| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLES |
Section of Endocrinology, Yale University, New Haven, Connecticut 06520; and Fred Hutchinson Cancer Research Center (K.C.), Seattle, Washington 98104
Address all correspondence and requests for reprints to: Dr. Karl Insogna, Section of Endocrinology, Yale University School of Medicine, P.O. Box 208020, New Haven, Connecticut 06520-8020. E-mail: karl.insogna{at}yale.edu
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Mature osteoclasts must move over the bone surface to initiate new sites of bone resorption, form an attachment ring at those sites, and develop a ruffled border across which protons and degradative enzymes are transported during the resorptive process (2). The factors that regulate osteoclast motility are poorly understood, but among them is the growth factor/cytokine colony-stimulating factor-1 (CSF-1), which is produced in osteoblastic cells in response to osteotropic hormones (3). When isolated osteoclasts are exposed to a gradient of CSF-1, they move up that gradient (4). When osteoclasts are exposed to CSF-1 in a nongraded fashion, they exhibit a spreading response (5, 6, 7). CSF-1 signals through the receptor tyrosine kinase, c-fms, which is highly expressed on mature osteoclasts (3). After ligand binding and receptor autophosphorylation, src homology 2 (SH2) domain-containing molecules are recruited to c-fms and mediate downstream signaling events. Among the molecules that associate with c-fms following CSF-1 stimulation is the nonreceptor tyrosine kinase, c-src (8, 9). c-src plays an essential role in osteoclast function, as mice in which the src gene has been disrupted show normal osteoclast development but the mature cells fail to resorb bone, resulting in osteopetrosis (10). This defect may in part be explained by abnormal cytoskeletal organization, as src-deficient osteoclasts are unable to form ruffled borders (11) or develop normal attachment zones, which are prerequisites for bone resorption. We and others have reported that c-src is also required for CSF-1-induced cytoskeletal reorganization and spreading in osteoclasts (5, 12).
The signaling pathways that participate in CSF-1-induced cytoskeletal changes in osteoclasts have not been identified. However, the type I p85/p110 phosphatidylinositol 3-kinase (PI3-K), which has been implicated in the regulation of cytoskeletal function in a number of cell types (13, 14, 15), is also recruited to c-fms after CSF-1 stimulation (9, 16, 17). PI3-K appears to play a role in CSF-1-induced motility in osteoclasts, as inhibitors of PI3-K block CSF-1-induced osteoclast chemotaxis (4). Inhibition of PI3-K also abrogates ruffled border formation, osteopontin-induced osteoclast actin polymerization, and bone resorption by osteoclasts in vitro (18, 19, 20).
Taken together, these data suggest that both c-src and PI3-K are important for normal cytoskeletal function and bone resorption by osteoclasts. In the current study we investigated whether c-src and PI3-K form part of a common or overlapping signaling pathway that subserves CSF-1-induced spreading in osteoclasts.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MEM cell culture medium, FCS, BSA, wortmannin, LY294002,
glutathione agarose, human thrombin (2000 U/mg), lysozyme,
fluorescein-conjugated goat antimouse IgG antibody, and
phosphatidylinositol-4-monophosphate (PI-4-P) were obtained from
Sigma (St. Louis, MO).
[
-32P]ATP and enhanced chemiluminescence
(ECL) reagents were purchased from Amersham Pharmacia Biotech (Arlington Heights, IL). Calcitriol was obtained from
Wako Chemicals (Osaka, Japan). Protein G-agarose beads
were obtained from Calbiochem (San Diego, CA).
Phosphatidylinositol and
phosphatidylinositol-4,5-P2 were purchased from
Avanti Polar Lipids (Birmingham, AL). Antibodies to the
p85 subunit of PI3-K, c-fms, and phosphotyrosine were
obtained from Upstate Biotechnology, Inc. (Lake Placid,
NY), antibodies to glutathione-S-transferase (GST) were
obtained from Santa Cruz Biotechnology, Inc. (Santa Cruz,
CA), and antibodies to c-src were obtained from
Oncogene Research Products (Cambridge, MA). Horseradish
peroxidase (HRP)-conjugated goat antimouse IgG antibody was purchased
from Promega Corp., (Madison, WI). Recombinant human CSF-1
was a generous gift from Genetics Institute (Cambridge,
MA). The protein content of cell lysates was measured using the DC protein assay kit from Bio-Rad Laboratories, Inc. (Hercules, CA).
