(Received for publication, November 21, 1996, and in revised form, February 11, 1997)
From the Division of Cell Biology, La Jolla Institute for Allergy
and Immunology, San Diego, California 92121, Institute of Cytosignal Research, Inc., Tokyo 140, Japan,
and the § National Cancer Institute-Navy Medical
Oncology Branch, National Naval Medical Center,
Bethesda, Maryland 20889
Stimulation of the T cell antigen receptor
(TCR)·CD3 complex induces rapid tyrosine phosphorylation of Cbl, a
protooncogene product which has been implicated in intracellular
signaling pathways via its interaction with several signaling
molecules. We found recently that Cbl associates directly with a member
of the 14-3-3 protein family (14-3-3) in T cells and that the
association is increased as a consequence of anti-CD3-mediated T cell
activation. We report here that phorbol 12-myristate 13-acetate
stimulation of T cells also enhanced the interaction between Cbl and
two 14-3-3 isoforms (
and
). Tyrosine phosphorylation of Cbl was
not sufficient or required for this increased interaction. Thus,
cotransfection of COS cells with Cbl plus Lck and/or Syk family
protein-tyrosine kinases caused a marked increase in the
phosphotyrosine content of Cbl without a concomitant enhancement of its
association with 14-3-3. Phorbol 12-myristate 13-acetate stimulation
induced serine phosphorylation of Cbl, and dephosphorylation of
immunoprecipitated Cbl by a Ser/Thr phosphatase disrupted its
interaction with 14-3-3. By using successive carboxyl-terminal deletion
mutants of Cbl, the 14-3-3-binding domain was mapped to a serine-rich
30-amino acid region (residues 615-644) of Cbl. Mutation of serine
residues in this region further defined a binding motif distinct from
the consensus sequence RSXSXP, which was
recently identified as a 14-3-3-binding motif. These results suggest
that TCR stimulation induces both tyrosine and serine phosphorylation
of Cbl. These phosphorylation events allow Cbl to recruit distinct
signaling elements that participate in TCR-mediated signal transduction pathways.
Binding of antigenic peptides presented by major histocompatibility complex molecules to the T cell antigen receptor (TCR)1·CD3 complex induces a rapid increase in the activities of protein-tyrosine kinases (PTKs) of the Src and Syk families (1), which in turn function to propagate activation signals by phosphorylating multiple intracellular proteins in T lymphocytes, eventually leading to T cell activation, lymphokine production, and proliferation. One of the major PTK substrates in TCR/CD3-activated T cells is Cbl (2, 3). The corresponding protooncogene, c-cbl, is the cellular homologue of a transforming gene of Cas NS-1 retrovirus, which induces pro-B cell lymphomas and myeloid leukemias in mice (4). The 120-kDa product of c-cbl consists of a highly basic amino-terminal region, a Ring zinc finger motif, multiple proline-rich stretches, and contains several potential tyrosine phosphorylation sites (6, 7). Cbl associates with the Fyn (3, 5) and Zap-70 (8) kinases both in vitro and following T cell activation in vivo, and Zap-70 causes tyrosine phosphorylation of Cbl in an Lck- and Fyn-dependent manner indicating that Cbl may couple Zap-70 to downstream biochemical events during T cell activation (8). This idea is supported by the findings that the proline-rich domain of Cbl mediates constitutive associations with Src homology 3-containing signaling proteins, including the adaptor protein Grb2 (2, 9-11), and phosphorylated tyrosine residues in Cbl associate with the Src homology 2 domains of other signaling proteins, e.g. phosphatidylinositol 3-kinase (PI3-K) (2, 9, 11-13) and Crk (14-17) in an activation-dependent manner.
Recently, we found that Cbl interacts directly with 14-3-3 in T
cells (18). The 14-3-3 protein family, which is expressed in many
organisms and tissues, consists of highly conserved ~30-kDa isoforms
possessing a variety of biological activities (19-21). 14-3-3 proteins
were recently found to bind oncogene and protooncogene products such as
polyoma virus middle-T antigen (22), Raf-1 (23-25), Bcr-Abl (26),
PI3-K (27), protein kinase C (PKC, Ref. 28), and the cdc25 phosphatase
(29), implicating this family of proteins as regulators of
intracellular signaling pathways. More recently, it was reported that a
phosphorylated consensus sequence,
RSXXP, in Raf and other proteins,
represents a 14-3-3-binding motif (30). Interestingly, Cbl does not
contain this consensus sequence. Moreover, in contrast to Raf-1,
interaction of 14-3-3
with Cbl is markedly enhanced by T cell
stimulation (18), suggesting that the interaction of 14-3-3 with Cbl or
Raf is differentially regulated in T cells. However, the molecular
mechanism underlying this TCR-stimulated interaction between 14-3-3 and
Cbl remains unclear.
