(Received for publication, November 21, 1996)
From the Beirne Carter Center for Immunology Research
and the Department of Microbiology, University of Virginia,
Charlottesville, Virginia 22908 and the § Terry Fox
Laboratory, British Columbia Cancer Research Centre, British Columbia,
Vancouver V5Z 1L3, Canada
The adapter protein Shc has been implicated in
mitogenic signaling via growth factor receptors, cytokine receptors,
and antigen receptors on lymphocytes. Besides the well characterized
interaction of Shc with molecules involved in Ras activation, Shc also
associates with a 145-kDa tyrosine-phosphorylated protein upon
triggering via antigen receptors and many cytokine receptors. This
145-kDa protein has been recently identified as an 2
domain containing 5-nositol hosphatase
(SHIP) and has been implicated in the regulation of growth and
differentiation in hematopoietic cells. In this report, we have
addressed the molecular details of the interaction between Shc and SHIP
in vivo. During T cell receptor signaling, tyrosine
phosphorylation of SHIP and its association with Shc occurred only upon
activation. We demonstrate that the phosphotyrosine binding domain of
Shc is necessary and sufficient for its association with
tyrosine-phosphorylated SHIP. Through site-directed mutagenesis,
we have identified two tyrosines on SHIP, Tyr-917, and Tyr-1020, as the
principal contact sites for the Shc-phosphotyrosine binding domain. Our
data also suggest a role for the tyrosine kinase Lck in phosphorylation
of SHIP. We also show that the SH2 domain of SHIP is dispensable for
the Shc-SHIP interaction in vivo. These data have
implications for the localization of the Shc·SHIP complex and
regulation of SHIP function during T cell receptor signaling.
The adapter protein Shc is a key regulator of intracellular signaling events that lead to such varied biological processes as neuronal differentiation, lymphocyte proliferation, and cellular transformation via polyoma virus middle T antigen (1-6). Shc mediates these effects, at least in part, through the activation of Ras proteins following stimulation of many receptors, including the receptors for growth factors (2, 7-10), antigens (11, 12), and cytokines (13-18), as well as G protein-coupled receptors (19). Shc contains an amino-terminal phosphotyrosine binding (PTB)1 domain, a central collagen homology (CH) region, and a carboxyl-terminal Src homology 2 (SH2) domain but no apparent catalytic domain (7, 20, 21). In hematopoietic cells, Shc exists in two isoforms of 46 and 52 kDa. Upon activation of many receptors, Shc is tyrosine-phosphorylated and subsequently interacts with the SH2 domain of Grb2 (22, 23). Grb2, in turn, interacts with the Ras GTP/GDP exchange factor, mSOS (24-26). The complex of Shc·Grb2·mSOS becomes localized to the membrane through the association of Shc with the activated, tyrosine-phosphorylated receptors, where it leads to Ras activation (27).
Shc interaction with activated receptors can occur either via the SH2 or the PTB domain. While both domains bind phosphotyrosine-containing sequences, their specificities of recognition are different and require residues either COOH-terminal (for SH2) or NH2-terminal (for PTB) to the Tyr(P) (28-33). In hematopoietic cells, the Shc SH2 domain binds to components of the T cell receptor (TCR) (6, 11) and B cell receptor (34), while the PTB domain interacts with the receptors for interleukin (IL)-2 (35) and IL-3 (36). Structures of the Shc PTB and Shc SH2 domains bound to their respective phosphopeptides have revealed the molecular basis for their interactions.
Besides interaction with the activated receptors, Shc also associates
with a prominent 145-kDa tyrosine-phosphorylated protein upon
activation of several receptors on hematopoietic cells (such as the
TCR, B cell receptor, receptors for IL-2, IL-3, granulocyte-macrophage colony-stimulating factor (CSF), erythropoietin, steel factor, and
macrophage CSF) (12, 13, 17, 37, 38). Recently, this 145-kDa
phosphoprotein has been identified as a novel SH2 domain containing
5-inositol phosphatase, SHIP (39-41). SHIP has homology to several
previously identified 5
-inositol phosphatases that have been linked to
certain congenital metabolic disorders (42, 43). Moreover, SHIP has
been shown to dephosphorylate the 5
position of phosphatidylinositol
3,4,5-P3 and inositol 1,3,4,5-tetrakisphosphate (39-41).
