Molecular Endocrinology 14 (11): 1750-1775
Copyright © 2000 by The Endocrine Society
Constitutive and Growth Factor-Regulated Phosphorylation of Caveolin-1 Occurs at the Same Site (Tyr-14) in Vivo: Identification of a c-Src/Cav-1/Grb7 Signaling Cassette
Hyangkyu Lee,
Daniela Volonte,
Ferruccio Galbiati,
Puneeth Iyengar,
Douglas M. Lublin,
David B. Bregman,
Mark T. Wilson,
Roberto Campos-Gonzalez Boumediene Bouzahzah,
Richard G. Pestell,
Philipp E. Scherer and
Michael P. Lisanti
Department of Molecular Pharmacology and The Albert Einstein Cancer
Center (H.L., D.V., F.G., M.P.L.) Department of Cell Biology and
The Albert Einstein Cancer Center (P.I., P.E.S.) Department of
Pathology and The Albert Einstein Cancer Center (D.B.B.)
Departments of Developmental & Molecular Biology (DMB) and Medicine;
and the Albert Einstein Cancer Center (B.B., R.G.P.) Albert
Einstein College of Medicine Bronx, New York 10461
Department of Pathology (D.M.L.) Washington University School
of Medicine St. Louis, Missouri 63110
BD Transduction
Laboratories (M.T.W., R.C.-G.) Lexington, Kentucky 40511
 |
ABSTRACT
|
|---|
Caveolin-1 was first identified as a
phospho-protein in Rous sarcoma virus (RSV)-transformed chicken
embryo fibroblasts. Tyrosine 14 is now thought to be the principal site
for recognition by c-Src kinase; however, little is known about this
phosphorylation event. Here, we generated a monoclonal antibody (mAb)
probe that recognizes only tyrosine 14-phosphorylated caveolin-1. Using
this approach, we show that caveolin-1 (Y14) is a specific tyrosine
kinase substrate that is constitutively phosphorylated in Src- and
Abl-transformed cells and transiently phosphorylated in a regulated
fashion during growth factor signaling. We also provide evidence that
tyrosine-phosphorylated caveolin-1 is localized at the major sites of
tyrosine-kinase signaling, i.e. focal adhesions. By analogy
with other signaling events, we hypothesized that caveolin-1 could
serve as a docking site for pTyr-binding molecules. In support of this
hypothesis, we show that phosphorylation of caveolin-1 on tyrosine 14
confers binding to Grb7 (an SH2-domain containing protein) both
in vitro and in vivo. Furthermore, we
demonstrate that binding of Grb7 to tyrosine 14-phosphorylated
caveolin-1 functionally augments anchorage-independent growth and
epidermal growth factor (EGF)-stimulated cell migration. We discuss the
possible implications of our findings in the context of signal
transduction.
 |
INTRODUCTION
|
|---|
Caveolae are small omega-shaped indentations of the plasma
membrane that have been implicated in signal transduction and vesicular
transport processes (1, 2, 3, 4). Caveolae are found in most cell types but
are extremely abundant in terminally differentiated cell types:
adipocytes (5, 6, 7), simple squamous epithelia (type I pneumocytes and
endothelial cells) (8), smooth muscle cells (9), and fibroblasts
(10).
Caveolin, a 21- to 24-kDa integral membrane protein, is a principal
structural component of caveolae membranes in vivo (11, 12, 13, 14, 15).
Recent studies have shown that caveolin is only the first member of a
growing gene family of caveolin proteins; caveolin has been retermed
caveolin-1. Three different caveolin genes (Cav-1, Cav-2, and Cav-3)
encoding four different subtypes of caveolin have been described thus
far (2, 3, 4, 16). There are two subtypes of caveolin-1 (Cav-1
and
Cav-1ß) that differ in their respective translation initiation sites
(17). The tissue distribution of caveolin-2 mRNA and protein is
extremely similar to caveolin-1 (7, 18). In striking contrast,
caveolin-3 mRNA and protein are expressed mainly in muscle tissue
types (skeletal, cardiac, and smooth) (16, 19, 20).
Several independent lines of evidence suggest that caveolins may act as
scaffolding proteins within caveolae membranes. In support of this
assertion, 1) both the N-terminal and C-terminal domains of caveolin-1
face the cytoplasm (17, 21, 22, 23, 24); 2) caveolins exist within caveolae
membranes as high molecular mass oligomers (18, 20, 23, 25, 26); 3)
these caveolin oligomers have the capacity to self-associate in
vitro to form larger structures that resemble caveolae (23); 4)
caveolins copurify with cytoplasmic lipid-modified signaling molecules
including heterotrimeric G proteins, Src family tyrosine kinases, and
Ras-related GTPases (27, 28, 29, 30, 31); and 5) recombinant caveolin-1 interacts
directly with heterotrimeric G proteins (32), c-Src (33) and H-Ras
(30). Thus, we and others have proposed the "caveolae signaling
hypothesis," which states that caveolins may serve as oligomeric
docking sites for organizing and concentrating signaling molecules
within caveolae membranes (1, 2, 3, 4, 23).
It has been suggested that fatty acylation may represent a common
mechanism for targeting cytoplasmic signaling molecules to caveolae (1, 29, 34). In direct support of this hypothesis, many proteins (G protein
-subunits, Src-family tyrosine kinases, Ras-related GTPases, and
endothelial nitric oxide synthase) that copurify with caveolins
undergo myristoylation, palmitoylation, prenylation, or dual acylation
(1, 28, 35). These results indirectly suggest that acylation of
lipid-modified signaling molecules may be required for their caveolar
targeting in vivo.
Caveolins may facilitate this lipid-based targeting as: 1) caveolin-1
requires cholesterol for insertion into model lipid membranes (36, 37);
2) the membrane-spanning domain of caveolin-1 has a specific and
high-affinity for fatty acyl moieties (38, 39); and 3) the C-terminal
domain of caveolin-1 undergoes palmitoylation on three cysteines
(residues 133, 143, and 156) (22). However, palmitoylation of
caveolin-1 is not required for its membrane attachment or targeting to
caveolae (22), suggesting that this modification may serve another
roleperhaps in trapping other acylated proteins in the vicinity of
caveolin molecules.
Historically, caveolin-1 was first identified as a major tyrosine
phosphorylated protein in v-Src-transformed chicken embryo fibroblasts
(40). Based on this observation, Glenney and co-workers (40)
have proposed that caveolin-1 may represent a critical target during
cellular transformation. The functional consequences of the
phosphorylation of caveolin-1 on tyrosine are not yet known. At steady
state, caveolin-1 is not phosphorylated on tyrosine (6). This is in
contrast to v-Src-transformed cells where caveolin-1 is constitutively
phosphorylated on tyrosine (11). However, caveolin tyrosine
phosphorylation also occurs in normal cells but in a tightly regulated
fashion (41). If 3T3-L1 adipocytes are serum starved and rapidly
stimulated with insulin, caveolin transiently undergoes phosphorylation
on tyrosine (41). It has been suggested that insulin-stimulated
tyrosine phosphorylation of caveolin occurs indirectly via an
endogenous Src-family tyrosine kinase, rather than directly via the
insulin receptor. In support of this idea, caveolin-1 can be purified
as part of a hetero-oligomeric complex that contains c-Src and other
Src-family tyrosine kinases (27, 28, 30, 33).
Recently, we began to study the phosphorylation of caveolin-1 by Src
family tyrosine kinases both in vitro and in vivo
(42). Using purified recombinant components, we first reconstituted the
phosphorylation of caveolin-1 by Src kinase in vitro.