Cell spreading assay
Cell spreading was measured as previously described (5).
Briefly, osteoclasts were isolated by mechanical disaggregation from
the long bones of neonatal mice and allowed to settle on glass
coverslips in MEM containing 10% FCS for 1 h at 37 C. The
concentration of FCS was reduced to 2% for 2 h, after which the
cells were serum starved for 1.5 h. Cells were then pretreated for
30 min with medium alone (MEM, 0.1% BSA, and 20 mM
HEPES, pH 7.36) or with identical medium containing either wortmannin
or LY294002 in the doses described in the figure legends before
treatment with 2.5 nM CSF-1. The cells were photographed
immediately before and 10 min after the addition of CSF-1. Cell
spreading was assessed by scanning the images using an Arcus II flatbed
scanner (Agfa Division, Bayer Corp., Ridgefield Park, NJ),
and measuring the planar area of each cell using NIH Image 1.61
software. Previous studies in our laboratory had demonstrated
substantial increases in osteoclast cell area in response to CSF-1
after 10 min of treatment (5).
Confocal immunocytochemistry
Rat osteoclasts were isolated as previously described (5) and
allowed to attach to glass coverslips for 2 h in MEM containing
10% FCS, after which the FCS concentration was reduced to 2% for an
additional 2 h. Cells were then treated for 10 min with either
vehicle or 2.5 nM CSF-1 for 10 min, after which they were
fixed in 3.7% formaldehyde and permeabilized in PBS, containing 15%
FCS and 0.3% Triton X-100. After permeabilization, cells were
incubated with anti-p85 PI3-K antibody for 1 h at a final
concentration of 10 µg/ml, washed, and incubated with
fluorescein-conjugated goat antimouse IgG antibody at a final
concentration of 85 µg/ml. Immunostaining was analyzed using a
Carl Zeiss LSM 510 confocal imaging system (New York, NY).
Images of cells were compiled in Adobe Photoshop 5.0 (Abacus
Concepts, Berkeley, CA).
Cell culture
Osteoclast-like cells (OCLs) were generated as previously
reported (5) by coculturing murine bone marrow with osteoblast-like
cells, derived from sequential collagenase-dispase digestion of
neonatal murine calvariae. Cocultures were continued for 5–7 days in
MEM containing 10% FCS in the presence of
10-8 M calcitriol. Osteoblasts and
contaminating mononuclear cells were then removed by treatment with
0.1% EDTA for 5 min at 37 C. The resulting preparation was comprised
of 80–90% OCLs. OCLs possess the important phenotypic markers of
authentic osteoclasts, including strong staining for tartrate-resistant
acid phosphatase, expression of calcitonin receptors, and the ability
to resorb bone (21).
After purification, the OCLs were recovered at 37 C for 15–30 min in
MEM and 2% FCS. The medium was then aspirated, and the cells were
treated at room temperature with MEM, 0.1% BSA, and 20
mM HEPES, pH 7.36, containing either vehicle or 2.5
nM CSF-1, for the times indicated in the figure legends.
The reaction was stopped by removing the treatment medium and washing
the cells with ice-cold PBS. The cells were then immediately lysed by
scraping them in HNTG lysis buffer [50 mM HEPES (pH 7.5),
150 mM NaCl, 1% Triton, 10% glycerol, 1 mM
EGTA, and 1.5 mM MgCl2] containing
protease and phosphatase inhibitors (1 mM
phenylmethylsulfonylfluoride, 1 µg/ml pepstatin, 10 µg/ml
leupeptin, 10 µg/ml aprotinin, 1 mM sodium vanadate, and
500 mM NaF). The lysates were vortexed for 1 min, then
centrifuged at 14,000 x g for 10 min at 4 C. The
clarified lysates were stored at -70 C until used.
