(Received for publication, December 30, 1996, and in revised form, June 3, 1997)
From the Department of Medicine, The adhesion of platelets to sites of vascular
injury is critically dependent on the binding of subendothelial bound
von Willebrand factor (vWf) to the platelet surface glycoprotein
complexes, GP Ib-V-IX and GP IIb-IIIa (integrin
Cell adhesion processes play a fundamental role in inflammation,
immunity, embryogenesis, and hemostasis. These adhesion processes are
regulated by distinct families of cell surface receptors, including the
integrins, immunoglobulin gene family, cadherins, selectins, and the
leucine-rich glycoprotein gene family, of which the receptor for
vWf,1 GP Ib-V-IX, is the best
characterized member (1, 2). There is growing evidence that many of
these receptors not only mediate cell adhesion but also play an
important role in transducing signals that regulate cell morphology,
motility, growth, and differentiation (3, 4). The molecular basis by
which these receptors transmit signals has not been clearly defined,
although there is increasing evidence that the interaction between
these receptors and the cytoskeleton is critical for signal
transduction (4). For example, studies in human platelets and a range
of cultured cells have demonstrated that the ligation and clustering of
integrins on the cell surface lead to the cytoskeletal recruitment of
various structural proteins (talin, vinculin, Although considerable progress has been made in identifying several key
proteins involved in regulating the formation of cytoskeletal signaling
complexes, there is far less information on cellular proteins
negatively regulating these signaling events. A potential candidate
protein is the ubiquitous thiol protease, calpain. This enzyme has
previously been localized to focal adhesions in cultured cells and in
the cytoskeletal fraction of thrombin-stimulated platelets (18, 19),
where it cleaves a number of focal adhesion structural proteins (talin,
We have recently reported that vWf binding to GP Ib-V-IX induces the
cytoskeletal association of
pp60c-src and PI 3-kinase in
aggregated platelets (29). Given the critical role for this receptor in
the initial adhesion of platelets to the site of vascular injury, we
have investigated the possibility that vWf binding to GP Ib-V-IX can
also induce the formation of cytoskeletal signaling complexes in
spreading platelets. Our studies indicate that various structural
proteins and signaling enzymes associate with the cytoskeletal fraction
of platelets spread on a vWf matrix. The formation of these complexes
was associated with calpain activation and the cleavage of talin, FAK,
pp60c-src, and PTP-1B. Cleavage of
these focal adhesion proteins in spreading and aggregated platelets was
associated with a substantial reduction in protein-tyrosine
phosphorylation and a 50-80% decrease in the cytoskeletal content of
integrin Calpeptin was purchased from Biomol Research
Laboratories (Plymouth Meeting, PA). Arg-Gly-Asp-Ser (RGDS) peptide,
bovine serum albumin, and FITC-conjugated phalloidin were purchased
from Sigma. Ristocetin was supplied by Paesel and Lorei Inc. (Germany).
All other materials were from sources we have described previously (29,
30).
Anti-GP IIb monoclonal antibody (mAb) SZ22 and
anti-GP IIIa mAb SZ21 were from Immunotech (France). Anti-GP Ib mAb AK2
was a generous donation from Dr. Michael Berndt (Baker Medical Research Institute, Melbourne, Australia). Anti-talin mAb 8d4 and anti-actin mAb
were from Sigma. Anti-p85 (p85 subunit of PI-3 kinase) polyclonal antibody was from Upstate Biotechnology Inc.
Anti-pp60c-src mAb 327 was a
generous donation of Dr. Joan Brugge (University of Pennsylvania).
Anti-phosphotyrosine mAb PY20 was from ICN Biochemicals Inc. Anti-FAK
mAb and anti-PTP-1B mAb were from Transduction Laboratories. Anti-calpain (80 kDa) polyclonal antibody was kindly donated by Dr.
T. C. Saido (Metropolitan Institute of Medical Science, Tokyo, Japan).
Anti-mouse and rabbit peroxidase-conjugated IgG were from Silenus
Laboratories (Victoria, Australia).
Blood was obtained from healthy donors who had not taken
any anti-platelet medication in the preceding 2 weeks. Platelets were
isolated and washed as described previously (29). For platelet aggregation studies, washed platelets were resuspended in modified Tyrode's buffer (10-50 mM Hepes, pH 7.5, 12 mM NaHCO3, 137 mM NaCl, 2.7 M KCl, and 5 mM glucose) and then incubated at
37 °C for 30 min before stimulation. Platelets (3 × 108/ml) were activated using either purified vWf (10-50
µg/ml) in the presence of ristocetin (1 mg/ml) or thrombin (1 unit/ml) in the presence of Ca2+ (1 mM).
Aggregation was initiated by stirring the platelet reaction mixtures at
950 rpm at 37 °C, in a four-channel automated platelet analyzer
(Kyoto Daiichi, Japan). In some experiments platelets were preincubated
for 30 min with either vehicle alone (0.5% Me2SO), calpeptin (50 µg/ml), RGDS (1 mM), anti-GP IIIa mAb SZ21
(5 µg/ml) or anti-GP Ib mAb AK2 (5 µg/ml) before stimulation. In
other experiments the GP IIb-IIIa complex was disrupted irreversibly by
pretreating platelets with 5 mM EDTA for 90 min at
37 °C, as described previously (29).
