistatin. Inhibition of echistatin binding by antibodies was performed by preincubating membranes with preimmune serum, antihuman vitronectin receptor antibodies, or antihuman fibronectin antibodies for 30 min (25 C) before the addition of radiolabeled echistatin. In this experiment, IgG fractions were purified from all antisera using IgPure isolation kit (UNISYN, Milford, MA).
Cell fusion assay
pOCs (10,000 cells/well) were plated on 96-well culture plates
with 5 nM macrophage colony-stimulating factor (M-CSF) in
the presence or absence of graded concentrations of echistatin
(10-8–10-11 M) for 15 h,
then fixed and stained for TRAP activity. In another set of
experiments, pOCs were plated at increasing cell densities (3.0 x
103–3.0 x 105 cells/well) for 1 h
in the absence of M-CSF or for 15 h in the presence of M-CSF and
10 nM echistatin. The number of multinucleated
TRAP-positive cells with more than 10 nuclei was quantified using an
image-analyzing system (Empire Imaging Systems, Milford, NJ). Results
are expressed as a mean value of triplicate samples.
Cell migration assay
Osteoclast migration was assayed using a Boyden chamber-type
apparatus (Neuro Probe, Inc., Cabin John, MD) using the pOC
preparation, as reported previously with slight modification (36).
Before assay, pOCs were preincubated with the graded concentrations of
echistatin (10-7–10-11 M) in
0.1% BSA in MEM for 20 min. M-CSF (5 nM), a chemotactic
inducer of osteoclast migration (39), and 10% FBS in MEM were
placed in the bottom chamber. pOCs were then added to the upper chamber
at a density of 10,000 cells/well and allowed to migrate through a
polycarbonate filter (Neuro Probe) for 15 h in a humidified
incubator at 37 C. After cells on the top of the filter were wiped
away, cells migrating to the bottom of the filter were stained for TRAP
activity. The number of TRAP-positive cells was counted, using an
image-analyzing system (Empire Imaging Systems). Results are expressed
as a mean value of triplicate or quadruplicate samples.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
,25-(OH)2D3 in the coculture of mouse
osteoblasts and bone marrow cells was previously reported (3, 40).
Multinucleated TRAP-positive cells appear on day 4, and their number
peaks around day 7. In this case, the presence and regulation of
vitronectin receptors vß3 and
vß5 were initially examined using Northern
analysis of total RNAs from the coculture. Two transcripts of the
v subunit, a major form at 9 kb and a minor form at 4
kb, were detected on day 2. Their levels gradually increased until days
5–6, when they peaked (3 times greater for the 9-kb transcript; Fig. 1
). The ß3 transcript (6.9
kb) was barely detected on days 2 and 3, increased markedly from days
4–6, then slightly decreased on day 7 (Fig. 1). In contrast, there was
no significant change in the expression of ß5 messenger
RNA (mRNA; 3.5 kb) during the entire coculture period (Fig. 1). As a
control, we also examined the expression of the fibronectin receptor in
this coculture system. The mRNA levels of both subunits
5 (4.4 kb) and ß1 (3.2 kb) did not change
during the coculture period (Fig. 1).
|
vß3 in OCLs
was also examined using enriched OCLs. OCLs were prepared by treating
the coculture with collagenase and dispase to remove the majority of
the osteoblastic MB1.8 cells. TRAP-positive cells that appear as
reddish stained cells are shown in the coculture on day 6, before and
after enzyme digestion (Fig. 2, A, C, and
E). Both TRAP-positive mononuclear and multinucleated cells attach
tightly to the plates and survive this treatment. The purity of the
enriched culture, quantitated by the number of TRAP-positive cells over
the total cell number, was approximately 71.2 ± 5.4%. As this
figure refers to multinucleated cells, it is an underestimate of the
purity of the enriched culture. When the area of the cells is measured,
the fraction of TRAP-positive cell surface is greater than 90% (data
not shown). The staining for TRAP activity results in severe quenching
of immunofluorescence; we therefore used parallel cultures of OCLs for
immunolocalization of the integrin vß3
with an anti-ß3 integrin monoclonal antibody. As shown in
Fig. 2B, in the coculture on day 6, we observed many mononuclear and
large multinucleated cells expressing vß3
integrins. After collagenase and dispase treatment, the
vß3-positive cells remained tightly
attached to the culture dishes (Fig. 2, D and F). These cells included
both mononuclear and multinucleated cells. The mononuclear cells
appeared to express vß3 at a higher level
than the multinucleated cells. In multinucleated cells, the highest
levels of integrin vß3 were found in the
microlamelopodia structure. A number of OCLs underwent fusion (Fig. 2, D and F), but vß3 integrin was not
concentrated at the site of fusion. These observations confirmed the
presence of the vß3 integrin in OCLs in
the coculture system.