Preparation of OCL plasma membranes
To examine the translocation of PI3-K, plasma membranes were
prepared as described previously (22). Briefly, CSF-1-treated and
untreated OCLs were harvested in hypotonic lysis buffer [10
mM Tris-HCl (pH 7.4), 1.5 mM
MgCl2, 1 mM dithiothreitol (DTT), 1
mM phenylmethylsulfonylfluoride, 0.5 µg/ml pepstatin, and
0.5 µg/ml leupeptin). Cell lysates were incubated on ice for 15 min
and disrupted by passage through a 27-gauge needle. The homogenates
were diluted 5-fold with lysis buffer and centrifuged for 10 min at
1,000 x g. The supernatant were then layered on top of
a 35% sucrose solution and centrifuged at 20,000 x g
for 1 h. The membrane fraction, located at the interface, was
removed and collected by centrifugation at 100,000 x g
for 1 h.
GST fusion proteins
GST fusion proteins containing the N-terminal [amino acids (aa)
3–332] fragment, the N-terminal SH2 domain (aa 300–435), the
C-terminal SH2 domain (aa 589–724), and the N- plus C-terminal SH2
domains (aa 300–724) of the p85 regulatory subunit of PI3-K were
generated as previously described (23). Escherichia coli
expressing GST or a GST fusion protein containing the SH3 domain of
human c-src (aa 84–145) (24) were provided by Dr. Richard
Rickles (ARIAD Pharmaceuticals, Cambridge, MA). E. coli
expressing a GST fusion protein encoding a mutant c-src SH3,
dl90–92, in which aa 90–92 have been deleted and which has
markedly impaired ability to bind proline-rich sequences (25), was
provided by Dr. Joan Brugge (Department of Cell Biology, Harvard
Medical School, Boston, MA).
To produce GST fusion proteins, bacterial cultures were induced with 0.1 mM isopropyl ß-D-thiogalactopyranoside, and the pelleted cells were lysed with lysozyme (0.5 mg/ml, 4 C). Fusion proteins were purified by adsorption to glutathione-agarose beads, and used to probe lysates of OCLs. Lysates were precleared by incubation for 1 h with 50 µg GST adsorbed to glutathione-agarose beads. After centrifugation, the pellet was discarded, and 2.5–5 µg of the relevant fusion protein were incubated with the supernatant for 3 h at 4 C. The resulting protein complexes were extensively washed in HNTG wash buffer [20 mM HEPES (pH 7.5), 150 mM NaCl, 0.1% Triton, and 10% glycerol], resolved by SDS-PAGE, and transferred to a nitrocellulose membrane, and the membrane probed with antibodies as outlined in the figure legends.
To prepare purified wild-type and dl90–92 c-src SH3 peptides, GST was cleaved from the appropriate GST c-src SH3 fusion protein adsorbed to glutathione agarose beads by incubating the latter with thrombin (0.5%, wt/wt) for 2 h at 37 C. The beads were then washed several times with a buffer containing 50 mM Tris-HCl (pH 7.5) and 150 mM NaCl. The supernatant, containing the purified peptide, was collected, dialyzed against 10 mM Tris using a 3500 mol wt cut-off membrane, and lyophilized. The lyophilized protein was resuspended at a concentration of 100 µM in HNTG wash buffer and stored at -70 C until further use.
Proline-rich peptides of p85
Synthetic peptides corresponding to the proline-rich peptide
sequences of p85 (aa 80–104 and 299–318) were provided by Dr. J.
Cambier (Department of Pediatrics, National Jewish Medical and Research
Center, Denver, CO) (26). The peptides were covalently coupled to
cyanogen-activated Sepharose beads (Amersham Pharmacia Biotech) at 4 mg/ml, according to the manufacturer’s protocol.
Fifteen micrograms of each peptide were used to probe lysates of
vehicle- and CSF-1-treated OCLs; the resulting protein complexes were
resolved by SDS-PAGE and immunoblotted with an antibody to c-Src.
Immunoprecipitation and Western blotting
Equivalent amounts of protein (200–400 µg) from cell lysates
were incubated with antibody overnight at 4 C, and the immune complexes
were captured using protein G-agarose beads. In some experiments,
antibodies had been previously covalently conjugated to protein
G-agarose beads. The beads were washed three times with HNTG wash
buffer containing the same cocktail of protease and phosphatase
inhibitors as described above and boiled in 2 x Laemmli sample
buffer for 5 min. Immunoprecipitated proteins were resolved by SDS-PAGE
and transferred to a nitrocellulose membrane. The membrane was
sequentially probed with the relevant primary antibody for 3 h at
room temperature at the concentration specified by the supplier,
followed by a 1-h incubation with a HRP-coupled secondary antibody at
room temperature. Detection was achieved using ECL. Some membranes were
reprobed after stripping with 62.5 mM Tris-HCl (pH 6.7),
2% SDS, and 100 mM 2-mercaptoethanol for 30 min at 50
C.