Glass coverslips were coated
with vWf (50 µg/ml) for 2 h at room temperature and then blocked
with 0.1% bovine serum albumin for 60 min. Washed platelets (1 × 108/ml) were applied to the coverslips for 60 min in the
presence of the platelet inhibitor theophylline (1 mM) to
allow platelet adhesion without spreading. The nonadherent platelets
were aspirated and the adherent cells washed twice with
theophylline-free phosphate-buffered saline (PBS) (20 mM
Na2HPO4, 4 mM
NaH2PO4, pH 7.4, 150 mM NaCl). Gentle washing was essential to minimize detachment of
theophylline-treated platelets from the vWf matrix. After platelets had
spread for the indicated times (0, 10, 20, 40, and 60 min), the cells
were washed three times with PBS, fixed for 30 min in 3.7%
formaldehyde in PBS, then permeabilized with 0.1% Triton X-100 in PBS
for 30 min. The permeabilized cells were stained with FITC-conjugated phalloidin (1:1,000 dilution) for 60 min in the dark, then washed three
times with PBS. The coverslips were mounted onto glass slides with
p-phenylenediamine (1 mg/ml) containing 90% glycerol, 10% PBS (v/v), then visualized with a fluorescence microscope (× 310) (Olympus).
To
obtain sufficient platelet protein for immunoblot analysis, platelet
adhesion studies were performed on polystyrene tissue culture dishes
(100 mm) (Nunc). These dishes were coated with vWf and the washed
platelets applied to these dishes in a manner identical to that
described above. After platelets had spread for the indicated times on
these dishes, the cells were washed three times, lysed with 1% Triton
X-100 lysis buffer (10 mM Tris, pH 7.2, 1% Triton X-100,
158 mM NaCl, 2 mM EDTA, 1 mM
phenylmethylsulfonyl fluoride, 5 µM calpain inhibitor I,
1 mM Na3VO4, and 2 mM
benzamidine), then scraped from the dishes with a rubber policeman.
Triton X-100-soluble and -insoluble (cytoskeleton) extracts were
isolated from the adherent platelets in a manner identical to that used
in aggregated platelets (30). The cytoskeletal proteins were extracted
from the Triton X-100-insoluble material with radioimmunoprecipitation assay buffer (10 mM Tris, pH 7.2, 1% Triton X-100, 1%
sodium deoxycholate, 0.1% sodium dodecyl sulfate (SDS), 158 mM NaCl, 2 mM EDTA, 1 mM phenylmethylsulfonyl fluoride, 5 µM calpain inhibitor I,
1 mM Na3VO4 and 2 mM
benzamidine) for 60 min at 4 °C. Radioimmunoprecipitation assay
buffer-insoluble material was removed by centrifugation at 15,000 × g for 5 min. Whole cell lysates were prepared by lysing adherent platelets with radioimmunoprecipitation assay buffer for
1 h at 4 °C.
Equal quantities of platelet extracts
were separated by SDS-polyacrylamide gel electrophoresis under reducing
conditions according to the method of Laemmli (31), then transferred to
polyvinylidene difluoride membranes. Western blots were performed as
described by Towbin et al. (32), using specific primary
antibodies, followed by horseradish peroxidase-conjugated secondary
antibodies, and finally developed using Enhanced Chemiluminescence
(DuPont). When studies on tyrosine phosphorylation of platelet proteins
were performed, cells were lysed in an equal volume of Laemmli reducing buffer (50 mM Tris, pH 6.8, 4 mM EDTA, 2 mM Na3VO4, 2 mM
phenylmethylsulfonyl fluoride, 4 mM benzamidine, 10%
glycerol, 4% SDS, 10% vWf was purified from cryoprecipitate
as described by Montgomery and Zimmerman (33). Protein concentrations
were determined using the Bio-Rad protein assay with bovine serum
albumin as a standard.
vWf is essential for effective platelet adhesion at
sites of vascular injury, especially under conditions of high shear
stress. Stable platelet adhesion under high shear requires coordinated binding of immobilized vWf to both the GP Ib-V-IX and GP IIb-IIIa (integrin Association of integrin
We have reported recently that vWf-induced platelet aggregation is
associated with the cytoskeletal translocation of
pp60c-src and PI 3-kinase (29). We
therefore examined whether the changes in the actin filament network
induced by vWf binding to GP Ib-V-IX in spreading platelets was
associated with the cytoskeletal recruitment of signaling enzymes. To
obtain sufficient amounts of cytoskeletal associated protein for
immunoblot analysis, platelet adhesion studies were performed on Petri
dishes coated with purified vWf. Washed platelets (3 × 108/ml) were applied to this matrix under conditions
identical to those used in Fig. 1A. At various time points
(0, 10, 20, 40, and 60 min) the adherent platelets were lysed with
Triton X-100 lysis buffer and cytoskeletal proteins isolated as
described under "Experimental Procedures." We confirmed that the
Triton X-100-insoluble residues remaining on the matrix were
cytoskeletal structures by performing immunofluorescence with
FITC-conjugated phalloidin. As demonstrated in Fig. 1B,
these Triton X-100-resistant structures demonstrated an overall similar
pattern of immunofluorescence to the permeabilized platelets in Fig.
1A (60 min), except for the loss of actin filament networks
lining the surface membrane. SDS-polyacrylamide gel electrophoresis
analysis of the proteins extracted from the matrix revealed a protein
pattern similar to that reported previously for cytoskeletal extracts
from spreading platelets (35), with actin constituting greater than
65% of the total cytoskeletal protein (data not shown). An alternative technique for isolating cytoskeletal proteins in which the Triton X-100-insoluble residue from adherent platelets was extracted with
SDS-containing buffers yielded identical results (data not shown).