Microsequencing of Src-phosphorylated caveolin-1 revealed that
phosphorylation occurs within the extreme N-terminal region of
full-length caveolin-1, between residues 626. This region contains
three tyrosine residues at positions 6, 14, and 25. Deletion
mutagenesis demonstrated that caveolin-1 residues 121 are sufficient
to support this phosphorylation event, implicating tyrosine 6 and/or
14. In vitro phosphorylation of caveolin-1-derived synthetic
peptides and site-directed mutagenesis directly showed that tyrosine 14
is the principal substrate for Src kinase (42). In support of these
observations, tyrosine 14 is the only tyrosine residue within
caveolin-1 that bears any resemblance to the known recognition motifs
for tyrosine kinases (KYVDSEGHLpY;
[RK] - x(2, 3, 4) - [DE] - x(2, 3) - pY). To confirm or refute the
in vivo relevance of these in vitro studies, we
analyzed the tyrosine phosphorylation of endogenous caveolin-1 in
v-Src-transformed NIH 3T3 cells. In vivo, two isoforms of
caveolin-1 are known to exist: Cav-1
contains residues 1178 and
Cav-1ß contains residues 32178. Only Cav-1
underwent tyrosine
phosphorylation in v-Src-transformed NIH 3T3 cells, although Cav-1ß
is well expressed in these cells. As Cav-1ß lacks residues 131 (and
therefore tyrosine 14), these in vivo studies indirectly
demonstrated the validity of our in vitro studies (42).
Here, we directly examine the phosphorylation of caveolin-1 on tyrosine
14 in vivo. For this purpose, we generated and extensively
characterized a novel phospho-specific monoclonal antibody (mAb) probe
that selectively recognizes only tyrosine 14-phosphorylated caveolin-1.
Our results provide in vivo support for the hypothesis that
certain tyrosine-kinase-mediated transmembrane signaling events are
initiated at caveolae membranes and that caveolin-1 may function as a
signaling molecule.
 |
RESULTS
|
|---|
Characterization of a mAb Probe Specific for Tyrosine
14-Phosphorylated Caveolin-1
We previously showed that tyrosine 14 is the principal
site of caveolin-1 phosphorylation by c-Src in vitro (42).
To detect this phosphorylation event in vivo, we have
generated a novel mAb probe that recognizes only caveolin-1
phosphorylated on tyrosine 14 (see Materials and Methods). A
tyrosine-phosphorylated caveolin-1 peptide [SEGHL(pY)TVPI, residues
918] was used to immunize mice. Three positive clones were obtained
(cl 34A, 49, and 56). Clone 56 gave the strongest immunoreactivity by
Western blot analysis and was selected for detailed
characterization.
Figure 1
shows the
selectivity of antiphosphocaveolin-1 IgG. A
glutathione-S-transferase (GST)-fusion protein carrying the
N-terminal domain of caveolin-1 (residues 1101) was purified from
normal bacteria (BL21) or from a bacterial strain that harbors a
tyrosine kinase (TKB1). Note that clone 56 recognized only
tyrosine-phosphorylated caveolin-1 produced in the TKB1 strain, despite
equal protein loading (Fig. 1A
). We also reconstituted this tyrosine
phosphorylation event in vivo. Cos-7 cells were transiently
transfected with caveolin-1 and c-Src, alone or in combination. Figure 1B
shows that antiphosphocaveolin-1 IgG only recognizes caveolin-1 when
it is coexpressed with c-Src. Tyrosine-phosphorylated caveolin-1
migrated as multiple bands, with a major band at approximately 2428
kDa. Normally, nonphosphorylated caveolin-1 migrates at approximately
2224 kDa. In addition, we find that when tyrosine 14 is mutated to
alanine (Y14A), preventing phosphorylation at this site, this
immunoreactivity is completely abolished (Fig. 1B
). We further tested
the specificity of antiphosphocaveolin-1 IgG by peptide competition
with caveolin-1-derived peptides (see Table 1
and Fig. 1C
). Immunoreactivity was
abolished by preincubation with a 100-fold molar excess of the
antigenic peptide (PY14). Importantly, no inhibitory effect was
observed with the nonphosphorylated version of the same peptide (Y14)
or two irrelevant phosphopeptides (PY100 and PY148). In addition, when
caveolins-1, -2, and -3 were cotransfected with activated c-Src
(Y529F), only caveolin-1 was detectable with antiphosphocaveolin-1 IgG
(data not shown). These results demonstrate the high selectivity of
this novel mAb probe for tyrosine 14-phosphorylated caveolin-1.

View larger version (16K):
[in this window]
[in a new window]
|
Figure 1. Characterization of a Mouse mAb Probe (cl 56) That
Only Recognizes Tyrosine 14-Phosphorylated Caveolin-1
A tyrosine-phosphorylated caveolin-1 peptide (SEGHL(pY)TVPI, residues
918) was used to immunize mice and generate a
phosphocaveolin-1-specific mAb probe. A, GST-caveolin-1 fusion
proteins. A GST-fusion protein carrying the N-terminal domain of
caveolin-1 (residues 1101) was purified from normal bacteria (BL21)
or from a bacterial strain harboring a tyrosine kinase (TKB1). Note
that clone 56 recognized only tyrosine- phosphorylated caveolin-1
produced in the TKB1 strain, despite equal protein loading. B,
Caveolin-1 Y14A mutant. Cos-7 cells were transiently transfected with
caveolin-1 and c-Src, alone or in combination. Note that
antiphosphocaveolin-1 IgG only recognizes caveolin-1 when it is
coexpressed with c-Src. Tyrosine-phosphorylated caveolin-1 migrated as
multiple bands, with a major band at approximately 2425 kDa.
Importantly, when tyrosine 14 is mutated to alanine (Y14A), this
immunoreactivity is completely abolished. Immunoblotting with mAb 2297
that recognizes total caveolin-1 is shown as a control for equal
loading. The mobility of caveolin-1 (Y14A) may be shifted due to a
conformational change. C, Peptide competition.Cos-7 cells were
transiently transfected with caveolin-1 and c-Src in combination and
subjected to preparative SDS-PAGE. After transfer, the nitrocellulose
sheet was cut into strips and incubated with antiphosphocaveolin-1 IgG
alone or in combination with peptides. The sequences of these four
caveolin-1-based peptides (PY14, Y14, PY100, and PY148) are detailed in
Table 1 .
|
|
Src-Induced Phosphorylation of Caveolin-1: A Requirement for Lipid
Modification of c-Src and Caveolin-1
To examine whether other tyrosine residues are phosphorylated to a
significant extent, we cotransfected c-Src with caveolin-1 (WT or
Y14A). Total caveolin-1 was retrieved using IgG directed against the
caveolin-1 C terminus. These immunoprecipitates were then
probed with a well characterized mAb that recognizes generic
phosphotyrosine (mAb PY20). Note that when tyrosine 14 is mutated to
alanine (Y14A), reactivity of caveolin-1 with PY20 is completely
abolished (Fig. 2A
). These results
suggest that tyrosine 14 is the primary site of phosphorylation or that
phosphorylation at tyrosine 14 is required before other sites can be
phosphorylated.

View larger version (24K):
[in this window]
[in a new window]
|
Figure 2. Wild-Type and Mutationally Activated Src
Phosphorylate Caveolin-1 on Tyrosine 14
A, IP/Western of caveolin-1 (WT and Y14A). Cos-7 cells were transiently
cotransfected with caveolin-1 (WT or Y14A), alone or in combination
with c-Src. Thirty-six hours post transfection, the cells were
processed for immunoprecipitation. Total caveolin-1 was retrieved using
IgG directed against the caveolin-1 C terminus (pAb H-97).
