(Received for publication, November 18, 1996, and in revised form, May 20, 1997)
From the Molecular Oncology Group, The Tpr-Met oncoprotein consists of the catalytic
kinase domain of the hepatocyte growth factor/scatter factor receptor
tyrosine kinase (Met) fused downstream from sequences encoded by the
tpr gene. Tpr-Met is a member of a family of tyrosine
kinase oncoproteins generated following genomic rearrangement and has
constitutive kinase activity. We have previously demonstrated that a
single carboxyl-terminal tyrosine residue, Tyr489, is
essential for efficient transformation of Fr3T3 fibroblasts by Tpr-Met
and forms a multisubstrate binding site for Grb2, phosphatidylinositol 3 The receptor for hepatocyte growth factor/scatter factor
(HGF/SF),1 the Met receptor
tyrosine kinase (RTK), is expressed primarily in epithelial and
endothelial cells in vivo and in vitro where it
mediates the pleiotropic biological responses of HGF/SF (1-5). HGF/SF
is a mitogen for primary hepatocytes, and stimulates scatter, invasion,
and branching tubulogenesis of epithelial cells (6-9). A critical role
for both HGF/SF and the Met RTK in development was demonstrated by the
embryonic lethality of mice lacking genes encoding either HGF/SF or the
Met RTK (10, 11). Moreover, amplification and overexpression of the Met
gene is a frequent event in many human tumors (2, 12-16). Thus,
activation of the Met RTK is also implicated in neoplasia. Consistent
with this, the Met RTK was originally isolated as an oncogene, Tpr-Met
(14). Oncogenic activation of the Met receptor occurred following
translocation of sequences encoding the amino terminus of the Tpr gene
located on chromosome one upstream of sequences encoding the
cytoplasmic domain of the Met RTK located on chromosome seven (15). The product of the resulting chimeric gene, Tpr-Met, is a constitutively activated, transforming kinase (16, 17).
To define signal transduction pathways required for transformation of
fibroblasts by Tpr-Met, we have shown that a single carboxyl-terminal
tyrosine residue, Tyr489, is phosphorylated (18) and
essential for the association of Tpr-Met with the Grb2 adaptor protein,
phospholipase C The generation of the various Tpr-Met mutant
proteins has been described previously (18-20).
All cell lines were
maintained in Dulbecco's modified Eagle's medium containing 10%
fetal bovine serum (Life Technologies, Inc.) and antibiotics. Cell
lines expressing either wild-type or mutant forms of Tpr-Met were
generated by ecotropic retroviral infection of parental Fischer rat 3T3
(Fr3T3) fibroblasts as described previously (22).
Antibodies which recognize Tpr-Met were
generated against a carboxyl-terminal peptide of the Tpr-Met protein as
described previously (22). Antibodies which recognize Gab1 were
generated against bacterially expressed GST-Gab1 as described
previously (23). Antibodies to Cbl were purchased from Santa Cruz
Biotechnology (Santa Cruz, CA). Antibodies to hemagglutinin were
purchased from Berkeley Antibody Company. For anti-phosphotyrosine
immunoblotting, recombinant horseradish peroxidase-conjugated
antiphosphotyrosine Fab fragment, RC20H, from Transduction
Laboratories, was utilized.
Generation of cell lysates, immunoprecipitation, and
Western blotting have been described previously (19).
Bacteria expressing GST
proteins fused to either full-length Grb2, SH2, SH3(N)-SH2, or
SH2-SH3(C) domains of Grb2 were kindly provided by Drs. Mike Moran,
Alain Charest, and Michel Tremblay. Fusion proteins were produced by
isopropyl- 293
cells (8 × 105) were transfected with 5 µg of
expression plasmid DNA encoding either wild-type or mutant forms of the
Tpr-Met proteins utilizing calcium phosphate coprecipitation. In Fig. 5B, 2 µg of an expression plasmid encoding
hemagglutinin-tagged Gab1 were included. The cells were maintained for
3 days in Dulbecco's modified Eagle's medium, 10% fetal bovine
serum, at which time they were harvested in 0.5% Triton X-100 lysis
buffer. COS-1 cells (106) were transfected with 6 µg of
expression plasmid DNA encoding either wild-type or mutant forms of the
Tpr-Met proteins utilizing a modification of the DEAE-dextran protocol
(19). The cells were maintained for 60 h in Dulbecco's modified
Eagle's medium, 10% fetal bovine serum, after which they were
harvested in 0.5% Triton X-100 lysis buffer. Tpr-Met proteins from
approximately 20% of each lysate were immunoprecipitated with antibody
144, the immune complexes collected on protein A-Sepharose, and washed with 0.5% Triton X-100 lysis buffer. In vitro
association/kinase assays were carried out as described (19, 25). For
denaturation and reimmunoprecipitation following the in
vitro association/kinase assay, the immune complexes were boiled
for 5 min in 900 µl of buffer composed of 0.4% SDS, 50 mM triethanolamine, 100 mM NaCl, 2 mM EDTA, and 2 µM
The carboxyl terminus of the Tpr-Met oncoprotein contains three
tyrosine residues, Tyr482, Tyr489, and
Tyr498, two of which are followed by consensus binding
sites for several SH2 binding proteins (Fig.