|
vß3 integrins
in mononuclear and multinucleated OCLs was further supported by the
comparison of vß3 integrin expression in
the enriched OCLs vs. all of the cocultured cells by
Northern analysis. Enrichment of OCLs increased mRNA levels of both
v and ß3 integrins by about 5-fold and 10
fold, respectively, whereas ß5 mRNA levels were reduced
7-fold (Fig. 3). The level of ß-actin
mRNA was not changed, which was used as a reference for quantitation in
this experiment.
|
shows
reflection light microscopy images, in which yellow dots represent
silver grains in the autoradiography of [125I]echistatin
binding, side by side with phase contrast images of the same fields.
TRAP-positive mononuclear cells appeared as early as day 3 in the
coculture, and some of them showed weak binding of
[125I]echistatin (data not shown). On day 4, significant
echistatin binding colocalized with many TRAP-positive mononuclear
cells (Fig. 4, A and B). In the early stage of the coculture, a small
percentage (<10%) of TRAP-positive mononuclear cells did not bind
echistatin. On day 5, [125I]echistatin bound to both
mononuclear and some small multinucleated cells (Fig. 4, C and D). On
day 7, the majority of echistatin binding colocalized with large
multinucleated cells (Fig. 4, E and F). The presence of excess
unlabeled echistatin completely competed radiolabeled echistatin
binding to these cells (Fig. 4, G and H), suggesting that the target
cells of echistatin are mononuclear and multinucleated OCLs.
|
vß3 integrin, this integrin was
a candidate for the echistatin receptor. This was further tested by
measuring echistatin binding to membrane fractions prepared from the
cocultured cells. As shown in Fig. 5A, echistatin bound specifically to the membrane fraction prepared from
enriched OCLs. In contrast, equal amounts of membrane prepared from the
nonadhering cells removed by enzyme digestion did not bind echistatin
significantly. Most importantly, when MB1.8 cells were cultured in the
presence of 1,25-(OH)2D3 alone, their
membranes exhibited little echistatin binding. The specificity of
echistatin binding to vß3 in OCLs was
demonstrated when membrane fractions of enriched OCLs were preincubated
with either polyclonal antibodies raised against
vß3 or fibronectin receptor
5ß1 before [125I]echistatin
binding. As shown in Fig. 5B, anti-vß3
antibodies specifically blocked echistatin binding to osteoclast
membranes, whereas anti-5ß1 antibodies had
no effect.
|
). Under nonreducing conditions, one was
approximately 135 kDa and the other was 89 kDa (Fig. 6A, lane 2). Under
reducing conditions, these proteins shifted to 123 and 104 kDa,
respectively (Fig. 6B, lane 2). Several faint bands around 75 kDa and
above 200 kDa were also detected. All cross-linked products were
completely competed by excess unlabeled echistatin (Fig. 6, A and B,
lane 3), suggesting specific binding of echistatin to these proteins.
When the cross-linking products were immunoprecipitated with
anti-vß3 antibodies, the major
immunoprecipitated proteins had the same apparent molecular masses
(Fig. 6, A and B, lane 4). These results strongly suggest that
echistatin binds to the vß3 integrins in
OCLs in the coculture system. We, therefore, examined the effect of
echistatin on the coculture system to investigate the involvement of
vß3 integrins in osteoclast formation
in vitro.
|
). As the formation of multinucleated
cells is not observed until day 4 (Fig. 4
), echistatin was added during
days 0–4 and days 4–7. Addition of echistatin for the first 4 days
had no effect on the subsequent formation of multinucleated cells;
however, the presence of echistatin during the last 4 days completely
inhibited the formation of multinucleated OCLs (Fig. 7B). These results
suggest that echistatin may block the fusion of mononuclear osteoclasts
into multinucleated osteoclasts.
|
vß3-mediated osteoclast membrane binding.