Immunoprecipitates were then probed with a mAb that recognizes
phosphotyrosine (mAb PY20). Note that when tyrosine 14 is mutated to
alanine (Y14A), reactivity of caveolin-1 with PY20 is completely
abolished. LC denotes the IgG light chain. B, Activated and kinase-dead
c-Src mutants. Cos-7 cells were transiently transfected with caveolin-1
and c-Src [WT, activated (Y529F), and kinase-dead (K297R)]. Tyrosine
14- phosphorylated caveolin-1 was detected by Western blot analysis
with antiphosphocaveolin-1 IgG. Activated c-Src phosphorylated
caveolin-1 to a much greater extent than wild-type, and kinase-dead
c-Src showed no activity. Immunoblotting with mAb 2297 that recognizes
total caveolin-1 is shown as a control for equal loading. C, v-Src
cells/WB. Lysates were prepared from normal and v-Src-transformed NIH
3T3 cells and subjected to immunoblot analysis. Note that
antiphosphocaveolin-1 IgG recognizes tyrosine 14-phosphorylated
caveolin-1 in NIH 3T3 cells that stably overexpress the v-Src. In
contrast, little or no immunoreactivity was observed in normal NIH 3T3
cells. A comparison between normal and v-Abl-transformed NIH 3T3 cells
is also shown. Each lane contains equal amounts of total protein.
Immunoblotting with mAb 2297 that recognizes total caveolin-1 is also
shown. D, v-Src cells/IP. v-Src-transformed NIH 3T3 cells were lysed
and immunoprecipitated with an excess of mAb 2234 that recognizes total
caveolin-1 (residues 121) or an excess of mAb 56 that recognizes
phosphocaveolin-1 (residues 918). After SDS-PAGE and transfer to
nitrocellulose, the amount of caveolin-1 precipitated was estimated by
Western blotting with rabbit anticaveolin-1 IgG (N-20). Note that that
only a small fraction ( 5%) of caveolin-1 undergoes phosphorylation
on tyrosine 14.
|
|
We next compared the activity of three different forms of
human c-Src [WT, activated (Y529F), and kinase-dead (K297R)]. As
predicted, activated c-Src phosphorylated caveolin-1 to a much greater
extent than wild type, and kinase dead c-Src showed no activity in this
in vivo assay system (Fig. 2B
). Figure 2C
(upper
left) shows that antiphosphocaveolin-1 IgG also recognizes
tyrosine 14-phosphorylated caveolin-1 in NIH 3T3 cells that stably
overexpress v-Src; little or no immunoreactivity was observed in normal
NIH 3T3 cells. These findings are consistent with the previous
observation that the overall tyrosine phosphorylation of caveolin-1 is
dramatically elevated in Rous sarcoma virus (RSV)-transformed chicken
embryo fibroblasts (11, 40); however, this observation predated the
identification of the major site of phosphorylation as tyrosine 14
(42). As we previously noted, total caveolin-1 protein levels are
decreased approximately 3- to 4-fold in v-Src-transformed NIH 3T3 cells
(Fig. 2C
, lower) due to a reduction in the level of
caveolin-1 mRNA (43). Similarly, antiphosphocaveolin-1 IgG recognizes
tyrosine 14-phosphorylated caveolin-1 in v-Abl-transformed NIH 3T3
cells (Fig. 2C
). This is despite the fact that caveolin-1 mRNA and
protein levels are dramatically down-regulated to almost undetectable
levels in these cells, as we have shown previously (43). In support of
this observation, we have previously pointed out that tyrosine 14
within caveolin-1 matches the consensus for phosphorylation by v-Abl
and v-Src tyrosine kinases (42).
To roughly estimate the percentage of total cellular
caveolin-1 that undergoes phosphorylation on tyrosine 14, we employed
an immunoprecipitation approach. v-Src- transformed NIH 3T3 cells were
lysed and subjected to immunoprecipitation with mAb 2234, which
recognizes total caveolin-1 (residues 121), or mAb 56, which
recognizes phosphocaveolin-1 (residues 918). The amount of caveolin-1
precipitated was estimated by Western blotting with rabbit
anticaveolin-1 IgG. Using this approach, Fig. 2D
shows that only a
fraction (
5%) of total caveolin-1 undergoes phosphorylation on
tyrosine 14.
Membrane attachment of c-Src is mediated in part by N-terminal
myristoylation of the protein product (44). Interestingly, this
modification is required for the targeting of c-Src to caveolae
membrane domains (45). Caveolin-1 also undergoes lipid modification
(22). However, the functional role of caveolin-1 palmitoylation remains
largely unknown. Thus, we next examined the possible requirement of
these lipid modifications for the tyrosine phosphorylation of
caveolin-1.
To evaluate the role of c-Src myristoylation, Cos-7 cells were
cotransfected with caveolin-1 and c-Src [wild-type (WT) or
myristoylation-deficient (Myr-minus)]. Figure 3A
shows that myristoylation of c-Src is
required for phosphorylation of caveolin-1 on tyrosine 14. These
results are consistent with the observation that myristoylation of
c-Src is required for its targeting to caveolae (45).

View larger version (9K):
[in this window]
[in a new window]
|
Figure 3. Lipid Modification Is Required for Src-Induced
Phosphorylation of Caveolin-1 on Tyrosine 14
A, Myristoylation minus c-Src. Cos-7 cells were cotransfected with
caveolin-1 and c-Src (WT or Myr-minus). Note that myristoylation of
c-Src is required for phosphorylation of caveolin-1 on tyrosine 14. B,
Palmitoylation-deficient caveolin-1. To evaluate the role of caveolin-1
palmitoylation, we cotransfected Cos-7 cells with wild-type c-Src and
caveolin-1 (WT or Pal-minus). Note that palmitoylation of caveolin-1 is
required for or greatly facilitates phosphorylation of caveolin-1 on
tyrosine 14. In panels A and B, immunoblotting with mAb 2297 that
recognizes total caveolin-1 is shown as a control for equal loading.
|
|
To evaluate the role of caveolin-1 palmitoylation, we cotransfected
wild-type c-Src with caveolin-1 [WT or palmitoylation-deficient
(Pal-minus)]. In palmitoylation-deficient caveolin-1, the three
acceptor sites for palmitoylation have been mutated from cysteine to
serine (C133, 143, 156S) (22). Mutation of these residues, thereby
creating a palmitoylation-deficient form of caveolin-1, does not
prevent proper oligomerization of caveolin-1 or its targeting to
caveolae membranes (22, 46). Interestingly, Fig. 3B
shows that
palmitoylation of caveolin-1 is required for or greatly facilitates
phosphorylation of caveolin-1 on tyrosine 14. This is consistent with
the recent finding that palmitoylation of caveolin-1 is required for
its stable association with another myristoylated protein,
Gi1 (a G protein
-subunit), as measured
via coimmunoprecipitation (47). Thus, lipid modification of both c-Src
and caveolin-1 is required for this reciprocal interaction, as measured
using tyrosine phosphorylation of caveolin-1. These observations
provide a novel function for caveolin-1 lipid modification,
i.e. for association with c-Src kinase. In addition, one
advantage of using tyrosine phosphorylation of caveolin-1 as a method
to detect an interaction with c-Src is that this reaction occurs
in vivo and is, therefore, less subject to possible
postlysis artifacts.
Colocalization of Tyrosine 14-Phosphorylated Caveolin-1 with
Markers of Focal Adhesions in Cells Expressing Activated Src
To examine the localization of tyrosine-phosphorylated caveolin-1,
we cotransfected Cos-7 cells with caveolin-1 alone and in combination
with either wild-type c-Src or activated c-Src (Y529F). These cells
were then doubly immunostained with antiphosphocaveolin-1 mouse mAb (cl
56) and an anticaveolin-1 rabbit polyclonal antibody (pAb)
(N-20). These bound primary antibodies were visualized by using
distinctly tagged fluorescent secondary antibodies (see Materials
And Methods).
Figure 4
shows the distribution of
caveolin-1 phosphorylated on tyrosine 14. In cells transfected with
caveolin-1 alone, little or no immunostaining with
antiphosphocaveolin-1 was observed (Fig. 4A
). In cells cotransfected
with caveolin-1 plus wild-type c-Src, immunostaining with
antiphosphocaveolin-1 appeared as large fluorescent dots in the
center and along the cell periphery (Fig. 4B
). Immunostaining with
anticaveolin-1 also appeared punctate, but the dots were of a much
smaller size. These two labeling patterns appeared distinct, suggesting
that tyrosine-phosphorylated caveolin-1 is present in a different
region of the cell or that only a subpopulation of caveolae are
tyrosine phosphorylated. In cells cotransfected with caveolin-1 plus
activated c-Src (Y529F), immunostaining with antiphosphocaveolin-1 also
appeared as large dots, but these dots were confined to the cell
periphery and appeared to coincide with focal contacts or adhesions
(Fig. 4C
). A very similar staining pattern was observed in
v-Src-transformed NIH 3T3 cells (Fig. 5A
).