Far Western blotting
Immunoprecipitates of p85 were resolved by SDS-PAGE and
transferred to a nitrocellulose membrane. The membrane was probed for
2 h with 100 µg/ml of either purified GST c-src SH3
fusion protein or GST alone, then for 1 h with a monoclonal
antibody to GST, and then for 1 h with a HRP-conjugated goat
antimouse IgG antibody before ECL detection.
PI 3-kinase assay
PI3-K activity was measured by assessing the incorporation of
32P into phosphatidylinositol (PI) to yield
phosphatidylinositol-3-monophosphate. PI was stored at 2 mg/ml at -70
C in chloroform until used, at which time the chloroform was evaporated
using N2 gas. The dried phospholipid was
sonicated in a buffer containing 20 mM HEPES, 0.4
mM EDTA, and 0.4 mM
Na3PO4, pH 7.1, to yield a
final concentration of 0.2 mg/ml. In experiments designed to determine
the effect of CSF-1 on c-fms-associated PI3-K activity,
c-fms was immunoprecipitated from lysates of untreated and
CSF-1-treated OCLs with 5 µg of a polyclonal antibody to
c-fms. The pellets were washed twice with each of three
ice-cold buffers [wash 1: 1% Nonidet P-40 and 100
µM NaVO4 in PBS; wash 2:
100 mM Tris (pH 7.4), 0.5M
LiCl2, 100 µM
NaVO4, and 0.1 mM DTT; wash
3: 10 mM Tris (pH 7.1), 100
mM NaCl, 0.1 mM DTT, and
100 µM NaVO4]. After the
last wash, 50 µl of 0.2 mg/ml PI were added to each pellet, and after
5 min at room temperature, 10 µl [-32P]ATP
(10 mCi/ml) were added. After 10 min, the reaction was stopped by the
addition of 15 µl 4 N HCl and placing the
samples on ice. Phospholipid products were extracted sequentially in
1:1 chloroform-methanol and then in 1:1 methanol-1
N HCl. After the second extraction, 50 µl of
the organic phase were spotted onto silica gel plates (VWR
Scientific, Boston, MA) and resolved by TLC
(chloroform-methanol-water-ammonium hydroxide, 9:7:1.7:0.3). Detection
and quantitation of phosphatidylinositol-3-phosphate were performed by
autoradiography and liquid scintillation counting of scrapings from the
chromatography plate. Phosphatidylinositol-4-phosphate was run as a
standard in each experiment and was detected by exposing the plate to
iodine vapor.
To verify that the PI-kinase activity we observed in c-fms immunoprecipitates after CSF-1 treatment was specifically attributable to PI3-K, we also performed the assay as described above, using the specific PI3-K substrate phosphatidylinositol-4,5-biphosphate. The assay protocol was exactly as described above, except that the silica gel plates were pretreated with 1.2% potassium oxalate in methanol and water (2:3) before being used. The reaction product was visualized by autoradiography, scanned by VistaScan (Umax Technologies, Inc., Fremont, CA), and quantitated using MacBAS V2.31 (Fuji Photo Film Co. Ltd. Medical Systems U.S.A., Stamford, CT).
In experiments to assess the effects of purified wild-type and dl90–92 c-src SH3 peptides on PI3-K activity, OCL lysates were immunoprecipitated with antibody to p85, and the immunoprecipitates were washed extensively with HNTG wash buffer and incubated overnight at 4 C with the indicated concentration of peptide. PI3-K activity was then assessed as described above.