Echistatin inhibited the formation of multinucleated OCLs with an
IC50 of 0.7 nM, whereas
A24-echistatin, in which the arginine residue in RGD is
substituted with an alanine, had no effect at 1 mM (Table 1). Further structure-activity
relationship studies showed that the oxidation of methionine at
position 28 [M(O)28-echistatin] or substitution with
phenylalanine for aspartic acid at position 27 or of leucine for
methionine at position 28 also reduced the affinity of echistatin
analogs to its receptor 400- and 20-fold, respectively. Surprisingly,
deletion of only six residues from the carboxyl-terminus of echistatin
[echi(1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43)] resulted in a 10-fold decrease in echistatin affinity
to the receptor, indicating that changes in the sequence proximal to
and distal to the RGD moiety result in a significant loss in receptor
affinity. This suggests that the RGD moiety alone was necessary, but
not sufficient, to maintain the high affinity and selectivity of
echistatin binding to vß3 integrins.
Furthermore, a direct role of vß3
integrins in the formation of OCLs was suggested by the similar
affinities (IC50 values) for the binding of echistatin
analogs to OCL membranes and their inhibition of TRAP-positive
multinucleated cell formation (Table 1).
|
, echistatin inhibited M-CSF-induced pOC
fusion dose dependently (IC50 = 0.6 nM),
whereas A24-echistatin did not. This IC50 is in
agreement with the inhibitory effect of echistatin on the formation of
multinucleated OCLs in the coculture system. In the absence of M-CSF,
pOCs alone could not survive for more than 4 h (data not shown);
we therefore could not determine the effect of echistatin on pOC fusion
in the absence of this growth factor.
|
, echistatin inhibited
pOC migration induced by M-CSF in a dose-dependent manner with an
IC50 of 1.0 nM. Similar to the inhibitory
effect of echistatin on the formation of multinucleated OCLs,
inhibition of pOC migration was RGD dependent, as demonstrated by the
A24-echistatin analog. Taken together, the above
observations suggested that the blocking effect of echistatin on OCL
formation may be due to the inhibition of pOC migration.
|
). Furthermore,
echistatin (10 nM) had no effect on pOC fusion at the high
cell density (Fig. 10B) compared with the complete inhibition of
osteoclast multinucleation at the lower cell density (Fig. 8). This
result suggested that vß3 integrins play a
major role in the migration of osteoclast precursors that precedes the
fusion of pOCs in vitro.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
2ß1 integrin
(very late antigen-2), and the vitronectin receptor,
vß3 integrin (9). The
vß3 integrin was proposed to play a role
in osteoclast migration, attachment, and, therefore, bone resorption
(16, 17, 18, 19). Mononuclear osteoclast precursors also express
vß3 integrin, but its function in these
cells is not clear (42, 43). In this report, we demonstrate that
vß3 integrin is the major vitronectin
receptor in mouse multinucleated and mononuclear OCLs. We also present
data that strongly suggest the involvement of this receptor in one of
the later phases of osteoclast maturation in vitro, the
formation of multinucleated osteoclasts from mononuclear
osteoclasts.
First, we confirmed that the expression of
vß3 integrin is up-regulated during
osteoclastogenesis in a coculture system and is highly increased in the
enriched preparation of OCLs, examined by mRNA levels and
immunolocalization, which suggests that this receptor is expressed in
OCLs. The expression of the ß5 integrin subunit mRNA was
detected in cocultured cells (OCLs and MB1.8 cells), but its level was
not changed during the coculture period and was drastically reduced in
the enriched preparation of OCLs, indicating that
vß5 is not a major integrin in murine
osteoclasts and appears to be the vitronectin receptor in MB1.8 cells.
These findings support previous observations showing that
vß5 is expressed mainly in osteoblastic
cells (10). Another RGD- dependent integrin, the fibronectin
receptor 5ß1, was also expressed in the
cocultured cells, and its level was not changed in the enriched OCL
preparation (data not shown), suggesting its expression in both cell
types.