1) (26, 27). Consistent with
Tyr489 being the only detectably phosphorylated residue in
the carboxyl terminus (18), mutation of this residue to a conserved Phe
residue results in a dramatic decrease in transforming activity of the Y489F mutant protein (19, 28). Utilizing carboxyl-terminal tyrosine
mutant Tpr-Met proteins, Y482F, Y489F, and Y482F/Y489F, as well as a
mutant that has selectively lost the ability to associate with Grb2,
N491H, we have demonstrated that pathways downstream of Grb2 and Shc
are required for transformation of Fr3T3 cells by Tpr-Met (20). Thus,
characterization of the signaling pathways downstream of Grb2 and Shc
will permit further definition of the mechanism by which Tpr-Met
transforms Fr3T3 fibroblasts.
Using a
sensitive in vitro association/kinase assay to detect
proteins which associate with Tpr-Met, we have previously identified a
highly phosphorylated protein of 110 kDa whose association with and/or
phosphorylation by Tpr-Met was dependent upon the presence of
Tyr489 (19) (Fig.
2A, lane 5). When
compared with wild-type Tpr-Met (Fig. 2A, lane
2), a Y482F/Y489F Tpr-Met mutant failed to associate with and/or
phosphorylate the 110-kDa protein (Fig. 2A, lane
6), whereas the Y482F Tpr-Met mutant retained the ability to
associate with and/or phosphorylate this protein (Fig. 2A,
lane 4). Interestingly, the N491H Tpr-Met mutant, which no
longer binds to Grb2, was also severely impaired in its ability to
associate with and/or phosphorylate the 110-kDa protein (Fig.
2A, lane 7). This suggested that the association
of the 110-kDa protein with Tpr-Met was indirect, mediated by the Grb2
adaptor protein, or direct and, like Grb2, dependent upon the Asn
residue two amino acids downstream of tyrosine 489.
To investigate whether the 110-kDa protein could associate directly
with Grb2, the phosphorylated proteins from an in vitro association/kinase assay were denatured by boiling in buffer containing 0.4% SDS and 2 µM The Cbl protooncogene product, which is similar in size to
the 110-kDa protein, is phosphorylated on tyrosine residues following stimulation of several receptor tyrosine kinases (29, 30) and in v-Src
and v-Abl transformed fibroblasts (29, 31). As demonstrated by
phosphotyrosine immunoblotting, Cbl was highly phosphorylated in the
stable cell lines expressing wild-type or the highly transforming Y482F
Tpr-Met mutants (Fig. 3A,
lanes 2-5). However, Cbl was poorly phosphorylated in
stable cell lines expressing the weakly transforming N491H, Y489F or
the nontransforming Y482F/Y489F Tpr-Met mutants (Fig. 3A,
lanes 6-11), although similar amounts of protein were
immunoprecipitated (Fig. 3B). Thus, in a manner similar to
the 110-kDa protein, the increased level of Cbl phosphorylation in
stable cell lines expressing Tpr-Met was dependent upon
Tyr489 and more specifically, the Asn residue two amino
acids downstream of Tyr489.
These data suggested that Cbl was binding Tpr-Met either directly,
recognizing the consensus binding site, YVNV, present at position 489, or indirectly via another protein that required Tyr489 for
association. Cbl contains several proline-rich repeats that mediate its
association with SH3 domain-containing proteins, including Grb2
(32-34). Because Cbl was not highly phosphorylated in cell lines
expressing Tpr-Met mutants unable to bind Grb2, the possibility that
Grb2 was functioning as an adaptor coupling the Tpr-Met oncoprotein with Cbl was investigated. A GST-Grb2 fusion protein was able to bind
to Cbl present in lysates prepared from both Fr3T3 fibroblasts and cell
lines expressing wild-type or the mutant Tpr-Met proteins (Fig.