Statistical analyses
All statistical analyses were performed using the Oxstat
statistical package (Medstat Ltd., Nottingham, UK). Between-group
differences in the change from baseline in cell area in the spreading
experiments and changes in c-fms-associated PI3-K activity
in response to CSF-1 were assessed by one-way ANOVA with Dunnett’s
multiple comparisons test. Comparison between the effects of the
wild-type and dl90–92 c-src SH3 peptides on
PI3-K activity was made using Student’s t test for unpaired
samples. Data are presented as the mean ±
SEM.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
, A, top panel,
and B). Pretreatment of cells with either of the PI3-K inhibitors,
wortmannin (Fig. 1A, middle panel) or LY294002 (Fig. 1A, bottom panel), prevented CSF-1-induced spreading. Thus, the
mean changes in cell area in osteoclasts exposed to CSF-1 after
pretreatment with 100 nM (n = 20) and 500
nM (n = 12) wortmannin were 2 ± 5%
and 2 ± 8%, respectively; those in cells pretreated with 25
µM (n = 12) and 50
µM (n = 10) LY294002 were 7 ± 11%
and -6 ± 4%, respectively (Fig. 1B). The CSF-1-induced change
in cell area was significantly reduced in the presence of either
inhibitor at each dose (P < 0.05 vs.
vehicle for each treatment condition).
|
, right lane, CSF-1
treatment led to c-fms autophosphorylation. This was
accompanied, as shown in Fig. 2B, right lane, by the
association of p85 with c-fms, as detected in
immunoprecipitates of c-fms. In contrast, there was neither
autophosphorylation of c-fms nor an association between
c-fms and p85 in untreated cells (Fig. 2, A and B,
left lanes). p85 did not become tyrosine phosphorylated
after association with c-fms (data not shown).
|
). The increase in c-fms-associated PI3-K kinetic activity
was apparent by 2 min after the addition of CSF-1 and was maintained at
10 min. That the CSF-1-induced increase in c-fms-associated
PI-kinase activity was attributable specifically to PI3-K was confirmed
by separate experiments in which the specific PI3-K substrate
PI-4,5-P2 was added to immunoprecipitates of
c-fms from OCLs treated with vehicle of CSF-1 (Fig. 2D).
To determine whether the association of p85 with c-fms was
accompanied by a redistribution of PI3-K within osteoclasts, cells were
treated with CSF-1 and analyzed by confocal microscopy. As shown in
Fig. 3, A–D, CSF-1 induced a distinct
change in the pattern of staining for PI3-K, with an increase in
staining intensity at the periphery of cells, consistent with
translocation to the cell membrane after activation of
c-fms. This observation was confirmed by Western blot
analysis of the p85 content in cell membranes isolated from OCLs. As
demonstrated in Fig. 3E, there was very little p85 detected by
immunoblotting in the membrane fractions isolated from vehicle-treated
cells (top panel, lane 1). In contrast, after CSF-1
treatment for 10 min, p85 was clearly present in the cell membrane
fraction (top panel, lane 2). Immunoblotting for the
epidermal growth factor (EGF) receptor confirmed that equivalent
amounts of membrane protein were present in both lanes (bottom
panel).
|
, left panel, immunoprecipitation of p85 from cells treated
with CSF-1, but not vehicle, revealed several coimmunoprecipitating
proteins on phosphotyrosine blotting, including those with approximate
molecular masses of 160, 120, 100, and 60 kDa. The 60-kDa
phosphoprotein was identified as c-src, by stripping the
corresponding membrane and probing it with a monoclonal antibody to
c-src (Fig. 4, right panel). The 160-kDa
phosphoprotein in the anti-p85 immune complex was identified by Western
blotting to be c-fms (data not shown).
|
). One comprised the
N-terminal fragment of p85 and contained the SH3 domain and
proline-rich sequences of p85 (aa 3–332), the second included both SH2
domains (aa 300–724), the third included the N-terminal SH2 domain (aa
300–435), and the fourth included the C-terminal SH2 domain (aa
589–724). As shown in Fig. 5B, c-src was found to associate
with the fusion proteins containing both of the p85 SH2 domains (lane
3) as well as with the fusion proteins containing each SH2 domain alone
(lanes 4 and 5). The association between c-src and the N-SH2
domain of p85 (lane 4) was very weak and much less than that between
c-src and the C-SH2 domain (lane 5). c-src also
associated with a fusion protein encoding the N-terminal 330 aa of p85
(lane 2). As the N-terminal fragment of p85 contains both of its
proline-rich sequences, we next determined whether these sequences were
responsible for the observed interaction of c-src with this
peptide. OCL lysates were therefore probed with synthetic peptides
containing the p85 proline-rich sequences (aa 80–104 and 299–318). As
shown in Fig. 5B, c-src associated with each of the p85
proline-rich peptides in a CSF-1-dependent manner. Thus, both the 80-
to 194-aa and 299- to 318-aa peptides bound c-src in the
basal state (lanes 1 and 3), and this association was increased after
CSF-1 treatment (lanes 2 and 4).