The possible role of vß3 integrin in
osteoclast formation in this system was studied by using the snake
venom echistatin after demonstration of its specific and selective
binding to the vitronectin receptor in OCLs. Echistatin was previously
colocalized with vß3 integrin in both
chicken and rat osteoclasts (25) and was shown to inhibit osteoclast
attachment to bone (44) and bone resorption in vitro (15)
and in vivo (22, 23, 24). In this study, we demonstrate that
both multinucleated and mononuclear OCLs are major echistatin-binding
cells, and no binding to osteoblasts or other bone marrow-derived cells
was significantly detected. This finding was further supported by the
specific binding of echistatin to membranes prepared from enriched
OCLs. Furthermore, the echistatin receptor in OCLs was shown to be the
vß3 integrin, as echistatin binding to
osteoclast membranes was blocked by
anti-vß3 integrin polyclonal antibodies
and was not affected by antifibronectin receptor antibodies. To further
substantiate that vß3 integrin is the
major echistatin receptor in this coculture system, we show that
virtually all the [125I]echistatin cross-linked proteins
in OCLs were immunoprecipitated by
anti-vß3 integrin antibodies. We also
ruled out directly the possible involvement of two other RGD-dependent
integrins, vß5 and
5ß1, in echistatin binding to OCLs.
Northern analysis reveals that osteoblasts express the majority of
vß5 and 5ß1
mRNA in the coculture, and the membrane fraction prepared from
osteoblasts did not bind echistatin. Furthermore, separate observations
in our laboratory show that echistatin does not inhibit cell attachment
to vitronectin via vß5 integrin or to
fibronectin via 5ß1 integrin (unpublished
observations). Taken together, these findings strongly suggest that in
this coculture system, echistatin binds specifically to
vß3 integrin, which is highly expressed in
TRAP-positive mononuclear and multinucleated cells. We concluded
therefore that its effects in this coculture are related to the
function of vß3 integrin.
Echistatin inhibited the formation of TRAP-positive multinucleated
cells (IC50 = 0.7 nM), but did not inhibit the
appearance of TRAP-positive mononuclear cells on days 4–5 of the
coculture, suggesting that echistatin inhibits osteoclast
multinucleation. This effect was reversible, as multinucleation
proceeded again after echistatin washout (data not shown). At the
concentrations that inhibit osteoclast formation, echistatin did not
cause detachment of either the osteoblastic MB1.8 cells or pOCs from
the culture dishes, even when echistatin was added to the cocultures
from the time of initial plating. This finding indicates that the
effects of echistatin are not due to cell removal or cellular toxicity
of osteoclast progenitors or MB1.8 cells. In fact, expression of
ß3 integrin was not detected in the coculture until day
4, coincident with the appearance of prefusion osteoclasts.
Furthermore, the addition of echistatin to the coculture from the time
of plating did not decrease the appearance of mononuclear osteoclast
precursors. By morphological analysis, we observed that pOCs attached
tightly to the surrounding MB1.8 cells. Interestingly, the
multinucleated TRAP-positive cells formed underneath the monolayer of
MB1.8 cells, as shown in Fig. 2. This suggests that in an intact
coculture system, adhesion of pOCs and OCLs might be mediated by other
adhesion receptors, including the collagen receptor
2ß1 and the fibronectin receptors
(11).
We also confirmed the inhibitory effect of echistatin on osteoclast
multinucleation, using pOCs that are virtually pure mononuclear
osteoclasts. Echistatin inhibited M-CSF-induced fusion of pOCs in a
dose-dependent manner (IC50 = 0.6 nM). The
effects of echistatin were RGD dependent, as documented by substituting
alanine for arginine. It has been previously reported that
RGD-containing peptides inhibit bone resorption in rat bone organ
culture (45) and block the appearance of TRAP-positive multinucleated
osteoclasts in calcified matrix, whereas the TRAP-positive mononuclear
cells were unaffected in the periosteum. The researchers suggested that
integrins, possibly vß3, play an important
role in osteoclast multinucleation preceding the actual resorption
activity. This observation is consistent with our findings and provides
ex vivo evidence that supports the hypothesis that
vß3 integrin may play a role in the later
phase of osteoclast formation. The IC50 of the inhibitory
effect of echistatin on multinucleation of osteoclasts is similar to
that for echistatin inhibition of bone resorption by rat osteoclasts
(0.1 nM) (15) and for the shape changes (retraction)
produced in rat or chick osteoclasts on serum-coated glass (0.78
nM) (44). This is also consistent with echistatin binding
to the membranes of OCLs (Kd = 0.6 nM) and
further supports the hypothesis that echistatin interferes with
vß3 integrin function in this system. By
comparison, the IC50 for echistatin inhibition of platelet
aggregation is 30 nM (21).