3C, lanes 3-10). Furthermore, while full-length
Grb2 (Fig. 3D, lane 1), as well as the
SH3(N)-SH2, and SH3(N)-SH3(C) fusion proteins (Fig. 3D,
lanes 2 and 4) all retained the ability to associate with Cbl from wild-type Tpr-Met-transformed Fr3T3 cells, fusion proteins encoding the SH2-SH3(C) or the SH2 domain alone of Grb2
did not bind Cbl well (Fig. 3D, lanes 3 and
5). Each of the GST fusion proteins was present in
approximately the same amount (Fig. 3D, bottom).
Thus, these data demonstrate that the association of Cbl with Grb2 was
mediated by the amino-terminal SH3 domain of Grb2. This is consistent
with the ability of GST-Grb2 to bind Cbl irrespective of the
phosphorylation status of Cbl (Fig. 3C). However, as shown
in Fig. 2B, the association between Grb2 and the 110-kDa
protein was mediated primarily by the carboxyl-terminal SH3 domain of
Grb2, suggesting that the 110-kDa protein was not Cbl. Furthermore, Cbl
was not readily coimmunoprecipitated with Tpr-Met in Fr3T3 cells
transformed by Tpr-Met (data not shown) or when Tpr-Met was
overexpressed in 293 cells, which express high levels of endogenous Cbl
(32) (Fig. 4A, lanes
2 and 3). In contrast, consistent with previous reports
(32, 35, 36) the EGF receptor coimmunoprecipitated with Cbl following
EGF stimulation of 293 cells (Fig. 4A, lanes 4 and 5, and data not shown). Thus, while Cbl
coimmunoprecipitated with the EGF receptor it did not detectably
coimmunoprecipitate with Tpr-Met, even though abundant Tpr-Met protein
was present (Fig. 4B, lanes 1-3) and Cbl was
phosphorylated on tyrosine residues in 293 cells expressing wild-type
Tpr-Met (Fig. 4A, lanes 1 and 2).
Moreover, when compared in the in vitro association/kinase
assay, a lysate depleted of Cbl protein (Fig. 4C) contained
abundant levels of the phosphorylated 110-kDa protein compared with a
nondepleted lysate (Fig. 4D, lanes 2 and
3). Thus, while Cbl is phosphorylated on tyrosine residues
in cells transformed by Tpr-Met, it does not correspond to the 110-kDa
protein that associates stably with and is tyrosine phosphorylated by
Tpr-Met.
Gab1, a Grb2-binding protein of 115 kDa, was recently
isolated and shown to be a multisubstrate docking protein that is
tyrosine-phosphorylated following stimulation of cells with EGF and
insulin (23). Because both Gab1 and the 110-kDa protein associate with
Grb2 and are similar in size, the ability of Gab1 to associate with
Tpr-Met was determined. In the in vitro association assay
Gab1 expressed in Fr3T3 fibroblasts associated with the wild-type
Tpr-Met oncoprotein (Fig. 5A,
lane 1) and to a lesser extent with the highly transforming Y482F mutant Tpr-Met protein (Fig. 5A, lane 3).
However, it did not associate well with the poorly transforming Y489F,
N491H, or the nontransforming Y482F/Y489F mutant Tpr-Met proteins, none of which bind Grb2 (Fig. 5A, lanes 4-6).
Similarly, in 293 cells overexpressing Tpr-Met and Gab1, both the
wild-type and the highly transforming Y482F mutant Tpr-Met proteins
coimmunoprecipitated with Gab1 (Fig. 5B, lanes 1 and 3), whereas the nontransforming Y482F/Y489F Tpr-Met
mutant associated poorly with Gab1 (Fig. 5B, lane
5), to the same extent as the nontransforming K241A Tpr-Met mutant
(Fig. 5B, lane 2). Interestingly, compared with
wild-type Tpr-Met, when overexpressed, the poorly transforming Y489F
and N491H mutants showed weak association with Gab1 (Fig.
5B, lanes 4 and 6). Each of the
Tpr-Met proteins was expressed at approximately equal levels (Fig.
5C, lanes 1-6).