In the present study, we demonstrate that PMA stimulation also enhances the association of Cbl with 14-3-3, and that tyrosine phosphorylation of Cbl is dispensable for this interaction. Furthermore, we show that PMA induces serine phosphorylation of Cbl and that Ser/Thr dephosphorylation of Cbl abolishes its association with 14-3-3. Finally, our experiments define a novel 14-3-3-binding serine-rich motif in Cbl. Our findings suggest that TCR stimulation activates both PTKs and Ser/Thr kinases which can phosphorylate Cbl, leading to recruitment of distinct proteins that serve to propagate TCR-mediated signals.
Polyclonal rabbit anti-Cbl (c-15), -Lck,
-Zap-70, -Syk, or -Raf-1 antibodies were from Santa Cruz Biotechnology
(Santa Cruz, CA). Anti-phosphotyrosine (Tyr(P)) monoclonal antibody
(mAb) 4G10 and polyclonal anti-PI3-K (p85) or -GST antibodies were from
Upstate Biotechnology (Lake Placid, NY). The anti-14-3-3 mAb was
described previously (27). Anti-hemagglutinin (HA) mAb (12CA5) was from Boehringer Mannheim. Anti-Grb2 was from Transduction Laboratories (Lexington, KY). Anti-human IgG was from Dako (Denmark). An anti-CD3 mAb, OKT3, was purified from hybridoma culture supernatants by using
protein A-Sepharose affinity chromatography. Horseradish peroxidase-conjugated F(ab)
2 fragments of donkey
anti-rabbit IgG or sheep anti-mouse IgG was from Amersham Corp.
Simian virus 40 T antigen-transfected human leukemic Jurkat T cells (Jurkat-TAg) were grown in RPMI 1640 medium (Life Technologies, Inc.) supplemented with 10% fetal bovine serum and antibiotics. Cells were resuspended (2 × 107/ml) in 0.5 ml of medium, equilibrated at 37 °C for 5 min, and activated with either OKT3 (4 µg/ml) for 5 min or with PMA (50 ng/ml) for 15 min. Stimulation was terminated by adding 0.5 ml of 2 × Nonidet P-40 lysis buffer (2% Nonidet P-40, 40 mM Tris-HCl, pH 7.5, 300 mM NaCl, 10 mM EDTA, 10 mM NaPP, 10 mM NaF, 4 mM Na3VO4, 20 µg/ml each aprotinin and leupeptin). Cells were lysed for 10 min at 4 °C, and insoluble material was removed by centrifugation at 15,000 × g (4 °C for 10 min).
COS-1 cells were cultured in Iscove's modified Dulbecco's medium (Life Technologies, Inc.) containing 10% heat-inactivated fetal bovine serum at 37 °C in 5% CO2. Jurkat-TAg or COS-1 cells were transiently transfected with an optimal amount of plasmid DNA (usually 10 µg) by electroporation as described previously (31). To stimulate COS-1 cells, PMA (100 ng/ml) was added at 37 °C for 30 min.
PlasmidsThe human Cbl, v-cbl, 70Z/3 (a 17-amino
acid deletion of Cbl; Ref. 32) cDNAs, and several carboxyl-terminal
deletion mutants of Cbl were subcloned into the pEFneo (33) mammalian
expression vector (as described in Ref. 34). To construct cDNAs
encoding Cbl proteins with additional carboxyl-terminal deletions,
200-300-base pair fragments corresponding to defined Cbl sequences
were generated by PCR amplification. The PCR products were cloned into
the TA cloning vector (Invitrogen). The cDNAs were then ligated
back to the original pEFneo-70Z/3 cDNA which has been digested with BglII and XbaI to remove a 3-coding sequence.
The residue numbers of the different truncation mutants correspond to
the sequence of wild-type Cbl. A fragment encoding Cbl residues
615-644 or mutants thereof was expressed by PCR amplification of the
corresponding cDNA fragment and ligated in-frame to a sequence
encoding the Fc fragment of human IgG1 (33) in pEFneo. DNA
sequences were verified in all instances by direct sequencing.
The 14-3-3 cDNA was cloned by PCR using a cDNA library from
U937 cells. The coding sequence was verified by DNA sequencing, and the
cDNA was subcloned in-frame into a bacterial expression vector,
pGEX-4T-2 (Pharmacia Biotech Inc.). The GST-14-3-3
fusion protein
was expressed and purified as described previously (18, 27). The
construction of the mammalian expression vector for Cbl-b (35) will be
described elsewhere. The Lck, Syk, and Zap-70 expression constructs
have been described (36).
Cell lysates were incubated with 10 µg of GST, GST-14-3-3, or GST-14-3-3
fusion proteins (27) for
2 h at 4 °C, followed by the addition of 40 µl of
glutathione-Sepharose beads. After 1 h at 4 °C, the binding
mixtures were washed extensively in 1 × Nonidet P-40 lysis buffer
and used for further analysis.