View larger version (69K):
[in this window]
[in a new window]
|
Figure 4. Localization of Tyrosine 14-Phosphorylated
Caveolin-1 in Cos-7 Cells Transiently Transfected with c-Src
To examine the localization of tyrosine-phosphorylated
caveolin-1, we cotransfected Cos-7 cells with caveolin-1 and either WT
or activated c-Src. Cells were doubly immunostained with
antiphosphocaveolin-1 mouse mAb (cl 56; at left) and an
anticaveolin-1 rabbit pAb (N-20; at right). Bound
primary antibodies were visualized by using distinctly tagged
fluorescent secondary antibodies. A, Caveolin-1 alone.In Cos-7 cells
transfected with caveolin-1 alone, little or no immunostaining with
antiphosphocaveolin-1 was observed. B, Caveolin-1 plus wild-type c-Src.
In cells cotransfected with caveolin-1 plus WT c-Src, immunostaining
with antiphosphocaveolin-1 appeared as large fluorescent
dots in the central region of the cell and along the cell
periphery. C, Caveolin-1 plus activated c-Src. In cells cotransfected
with caveolin-1 plus activated c-Src (Y529F), immunostaining with
antiphosphocaveolin-1 appeared as large dots confined to
the cell periphery that were reminiscent of focal contacts or
adhesions. In panels A-C, immunostaining with anticaveolin-1 IgG
appeared punctate (right) and did not coincide with the
labeling observed with antiphosphocaveolin-1.
|
|

View larger version (29K):
[in this window]
[in a new window]
|
Figure 5. Tyrosine 14-Phosphorylated Caveolin-1 Is Localized
in Close Proximity to Focal Adhesions in NIH 3T3 Cells Stably
Expressing v-Src
Cells were doubly immunostained with two distinct primary
antibodies as detailed below. Bound primary antibodies were visualized
by using distinctly tagged fluorescent secondary antibodies. In panels
A, B, and D, images were acquired with a MR600 confocal fluorescence
microscope (Bio-Rad Laboratories, Inc.). In panels C and
E, images were acquired with a IX80 microscope (Olympus Corp.) with a 60x Plan Neofluar objective and a Photometrics
cooled CCD camera with a 35-mm shutter. A, Antiphosphocaveolin-1 (mAb)
and anticaveolin-1 (pAb). Note that staining with antiphosphocaveolin-1
appeared as large dots, but these dots were confined to
the cell periphery and appeared to coincide with focal contacts or
adhesions. B, Antiphosphocaveolin-1 (mAb) and antiphospho tyrosine (pAb). Note that
tyrosine-phosphorylated caveolin-1 is colocalized with the major sites
of tyrosine phosphorylation. These sites of tyrosine phosphorylation
activity are known to correspond to focal adhesions (48 ). C,
Anti-phospho-caveolin-1 (mAb) and antipaxillin. Note that caveolin-1
and paxillin colocalize to a significant extent. These results suggest
that caveolae in close proximity to focal adhesions are preferentially
phosphorylated in v-Src-transformed cells. D, Anticaveolin-1 (pAb) and
antiphosphotyrosine (PY20 mAb). Note that little or no colocalization
is observed. These results indicate that the bulk of total caveolin-1
is not localized in proximity to the major sites of tyrosine
phosphorylation in vivo. E, Color overlay. A
color version of panel C is shown along with the merged
image to better demonstrate the colocalization of phosphocaveolin-1 and
paxillin.
|
|
To investigate the possibility that tyrosine-phosphorylated caveolin-1
was localized in close proximity with focal adhesions, we next
performed a series of double labeling experiments with two widely used
markers for focal adhesions: 1) antiphosphotyrosine IgG and 2)
antipaxillin IgG (48). v-Src-transformed NIH 3T3 cells were doubly
immunostained with antiphosphocaveolin-1 (mAb 56) and either
antiphosphotyrosine IgG (pAb) or antipaxillin IgG [lissamine rhodamine
B sulfonyl chloride (LRSC)-conjugated mAb]. Double labeling with
antiphosphocaveolin-1 and antiphosphotyrosine revealed that
tyrosine-phosphorylated caveolin-1 is colocalized with the major sites
of tyrosine phosphorylation in v-Src- transformed NIH 3T3 cells (Fig. 5B
). These sites of tyrosine phosphorylation activity are known to
correspond to focal adhesions (48). Similar results were obtained when
these cells were double labeled with antiphosphocaveolin-1 and
antipaxillin; tyrosine- phosphorylated caveolin-1 and paxillin were
found to colocalize to a significant extent (Fig. 5C
). A color
overlay is also shown to better illustrate this point (Fig. 5E
).
These results 1) suggest that caveolae in close proximity to focal
adhesions are preferentially phosphorylated in v-Src-transformed cells,
and 2) provide visual evidence that tyrosine-phosphorylated caveolin-1
is localized at the major sites of tyrosine kinase signaling in
vivo.
In contrast, when v-Src-transformed NIH 3T3 cells were double-labeled
with anticaveolin-1 (N-20 pAb) and antiphosphotyrosine (PY20 mAb),
little or no colocalization was observed (Fig. 5D
). These results
indicate that in v-Src-transformed cells the bulk of total caveolin-1
is not localized in proximity to the major sites of tyrosine
phosphorylation in vivo. This is consistent with the
observation that only a fraction of total cellular caveolin-1 (
5%)
is tyrosine phosphorylated in v-Src-transformed cells (Fig. 2D
).
Tyrosine-Phosphorylated Caveolin-1 Correctly Forms High Molecular
Mass Oligomers and Is Targeted to Caveolae-Enriched Membrane
Domains
As tyrosine-phosphorylated caveolin-1 was localized in close
proximity to focal adhesions, we next assessed the biochemical
properties of tyrosine-phosphorylated caveolin-1 using established
assay systems. These approaches have been used previously to
characterize the properties of total caveolin-1. Caveolin-1 normally
forms stable high molecular mass homooligomeric complexes (7, 18). Once
these homooligomers reach the plasma membrane, they become Triton
insoluble due to their incorporation into caveolae membranes (49). This
resistance to detergent solubilization is thought to reflect the local
lipid microenvironment in which these caveolin homooligomers are
embedded. In contrast, caveolin-1 associated with the Golgi complex
remains Triton soluble (50). Thus, we next examined the effects of the
tyrosine phosphorylation of caveolin-1 on its 1) Triton insolubility;
2) oligomeric state; and 3) caveolar targeting, using normal and v-Src-
transformed NIH 3T3 cells.
In both normal and v-Src-transformed NIH 3T3 cells, total caveolin-1
was >95% Triton insoluble (not shown), formed high molecular mass
oligomers of the correct size, and was targeted to caveolae-enriched
membrane fractions (Fig. 6
, A and B).
Interestingly, virtually identical results were obtained with
caveolin-1 phosphorylated on tyrosine 14, indicating that caveolin-1
remains caveolae associated even after tyrosine phosphorylation in
v-Src-transformed cells. As the subcellular distribution of tyrosine-
phosphorylated caveolin-1 coincides with focal adhesions by
fluorescence microscopy (Fig. 6
), these biochemical results are
consistent with the idea that caveolae in close proximity to focal
adhesions are preferentially targeted for phosphorylation in v-Src-
transformed cells.