In a manner similar to the 110-kDa protein (Fig. 2B), Gab1
from serum-starved Fr3T3 cells (Fig. 5D) or Tpr-Met
transformed Fr3T3 cells (Fig. 5E) associated predominantly
with GST-Grb2 fusion proteins containing the carboxyl-terminal SH3
domain of Grb2 (Fig. 5, D and E, lanes
1, 3, and 4) and not detectably with a
fusion protein encoding only the SH2 domain of Grb2 (Fig. 5,
D and E, lanes 5). Thus, because both
Gab1 and the 110-kDa protein bound preferentially to the
carboxyl-terminal SH3 domain of Grb2, were impaired in their ability to
interact with Tpr-Met mutants that did not bind Grb2 (Y489F and N491H,
Figs. 2A and 5, A and B) and were
similar in size, we examined whether the 110-kDa protein was Gab1. In
the in vitro association/kinase assay, using a lysate immunodepleted with anti-Gab1 antiserum, the 110-kDa Tpr-Met substrate was no longer detected (Fig. 5F, lane 4).
However, a lysate treated similarly, but without addition of anti-Gab1
antiserum, retained the 110-kDa protein (Fig. 5F, lane
3). Thus, the 110-kDa Tpr-Met substrate corresponds to the
Grb2-associated docking protein, Gab1.
Using a series of mutant Tpr-Met oncoproteins, we have previously
demonstrated that, in the absence of direct Grb2 association and Shc
phosphorylation, Tpr-Met fails to transform Fr3T3 fibroblasts (20).
Thus, the identification of pathways downstream of these two adaptor
molecules is crucial in defining the mechanism by which Tpr-Met
transforms fibroblasts. We have shown here that tyrosine
phosphorylation of the Cbl protooncogene product and the multisubstrate
docking protein, Gab1, correlates with the ability of the Tpr-Met
oncoprotein to associate with the Grb2 adaptor protein and to transform
cells.
The c-Cbl protooncogene
product is a 120-kDa phosphoprotein that is tyrosine-phosphorylated in
cells transformed by v-Src, v-Abl, or Bcr-Abl and following activation
of multiple receptor tyrosine kinases (29-31, 33, 35, 37). Consistent
with tyrosine phosphorylation of Cbl playing a role in cellular
transformation, Cbl was phosphorylated on tyrosine residues in cells
expressing only the highly transforming wild-type or Y482F mutant
Tpr-Met oncoproteins (Figs. 3A and 4A). The level
of Cbl phosphorylation was reduced in cells expressing the weakly
transforming Tpr-Met mutants, Y489F and N491H, which do not bind to
Grb2 (Fig. 3A) (19-21). Thus, the dramatic decrease in Cbl
phosphorylation in cells expressing the N491H or Y489F Tpr-Met mutants
suggests that either Grb2 or the Asn residue two amino acids downstream
of Tyr489 is required for subsequent phosphorylation of
Cbl. Consistent with the former, Cbl contains proline-rich sequences
that interact with SH3 domain-containing proteins (33, 38-40) and a
GST-Grb2 fusion protein was able to bind Cbl from Fr3T3 cells and all
Tpr-Met expressing cell lines irrespective of the phosphorylation
status of Cbl (Fig. 3C). Moreover, an interaction between
wild-type Tpr-Met and Cbl was not readily detected in
Tpr-Met-transformed fibroblasts (data not shown) or in cotransfection
assays where an association between Cbl and the EGF receptor was
observed (Fig. 4A). Thus, while the exact mechanism by which
Cbl is phosphorylated in Tpr-Met-transformed fibroblasts is unclear,
the level of Cbl phosphorylation correlated with the ability of the
mutant Tpr-Met proteins to transform cells.
In addition to Cbl, the
highly transforming wild-type Tpr-Met oncoprotein induced tyrosine
phosphorylation of a 110-kDa protein both in vivo (19) and
in vitro (Fig. 2A). Like Cbl, phosphorylation of
this protein was dependent upon the ability of Tpr-Met to associate with Grb2. However, unlike Cbl, which bound primarily to GST-Grb2 fusion proteins containing the amino-terminal SH3 domain of Grb2 (Fig.
3D), the 110-kDa protein bound primarily to GST-Grb2 fusion proteins containing the carboxyl-terminal SH3 domain of Grb2 (Fig. 2B). Moreover, unlike Cbl, a stable association between the
110-kDa protein and Tpr-Met was detected (Fig. 2A) and a
lysate prepared from Fr3T3 fibroblasts immunodepleted of Cbl, still
contained abundant levels of the 110-kDa phosphoprotein (Fig.