|
The c-src SH3 domain binds directly to p85
The association between c-src and the
proline-rich peptide sequences of p85 suggested that the
c-src SH3 domain mediates at least in part the association
between c-src and PI3-K that occurs after CSF-1 treatment of
OCLs. To examine this question more directly, a GST fusion protein
containing the SH3 domain of c-src was used to probe OCL
lysates. As shown in Fig. 5C, lanes 2 and 3, p85 was among the proteins
that associated with the c-src SH3 domain. This association
was not CSF-1 dependent (compare lanes 2 and 3). To determine whether
the association between p85 and the c-src SH3 domain was
direct, Far Western blotting was undertaken. As shown in Fig. 5D, GST
c-src SH3 bound directly to p85 immunoprecipitated from OCL
lysates (left panel, lane 2), whereas GST alone did not
(right panel), demonstrating a direct interaction between
the c-src SH3 domain and p85. GST c-src SH3
showed no specific binding when the antibody to p85 was replaced with
nonimmune serum (left panel, lane 1).
The c-src SH3 domain stimulates PI3-K
To examine whether the direct association between the
c-src SH3 domain and p85 is accompanied by a change in PI3-K
kinetic activity, PI3-K was immunoprecipitated from resting OCLs and
treated in vitro with purified c-src SH3 peptide.
As shown in Fig. 6A, purified
c-src SH3 peptide induced a dose-dependent increase in PI3-K
activity, to a level 3.8 times control values at a concentration of 100
µM. This stimulation is not the result of a
nonspecific peptide effect, as the c-src SH3
dl90–92 mutant, which associates with p85 less avidly than
the wild-type peptide (Fig. 6C), activates PI3-K correspondingly less
strongly (Fig. 6B). Thus, as shown in Fig. 6C (lane 3), in a GST
pull-down assay, the c-src SH3 dl90–92 mutant
was only able to bring down approximately 50% of the amount of PI3-K
brought down by the native peptide (lane 2). As shown in Fig. 6B, corresponding to this diminished affinity for PI3-K, the
c-src SH3 dl90–92 mutant was significantly less
effective at activating the enzyme.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Recent evidence has suggested that PI3-K plays a role in osteoclastic bone resorption (18, 20), cytoskeletal function (15, 19), and, specifically, CSF-1-induced osteoclast chemotaxis (3). The present study demonstrates that PI3-K activity is required for CSF-1-induced osteoclast spreading. We found that both of the specific PI3-K inhibitors, wortmannin and LY294002, each of which inhibits the enzyme by different mechanisms (29, 30), abrogated the spreading response to CSF-1 in isolated murine osteoclasts. Further, our data demonstrate that PI3-K and c-src participate in the signaling complex assembled in osteoclasts in response to CSF-1. Thus, we found that after CSF-1 treatment of OCLs, the PI3-K activity associated with c-fms is increased, and c-src is present in immunoprecipitates of the 85-kDa regulatory subunit of PI3-K. The association between c-src and p85 appears to be mediated by two distinct molecular interactions: one between the p85 SH2 domains and phosphotyrosine residues on c-src, and the second between the p85 proline-rich sequences and the c-src SH3 domain. Our in vitro Far Western studies suggest that the latter interaction is a direct one, and that it is accompanied by activation of PI3-K. It remains to be determined whether the interaction of the p85 SH2 domains and c-src is direct or whether it influences the enzymatic activity of p85. Taken together, the current data suggest that c-src may play a role in CSF-1-induced activation of PI3-K in OCLs, and that c-src and PI3-K may participate in a common signaling pathway that subserves CSF-1-induced motility in osteoclasts.