View larger version (15K):
[in this window]
[in a new window]
|
Figure 6. Tyrosine 14-Phosphorylated Caveolin-1 Correctly
Forms High Molecular Mass Oligomers and Is Targeted to
Caveolae-Enriched Membrane Domains
In both panels, tyrosine phosphorylated caveolin-1 was detected with
mAb 56 and total caveolin-1 was visualized using mAb 2297. A, Velocity
gradient analysis. Normal and v-Src-transformed NIH 3T3 cells were
solubilized, loaded atop a 540% sucrose density gradient, and
subjected to centrifugation for 10 h. Twelve fractions were
recovered and a 20 µl aliquot from each fraction was analyzed by
SDS-PAGE/Western blotting. Arrows mark the positions of
molecular mass standards. Note that caveolin-1 behaves as a high
molecular mass oligomer of approximately 150300 kDa. B, Caveolar
targeting. Normal and v-Src-transformed NIH 3T3 cells were homogenized
in a buffer containing 1% Triton X-100 and subjected to sucrose
density gradient centrifugation. Twelve 1-ml fractions were collected,
and an aliquot of each fraction ( 20 µl) was analyzed by
SDS-PAGE/Western blotting. As expected, caveolin-1 is highly enriched
in fractions 45, which represent the caveolae-enriched membrane
fractions. Note that both total caveolin-1 and tyrosine 14-
phosphorylated caveolin-1 form high molecular mass oligomers of the
correct size and are targeted to caveolae-enriched membrane fractions
in normal and v-Src-transformed cells.
|
|
Insulin-Stimulated Phosphorylation of Caveolin-1 on Tyrosine 14 in
3T3-L1 Adipocytes
Caveolin-1 is abundantly expressed in mature adipocytes, and its
mRNA and protein expression are induced approximately 20- to 25-fold
during differentiation of 3T3-L1 fibroblasts to adipocytes (6). In
parallel with an induction of caveolin-1 expression, a 10-fold increase
in the number of plasmalemmal caveolae is also observed (5). These
results suggest that caveolin-1 protein and caveolae organelles play an
important role in adipocyte signal transduction. In accordance with
this idea, the insulin receptor has been localized to adipocyte
caveolae (51, 52, 53, 54), and caveolin undergoes tyrosine phosphorylation in
response to insulin stimulation, but not other growth factors, such as
EGF and platelet-derived growth factor (PDGF) (41, 55).
This insulin-stimulated tyrosine phosphorylation of caveolin is thought
to occur via an endogenous member of the Src family of tyrosine kinases
(41). However, it remains unknown which tyrosine residue is
phosphorylated in response to insulin stimulation. Also, as caveolins-1
and -2 form a tight heterooligomeric complex and stably
coimmunoprecipitate (18, 26, 56, 57), it remains unclear whether
caveolin-1 or caveolin-2 is actually the target of tyrosine
phosphorylation in adipocytes.
To address these issues, we treated 3T3-L1 adipocytes with
insulin or a variety of other growth factors. Figure 7
shows that a brief treatment (10 min)
of 3T3-L1 adipocytes with insulin dramatically stimulated
phosphorylation of caveolin-1 on tyrosine 14, while the other factors
had little or no effect (see panels A and C). Interestingly, 3T3-L1
adipocytes showed an increased basal level of tyrosine phosphorylation
of caveolin-1, as compared with Cos-7 cells or NIH 3T3 cells. Virtually
identical results were obtained with 3T3-L1 fibroblasts, indicating
that this event also occurs in the preadipocyte (Fig. 7A
). Tyrosine
phosphorylation of caveolin-1 was concentration dependent and increased
from 5150 nM insulin (not shown). Insulin-stimulated
phosphorylation of caveolin-1 on tyrosine 14 was apparent after only 1
min of treatment, reached maximal levels at 5 min, and declined to
basal levels by 120 min (Fig. 7B
). These results are consistent with
the idea that tyrosine phosphorylation of caveolin-1 is an early event
in insulin receptor signal transduction.

View larger version (24K):
[in this window]
[in a new window]
|
Figure 7. Insulin-Stimulated Phosphorylation of Caveolin-1 on
Tyrosine 14 in 3T3-L1 Adipocytes
In panels AC, cells were lysed and subjected to immunoblot analysis
with antiphosphocaveolin-1 IgG. Immunoblotting with mAb 2297 that
recognizes total caveolin-1 is shown as a control for equal loading. A,
Growth-factor stimulation. 3T3-L1 adipocytes (left) or
fibroblasts (right) were serum starved for 3 h and
subsequently treated with insulin (150 nM) or a variety of
other growth factors (PDGF (10 ng/ml), bFGF (5 ng/ml), TNF (10
nM), and IL6 (5 nM)). Note that a brief
treatment (10 min) of 3T3-L1 adipocytes with insulin dramatically
stimulated phosphorylation of caveolin-1 on tyrosine 14, while the
other factors had little or no effect (left). Virtually
identical results were obtained with 3T3-L1 preadipocytes
(right). B, Insulin time course. 3T3-L1 adipocytes were
serum starved for 3 h and subsequently treated with insulin. Note
that insulin (150 nM)-stimulated tyrosine phosphorylation
of caveolin-1 was apparent after only 1 min of treatment, reached
maximal levels at 5 min, and declined to basal levels by 120 min. C,
Insulin vs. EGF stimulation. 3T3-L1 adipocytes were
serum starved for 3 h and subsequently treated with EGF (100
ng/ml) or insulin (150 nM). Note that a brief treatment (5
min) of 3T3-L1 adipocytes with insulin stimulated phosphorylation of
caveolin-1 on tyrosine 14, while treatment with EGF had no effect.
|
|
Importantly, these studies directly demonstrate that caveolin-1
undergoes phosphorylation on tyrosine 14 in response to insulin
stimulation. Thus, these results provide the first evidence that both
constitutive (via activated Src) and regulated (via insulin)
phosphorylation of caveolin-1 converge on the same site,
i.e. tyrosine 14.
EGF-Stimulated Phosphorylation of Caveolin-1 on Tyrosine 14 in A431
Cells
A431 cells are a human epidermoid carcinoma-derived cell line
that has been used extensively to study EGF-mediated signal
transduction. As A431 cells are known to express both EGF receptor
(EGF-R) (58) and caveolin-1 (21, 59), we used these cells to evaluate
the effects of EGF stimulation on the tyrosine phosphorylation of
caveolin-1. For this purpose, A431 cells were serum starved and treated
with EGF (100 ng/ml) for various times. As a positive control for these
studies, we used an antibody that only detects the activated
phosphorylated form of the EGF-R (58). Figure 8A
shows that a brief treatment (5 min)
of A431 cells with EGF activated the EGF-R and dramatically stimulated
phosphorylation of caveolin-1 on tyrosine 14. This is in contrast to
our results with 3T3-L1 adipocytes (Fig. 7C
), which do not show
tyrosine phosphorylation of caveolin-1 in response to EGF, despite the
fact that these cells are known to express the EGF-R and are EGF
responsive. Thus, growth factor-stimulated tyrosine phosphorylation of
caveolin-1 may be dependent on cell-type specific coupling factors or
simply dependent on the relative expression levels of different growth
factor receptors in a given cell type.

View larger version (94K):
[in this window]
[in a new window]
|
Figure 8. EGF-Stimulated Phosphorylation of Caveolin-1 on
Tyrosine 14 in A431 Cells
A, EGF time course. A431 cells were serum-starved for 1216 h and
subsequently treated with EGF (100 ng/ml) for 04 h. Cells were lysed
and subjected to immunoblot analysis with antiphosphocaveolin-1 IgG. As
a positive control, we used an antibody that only detects the activated
phosphorylated form of the EGF-R (mAb 74). Note that a brief treatment
(5 min) of A431 cells with EGF dramatically activated EGF-R and
stimulated the phosphorylation of caveolin-1 on tyrosine 14.
Immunoblotting with phospho-independent antibodies to EGF-R (mAb 13)
and caveolin-1 (mAb 2297) was used to visualize total EGF-R and total
caveolin-1 levels. Each lane contains an equal amount of total protein.