4D), providing further evidence that the 110-kDa protein did
not correspond to Cbl.
Recently a 115-kDa, Grb2 associated protein, Gab1, was identified and
shown to be tyrosine-phosphorylated following stimulation of cells with
EGF and insulin (23). Our data demonstrate that the highly
phosphorylated 110-kDa Tpr-Met substrate in Fr3T3 fibroblasts is Gab1.
Like the 110-kDa Tpr-Met substrate, Gab1 associated primarily with the
carboxyl-terminal SH3 domain of Grb2 (Figs. 2B and 5, D, and E) and neither Gab1 nor the 110-kDa
Tpr-Met substrate associated efficiently with Tpr-Met mutant proteins
that failed to bind Grb2 directly (Figs. 2A and 5,
A and B). Furthermore, a lysate prepared from
Fr3T3 fibroblasts and immunodepleted of Gab1 contained no 110-kDa
protein (Fig. 5F), consistent with the 110-kDa protein being
Gab1. While the majority of Gab1 association with Tpr-Met was dependent
upon an intact Grb2 binding site at Tyr489, we also
consistently observed low, but detectable binding of Gab1 to the Y489F
and N491H Tpr-Met mutants that fail to bind Grb2 (Fig. 5, A
and B). This may represent direct binding of Gab1 to Y482 of
Tpr-Met. Interestingly, a direct interaction between Gab1 and Y1349 in
the Met receptor (which corresponds to Y482 in Tpr-Met) was identified
using the yeast two hybrid system (41). This is consistent with our
data; however, we demonstrate here that the majority of Gab1
association with Tpr-Met is mediated by Tyr489 and Grb2 and
not by direct association with Tyr482. Importantly, the
near loss of Gab1 phosphorylation and reduced association of Gab1 with
the Y489F and N491H Tpr-Met mutants correlated with the poor
transforming activity of these mutant proteins.
From genetic and biochemical data, Grb2
couples receptor tyrosine kinases to the Ras pathway via an interaction
with SOS, a guanine nucleotide exchange factor for Ras (42-48).
Activation of Ras occurs following association of Grb2 with either the
activated RTK or another adaptor protein, Shc (42, 49-51). In cell
lines expressing the poorly transforming Tpr-Met mutants that fail to bind Grb2, Shc is still phosphorylated and coupled to Grb2, yet these
mutant oncoproteins activate Ras-dependent pathways at
50-60% the efficiency of the wild-type Tpr-Met oncoprotein (20). The poor transforming activity of these Tpr-Met mutants (20% of wild-type) was inconsistent with the observed decrease in Ras dependent signaling suggesting that other Grb2 dependent pathways, in addition to Ras, were
required for efficient cell transformation by Tpr-Met. In support of
this, our data demonstrate that phosphorylation of Cbl and Gab1 was
dependent upon Grb2 association with Tyr489 of Tpr-Met.
From sequence analysis, both Cbl and Gab1 contain proline rich
sequences that interact with SH3 domain-containing proteins such as
Grb2 and several tyrosine residues which, when phosphorylated,
associate with multiple signaling molecules (23, 38).
Functionally, tyrosine phosphorylation of both Gab1 and Cbl is
associated with an altered growth response. Constitutively phosphorylated variants of Cbl transform fibroblasts (31) and NIH3T3
cells overexpressing Gab1 have an increased growth rate in 1% serum
and grow in soft agar in the presence of EGF or insulin whereas control
cells do not (23). Although the function of Cbl and/or Gab1 is as yet
unknown, Tpr-Met mutants impaired in their ability to couple with Grb2
and to transform fibroblasts fail to induce high levels of
phosphorylation of Cbl or Gab1 consistent with one or both of these
proteins functioning as important signal transducers for transformation
by Tpr-Met.
We thank members of the Park laboratory for
helpful discussions and Drs. Alain Charest, Michel Tremblay, and Mike
Moran for GST fusion proteins.
ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
FOOTNOTES
ACKNOWLEDGEMENTS
REFERENCES
kinase, phospholipase C
, SHP2, and an unknown protein of 110 kDa.