Immunoblot analysis of the cytoskeletal extracts from spreading
platelets revealed a time-dependent increase in the
cytoskeletal content of integrin We noted consistently that the
cytoskeletal associated forms of talin, FAK, and PTP-1B underwent
proteolytic modification in spreading platelets. Talin was proteolyzed
to a 190-kDa fragment, FAK to a 90-kDa fragment, and PTP-1B to a 42-kDa
fragment. This pattern of proteolysis is identical to that mediated by
calpain in thrombin and ionophore A23187-aggregated platelets (21,
23-26). We did not, however, detect pp60c-src
proteolytic fragments in these cytoskeletal extracts even though this
kinase has been identified previously as a calpain substrate (25).
Previous studies in ionophore A23187-stimulated platelets have
demonstrated that the proteolysis of
pp60c-src leads to its relocation
from the particulate to the soluble fraction of platelets (25). We
therefore examined the Triton X-100-soluble extracts from spreading
platelets for evidence of pp60c-src
proteolysis. As demonstrated in Fig. 2,
proteolytic fragments of pp60c-src,
FAK, PTP-1B, and talin were evident in these Triton X-100-soluble extracts. The cleavage of each of these proteins was specifically inhibited by pretreating platelets with the calpain inhibitors calpeptin (Fig. 2) or E64d, or by chelating extracellular calcium with
EGTA/Mg2+ (data not shown). Inhibiting calpain under each
of these experimental conditions had no inhibitory effect on the rate
or extent of platelet spreading (data not shown).
All reports to date have indicated that calpain activation by
physiological agonists is a postaggregation event critically dependent
on fibrinogen binding to integrin Previous studies have demonstrated that irreversible platelet adhesion
onto a vWf matrix is critically dependent on both GP Ib-V-IX and
integrin
To examine the role of integrin
Although ligand binding to integrin
The ability of calpain to regulate the
subcellular distribution and enzyme activity of multiple nonreceptor
tyrosine kinases and phosphatases in human platelets has suggested a
potentially important role for this protease in regulating the level of
tyrosine phosphorylation within the cell (28). We investigated the
changes in protein-tyrosine phosphorylation in platelets spread on a
vWf matrix. As demonstrated in Fig.
6A, only a small number of
platelet proteins (60 and 64 kDa) were tyrosine phosphorylated in
adherent nonspread platelets, as assessed by anti-phosphotyrosine
immunoblot analysis. In contrast, the pattern of protein-tyrosine
phosphorylation in spread platelets was dramatically different, with
increased phosphorylation of the 60- and 64-kDa proteins and the
appearance of 118-, 107-, 100-, 92-, 74-, 70-, 56-, and 49-kDa
phosphorylated proteins. Pretreating platelets with calpeptin resulted
in a substantial increase in the level of phosphorylation of the 107-, 74-, 70-, and 64-kDa proteins, a modest increase in the phosphorylation of the 92-, 86-, and 60-kDa proteins, and a sharp reduction in the
level of phosphorylation of the 49-kDa protein. Similar changes in
protein-tyrosine phosphorylation were also observed in vWf-aggregated platelets (Fig. 6B). Time course studies revealed that the
tyrosine phosphorylation of platelet proteins in vWf-aggregated
platelets was a transient phenomenon with maximal phosphorylation
observed within 1 min of stimulation followed by progressive
dephosphorylation between 2 and 10 min. As with spreading platelets,
pretreating these cells with calpeptin before vWf activation led to a
higher and more sustained pattern of tyrosine phosphorylation (Fig.
6B).
Previous studies have revealed that many of the proteins phosphorylated
on tyrosine residues are located within the cytoskeleton (8). We
therefore investigated the time course for PTP-1B cytoskeletal translocation and cleavage in vWf-aggregated platelets and correlated this with changes in tyrosine phosphorylation. As demonstrated in Fig.
6C, PTP-1B translocation occurred rapidly within 45 s of platelet activation, whereas cleavage was not observed until after 2 min of stimulation. The time course for PTP-1B cleavage correlated well
with the calpain-regulated dephosphorylation of platelet proteins,
suggesting a potential role for this phosphatase in this
dephosphorylation process.
Given the potential importance
of FAK and pp60c-src in regulating
the formation of cytoskeletal signaling complexes, we investigated the
possibility that calpain activation may lead to reduced formation and/or stability of cytoskeletal signaling complexes in vWf-stimulated platelets. We compared the cytoskeletal content of integrin
The stable adhesion of platelets to the site of vascular injury,
particularly under conditions of high shear stress, is critically dependent on the coordinated binding of GP Ib-V-IX and integrin Previous studies in suspension-activated platelets have clearly
established a central role for integrin
We have demonstrated under a variety of different experimental
conditions that the activation of calpain by vWf requires ligand binding to both GP Ib-V-IX and integrin
Studies of calpain activation in human platelets have been confined to
suspension-activated platelets. We have demonstrated here that calpain
activation occurs in surface-activated platelets, in the absence of
added exogenous agonists or platelet aggregation. Several lines of
evidence support the hypothesis that the observed proteolysis of talin,
FAK, PTP-1B, and pp60c-src occurred
in vivo and was not a post-lysis phenomenon. First, proteolysis of these substrates was not only observed when spreading platelets were lysed with Triton X-100, but also when the cells were
directly solubilized in SDS reducing buffer and boiled immediately. Second, proteolysis was prevented by pretreating platelets with membrane-permeable calpain inhibitors such as calpeptin, E64d, and MDL
and not by membrane-impermeable inhibitors such as leupeptin, E64, or
calpastatin. Finally, pretreating platelets with the membrane-permeable inhibitors of calpain resulted in increased cytoskeletal content of
integrin The observation that calpain is activated in spreading platelets
challenges the view that calpain activation only occurs as a
post-platelet aggregation event. The ability of this protease to
regulate the cytoskeletal content of a range of focal adhesion structural proteins and signaling enzymes raises the possibility that
this protease may regulate cell matrix-adhesive processes. Inhibiting
calpain activation with a number of different calpain inhibitors,
including calpeptin, calpain inhibitor-1, E64d, and MDL, has had no
demonstrable inhibitory effect on the rate of platelet adhesion or
spreading on a vWf matrix in our static adhesion assays. These
conditions do not, however, mimic the situation in vivo, in
which platelets adhere and spread at sites of vascular injury over a
wide range of shear stresses. It is possible under these conditions
that the calpain-mediated reduction in cytoskeletal-associated adhesion
complexes may lead to alterations in the stability of adhesion
contacts. Future platelet adhesion studies under a variety of flow
conditions will be required to resolve these issues.