B, Immunolocalization. Confluent A431 cells were serum starved and
incubated in the presence or absence of EGF (100 ng/ml) for 5 min. Note
that tyrosine-phosphorylated caveolin-1 accumulated at the plasma
membrane and was highest in areas of cell-cell contact. Importantly,
little or no immunostaining was observed using antiphosphocaveolin-1
IgG if A431 cells were not treated with EGF. C, Colocalization with
paxillin. Subconfluent A431 cells were serum-starved and subsequently
incubated in the presence of EGF (100 ng/ml) for 5 min. Double labeling
with antiphosphocaveolin-1 and antipaxillin revealed that caveolin-1
and paxillin colocalize to a significant extent.
Arrowheads point to areas of colocalization. These
results suggest that caveolae in close proximity to focal adhesions are
preferentially phosphorylated in EGF-stimulated A431 cells.
|
|
In confluent A431 cells, tyrosine 14-phosphorylated caveolin-1 was
localized at the cell surface and was only detectable after EGF
stimulation (Fig. 8B
). In subconfluent EGF-stimulated A431 cells, it
was apparent that tyrosine-phosphorylated caveolin-1 was localized in
close proximity to focal adhesions. Importantly, this suspicion was
confirmed by performing double labeling with antibodies directed
against paxillin (Fig. 8C
). Thus, in both v-Src-transformed NIH 3T3
cells and EGF-stimulated A431 cells, tyrosine 14- phosphorylated
caveolin-1 is localized near focal adhesions.
Vanadate Treatment Induces the Accumulation of Tyrosine
14-Phosphoryated Caveolin-1 in Normal NIH 3T3 Cells
It remains unknown what ligands might stimulate tyrosine
phosphorylation of caveolin-1 in other cell types, such as NIH 3T3
cells. For example, treatment of NIH 3T3 cells with various growth
factors [EGF, PDGF, and basic fibroblast growth factor (bFGF)] had no
apparent effect on the tyrosine phosphorylation of caveolin-1 (not
shown). To investigate whether caveolin-1 undergoes tyrosine
phosphorylation in normal NIH 3T3 cells, these cells were treated with
vanadate (100 µM), a tyrosine phosphatase inhibitor. If
caveolin-1 undergoes tyrosine phosphorylation, vanadate should prevent
dephosphorylation and allow detection of accumulated
tyrosine-phosphorylated intermediates. Figure 9A
shows that treatment with vanadate for
increasing times leads to the accumulation of tyrosine 14-
phosphorylated caveolin-1. As in v-Src-transformed cells, several
caveolin-1 bands of approximately 2228 kDa were apparent (compare
Figs. 2C
and 9A
).

View larger version (14K):
[in this window]
[in a new window]
|
Figure 9. Vanadate-Treatment Induces the Accumulation of
Tyrosine 14-Phosphoryated Caveolin-1 in Normal NIH 3T3 Cells
To investigate whether caveolin-1 undergoes tyrosine
phosphorylation in normal NIH 3T3 cells, these cells were treated with
vanadate (100 µM), a tyrosine phosphatase inhibitor. A,
Western blot analysis. Note that treatment with vanadate for increasing
times leads to the accumulation of tyrosine 14-phosphorylated
caveolin-1. As in v-Src- transformed cells, several caveolin-1 bands of
approximately 2228 kDa are apparent. B, Immunofluorescence. NIH 3T3
cells were treated with vanadate for 2 h or left untreated. Note
that tyrosine-phosphorylated caveolin-1 accumulated at the plasma
membrane and was highest in areas of cell-cell contact (upper
panels). Importantly, little or no immunostaining was observed
using antiphosphocaveolin-1 IgG if normal NIH 3T3 cells were not
treated with vanadate. In contrast, immunostaining of vanadate-treated
NIH 3T3 cells with anti-caveolin-1 IgG demonstrated that caveolin-1 was
predominantly localized to a perinuclear compartment and at areas of
cell-cell contact (lower panels). Thus, tyrosine
14-phosphorylated caveolin-1 accumulated predominantly at the plasma
membrane and did not strictly follow the distribution of total
caveolin-1 in vanadate-treated NIH 3T3 cells. Images were acquired with
an MR600 confocal fluorescence microscope(Bio-Rad Laboratories, Inc.).
|
|
We also examined the cellular distribution of tyrosine-phosphorylated
caveolin-1 in vanadate-treated NIH 3T3 cells. Tyrosine-phosphorylated
caveolin-1 accumulated at the plasma membrane and was highest in areas
of cell-cell contact (Fig. 9B
, upper panels). Importantly,
little or no immunostaining was observed using antiphosphocaveolin-1
IgG if normal NIH 3T3 cells were not treated with vanadate. In
contrast, immunostaining of vanadate-treated NIH 3T3 cells with
anticaveolin-1 IgG demonstrated that caveolin-1 was predominantly
localized within a perinuclear compartment (consistent with Golgi
localization) and at areas of cell-cell contact (Fig. 9B
, lower
panels). Thus, tyrosine-phosphorylated caveolin-1 accumulated
predominantly at the plasma membrane and did not strictly follow the
distribution of total caveolin-1 in vanadate-treated NIH 3T3 cells.
Purified Caveolae-Enriched Membranes Contain a Kinase That
Phosphorylates Caveolin-1 on Tyrosine 14 in Vitro
Purified caveolae membranes are dramatically enriched in
caveolin-1 and have been shown to contain tyrosine kinase activity (6, 27, 41, 60, 61). Also, a number of Src-family tyrosine kinases (c-Src,
c-Yes, Lyn, Fyn, Lck, and c-Fgr) and receptor tyrosine kinases (Ins-R,
EGF-R, and PDGF-R) copurify with caveolae and caveolin-1 under these
conditions (1, 27, 28, 30, 31, 61, 62, 63, 64, 65, 66, 67, 68). Thus, we examined the ability
of caveolin-1 to be phosphorylated on tyrosine 14 using purified
caveolae-enriched membrane domains.
Caveolae-enriched domains were purified from murine lung tissue
using an established protocol (28) and incubated in kinase reaction
buffer in the presence or absence of exogenous ATP (1 mM).
Figure 10A
(left panel)
shows that addition of ATP dramatically induced phosphorylation of
caveolin-1 on tyrosine 14. Thus, a kinase activity with the same
specificity as is observed in vivo is present within
purified caveolae membranes and copurifies with caveolin-1. Other
tyrosine-phosphorylated proteins were visualized by immunoblotting with
an antiphosphotyrosine mAb (PY20); the major tyrosine-phosphorylated
proteins appeared in the 50- to 60-kDa and the 200-kDa range,
consistent with the autophosphorylation of Src-family tyrosine kinases
and receptor tyrosine kinases, respectively (Fig. 10A
, right
panel). [It should be noted that a 22- to 25-kDa band
corresponding to caveolin-1 was also detectable with PY20 on longer
exposures (not shown)].

View larger version (14K):
[in this window]
[in a new window]
|
Figure 10. Purified Caveolae-Enriched Membranes Contain
a Kinase That Phosphorylates Caveolin-1 on Tyrosine 14 in
Vitro
A, In vitro kinase assay. Caveolae-enriched
domains were purified from murine lung tissue using an established
protocol (28 ) and incubated in kinase reaction buffer in the presence
or absence of exogenous ATP (1 mM). After 10 min at 25 C,
the reaction was halted by the addition of SDS sample buffer/boiling
and samples were subjected to Western blotting. Note that addition of
ATP dramatically induced phosphorylation of caveolin-1 on tyrosine 14
(left panel). The total pattern of
tyrosine-phosphorylated proteins was visualized by immunoblotting with
an antiphosphotyrosine mAb (PY20) (right panel).