A mutant Tpr-Met protein that selectively fails to bind Grb2 has
reduced transforming activity, implicating pathways downstream of Grb2
in Tpr-Met mediated cell transformation. We show here that the 110-kDa
Tpr-Met substrate corresponds to the recently identified
Grb2-associated protein, Gab1. Moreover, we show that tyrosine
phosphorylation of the Cbl protooncogene product as well as Gab1
required Tyr489 and correlated with the ability of Tpr-Met
to associate with Grb2 and to transform cells, providing evidence that
pathways downstream of Gab1 and/or Cbl may play a role in
Tpr-Met-mediated cell transformation.
, and the tyrosine phosphatase, SHP2 (19, 20). This
tyrosine is also required for the activation of phosphatidylinositol
3
-kinase and the tyrosine phosphorylation of and/or association with
an unknown protein of 110 kDa (19). A Y489F Tpr-Met mutant transforms cells at 20% of the efficiency of the wild-type Tpr-Met oncoprotein; however, because the association of multiple signaling molecules with
Tpr-Met is dependent upon Tyr489, it was unclear which of
these were required for transformation. A mutant that selectively fails
to associate with Grb2, yet retains the ability to associate with
phospholipase C
, phosphatidylinositol 3
-kinase, and SHP2,
transforms cells with the same efficiency as the Y489F mutant (20, 21).
Thus, association of Tpr-Met with the Grb2 adaptor protein is essential
for efficient transformation of fibroblasts by Tpr-Met. The
identification of signaling pathways downstream of Grb2 will help
define the mechanism by which Tpr-Met transforms Fr3T3 fibroblasts. In
the present study we demonstrate that association of Grb2 with Tpr-Met
is essential for the efficient tyrosine phosphorylation of the Cbl
protooncogene product, and a previously characterized 110-kDa protein
which we show corresponds to the Grb2-associated docking protein,
Gab1.
Mutagenesis
-D-thiogalactopyranoside induction and
purification on glutathione-agarose beads (24). Approximately 0.5-1.0
µg of protein was used in the in vitro association experiments.
-mercaptoethanol.
Following addition of 125 µl of Tris-Cl (pH 7.4) and 250 µl of 10%
Triton X-100, the lysates were incubated with GST-Grb2 fusion proteins
or anti-Tpr-Met antibody 144 for 3 h at 4 °C and washed with
0.5% Triton X-100 lysis buffer. Samples were then suspended in Laemmli
sample buffer, boiled for 10 min, and subjected to 9% SDS-PAGE.
Proteins were visualized by autoradiography.
Fig. 5.
The highly phosphorylated 110-kDa protein
corresponds to Gab1. A, wild-type Tpr-Met (lane
1) or the various Tpr-Met mutant proteins (lanes 2-6)
from transiently transfected COS-1 cells were immunoprecipitated and
activated by phosphorylation with nonradioactive ATP. Extracts from
serum-starved Fr3T3 cells were added and incubated for 3 h. The
oncoprotein complexes were washed three times, resolved by 9% SDS-PAGE
and immunoblotted with anti-Gab1 serum. The positions of Tpr-Met, Gab1,
and molecular weight markers are indicated. B and
C, lysates from 293 cells transiently transfected with
expression plasmids encoding either wild-type or the mutant forms of
Tpr-Met and hemagglutinin (HA)-tagged murine Gab1 were
immunoprecipitated with anti-HA antibody (B), or
anti-Tpr-Met antibody (C), resolved by SDS-PAGE and
immunoblotted with anti-Tpr-Met antibody. Lysates from parental Fr3T3
cells (D) or wild-type Tpr-Met transformed Fr3T3 cells
(E) were incubated with GST proteins encoding full-length
Grb2 (lane 1), and the SH3(N)-SH2 (lane 2),
SH2-SH3(C) (lane 3), SH3(N)-SH3(C) (lane 4), or
SH2 (lane 5) domains of Grb2 coupled to
glutathione-Sepharose. Complexes were washed, resolved by SDS-PAGE, and
immunoblotted with anti-Gab1 antiserum. The amount of each GST fusion
protein is shown in Fig. 3D, bottom. F, lysis
buffer (lane 1), an untreated lysate (lane 2), a
protein A-Sepharose-treated lysate (lane 3), or a
Gab1-depleted lysate (lane 4) were then added to wild-type Tpr-Met from transiently transfected COS-1 cells that had been immunoprecipitated and activated by phosphorylation with nonradioactive ATP. The protein complexes were incubated in kinase buffer with [-32P]ATP, resolved by 9% SDS-PAGE and visualized by
autoradiography. The positions of Tpr-Met, the 110-kDa protein, and
molecular weight markers are indicated.