Although considerable progress has been made in identifying the key
structural elements and signaling molecules involved in the assembly of
cytoskeletal signaling complexes (4), there is far less information
available regarding potential negative regulators of these processes.
The studies presented here, as well as recent observations
demonstrating calpain-mediated down-regulation of the kinase activity
and cytoskeletal attachment of
pp60c-src and FAK (23, 25), suggest
a potentially important role for this protease in the assembly and/or
stability of cytoskeletal signaling complexes. The close proximity of
calpain to focal adhesion contacts (18), the regulation of its protease
activity by integrins (36), and its ability to regulate proteolytically
several key proteins involved in adhesion receptor signal transduction
place this enzyme in a pivotal position to modulate downstream signals induced by cell-cell and cell-matrix interactions.
The first two authors contributed equally to this work. We thank Dr. T. C. Saido for the generous
supply of anti-calpain antibodies.
ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
FOOTNOTES
ACKNOWLEDGEMENT
REFERENCES
IIb
3). There is growing evidence that the binding of vWf to these receptors is not only essential for
stable platelet adhesion but is also important for the transduction of
activation signals required for changes in platelet morphology, granule
secretion, and platelet aggregation. In this study we have investigated
signaling events induced by vWf binding to GP Ib-V-IX in both spreading
and aggregated platelets. The adhesion of platelets to vWf resulted in
dramatic actin filament reorganization, as assessed by
immunofluorescence with fluorescein isothiocyanate-conjugated phalloidin, and the cytoskeletal recruitment of various structural proteins (talin and integrin
IIb
3) and
signaling enzymes (pp60c-src, focal
adhesion kinase (FAK), phosphatidylinositol 3-kinase (PI 3-kinase), and
protein-tyrosine phosphatase (PTP)-1B). Time course experiments in both
spreading and aggregated platelets revealed that talin, FAK, and PTP-1B
were proteolyzed after translocation to the cytoskeleton. The
proteolysis of these proteins was dependent on the presence of
extracellular calcium and was specifically inhibited by pretreating
platelets with the membrane-permeable calpain inhibitors calpeptin,
E64d, and MDL 28,170, but not with the membrane-impermeable inhibitors
leupeptin, E64, and calpastatin. The cytoskeletal translocation of
signaling enzymes in vWf-stimulated platelets was abolished by
pretreating platelets with an anti-GP Ib-V-IX antibody but was
unaffected by blocking ligand binding to integrin
IIb
3. In contrast, calpain activation in
vWf-stimulated platelets required ligand binding to both GP Ib-V-IX and
integrin
IIb
3. The activation of calpain
in both spreading and aggregated platelets resulted in a substantial
decrease in the level of tyrosine phosphorylation of multiple platelet
proteins and was associated with a 50-80% reduction in the amount of
cytoskeletal associated talin, integrin
IIb
3, PI 3-kinase, FAK,
pp60c-src, and PTP-1B. These
studies suggest a potentially important role for calpain in regulating
the formation and/or stability of cytoskeletal signaling complexes in
vWf-stimulated platelets. Furthermore, they demonstrate distinct roles
for GP Ib-V-IX and integrin
IIb
3 in
vWf-induced signal transduction.
-actinin, tensin,
paxillin) and signaling molecules (FAK, Src kinases, c-CSK, PI
3-kinase, protein kinase C, phospholipase C
, rasGAP, Grb2, SOS,
SHP1, PTP-1B, Rho, Ras, Raf, MEK kinases, ERK1, ERK2 and calpain)
(5-15). There is growing evidence that tyrosine phosphorylation
events, mediated by the FAK and Src family kinases, play a major role
in recruiting these signaling molecules to the cytoskeleton by
providing high affinity binding sites for SH2-containing signaling
molecules (5, 6, 15-17).
3 integrin) (20-22) and several signaling enzymes (FAK,
pp60c-src, PTP-1B, phospholipase
C
, protein kinase C) (23-27). Recent studies in thrombin-stimulated
platelets have revealed that calpain cleavage of FAK and
pp60c-src results in a reduction in
their autokinase activity (23, 25), whereas PTP-1B cleavage is
associated with enhanced phosphatase activity (24), raising the
possibility that calpain may indirectly regulate the level of
protein-tyrosine phosphorylation within the cell (28). Furthermore, the
cleavage of talin, pp60c-src, and
FAK is associated with their subcellular relocation (23, 25),
suggesting a potentially important role for calpain in regulating the
stability of integrin signaling complexes.