The precise role of SHIP in regulating inositol lipids in
vivo and the influence of these phosphoinositides (generated by
SHIP-mediated dephosphorylation) on downstream signaling events remain
unclear. SHIP has been implicated in inhibitory signaling via Fc
RIIb
on mast cells (44) and mitogenic signaling via macrophage CSF (41).
Since 3
-OH-phosphorylated inositol lipids are generated by the
triggering of many receptors (through the action of the enzyme
phosphatidylinositol 3-kinase) (45), the dephosphorylation of
phosphatidylinositol 3,4,5-P3 by SHIP could down-regulate
signals via phosphatidylinositol 3,4,5-P3; alternatively,
phosphatidylinositol 3,4-P2, the product of the SHIP-mediated dephosphorylation, could lead to activation of kinases such as Akt (45). However, the mechanisms by which SHIP regulates signaling via these receptors are not yet understood.
In vivo, SHIP may be regulated through alteration of its catalytic activity or through its association with other proteins such as Shc. The initial studies reported that, during macrophage CSF and IL-3 signaling (39, 41), there was no detectable alteration of the catalytic activity of SHIP following receptor activation. This suggested that the interaction of SHIP with Shc and, in turn, the subcellular localization of the Shc·SHIP complex may be quite important in SHIP-mediated regulation of specific phosphoinositides. The details of the in vivo molecular interaction between Shc and SHIP have not been elucidated. Some in vitro studies indicated a direct binding of the PTB domain to phosphorylated SHIP (21, 37). However, Liu et al. (46), based on in vitro studies, have proposed that the SH2 domain of SHIP may also reciprocally bind to phosphorylated Shc. In this report, we have addressed the in vivo requirements for the interaction between Shc and SHIP during T cell receptor signaling. Our data show that the Shc-PTB domain and specific tyrosine residues within the COOH terminus of SHIP mediate their association in vivo.
The murine T cell hybridoma (BYDP) has been described
previously (47) and was grown in RPMI 1640 medium supplemented with 10% fetal bovine serum, 2 mM L-glutamine,
penicillin, streptomycin, and 2-mercaptoethanol (2 × 105 M). COS-7 cells were maintained in
Dulbecco's modified Eagle's medium supplemented with 10% fetal
bovine serum, 2 mM L-glutamine, and
antibiotics.
DNA encoding GST-tagged full-length (FL) wt Shc,
FL Shc R175Q, PTB domain alone (Shc-PTB), or the CH and SH2 domains of
Shc (Shc CH-SH2) were generated by polymerase chain reaction and
subcloning of DNA fragments into the pEBG vector, as described
previously (36, 48). GST-tagged SHIP constructs were also generated
using the same strategy, and a polymerase chain reaction-based
mutagenesis was used to generate the various mutants (36, 49). To
generate hemagglutinin (HA)-tagged constructs, oligonucleotides
encoding three HA tags in tandem were cloned into the pEBB vector (48), and the DNA encoding different SHIP regions were subcloned in-frame at
the 3 end of the tag. All constructs were sequenced, and the presence
of appropriate mutations was confirmed. The Lck cDNA cloned into
pEF vector was kindly provided by Dr. Tomas Mustelin.
Antibody specific for the SH2 domain of SHIP has been described previously (39, 44). Antibodies specific for Shc and GST were purchased from Transduction Laboratories (Kentucky) and Santa Cruz Biotechnology (Santa Cruz, CA), respectively. Horseradish peroxidase-labeled anti-Tyr(P) antibody (RC20H) was purchased from Transduction Laboratories. Anti-HA antibody 12CA5 was obtained from Boehringer Mannheim.