[View Larger Version of this Image (44K GIF file)]
Fig. 1.
Schematic representation of the Tpr-Met
oncoprotein. The hatched boxes represent the kinase
domain and the shaded boxes represent a leucine zipper
dimerization motif that is required for constitutive activation of
Tpr-Met. The sequences surrounding the carboxyl-terminal tyrosines
Tyr482 and Tyr489 of Tpr-Met are shown along
with the transformation efficiency of the Y482F, Y489F, Y482F/Y489F,
and N491H Tpr-Met mutants.
[View Larger Version of this Image (29K GIF file)]
Fig. 2.
The 110-kDa protein is not associated with
and/or phosphorylated by the N491H Tpr-Met mutant in
vitro. A, wild-type Tpr-Met (lanes 1 and
2) or the various Tpr-Met mutant proteins (lanes
3-7) from transiently transfected COS-1 cells were
immunoprecipitated and activated by phosphorylation with nonradioactive
ATP. Extracts from serum-starved Fr3T3 cells (lanes 2-7) or
an equivalent volume of lysis buffer (lane 1) were then
added. To visualize associated proteins, the oncoprotein complexes were
washed several times and then incubated in kinase buffer with
[-32P]ATP. The complexes were resolved by 8% SDS-PAGE
and visualized by autoradiography. The positions of Tpr-Met, the
110-kDa protein, and the molecular weight markers are indicated.
B, following incubation with [
-32P]ATP,
Tpr-Met and associated proteins were boiled 5 min in buffer containing
0.4% SDS and 2 µM
-mercaptoethanol (see "Materials and Methods"). After adding Tris-Cl and Triton X-100, the dissociated and denatured proteins were incubated with GST proteins encoding full-length Grb2 (lane 1), and the SH3(N)-SH2 (lane
2), SH2-SH3(C) (lane 3), SH3(N)-SH3(C) (lane
4), or SH2 (lane 5) domains of Grb2 bound to
glutathione-Sepharose or antibody 144 that recognizes Tpr-Met
(lane 6). Following several washes with lysis buffer, the
proteins were resolved by 9% SDS-PAGE and visualized by
autoradiography.
[View Larger Version of this Image (32K GIF file)]
-mercaptoethanol. Following
addition of Tris-Cl and Triton X-100, the dissociated and denatured
proteins were incubated with anti-Tpr-Met antibody 144 (22) or GST
proteins encoding either full-length Grb2 or various domains of Grb2.
No detectable 110-kDa protein was coimmunoprecipitated with Tpr-Met, demonstrating that the protein complexes had been efficiently dissociated (Fig. 2B, lane 6). However, following
incubation of the dissociated, denatured proteins with a nondenatured
GST-Grb2 fusion protein, association of the 110-kDa protein with
full-length Grb2 was detected (Fig. 2B, lane 1)
demonstrating the the 110-kDa protein could associate directly with
Grb2. The 110-kDa protein also associated to the same extent with the
SH2-SH3(C) Grb2 fusion protein (Fig. 2B, lane 3)
and to a lesser extent with the SH3(N)-SH2, and SH3(N)-SH3(C) Grb2
fusion proteins (Fig. 2B, lanes 2 and
4). It did not associate with a fusion protein encoding only
the SH2 domain of Grb2 (Fig. 2B, lane 5).
Tpr-Met, on the other hand, associated with full-length Grb2 (Fig.
2B, lane 1) as well as the SH3(N)-SH2,
SH2-SH3(C), and SH2 fusion proteins (Fig. 2B, lanes
2, 3, and 5) but not with the SH3(N)-SH3(C)
Grb2 fusion protein. Neither Tpr-Met nor the 110-kDa protein associated
with GST (data not shown). These data demonstrate that the 110-kDa protein associated primarily with the carboxyl-terminal SH3 domain of
Grb2 and suggest that the Grb2 adaptor protein acts to couple the
110-kDa protein with Tpr-Met.
Fig. 3.