IIb
3, talin, FAK,
pp60c-src, PTP-1B, and PI 3-kinase.
These observations suggest a potentially important role for calpain in
modulating the formation and/or stability of cytoskeletal signaling
complexes in human platelets.
Materials
-mercaptoethanol, 0.002% bromphenol blue)
and boiled immediately for 5 min before SDS-polyacrylamide gel
electrophoresis.
Formation of Cytoskeletal Signaling Complexes in Spreading
Platelets
IIb
3) complexes on the platelet
surface (34). Platelet adhesion is followed by dramatic cytoskeletal
reorganization, leading to platelet shape change, spreading, and
aggregation. In preliminary studies, we examined the time course for
platelet spreading on vWf-coated glass coverslips. Washed platelets
(1 × 108/ml) were applied to vWf-coated coverslips in
the presence of the platelet activation inhibitor theophylline (1 mM) to allow platelet adhesion without spreading, as
described under "Experimental Procedures." Platelet adhesion to vWf
under these conditions is potentially reversible, therefore gentle
washing conditions were required to minimize the displacement of
platelets from the matrix. After removal of theophylline, adherent
platelets were allowed to spread for the indicated time points (0, 10, 20, 40, and 60 min), before fixation with formaldehyde,
permeabilization with Triton X-100, and staining with FITC-conjugated
phalloidin to visualize changes in the actin filament organization. As
demonstrated in Fig. 1A,
within 10 min of binding to the vWf matrix, multiple filopodial
extensions (1-4 µM in length) projected from the
platelet surface. By 20 min, broad lamellipodia sheets began extending between these filopodial projections. These lamellipodia continued to
radiate out from the cell center until a confluent platelet monolayer
had covered the vWf matrix (60 min).
Fig. 1.
IIb
3 (GP IIb-IIIa), talin,
pp60c-src, and FAK with the
cytoskeletal fraction of spreading platelets. Washed platelets
(1 × 108/ml) were applied to a vWf matrix in the
presence of theophylline to allow platelet adhesion without spreading.
After 60 min, nonadherent platelets were washed from the coverslips and
theophylline removed to allow spreading. At the indicated time points
(0, 10, 20, 40, and 60 min), spreading platelets were fixed,
permeabilized with Triton X-100, and then stained with FITC-conjugated
phalloidin to visualize actin filaments (magnification × 310).
Panel A, prominent filopodial extensions were evident after
10 min (arrow). Broad lamellipodia sheets were apparent by
20 min (arrow), and by 60 min the lamellipodia had covered
the matrix surface. Panel B, immunofluorescent staining with
FITC-conjugated phalloidin demonstrating an intact actin filament
network after Triton X-100 extraction of spread platelets. These images
are representative of the total area. Panel C, immunoblot
analysis of cytoskeletal extracts from spreading platelets. Equal
amounts of cytoskeletal protein (30 µg) from each of the time points
examined were separated by SDS-polyacrylamide gel electrophoresis,
transferred to polyvinylidene difluoride membranes, and immunoblotted
with antibodies against GP IIb, talin, pp60c-src, and FAK. Coomassie
Brilliant Blue staining of the polyvinylidene difluoride membrane
demonstrated the cytoskeletal F-actin content. The cytoskeletal content
of GP IIb, talin, pp60c-src, FAK,
and actin was quantitated by performing densitometry on the immunoblots
or after staining the polyvinylidene difluoride membranes with
Coomassie Brilliant Blue (line graph). The proteolytic fragments of talin (190 kDa) and FAK (90 kDa) are highlighted with
arrows. These results are from one experiment,
representative of four.
[View Larger Version of this Image (39K GIF file)]
IIb
3 (GP
IIb), talin, pp60c-src, FAK (Fig.
1C), the p85 subunit of PI 3-kinase, and PTP-1B (data not
shown). Several lines of evidence suggest that this increase represents
specific recruitment and enrichment of these proteins within the
cytoskeleton rather than simply a nonspecific increase caused by
changes in the total amount of cytoskeletal protein. First, at each of
the time points examined, equal amounts of cytoskeletal protein (30 µg) were used for immunoblot analysis. Second, the increase in these
proteins was disproportionate to changes in the cytoskeletal actin
content. In fully spread platelets (60 min) the ratio of integrin
IIb
3, talin,
pp60c-src, and FAK to actin was
substantially greater (at least 10-fold) than at earlier time points of
platelet spreading (10 min) (line graph, Fig.
1C). Finally, changes in the level of other cytoskeletal associated proteins, such as GP Ib, did not correlate with the increases in these other proteins (data not shown).
Fig. 2.
Proteolytic fragments of FAK,
pp60c-src, PTP-1B, and talin within
the Triton X-100-soluble fraction of spread platelets. Washed
platelets (3 × 108/ml) were preincubated with
Me2SO or calpeptin (50 µg/ml) for 30 min before their
application to the vWf matrix. Platelets were allowed to spread for the
indicated time points before lysis. The Triton X-100-soluble extracts
were immunoblotted with antibodies against FAK,
pp60c-src, PTP-1B, and talin. Note that the
PTP-1B immunoblots were overexposed to demonstrate the presence of the
42-kDa fragment. This either reflects poor recognition of the PTP-1B
cleavage product by the monoclonal antibody or is caused by rapid
cleavage of the 42-kDa fragment to lower molecular forms, not
recognized by the primary antibody. We consistently observed an
approximately 50% reduction in intact PTP-1B (50 kDa) in spreading and
aggregated platelets. These results are from one experiment,
representative of five.