Our finding that PI3-K activity is required for CSF-1-induced
cytoskeletal reorganization in osteoclasts is consistent with a growing
body of evidence that this phospholipid kinase is involved in the
regulation of the actin cytoskeleton (13). The mechanism(s) by which
PI3-K influences cytoskeletal function is not well defined, but several
possibilities exist. PI3-K may lie upstream of the small, GTP-binding
proteins Rac (31, 32) and Rho (32), which have been implicated in
membrane ruffling and actin stress fiber assembly, respectively (33, 34). The phospholipid products of PI3-K, in particular
phosphatidylinositol-3,4,5-triphosphate, are capable of directly
influencing the function of downstream proteins by binding to their
pleckstrin homology domains (35). Recent evidence suggests that
phosphatidylinositol-3,4,5-triphosphate may stimulate cell motility by
regulating protein kinase C activity (36). Alternatively, PI3-K may
influence cytoskeletal function by altering the levels of one of its
substrates, phosphatidylinositol-4,5-biphosphate, which has been shown
to bind to and thereby influence the activity of actin-regulating
proteins such as gelsolin (37), -actinin, and vinculin (38). Indeed,
recent studies have demonstrated an association between PI3-K activity
and the actin capping protein gelsolin in avian osteoclasts exposed to
osteopontin (39). Further, the association between PI3-K activity and
gelsolin is blocked by antisense oligonucleotides to c-src
(39). We did not detect a CSF-1-dependent association between gelsolin
and p85 in the current studies (data not shown), suggesting that the
mechanisms by which CSF-1 and osteopontin influence cytoskeletal
function in osteoclasts may differ with regard to the events
subsequent to PI3-K activation.
Our data provide additional evidence that PI3-K participates in signaling pathways involving members of the src family of nonreceptor tyrosine kinases. Our finding that c-src and p85 coimmunoprecipitate in a CSF-1-dependent fashion from osteoclast-like cells is in agreement with similar findings in thrombin-stimulated platelets (40) and insulin-like growth factor I-stimulated glial cells (41). Osteoclasts, platelets, and neuronal cells have in common the expression of high levels of c-src (42), but only in the osteoclast is c-src essential for normal cell function (10). Our in vitro evidence suggests that c-src binds directly to p85 via its SH3 domain. This finding is consistent with those of previous studies, which demonstrated binding of the SH3 domains of src family members fyn (43, 44, 45), lyn (46), and lck (44, 47) to p85. Further, the SH3 domains of fyn and lyn have been shown to activate PI3-K in vitro (26). These data suggest that src family kinases, and specifically their SH3 domains, may participate in the activation of PI 3-K and subsequent signaling events after growth factor stimulation.
We observed a CSF-1-dependent increase in the association between c-src and the proline-rich peptides of p85, suggesting that CSF-1 induces a change in c-src, which increases the ability of the c-src SH3 domain to bind p85. The recent elucidation of the crystal structure of c-src provides a potential physical explanation for this observation, as activation of c-src by dephosphorylation of its regulatory site desequesters the binding surface of the SH3 domain (48). Thus, CSF-1 treatment of OCLs, which increases c-src kinase activity (5), can promote increased access to the c-src SH3 domain for interacting proteins.
We detected, but were unable to identify, two proteins that become tyrosine phosphorylated and associate with p85 in OCLs after exposure to CSF-1. A 100-kDa protein was the major tyrosine-phosphorylated protein in the CSF-1-induced p85 immunoprecipitate. Its molecular weight and CSF-1-dependent coimmunoprecipitation with p85 suggest that it may be the same protein as that recently reported in anti-p85 immunoprecipitates from a myeloid cell line overexpressing c-fms (23) and recently cloned as a 97-kDa SHP2-binding protein (49). The 120-kDa protein we found to associate with p85 in a CSF-1-dependent fashion did not react with antisera to c-Cbl, focal adhesion kinase, actin filament-associated protein, pp120cas, pp130cas, or pyk2 (data not shown).
In summary, we have shown that both c-src (5) and PI3-K activity are required for osteoclasts to spread in response to CSF-1, that c-src coimmunoprecipitates with the p85 regulatory subunit of PI3-K from lysates of OCLs in a CSF-1-dependent manner, and that the c-src SH3 domain binds directly to the proline-rich sequences of p85. The binding of the c-src SH3 domain to p85 is accompanied by an increase in PI3-K activity. Our results provide evidence that c-src and PI3-K participate in a common signaling pathway that subserves CSF-1-induced osteoclast motility.
|
Footnotes |
|---|
2 Recipient of a postdoctoral fellowship from the Health Research
Council of New Zealand.