The formation of multinucleated osteoclasts is proposed to involve at
least three basic events: migration of precursor cells, homotypic
cell-cell recognition, and membrane fusion in conjunction with the
structural reorganization of the multinucleated cell. The role of
vß3 integrin in cell migration has been
extensively characterized in endothelial cells, smooth muscle cells,
and neutrophils (46, 47, 48, 49). By analogy, a likely target of echistatin
action is the migration of mononuclear osteoclasts, although effects on
others events cannot be excluded. As several lines of evidence,
including this study, have shown that mononuclear osteoclasts in
culture express vß3 integrin (27, 41, 42),
we examined the effect of echistatin on the migration of mononuclear
osteoclasts stimulated by M-CSF, which was reported to induce
osteoclast migration (39). Indeed, echistatin inhibited pOC migration
in a RGD- and dose-dependent manner, similar to that found to inhibit
the formation of multinucleated OCLs in the coculture system. Simpson
and Horton (50) reported that vitronectin receptors are expressed by
periosteal mononuclear cells and suggested that these cells home to
bone and fuse to form osteoclasts. Based on the above observations, we
propose that one of the major roles of the integrin
vß3 is mediation of osteoclast
migration.
The following observations favor a role for
vß3 integrins in the migration, rather
than the fusion, of osteoclast precursors. First, in migration studies,
echistatin treatment decreases the number of mononuclear TRAP-positive
cells on the membrane surface facing the lower Boyden chamber; second,
increased pOC cell density induces their spontaneous fusion, suggesting
that the proximity of these cells can result in fusion. The fusion of
osteoclast precursors at high cell density appears to be insensitive to
echistatin treatment. Therefore, we suggest that in the coculture
system, the primary effect of vß3 integrin
on osteoclast formation is based on its role in the migration of
osteoclast precursors, rather than direct involvement in the fusion
process.
This inhibitory effect of echistatin on osteoclastogenesis is different from that of osteoprotegrin, which was identified as a soluble receptor of TRANCE/RANK ligand/osteoclast differentiation factor (51, 52, 53, 54). Osteoprotegrin inhibits the appearance of TRAP-positive cells in the coculture system, whereas echistatin does not inhibit the differentiation process from hemopoietic progenitors to TRAP-positive cells, but blocks the later step, that is their fusion into multinucleated TRAP-positive cells.
Interestingly, we recently reported that echistatin inhibits bone
resorption in vivo in estrogen-deficient mice and rats (23)
and in mice with secondary hyperparathyroidism (24) without decreasing
osteoclast number on the bone surface. This observation indicates that
in vivo echistatin does not inhibit the multinucleation
of osteoclasts. In addition, no gross change was detected in
osteoclastic ultrastructure, including the organization of the ruffled
border and sealing zone, in echistatin-treated mice (24). These
observations indicated that in vivo
vß3 integrin plays a rate-limiting role in
osteoclast functions other than in cell differentiation and cell
adhesion. The present study leads by analogy to the supposition that
vß3 integrins may mediate the migration
not only of osteoclast precursors, but mature osteoclasts as well.
Inhibition of mature osteoclast migration during the resorption cycle
would result in less efficient bone resorption. The echistatin effect
on cell migration could thus explain both the in vitro and
in vivo effects on osteoclasts. However, it should be borne
in mind that in vivo there is no reduction in osteoclast
number, suggesting that at the given concentrations, echistatin does
not inhibit the in vivo preosteoclast recruitment.
The ligand of vß3 integrin with which
echistatin competes remains unknown. Echistatin inhibits the binding of
vß3 integrin to all of its natural
ligands, including osteopontin and bone sialoprotein, two
RGD-containing extracellular matrix proteins abundant in bone (55).
Osteopontin is produced by both osteoblasts (56) and osteoclasts (57, 58). However, in preliminary experiments, soluble osteopontin or
antiosteopontin antibodies did not inhibit the formation of
multinucleated OCLs (data not shown).