Immunoblotting with mAb 2297 that recognizes total caveolin-1 is shown
as a control for equal loading (middle panel). B,
Inhibitor profile. Note that pretreatment of caveolae-enriched
membranes with either AG1478 or A9 (5 µM) had little or
no effect (upper panel). In contrast, pretreatment with
PP2 (5 µM) dramatically inhibited the phosphorylation of
caveolin-1 on tyrosine 14. Immunoblotting with mAb 2297 that recognizes
total caveolin-1 is shown as a control for equal loading (lower
panel).
|
|
To explore whether caveolin-1 was tyrosine phosphorylated
in vitro by either a receptor tyrosine kinase (such as
EGF-R) or a Src family tyrosine kinase, we employed a panel of specific
tyrosine kinase inhibitors. Tyrophostin AG1478 and tyrophostin A9
preferentially inhibit the receptor tyrosine kinases, EGF-R and
PDGF-R, respectively, while PP2 is a selective inhibitor of Src-family
tyrosine kinases (69, 70, 71). Figure 10B
shows that pretreatment of
caveolae-enriched membranes with either AG1478 or A9 (5
µM) had little or no effect. In contrast,
pretreatment with PP2 (5 µM) dramatically
inhibited the phosphorylation of caveolin-1 on tyrosine 14. These
results are consistent with the idea that an endogenous Src family
kinase that is localized to caveolae membranes phosphorylates
caveolin-1 both in vivo and in vitro.
Identification of Grb7 as a pY14-Caveolin-1 Binding Partner
in Vitro and in Vivo
One known function of tyrosine phosphorylation is to confer
binding to SH2 domain- containing proteins. To identify SH2 domain
proteins that bind to tyrosine-phosphorylated caveolin-1, we used an
established in vitro binding approach. Briefly, we prepared
nonphosphorylated and tyrosine-phosphorylated caveolin-1 from bacteria
(Fig. 1A
; see Materials and Methods). These purified
GST-Cav-1 (1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99, 100, 101) fusion proteins were then incubated with lysates
prepared from normal NIH 3T3 cells. After extensive washing, the bound
material was subjected to Western blot analysis with a panel of
antibodies directed against SH2 domain-containing proteins (listed
in Table 2
).
Figure 11A
shows that of the 20
proteins surveyed, only Grb7 showed any binding to
tyrosine-phosphorylated caveolin-1. Importantly, Grb7 bound only to
tyrosine-phosphorylated GST-Cav-1 (1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99, 100, 101), but not to GST alone, or
nonphosphorylated GST-Cav-1 (1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99, 100, 101) (left panel). Other
closely related proteins, such as Grb2 and Grb14, did not show any
binding activity, clearly demonstrating the specificity of this binding
event (right panel).

View larger version (24K):
[in this window]
[in a new window]
|
Figure 11. Tyrosine 14-Phosphorylated Caveolin-1 Binds Grb7
in a Phosphorylation-Dependent Manner
A, Grb7 binding in vitro. Nonphosphorylated and
tyrosine-phosphorylated GST-Cav-1 (1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 ) fusion proteins were
prepared and incubated with cell lysates derived from normal NIH 3T3
cells. After binding, washing, and elution, the eluates were subjected
to immunoblot analysis with antibodies directed against 20 different
SH2 domain-containing proteins (listed in Table 2 ). Left
panel, Note that only the binding of Grb7 was observed.
Importantly, Grb7 bound to tyrosine-phosphorylated GST-Cav-1 (1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 ),
but not to GST alone, or nonphosphorylated GST-Cav-1 (1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 ).
Right panel, Other closely related proteins, such as
Grb2 and Grb14, did not show any binding activity. B, Grb7 binding
in vivo. Left panel, Cos-7 cells were co-transfected
with cDNAs encoding caveolin-1 and Grb7, in the presence of c-Src WT or
its corresponding empty vector (pUSE). Forty-eight hours post
transfection, lysates were prepared and subjected to
immunoprecipitation with antibodies directed against Grb7. These
immunoprecipitates were then subjected to Western blot analysis with
anticaveolin-1 IgG or antiphosphocaveolin-1 IgG. Note that both
antibodies detect caveolin-1 coimmunoprecipitating with Grb7, but not
with beads alone or an irrelevant IgG; and this binding occurs only in
cells cotransfected with c-Src WT. Right panel, Note
that this binding is strictly dependent on tyrosine 14, as caveolin-1
(Y14A) fails to coimmunoprecipitate with Grb7. C, Importance of
tyrosine 14. Left panel, Cos-7 cells were cotransfected
with cDNAs encoding caveolin-1 and Grb-7, in the presence of c-Src WT.
Lysates were prepared and subjected to immunoprecipitation with
antibodies directed against Grb7. Immunoprecipitates were then
subjected to Western blotting with antiphosphocaveolin-1 IgG. Before
immunoprecipitation, competing peptides were added to these lysates
(detailed in Table 1 ). Note that the PY14 peptide completely blocks
this interaction, while the corresponding nonphosphorylated peptide
(Y14) or an irrelevant phosphopeptide (PY100) has no effect.
Right panel, Cos-7 cells were transiently transfected
with the cDNA encoding Grb-7. Lysates were prepared and incubated with
biotinylated phosphopeptides prebound to streptavidin beads. After
washing, these precipitates were subjected to Western blotting with
anti-Grb7 IgG. Note that the caveolin-1-derived phosphopeptide
containing tyrosine 14 (PY14) effectively pulls down Grb7, while
another irrelevant caveolin-1-derived phosphopeptide containing
tyrosine 100 (PY100) does not.
|
|
We next reconstituted this binding event in vivo. Cos-7
cells were cotransfected with the cDNAs encoding caveolin-1 and Grb-7,
in the presence of c-Src WT or its corresponding empty vector. Lysates
were subjected to immunoprecipitation with antibodies directed against
Grb7, followed by Western blotting with anticaveolin-1 IgG or
phosphocaveolin-1 IgG. Figure 11B
(left panel) shows that
caveolin-1 coimmunoprecipitates with Grb7, but not with beads alone or
an irrelevant IgG. Importantly, Grb7 binding only occurs in cells
cotransfected with c-Src WT. This binding is strictly dependent on
tyrosine 14, as caveolin-1 (Y14A) fails to coimmunoprecipitate with
Grb7 under these conditions (Fig. 11B
, right panel).
To further investigate the specific role of caveolin-1 tyrosine 14, we
next used a peptide-based approach. Cos-7 cells were cotransfected with
the cDNAs encoding caveolin-1 and Grb-7, in the presence of c-Src.
Lysates were prepared and competing peptides were added (detailed in
Table 1
), and they were immunoprecipitated with antibodies
directed against Grb7. These immunoprecipitates were subjected to
Western blotting with antiphosphocaveolin-1 IgG. Figure 11C
(left
panel) shows that addition of the PY14 peptide completely blocks
this interaction, while the addition of the corresponding
nonphosphorylated peptide (Y14) or an irrelevant phosphopeptide (PY100)
has no effect. These results again implicate phosphorylated tyrosine 14
in this binding event.
To assess whether phosphotyrosine 14 and its surrounding sequence is
sufficient for Grb7 binding, Cos-7 cells were transiently transfected
with the cDNA encoding Grb-7. Lysates were then incubated with
biotinylated phosphopeptides bound to streptavidin beads. After
washing, these precipitates were subjected to Western blot analysis
with anti- Grb7 IgG. Figure 11C
(right panel) shows that a
caveolin-1-derived phosphopeptide containing tyrosine 14 (PY14)
effectively pulls down Grb7, while another irrelevant
caveolin-1-derived phosphopeptide (PY100) does not. Thus, caveolin-1
phosphotyrosine 14 and its surrounding sequence are sufficient for Grb7
binding.
Binding of Grb7 to Tyrosine 14-Phosphorylated Caveolin-1
Functionally Stimulates Anchorage-Independent Growth and Cell
Migration
What is the function of tyrosine-phosphorylated caveolin-1 (Y14) ?