3 These two authors contributed equally to this work.
Received June 25, 1999.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-Actinin and vinculin are PIP2-binding
proteins involved in signalling by tyrosine kinase. J Biol Chem 269:1518–1522This article has been cited by other articles:
 |
|
 |
|
  B. B. Yaroslavskiy, A. C. Sharrow, A. Wells, L. J. Robinson, and H. C. Blair Necessity of inositol (1,4,5)-trisphosphate receptor 1 and {micro}-calpain in NO-induced osteoclast motility J. Cell Sci., August 15, 2007; 120(16): 2884 - 2894. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Faccio, S. Takeshita, G. Colaianni, J. Chappel, A. Zallone, S. L. Teitelbaum, and F. P. Ross M-CSF Regulates the Cytoskeleton via Recruitment of a Multimeric Signaling Complex to c-Fms Tyr-559/697/721 J. Biol. Chem., June 29, 2007; 282(26): 18991 - 18999. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Ishijima, K. Tsuji, S. R Rittling, T. Yamashita, H. Kurosawa, D. T Denhardt, A. Nifuji, Y. Ezura, and M. Noda Osteopontin is required for mechanical stress-dependent signals to bone marrow cells J. Endocrinol., May 1, 2007; 193(2): 235 - 243. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. B. Yaroslavskiy, Y. Zhang, S. E. Kalla, V. G. Palacios, A. C. Sharrow, Y. Li, M. Zaidi, C. Wu, and H. C. Blair NO-dependent osteoclast motility: reliance on cGMP-dependent protein kinase I and VASP J. Cell Sci., December 1, 2005; 118(23): 5479 - 5487. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Miyazaki, A. Sanjay, L. Neff, S. Tanaka, W. C. Horne, and R. Baron Src Kinase Activity Is Essential for Osteoclast Function J. Biol. Chem., April 23, 2004; 279(17): 17660 - 17666. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Yoshida, A. Kumar, Y. Koyama, H. Peng, A. Arman, J. A. Boch, and P. E. Auron Interleukin 1 Activates STAT3/Nuclear Factor-{kappa}B Cross-talk via a Unique TRAF6- and p65-dependent Mechanism J. Biol. Chem., January 16, 2004; 279(3): 1768 - 1776. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Wu, N. Asazuma, K. Satoh, Y. Yatomi, T. Takafuta, M. C. Berndt, and Y. Ozaki Interaction between von Willebrand factor and glycoprotein Ib activates Src kinase in human platelets: role of phosphoinositide 3-kinase Blood, May 1, 2003; 101(9): 3469 - 3476. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. P. Haynes, L. Li, D. Sinha, K. S. Russell, K. Hisamoto, R. Baron, M. Collinge, W. C. Sessa, and J. R. Bender Src Kinase Mediates Phosphatidylinositol 3-Kinase/Akt-dependent Rapid Endothelial Nitric-oxide Synthase Activation by Estrogen J. Biol. Chem., January 17, 2003; 278(4): 2118 - 2123. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Fan, Q. Li, D. Ross, and J. F. Engelhardt Tyrosine Phosphorylation of Ikappa Balpha Activates NFkappa B through a Redox-regulated and c-Src-dependent Mechanism Following Hypoxia/Reoxygenation J. Biol. Chem., January 10, 2003; 278(3): 2072 - 2080. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Faccio, M. Grano, S. Colucci, A. Villa, G. Giannelli, V. Quaranta, and A. Zallone Localization and possible role of two different alpha v beta 3 integrin conformations in resting and resorbing osteoclasts J. Cell Sci., July 15, 2002; 115(14): 2919 - 2929. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. V. Chibalin, M. V. Kovalenko, J. W. Ryder, E. Feraille, H. Wallberg-Henriksson, and J. R. Zierath Insulin- and Glucose-Induced Phosphorylation of the Na+,K+-Adenosine Triphosphatase {alpha}-Subunits in Rat Skeletal Muscle Endocrinology, August 1, 2001; 142(8): 3474 - 3482. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. R. Hauck, T. Hunter, and D. D. Schlaepfer The v-Src SH3 Domain Facilitates a Cell Adhesion-independent Association with Focal Adh |