TransfectionsTransient transfections into COS cells were performed with DEAE-dextran and chloroquine as described previously (36) with 1 µg each of the indicated DNA. Approximately 24 h post-transfection, the cells were starved overnight without serum and harvested, and the proteins were analyzed as described below. Transient transfections into BYDP cells (1 × 107/ml in growth medium) were performed by electroporation using 50 µg of the indicated DNA (Cell Porator, Life Technologies, Inc.) at 250 V and 1180 microfarads. The transfected cells (after approximately 24 h) were stimulated and processed as described below. BYDP cells stably expressing Shc proteins were generated by using the electroporation conditions described above with 20 µg of the relevant plasmids co-transfected with 5 µg of pMHneo plasmid for selection (50). Forty-eight hours post-transfection, the cells were plated out at 2 × 104 cells/well in 48-well plates in medium containing G418 (Life Technologies, Inc.). Individual clones were screened by Western blotting for expression of the proteins of interest, and positive clones that expressed roughly equivalent levels of the different proteins were used in further studies.
T Cell StimulationsBYDP cells (1-2 × 107/ml) were incubated with anti-TCR antibody (F23.1 at 1 µg/ml) and/or anti-CD4 antibody (OKT4D at 1 µg/ml) for 10 min on ice. Rabbit anti-mouse IgG was added (7.5 µg/ml) for cross-linking, and the cells were incubated for an additional 10 min on ice. The cells were stimulated at 37 °C for 2 min, washed, and lysed (1% Nonidet P-40, 50 mM Tris (pH 7.6), 150 mM NaCl, 1 mM Na3VO4, 10 mM NaF, 10 µg each of leupeptin, aprotinin, and pepstatin, and 2 mM phenylmethylsulfonyl fluoride). The lysates were spun at 14,000 × g for 10 min at 4 °C, and the proteins were precipitated with glutathione-Sepharose beads (for GST-tagged proteins) or with the indicated antibodies and protein A-Sepharose beads (Pharmacia Biotech Inc.). The beads were washed (with a buffer containing 0.1% Nonidet P-40, 20 mM HEPES, 150 mM NaCl, 10% glycerol, 1 mM Na3VO4, 10 mM NaF, and 10 µg each of leupeptin, aprotinin, and pepstatin), and the bound proteins were analyzed by SDS-polyacrylamide gel electrophoresis, transferred to nitrocellulose (Schleicher and Schuell), immunoblotted with the indicated antibodies, and developed using enhanced chemiluminescence (Amersham Corp.).
We have previously observed
(11, 35, 37) that a 145-kDa tyrosine-phosphorylated protein
co-precipitates with Shc upon TCR and IL-2 receptor stimulation in T
cells. We first determined if this p145 protein represents SHIP (17,
40, 41, 44). Immunoprecipitation of Shc following cross-linking of TCR
with CD4 and immunoblotting with antibodies specific for SHIP
demonstrated that the 145-kDa phosphoprotein is in fact SHIP (Fig.
1a). This band co-migrated with SHIP proteins
detected in direct anti-SHIP immunoprecipitates (Fig. 1a).
Two additional observations were made. First, SHIP associated with Shc
only upon activation. Second, very little tyrosine phosphorylation of
SHIP was detected in unstimulated T cell lysates. This is in contrast
to the previous observations in IL-3-dependent cells where
there was considerable basal phosphorylation of SHIP, which was
augmented by IL-3 stimulation (39). We observe a similar basal
phosphorylation of SHIP in the IL-2-dependent cell line
CTLL-20 (data not shown). Thus, different cell types may have
variations in the levels of basal tyrosine phosphorylation of SHIP,
which may have implications for SHIP functional activity. However, it
is important to note that despite the basal tyrosine phosphorylation,
SHIP associated with Shc only upon IL-2 and IL-3 stimulation (data not
shown and Ref. 39).