The intracellular signals that mediate the effects of echistatin remain
to be determined. Integrins activate multiple signaling pathways,
including the elevation of intracellular Ca2+, lipid
turnover, and tyrosine phosphorylation, leading to cytoskeletal
rearrangement and de novo gene expression (59). Adhesion of
neutrophils via ß2 integrins and of endothelial cells via
vß3 integrins induced a transient increase
in the intracellular Ca2+ concentration (47, 60, 61). In
isolated rat osteoclasts and OCLs generated in this coculture, soluble
ligands of vß3 integrin, including
echistatin, also increased the intracellular Ca2+
concentration (62, 63). In human OCLs, calcium-mediated intracellular
signaling was reported to be involved in osteoclast chemotaxis (64, 65). Further studies are required to clarify the role of changes in
intracellular Ca2+ and the involvement of other signal
transduction pathways in the effects of echistatin.
In conclusion, we have demonstrated that the target of echistatin in
osteoclasts is vß3 integrin and that this
integrin plays a crucial role in the multinucleation of osteoclasts
in vitro, possibly via cell migration. Many key questions
remain to be answered regarding the role of
vß3 integrin in osteoclast function
in vivo, because, as mentioned above, infusion of inhibitors
of vß3 to hypercalcemic or ovariectomized
rodents blocked bone resorption, but had little effect on the number of
osteoclasts on the bone surface (23, 24). These observations suggest
that inhibition via this integrin did not affect osteoclast
differentiation, at least within the short duration of treatment (up to
4 weeks). As efficient bone resorption requires osteoclast migration
from site to site, both in vitro and in vivo
echistatin effects can be explained by a direct role of the
vß3 integrin in osteoclast migration.
|
Acknowledgments |
|---|
Received April 28, 1998.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
vß3,
2ß1, and vß1
integrins. J Biol Chem 268:16737–16745vß3, in osteoclasts. A study using
confocal laser scanning microscopy. J Cell Sci 104:663–670[Abstract]
vß3 is present in the clear zone
and contributes to cellular polarization. Cell Tissue Res 286:507–515[CrossRef][Medline]
vß3 integrin regulates
5ß1-mediated cell migration toward
fibronectin. J Biol Chem 272:29380–29389v subunit to form a novel vitronectin receptor in a
human embryonic kidney cell line. J Biol Chem 265:5938–5941vß3. J Bone Miner Res 8:S378
vß3 integrin
ligands raise [Ca2+]i in rat osteoclasts and mouse-
derived osteoclast-like cells. Am J Physiol 266:C376–C381
This article has been cited by other articles:
 |
|
 |
|
  S. R. Wilson, C. Peters, P. Saftig, and D. Bromme Cathepsin K Activity-dependent Regulation of Osteoclast Actin Ring Formation and Bone Resorption J. Biol. Chem., January 23, 2009; 284(4): 2584 - 2592. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Koga, K. Yogo, K. Yoshikawa, H. Samori, M. Goto, T. Uchida, N. Ishida, and T. Takeya Physical and Functional Association of c-Src and Adhesion and Degranulation Promoting Adaptor Protein (ADAP) in Osteoclastogenesis in Vitro J. Biol. Chem., September 9, 2005; 280(36): 31564 - 31571. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I. Sato, N. Yamamoto, H. Yamazaki, S. Hashimoto, M. Hino, F. Sakai, and A. Fujie Prevention of the Cryptic Epitope SLAYGLR within Osteopontin Does Not Influence Susceptibility to Candida albicans Infection Antimicrob. Agents Chemother., July 1, 2005; 49(7): 3053 - 3055. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Miyamoto, F. Arai, O. Ohneda, K. Takagi, D. M. Anderson, and T. Suda An adherent condition is required for formation of multinuclear osteoclasts in the presence of macrophage colony-stimulating factor and receptor activator of nuclear factor kappa B ligand Blood, December 15, 2000; 96(13): 4335 - 4343. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I Nakamura, M. Pilkington, P. Lakkakorpi, L Lipfert, S. Sims, S. Dixon, G. Rodan, and L. Duong Role of alpha(v)beta(3) integrin in osteoclast migration and formation of the sealing zone J. Cell Sci., January 11, 1999; 112(22): 3985 - 3993. [Abstract] [PDF]
|
|
|
|
|
|
L. T. Duong, I. Nakamura, P. T. Lakkakorpi, L. Lipfert, A. J. Bett, and G. A. Rodan Inhibition of Osteoclast Function by Adenovirus Expressing Antisense Protein-tyrosine Kinase 2 J. Biol. Chem., March 2, 2001; 276(10): 7484 - 7492. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