[View Larger Version of this Image (39K GIF file)]
IIb
3
(36). Several lines of evidence suggest that the calpain activation
observed in spreading platelets is unlikely to be caused by the
formation of platelet aggregates in our static adhesion assays. First,
phase-contrast microscopy of platelets adherent to the vWf matrix on
Petri dishes failed to reveal platelet aggregate formation in 15 independent experiments. Furthermore, we have observed consistently
that the formation of small platelet aggregates in vWf-stimulated
nonstirred platelets is not associated with
pp60c-src
cleavage.2 Second, the
platelet adhesion studies were performed in the presence of the
platelet inhibitor theophylline, as described under "Experimental Procedures." In the presence of theophylline, platelets adhere to the
vWf matrix as single cells but do not spread or aggregate. All
nonadherent platelets were washed from the matrix before removing the
platelet inhibitor to ensure that no platelet aggregates form on the
adherent monolayer. Under these assay conditions, cytoskeletal signal
complex formation and calpain activation were identical to that
observed in platelets not pretreated with theophylline (data not
shown).
IIb
3 (37). Using monoclonal
antibodies against GP Ib-V-IX we were able to prevent platelet adhesion
to the vWf matrix, whereas antibodies against integrin
IIb
3 did not prevent platelet adhesion
but dramatically reduced platelet spreading. Similar observations were
found in Glanzmann's thrombasthenic platelets, which congenitally lack
integrin
IIb
3 (data not shown). These
observations limited our ability to investigate the relative roles of
GP Ib-V-IX and integrin
IIb
3 in inducing
calpain activation and the formation of cytoskeletal signaling
complexes in vWf-activated platelets, as these signaling events were
critically dependent on both platelet adhesion and spreading (data not
shown). We overcame these technical limitations by examining
cytoskeletal signal complex formation in suspension-activated
platelets. We have demonstrated previously in this system that vWf
binding to GP Ib-V-IX and subsequent platelet aggregation are
sufficient for pp60c-src
translocation to the cytoskeleton, independent of ligand binding to
integrin
IIb
3 (29). We investigated the
cytoskeletal translocation of the two calpain substrates, talin and
PTP-1B, in vWf-aggregated platelets. As with spreading platelets, both
talin and PTP-1B translocated to the cytoskeleton and underwent
proteolytic modification (Fig. 3). Both
the translocation and cleavage of PTP-1B and talin required platelet
stirring and aggregation and were abolished by anti-GP Ib-V-IX
antibodies (data not shown). Proteolysis of both substrates was also
abolished by pretreating the platelets with calpeptin (Fig. 3). A
potential limitation of these experimental conditions is the
requirement for a nonphysiological modulator, such as ristocetin, to
induce vWf binding to GP Ib-V-IX. The binding of vWf-ristocetin to GP
Ib-V-IX is irreversible and contrasts with the reversible binding of GP
Ib-V-IX to immobilized vWf (49). We have reported previously (29) that
bovine vWf, which binds human GP Ib-V-IX in the absence of
nonphysiological modulators, induces similar signaling events as
vWf-ristocetin, suggesting that these kinetic differences do not have a
major effect on signaling processes mediated by GP Ib-V-IX.
Fig. 3.
Calpain-mediated cleavage of talin and PTP-1B
in vWf-aggregated platelets. Washed platelets (3 × 108/ml) were preincubated with Me2SO or
calpeptin (50 µg/ml) for 30 min and then aggregated with vWf (10 µg/ml) and ristocetin (1 mg/ml) for 5 min. Cells were lysed and
fractionated into Triton X-100-soluble and cytoskeletal extracts as
described under "Experimental Procedures." Cytoskeletal proteins
were analyzed by immunoblotting with antibodies against talin or
PTP-1B. These results are from one experiment, representative of
five.
[View Larger Version of this Image (22K GIF file)]
IIb
3 in
regulating calpain activation and the formation of cytoskeletal signal
complexes in vWf-aggregated platelets, we disrupted ligand binding to
this adhesion receptor by preincubating platelets with EDTA for 90 min
at 37 °C or by blocking ligand binding with RGDS peptides or an
anti-integrin
IIb
3 monoclonal antibody.
As demonstrated in Fig. 4, abolishing
ligand binding to integrin
IIb
3 had no inhibitory effect on the ability of talin to translocate to the cytoskeleton, suggesting that cytoskeletal reorganization in response to vWf does not depend on signaling events linked to integrin
IIb
3 occupancy. Similar observations were
found with pp60c-src, p85, and
PTP-1B translocation (data not shown). Unfortunately we could not
examine FAK translocation under these assay conditions as ristocetin
nonspecifically precipitated this kinase. In contrast to translocation,
the cleavage of talin (Fig. 4) and PTP-1B (not shown) was abolished by
inhibiting ligand binding to integrin
IIb
3, indicating that integrin
IIb
3 postligand occupancy events are
essential for calpain activation in vWf-stimulated platelets. These
findings were observed consistently in both vWf-ristocetin and bovine
vWf-activated platelets.
Fig. 4.
Role of integrin
IIb
3 in regulating the cytoskeletal
translocation and cleavage of talin in vWf-aggregated platelets. Washed platelets were pretreated either with Me2SO,
calpeptin (50 µg/ml), RGDS peptide (1 mM), or anti-GP
IIIa mAb (
-GPIIIa) for 30 min at room temperature, or with 5 mM EDTA for 90 min (EDTA-90
) at 37 °C as
described under "Experimental Procedures." Platelets were
aggregated with vWf (50 µg/ml) and ristocetin (1 mg/ml) for the
indicated time points while stirring. Platelets were lysed with Triton
X-100, fractionated, and equal amounts of cytoskeletal protein
subjected to immunoblot analysis with an anti-talin mAb (8d4). These
results are from one experiment, representative of four.