We also observed that cross-linking of the TCR with the co-receptor CD4 enhances the tyrosine phosphorylation of SHIP. As shown in Fig. 1b, cross-linking of CD4 alone or cross-linking of TCR alone led to detectable tyrosine phosphorylation of SHIP, while activation via both receptors significantly enhanced the phosphorylation on SHIP. Since CD4 is physically associated with the tyrosine kinase Lck and the Lck kinase activity is enhanced by CD4 cross-linking, Lck may play a role in phosphorylation of SHIP, either directly or indirectly. Consistent with this notion, co-expression of Lck with SHIP leads to efficient tyrosine phosphorylation of SHIP in COS cells (see below).
Shc-PTB Domain Is Required for Interaction with SHIP in VivoShc contains two domains capable of interacting with
tyrosine-phosphorylated proteins: an amino-terminal PTB domain and a COOH-terminal SH2 domain (7, 20, 21). To determine which region(s) of
Shc is necessary for interaction with SHIP, we generated stable murine
T cell lines expressing either the Shc-PTB domain alone (GST-Shc-PTB)
or the CH and SH2 domains of Shc (GST-Shc CH-SH2), tagged at the amino
terminus with GST (Fig. 2a). Cells expressing
comparable amounts of the GST-tagged Shc proteins or control cells
expressing the GST tag alone were used in all of the studies described
below, and the data shown are representative of several independent
clones derived from multiple transfections. When cells expressing the
GST-Shc-PTB were activated via the T cell receptor and the cell lysates
were precipitated with glutathione-Sepharose beads, SHIP was
co-precipitated with Shc-PTB only upon activation (Fig. 2b).
In control cells, the GST tag alone did not precipitate SHIP. This is
consistent with our previous in vitro studies (35) using
bacterially expressed Shc proteins which had indicated a PTB-dependent interaction with SHIP. Comparison of cells
expressing the PTB or the CH-SH2 domains of Shc indicated that SHIP
co-precipitates only with the PTB domain and not with the CH or SH2
domains of Shc (Fig. 2c). Thus, we conclude that in
vivo, the Shc-PTB domain (and not the SH2 domain) is essential for
the interaction with SHIP during T cell receptor signaling.
SH2 Domain of SHIP Is Dispensable for Interaction with Shc
In the Shc·SHIP complexes that occur upon receptor activation, both Shc and SHIP are tyrosine-phosphorylated. With the identification of an SH2 domain in SHIP (39-41), a second interaction involving SHIP-SH2 binding to phosphorylated Shc has also been proposed (39). In fact, Damen et al. (39) have reported that a phosphopeptide containing the Tyr-317 site of Shc (attached to beads) can precipitate SHIP from cell lysates. While our above data using T cell lines expressing individual domains of Shc clearly demonstrated the requirement for the Shc-PTB domain, it did not exclude a parallel role for the SHIP-SH2 domain in binding to Shc. We examined if this interaction occurs in vivo and determined its relative significance compared with the Shc-PTB binding to phosphorylated SHIP.
To specifically test the role of SHIP-SH2, we needed a setting where the Shc proteins will be phosphorylated, yet the PTB domain would no longer be able to interact with SHIP. The rationale was that if SHIP-SH2 does play a role in binding to phosphorylated Shc, then SHIP would co-precipitate with a phosphorylated PTB-defective Shc mutant. We and others (36, 49, 51) have previously shown that mutation of Arg-175 within the PTB domain abrogates binding to Tyr(P)-containing proteins. Using a transient transfection approach in COS cells, we expressed HA-tagged wt SHIP proteins along with GST-tagged FL wt Shc or FL Shc R175Q. To obtain phosphorylation of the expressed Shc and SHIP proteins, we coexpressed the T cell tyrosine kinase Lck under these conditions. GST-Shc was precipitated, and the co-precipitation of HA-SHIP was examined by both anti-Tyr(P) and anti-HA immunoblotting.
As shown in Fig. 3, the HA-SHIP proteins were
co-precipitated efficiently by wt Shc but not by the R175Q Shc. This
again highlighted the role for the PTB domain in this interaction.