The Cbl protooncogene product is highly
phosphorylated in wild-type Tpr-Met transformed cells but at low levels
in cells expressing the N491H Tpr-Met mutant. Lysates from
parental Fr3T3 cells (lanes 1) or stable cell lines
expressing either wild-type (lanes 2 and 3) or
the various Tpr-Met mutant proteins (lanes 4-11) were
prepared and immunoprecipitated with an anti-Cbl serum, collected on
protein A-Sepharose, and washed three times with lysis buffer. The
proteins were resolved by 9% SDS-PAGE and immunoblotted with
recombinant RC20H anti-phosphotyrosine antibody (A) or
anti-Cbl serum (B). C, lysates from parental
Fr3T3 cells (lane 3) or stable cell lines expressing
wild-type Tpr-Met protein (lanes 2 and 4), the
Y482F (lane 5), N491H (lanes 6 and 7),
Y489F (lanes 8 and 9), or Y482F/Y489F (lane
10) Tpr-Met mutants were incubated with GST-Grb2-Sepharose
(lanes 3-10) or GST-Sepharose (lane 2).
Complexes were washed, resolved by SDS-PAGE, and immunoblotted with
anti-Cbl serum. An anti-Cbl immunoprecipitation was also included
(lane 1). D, top, lysates from
parental Fr3T3 cells were incubated with GST proteins encoding
full-length Grb2 (lane 1), and the SH3(N)-SH2 (lane
2), SH2-SH3(C) (lane 3), SH3(N)-SH3(C) (lane
4), or SH2 (lane 5) domains of Grb2 coupled to
glutathione-Sepharose. Complexes were washed, resolved by SDS-PAGE, and
immunoblotted with anti-Cbl serum. Bottom, the amount of GST
proteins used in the association assay were visualized by anti-GST
immunoblotting.
[View Larger Version of this Image (23K GIF file)]
Fig. 4.
The highly phosphorylated 110-kDa protein
does not correspond to Cbl. A, lysates were prepared from
293 cells either transiently transfected with expression plasmids
encoding wild-type Tpr-Met (lane 1, 1 µg; lane
2, 5 µg) or K241A, kinase dead mutant Tpr-Met (lane
3, 5 µg) or not transfected and mock stimulated (lane
4) or stimulated with EGF (lane 5). Proteins were
immunoprecipitated with anti-Cbl antibody, resolved by SDS-PAGE, and
immunoblotted with recombinant RC20H anti-phosphotyrosine antibody.
B, the supernatants from part A, lanes
1-3, were immunoprecipitated with anti-Tpr-Met antibody, resolved
by SDS-PAGE, and immunoblotted with anti-Tpr-Met antibody.
C, Fr3T3 cells were solubilized and depleted of Cbl by six
rounds of immunoprecipitations with antisera that recognizes the Cbl
protooncogene. D, lysis buffer (lane 1), a
nondepleted lysate (lane 2), or the Cbl-depleted lysate
(lane 3) were then added to wild-type Tpr-Met from
transiently transfected COS-1 cells that had been immunoprecipitated
and activated by phosphorylation with nonradioactive ATP. The protein
complexes were incubated in kinase buffer with
[-32P]ATP, resolved by 8% SDS-PAGE and visualized by
autoradiography. The positions of Tpr-Met, the 110-kDa protein, and
molecular weight markers are indicated.
[View Larger Version of this Image (33K GIF file)]
*
This research was supported in part by operating grants from
the National Cancer Institute of Canada and the Medical Research Council of Canada (to M. P.), American Cancer Society Grant CA69495, and National Institutes of Health Grant NS 34514 (to A. J. W.).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.
b
Recipient of a Fonds de la Recherche en Santé du
Québec Fellowship.
d
Recipient of a Royal Victoria Hospital Research
Institute Fellowship.
e
Recipient of a Steve Fonyo Research Studentship.
g
Recipient of a Royal Victoria Hospital Research
Institute Studentship.
j
Senior Scholar of National Cancer Institute of Canada. To
whom reprint requests should be addressed: Depts. of Medicine,
Oncology, and Biochemistry, McGill University, 687 Pine Ave., West,
Montreal PQ, Canada H3A 1A1. Tel.: 514-842-1231 (ext. 5834); Fax:
514-843-1478.
1
The abbreviations used are: HGF/SF, hepatocyte
growth factor/scatter factor; RTK, receptor tyrosine kinase; SH2, Src
homology domain 2; PAGE, polyacrylamide gel electrophoresis; GST,
glutathione S-transferase; EGF, epidermal growth
factor.
©1997 by The American Society for Biochemistry and Molecular Biology, Inc.