As tyrosine-phosphorylated caveolin-1 is localized at focal adhesions,
we speculated that it might play a role in regulating
anchorage-dependent growth and/or cell migration. To test this
hypothesis, we used 293T cells because they lack endogenous caveolin-1
expression and they undergo transfection with high efficiency, a
prerequisite for these cotransfection studies. In this cellular
context, we compared the activity of wild-type caveolin-1 with a mutant
form of caveolin-1 (Y14A) that is unable to undergo tyrosine
phosphorylation at residue 14.
Anchorage-Independent Growth Studies
Figure 12A
shows
that expression of caveolin-1 alone inhibited foci formation in 293T
cells. This is consistent with previous reports showing that
recombinant expression of caveolin-1 in other cell types inhibits
anchorage-independent growth (72, 73, 74). Expression of c-Src alone had no
effect on foci formation; it has been previously shown that
overexpression of c-Src is not sufficient to mediate cell
transformation (75). In contrast, coexpression of c-Src and caveolin-1
stimulated foci formation by approximately 2-fold, as compared with
mock-transfected or cells transfected with Src alone. These results
suggest that tyrosine phosphorylation of caveolin-1 can stimulate
anchorage-independent growth.

View larger version (11K):
[in this window]
[in a new window]
|
Figure 12. Binding of Grb7 to Tyrosine 14-Phosphorylated
Caveolin-1 Stimulates Both Anchorage-Independent Growth and Cell
Migration
293T cells were cotransfected with the cDNAs encoding c-Src, Grb7,
caveolin-1 (WT), and caveolin-1 (Y14A), alone or in combination. Cells
were then subjected to focus formation or cell migration assays. A,
Anchorage-independent growth. Note that coexpression of
c-Src and caveolin-1 stimulated foci formation by approximately 2-fold,
as compared with mock-transfected cells or cells transfected with Src
alone. In addition, coexpression of c-Src, caveolin-1 (WT), and Grb7
dramatically stimulated foci formation by approximately 3.5-fold, as
compared with mock-transfected cells, and by approximately 7-fold as
compared with cells cotransfected with c-Src and Grb7. In striking
contrast, coexpression of c-Src, caveolin-1 (Y14A), and Grb7 had no
effect on foci formation. B, Cell migration. Note that
coexpression of c-Src, caveolin-1 (WT), and Grb7 dramatically
stimulated cell migration by approximately 2- to 3-fold, as compared
with mock-transfected cells, cells transfected with caveolin-1 alone,
or cells cotransfected with c-Src and Grb7. In contrast,
coexpression of c-Src, caveolin-1 (Y14A), and Grb7 had little or no
effect on cell migration.
|
|
We next examined the effect of Grb7 binding to tyrosine-phosphorylated
caveolin-1. Expression of Grb7 alone or in combination with c-Src
failed to stimulate foci formation. However, coexpression of c-Src,
caveolin-1 (WT), and Grb7 dramatically stimulated foci formation by
approximately 3.5-fold, as compared with mock-transfected cells, and by
approximately 7-fold as compared with cells cotransfected with c-Src
and Grb7. In striking contrast, coexpression of c-Src, caveolin-1
(Y14A), and Grb7 had no effect on foci formation, as compared with
mock-transfected cells. Thus, the ability of the c-Src/Cav-1/Grb7
signaling cassette to confer anchorage independence requires
phosphorylation of caveolin-1 on tyrosine 14.
Cell Migration Studies
293T cells are known to express endogenous EGF-R and to undergo
EGF-stimulated cell migration. Thus, we next assessed the effect of the
Src/Cav-1/Grb7 signaling cassette on this process (Fig. 12B
).
Interestingly, coexpression of c-Src, caveolin-1 (WT), and Grb7
dramatically stimulated cell migration by approximately 2- to 3-fold,
as compared with mock-transfected cells, cells transfected with
caveolin-1 alone, or cells transfected with c-Src and Grb7 alone. In
contrast, coexpression of c-Src, caveolin-1 (Y14A), and Grb7 had little
or no effect on cell migration, as compared with mock-transfected
cells, and behaved the same as cells transfected with c-Src and Grb7
alone. Thus, binding of Grb7 to tyrosine 14-phosphorylated caveolin-1
augments both anchorage-independent growth and EGF-stimulated cell
migration in 293T cells.
It should also be noted that expression of caveolin-1 alone (either WT
or Y14A) had little or no effect on cell migration in 293T cells. This
is in contrast to our previous results employing MTLn3 cells, a
metastatic rat mammary adenocarcinoma cell line (74). In MTLn3 cells,
caveolin-1 expression blocked EGF-stimulated cell migration (74). Thus,
the negative regulatory effect of caveolin-1 on cell migration may be
cell type specific.
 |
DISCUSSION
|
|---|
The oncogene v-Src arose by viral transduction of the normal
cellular gene c-Src (76, 77). Thus, it is thought that viral tyrosine
kinases largely induce transformation by intercepting cell-regulatory
mechanisms that are normally under the control of tyrosine
phosphorylation. In support of this notion, v-Src and c-Src appear to
differ primarily in enzymatic activity, but not in their substrate
specificity (78, 79). This difference in enzymatic activity can be
attributed to the loss of tyrosine residue 527 within v-Src (residue
529 in human c-Src). Phosphorylation at this C-terminal site within
c-Src by CSK normally inactivates c-Src (80, 81).
Few physiologically relevant v-Src substrates are known. Caveolin-1 is
one of these substrates (11, 12, 13, 40). For example, caveolin-1
copurifies as a heterooligomeric complex with c-Src and other
Src-family tyrosine kinases (27, 28, 30, 35). This is consistent with
the general idea that v-Src phosphorylates the normal targets of c-Src
or related Src-family tyrosine kinases, but in an unregulated
fashion.
Here, we studied the phosphorylation of caveolin on tyrosine in
vivo by employing a novel phosphospecific mAb probe that
selectively recognizes tyrosine 14-phosphorylated caveolin-1. We
reconstituted this event by cotransfecting caveolin-1 with c-Src. We
observed that myristoylation of c-Src and palmitoylation of caveolin-1
are both required for Src-induced phosphorylation of caveolin-1 on
tyrosine 14. In cells transfected with activated forms of Src [either
c-Src (Y529F) or v-Src], tyrosine-phosphorylated caveolin-1 was
localized mainly in close proximity to focal adhesions, the major
cellular sites of tyrosine kinase-mediated signal transduction.
We also evaluated the tyrosine phosphorylation of caveolin-1 in other
cell types. In 3T3-L1 adipocytes, insulin dramatically increased the
phosphorylation of caveolin-1 on tyrosine 14, with maximal stimulation
at 5 min of treatment. In A431 cells, EGF stimulation greatly increased
the tyrosine phosphorylation of caveolin-1, which was localized near
focal adhesions. Incubation of NIH 3T3 cells with vanadate (a tyrosine
phosphatase inhibitor) drove the accumulation of
tyrosine-phosphorylated caveolin-1. Also, purified caveolae
membranes contain a kinase activity that phosphorylates caveolin-1 on
tyrosine 14 in vitro. This phosphorylation event was
inhibited by a selective inhibitor of Src-family tyrosine kinases
(PP2). Importantly, these results demonstrate that constitutive (via
activated Src) and regulated (via insulin, EGF, or vanadate treatment)
phosphorylation of caveolin-1 occur at the same site, i.e.
tyrosine 14, in vivo.
Caveolin-1 exists as a high molecular mass oligomer containing
approximately 1416 individual caveolin molecules (23, 25), and these
caveolin oligomers have the capacity to self-associate into larger
complexes (23). By analogy with other known tyrosine phosphorylation
events, we speculated that tyrosine-phosphorylated caveolin-1 could
potentially serve as an oligomeric docking-site for SH2-domain
signaling molecules (23, 42)much like activated growth factor
receptors that oligomerize, undergo tyrosine phosphorylation, and
recruit SH2 domain-containing proteins to the cytoplasmic surface of
the plasma membrane (Fig. 13
). In
support of this hypothesis, we show here that phosphorylation of
caveolin-1 on tyrosine 14 functionally confers binding to Grb7, an SH2
domain-containing adapt