[View Larger Version of this Image (18K GIF file)]
IIb
3
does not appear to be essential for cytoskeletal signal complex
formation in vWf-aggregated platelets, these results do not exclude the
possibility that this receptor may participate in the formation of
these complexes in the absence of bound ligand. To investigate further
the role of integrin
IIb
3 in vWf
signaling, we examined cytoskeletal signal complex formation and
calpain activation in Glanzmann's thrombasthenic platelets. As
demonstrated in Fig. 5, A and
B, pp60c-src, p85, talin,
and PTP-1B all translocated to the cytoskeletal fraction of
vWf-aggregated Glanzmann's platelets, strongly suggesting that
integrin
IIb
3 is not essential for this
signaling event. As expected, talin and PTP-1B were not proteolyzed in
these platelets. To confirm that the reduction in talin and PTP-1B
cleavage in Glanzmann's platelets was the result of the inability of
vWf to activate calpain, we immunoblotted whole cell lysates with an antibody that recognizes the full-length (80 kDa) inactive form of
calpain (Fig. 5C). Previous studies in thrombin-stimulated platelets have demonstrated that immunoblot analysis with this antibody
is a sensitive means of detecting calpain activation within the cell
(38, 39). Activation of calpain is associated with the autolytic
conversion of its large subunit (80 kDa) to 78- and 76-kDa active
forms, which are not recognized by this antibody. As demonstrated in
Fig. 5C, thrombin or vWf stimulation of normal platelets
leads to a 70-75% reduction in the amount of inactive calpain. In
contrast, no reduction in full-length calpain was observed in thrombin
or vWf-stimulated Glanzmann's platelets.
Fig. 5.
Calpain activation and cytoskeletal
association of pp60c-src, PI-3
kinase, talin, and PTP-1B in normal or Glanzmann's thrombasthenic platelets aggregated with vWf. Washed platelets (3 × 108/ml) from a healthy volunteer or from a patient with
Glanzmann's thrombasthenia were stirred in the presence of vWf (50 µg/ml) and ristocetin (1 mg/ml), or thrombin (1 unit/ml) for the
indicated time points. Aggregated platelets were lysed with Triton
X-100 and either analyzed as whole cell lysates or fractionated into Triton X-100-soluble or cytoskeletal extracts. Panel A,
immunoblot analysis of the cytoskeletal fraction from normal and
Glanzmann's platelets with antibodies against GP IIb,
pp60c-src, p85 subunit of PI-3
kinase (panel B) or against talin and PTP-1B. Note that the
additional band observed in the p85 immunoblot was a nonspecific band
not consistently observed in all experiments. Panel C, whole
cell lysates from platelets aggregated with vWf and ristocetin for 20 min were analyzed for calpain activation using an antibody that
selectively recognizes the 80-kDa inactive form of calpain. The extent
of calpain activation (histogram) was quantitated by
performing densitometry on the calpain immunoblots. The results
presented are from one experiment, representative of two independent
experiments performed in duplicate.
[View Larger Version of this Image (21K GIF file)]
Fig. 6.
Effect of calpain activation on
protein-tyrosine phosphorylation in vWf-stimulated platelets.
Washed platelets (1 × 108/ml) were preincubated with
Me2SO or calpeptin (50 µg/ml) for 30 min before the
performance of (panel A) platelet adhesion or (panel
B) platelet aggregation studies. Panel A, adherent
platelets were allowed to spread for the indicated time points before
lysis in SDS-sample buffer. Whole cell lysates were subjected to
immunoblot analysis with an anti-phosphotyrosine mAb (PY20).
Panel B, washed platelets (3 × 108/ml)
were aggregated with vWf (10 µg/ml) and ristocetin (1 mg/ml) for the
indicated time points in the presence or absence of calpeptin (50 µg/ml). Total cell lysates were subjected to anti-phosphotyrosine immunoblot analysis. Panel C, cytoskeletal extracts from
vWf-aggregated platelets were immunoblotted with anti-PTP-1B mAb. These
results were from one experiment representative of four.
[View Larger Version of this Image (42K GIF file)]
IIb
3, talin,
pp60c-src, FAK, and PI 3-kinase in
spreading platelets pretreated with calpeptin, E64d, or
Me2SO (Fig. 7). We
consistently observed in these studies a 2-4-fold higher content of
each of these proteins in spreading platelets pretreated with calpeptin
or E64d compared with Me2SO controls. This increase was
specific to these proteins as the cytoskeletal content of F-actin or GP
Ib remained essentially unchanged under the same experimental
conditions. We performed similar experiments in vWf-aggregated
platelets and observed a similar phenomenon (Fig. 3 and data not
shown). Collectively, these results suggest a potentially important
role for this protease in regulating cytoskeletal signaling complex
formation in vWf-stimulated platelets.
Fig. 7.
Effect of calpain activation on the
cytoskeletal content of GP IIb, talin,
pp60c-src, p85, FAK, and
actin. Washed platelets (1 × 108/ml) were
pretreated with Me2SO or calpeptin (50 µg/ml) for 30 min
and then applied to the vWf matrix in the presence of theophylline to
allow platelet adhesion without spreading. After 60 min, nonadherent cells were washed from the dish and the theophylline removed to allow
platelet spreading. At the indicated time points the adherent platelets
were lysed and the cytoskeletal proteins extracted. Immunoblot analysis
of equal quantities of these extracts was performed using antibodies
against GP IIb, talin, pp60c-src,
p85 subunit of PI 3-kinase, FAK, and actin. The cytoskeletal content of
each of these proteins (histogram) was quantitated by
performing densitometry on each of the immunoblots. Note that unlike
FAK and PTP-1B, the proteolysis of talin within the cytoskeletal fraction of platelets was not always completely inhibited by
pretreating these cells with calpeptin. These results were derived from
one experiment, representative of four.