Moreover, as shown in Fig. 3 (top panel), the Shc wt and the
Shc R175Q mutant were comparably tyrosine-phosphorylated. This
finding is consistent with our previous data that Tyr-317 appears to be
the major, if not the only, site of tyrosine phosphorylation on Shc
during T cell activation and that this site is efficiently
phosphorylated by Lck in vitro (37) and in
vivo.2 However, despite the
phosphorylation on the FL Shc R175Q, SHIP failed to interact with this
protein, thereby suggesting lack of a role for the SHIP-SH2 domain in
binding to phosphorylated Shc. Furthermore, when the Shc-PTB or CH-SH2
domains were coexpressed with HA-SHIP and Lck, the CH-SH2 domain (which
contains Tyr-317), despite its tyrosine phosphorylation, failed to
co-precipitate SHIP. On the other hand, the PTB domain, despite the
lack of phosphorylation, efficiently co-precipitated SHIP. Comparable
expression of Shc and SHIP proteins was confirmed by anti-GST or
anti-HA immunoblotting (Fig. 3, 2 bottom panels). Based on
these data we conclude that the SHIP-SH2 does not contribute to the
Shc-SHIP interaction in vivo. We cannot exclude very weak or
transient interactions (that do not withstand the precipitation
conditions) that may occur between SHIP-SH2 and phosphorylated Shc.
However, this clearly cannot be responsible for the strong interaction
that is observed between Shc and SHIP. Thus, our data demonstrate that
the interaction of Shc-PTB with tyrosine-phosphorylated SHIP is the
primary mode of interaction between these two proteins in
vivo.
Two Tyrosines (Tyr-917 and Tyr-1020) on SHIP Mediate Shc-PTB Binding
We then determined which tyrosine(s) on SHIP provided the
binding site(s) for the Shc-PTB. The preferred binding motif for the
Shc-PTB domain has been defined as XNXXY
(where
represents hydrophobic residues) through analysis of
phosphopeptide binding to Shc-PTB (28, 30-33). Based on this motif,
two tyrosines in the COOH-terminal region of SHIP, Tyr-917 and
Tyr-1020, were identified as putative Shc-PTB binding sites. To
determine if these two tyrosines on SHIP contribute to Shc binding,
individual or combined mutations of Tyr-917 and Tyr-1020 to
phenylalanine were performed (Fig. 4a). wt
SHIP, Y917F, Y1020F, or Y917F/Y1020F double mutant SHIP proteins (all
tagged with GST) were transiently expressed in the murine hybridoma
cells. After T cell activation, the expressed proteins were
precipitated using glutathione-Sepharose beads, and the tyrosine
phosphorylation status of these proteins was assessed by anti-Tyr(P)
immunoblotting. As shown in Fig. 4b, compared with wt SHIP,
the phosphorylation of Y917F and Y1020F proteins was decreased
moderately, while phosphorylation of the Y917F/Y1020F double mutant was
not detectable. These data suggested that these two tyrosines may
represent key phosphorylation sites on SHIP during T cell receptor
activation. However, we were unable to monitor co-precipitation of Shc
in these experiments due to the insufficient expression levels of the
transfected SHIP proteins (data not shown).
To test the role of these two tyrosines in greater detail, we
transiently expressed HA-tagged wt and mutant SHIP proteins along with
GST-tagged Shc-PTB in COS cells. The tyrosine kinase Lck was again
coexpressed to mediate phosphorylation of SHIP. The cells were lysed,
and the Shc-PTB was precipitated using glutathione-Sepharose beads. The
co-precipitating SHIP proteins were visualized by anti-HA immunoblotting. As shown in Fig. 4c, wt SHIP bound to Shc
quite efficiently. However, SHIP proteins with a mutation of either Tyr-917 or Tyr-1020 showed a significant decrease in binding to Shc-PTB, while the Y917F/Y1020F double mutation nearly abrogated binding to Shc-PTB. In contrast, deletion of either the SH2 domain of
SHIP (SH2) or a point mutation within the SH2 domain of SHIP (R34Q,
within the conserved FLVR sequence) had little effect on their
interaction with Shc-PTB. Anti-HA immunoblotting of total lysates
indicated that all of the mutant HA-SHIP proteins were expressed (Fig.