[View Larger Version of this Image (27K GIF file)]
IIb
3 to subendothelial bound vWf (34).
There is growing evidence that the interaction between vWf and these
adhesion receptors plays an important role in the initiation of
platelet activation, leading to platelet shape change, spreading,
granule release, and aggregation (29, 40-42). Despite the fundamental
importance of these adhesion events, the signaling processes linked to
these receptors, especially GP Ib-V-IX, remain poorly understood. The studies presented in this article reveal the following information regarding signaling processes induced by the binding of vWf to human
platelets. First, vWf is capable of inducing the cytoskeletal translocation of multiple structural proteins and signaling enzymes in
both spreading and aggregated platelets. The cytoskeletal association of these signaling molecules in aggregated platelets requires vWf
binding to GP Ib-V-IX but occurs independently of ligand binding to
integrin
IIb
3. Second, vWf stimulation of
platelets induces calpain activation, leading to the limited
proteolysis of talin, PTP-1B, FAK, and
pp60c-src. The activation of calpain
by vWf is distinct from the formation of cytoskeletal signaling
complexes in that it requires ligand binding to both GP Ib-V-IX and
integrin
IIb
3. Finally, the activation of
calpain in vWf-stimulated platelets results in a substantial decrease
in the level of tyrosine phosphorylation of multiple platelet proteins
and is associated with a 50-80% reduction in the cytoskeletal content
of integrin
IIb
3, talin,
pp60c-src, FAK, PTP-1B, and PI
3-kinase.
IIb
3 in regulating the formation of
cytoskeletal signaling complexes. Studies in thrombin-stimulated
platelets have suggested that fibrinogen binding to integrin
IIb
3 induces cytoskeletal reorganization
and the translocation of multiple signaling enzymes to the cytoskeleton (7-13). Ligand binding to integrin
IIb
3 per se does not appear to
be sufficient for the formation of these signaling complexes, as these
events only occur after stirring the platelet suspensions to induce
platelet aggregation (8). We have demonstrated a similar phenomenon in
vWf-stimulated platelets, whereby the binding of vWf to GP Ib-V-IX
induces the translocation of signaling enzymes when these cells are
stirred and allowed to aggregate (29). These observations raise the
possibility that vWf binding to GP Ib-V-IX induces cytoskeletal changes
similar to those observed following fibrinogen binding to integrin
IIb
3. Recent studies in Chinese hamster
ovary cells transfected with GP Ib
, GP Ib
, and GP IX support
these findings, as they demonstrate that binding of transfected cells
to vWf induces cell spreading and the extension of pseudopodia. These
changes in cell morphology appear to be regulated in part by the
interaction of the cytoplasmic tail of GP Ib
with the cytoskeleton,
as truncation of this tail resulted in a significant alteration in both
the morphology and adhesion of these cells (43).
IIb
3. A similar requirement for integrin
IIb
3 in calpain activation has been
reported previously in thrombin-stimulated platelets (36), although the
mechanism by which this receptor regulates calpain activity has yet to
be established. Studies in platelets have suggested an important role
for inositol phospholipids and calcium influx in inducing calpain
activation (21, 44). The more highly phosphorylated phosphoinositides
(phosphoinositide 4-phosphate and 4,5-bisphosphate) appear to be most
effective in lowering the calcium requirement for calpain activation
(44). Studies in cultured cells and platelets have demonstrated an
important role for integrins in regulating the formation of
phosphoinositide 4,5-bisphosphate and D-3 phosphoinositides (45, 46).
Furthermore, integrin
IIb
3 can regulate
the influx of calcium in response to a variety of physiological
agonists, including vWf (47, 48). Whether these integrin
IIb
3-regulated signaling processes are
directly responsible for calpain activation in vWf-stimulated platelets
will require further investigation.
IIb
3, talin, PTP-1B, FAK,
pp60c-src, and PI 3-kinase and
enhanced total cell protein-tyrosine phosphorylation, whereas
membrane-impermeable calpain inhibitors had no effect on these
signaling events.
*
This work was supported in part by a grant from the National
Health and Medical Research Council of Australia.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.
Recipient of an Australian postgraduate award. To whom
correspondence should be addressed: Dept. of Medicine, Monash Medical School, Box Hill Hospital, Arnold St., Box Hill, Victoria 3128, Australia. Tel.: 61-3-9895-0311; Fax: 61-3-9895-0332.
1
The abbreviations used are: vWf, von Willebrand
factor; GP, glycoprotein; PI 3-kinase, phosphatidylinositol 3-kinase;
PTP, protein-tyrosine phosphatase; RGDS peptide, Arg-Gly-Asp-Ser
peptide; FITC, fluorescein isothiocyanate; mAb, monoclonal antibody;
Me2SO, dimethyl sulfoxide; PBS, phosphate-buffered saline;
FAK, focal adhesion kinase.
2
Y. Yuan, S. Dopheide, C. Ivanidis, H. H. Salem,
and S. P. Jackson, unpublished observations.
©1997 by The American Society for Biochemistry and Molecular Biology, Inc.