4c, bottom panel). Direct anti-Tyr(P) blotting of
HA-SHIP precipitates indicated a diminished, yet detectable, phosphorylation of the Y917F/Y1020F double mutant. Although tyrosine phosphorylation of the Y917F/Y1020F double mutant was not detectable in
transient expression in T cells (Fig. 4b), phosphorylation of this protein was observed in COS cells (Fig. 4d). This
may be due to the apparently high activity of the Lck when expressed in
COS cells, leading to phosphorylation on tyrosines other than Tyr-917
and Tyr-1020 on SHIP. Also, the apparently increased phosphorylation of
the Y917F mutant (Fig. 4d) can be explained by the higher
expression of this protein in this experiment. All of the mutant
proteins had 5
-inositol phosphatase activity, indicating the mutations did not induce gross structural changes of the protein (data not shown). Thus, we conclude that Tyr-917 and Tyr-1020 represent the two
critical tyrosines within SHIP whose phosphorylation is required for
the efficient interaction with Shc in vivo.
The data presented in this report have several implications for
signaling via Shc and SHIP. In the case of FcRIIB, SHIP has been
shown directly to bind to the receptor via its SH2 domain (44).
However, we do not observe a direct binding of SHIP to either the TCR
complex or the interleukin-2 receptor (data not shown). Under these
conditions, the interaction of SHIP with proteins such as Shc may play
a much greater role in its localization. At a biochemical level, the
interaction between Shc and SHIP could occur via more than one
mechanism, i.e. Shc-PTB binding to phosphorylated SHIP
and/or the SHIP-SH2 binding to phosphorylated Shc. The data presented
in this report demonstrate that in T cells and in COS cells, the
Shc-PTB domain binding to phosphorylated SHIP (specifically Tyr-917 and
Tyr-1020) is necessary and sufficient for this interaction. Our data
also suggest that both sites (Tyr-917 and Tyr-1020) can contribute to
Shc binding. Although Liu et al. (46), based on peptide
competition studies, have indicated a SHIP-SH2 interaction with
phosphorylated Shc, we do not observe this interaction in vivo under the conditions we have tested. This suggested that the
SH2 domains of both proteins are "free" to interact with other molecules.
In the case of the TCR, Shc binds directly via its SH2 domain to the
TCR- chain (11); in this case, Shc, via its PTB domain, may recruit
SHIP to the proximity of the TCR. In our preliminary studies we have
failed to observe a co-precipitation of SHIP with components of the
TCR·CD3 complex, and the existence of such a trimeric complex during
T cell activation remains to be determined. Since the SHIP-SH2 domain
does not participate in binding to Shc, an interesting possibility is
that the SHIP-SH2 may be free to bind other proteins, which may
influence its function. In the case of the IL-2 receptor, Shc directly
binds to the IL-2 receptor
chain via its PTB domain (35).
Structural studies of the PTB domain indicate the existence of only a
single Tyr(P) binding pocket within the Shc-PTB (49). Thus, the
Shc·IL-2 receptor and Shc·SHIP complexes must exist separately and
may perform independent roles. Since Shc interacts via its PTB domain
with the
c chain (shared by IL-3, IL-5, and
granulocyte-macrophage CSF receptors) (36), similar independent
complexes may exist during signaling via these cytokines as well. Thus,
depending on the type of receptor, recruitment of SHIP to active
signaling complexes may occur by different mechanisms and, in turn, may
have different functional consequences.
We thank Drs. Ulrike Lorenz, Joanne Pratt, Steven Burakoff, Victor Engelhard, Lucia Rameh, and Lewis Cantley for helpful suggestions and Vivien Igras for technical help in early parts of this work.