Membranes were denatured in 6 M guanidine-HCl, dissolved in 50 mM Tris-HCl,
pH 8.0, 5 mM 2-mercaptoethanol, 2 mM EDTA for 1 h at room temperature, and renatured in the same buffer
overnight at 4 °C. Membranes were then incubated with GST-14-3-3
or GST alone (10 µg/ml) for 2 h at 4 °C, followed by anti-GST
antibody and enhanced chemiluminescence (ECL) detection system
(Amersham Corp.).
Lysates (1 × 107 cells) were mixed with antibodies for 2 h, followed by the addition of 40 µl of protein A/G Plus-Sepharose beads (Santa Cruz Biotechnology) for an additional hour at 4 °C. Immunoprecipitates (IPs) were washed four times with 1 × Nonidet P-40 lysis buffer and boiled in 30 µl of 2 × Laemmli's buffer. Samples were subjected to SDS-10% PAGE analysis and electrotransferred onto polyvinylidene difluoride membranes (Millipore). Membranes were immunoblotted with the indicated primary antibodies (1 µg/ml), followed by horseradish peroxidase-conjugated secondary antibodies. Membranes were washed and visualized by ECL. A minigel (9 × 8 cm) was used for the separation of Cbl, Raf-1, and PI3-K, and a 13.8 × 13-cm gel was used for the separation of 14-3-3 and Grb2. When necessary, membranes were stripped by incubation in 62.5 mM Tris-HCl, pH 6.7, 100 mM 2-mercaptoethanol, 2% SDS for 1 h at 70 °C with constant agitation, washed, and then reprobed with other antibodies as indicated.
32P Labeling and Phosphoamino Acid AnalysisCells were starved in phosphate-free medium for 2 h
and labeled with 32Pi (0.5 mCi/ml) for another
4 h. The cells were then treated with or without PMA (50 ng/ml)
for 15 min, lysed, and immunoprecipitated with anti-Cbl antibodies as
described above. The membrane containing SDS-PAGE-resolved proteins was
subjected to autoradiography, and the protein bands corresponding to
Cbl were cut out, hydrolyzed in 6 N HCl, and subjected to
two-dimensional thin layer chromatography analysis by electrophoresis
using a Hunter thin layer electrophoresis apparatus (C.B.S. Scientific
Co., Del Mar, CA) in the first dimension and by ascending
chromatography in isobutyric acid, 0.5 M NH4OH, 5:3 (v/v) in the second dimension (37). The thin layer plate was
exposed to an x-ray film at 70 °C. The positions of phosphoamino acids were determined by staining the cold phosphoamino acids included
in the mixture with ninhydrin.
Beads containing anti-Cbl IPs were
washed three times with Nonidet P-40 lysis buffer and twice with
protein phosphatase (PP) buffer containing 50 mM Tris-HCl,
pH 7.0, 0.1 mM EDTA, 5 mM dithiothreitol. The
beads were resuspended in 100 µl of PP buffer, and the reaction was
initiated by adding MnCl2 to a final concentration of 1 mM and 1 unit of recombinant PP1 (New England Biolabs
Inc.). After 1 h at 30 °C with constant shaking, the beads were
spun down and eluted with SDS-PAGE buffer. Proteins were analyzed by
SDS-10% PAGE and far Western blotting with GST-14-3-3.
To further define the nature of the enhanced
Cbl-14-3-3 association mediated by TCR/CD3 ligation (18), we evaluated
the ability of PMA, a phorbol ester which is known to activate PKC and
other Ser/Thr kinases, to modulate the association of Cbl with 14-3-3. Jurkat-TAg cells were either stimulated with OKT3 or PMA or left
unstimulated. Cell lysates were mixed with GST or GST-14-3-3, and
bound proteins, which were recovered with glutathione-Sepharose beads,
were subjected to SDS-PAGE and immunoblotting with an anti-Cbl
antibody. Consistent with our previous report (18), anti-CD3
stimulation of T cells greatly increased the amount of Cbl precipitated
by GST-14-3-3 (Fig. 1A, top
panel). However, PMA treatment similarly enhanced this
interaction. GST alone did not bind Cbl, confirming the specificity of
the association. Probing the same membrane with an antibody against
Raf, another 14-3-3-binding protein (23-25), revealed that, unlike
Cbl, similar amounts of Raf from resting and OKT3- or PMA-activated
cells were associated with GST-14-3-3
(Fig. 1A,
middle panel). The mobility shift of Raf observed following
PMA or anti-CD3 stimulation is consistent with its activation and most
likely results from its serine and/or threonine phosphorylation by PKC
(38). These results suggest that different mechanisms which regulate
the interaction between 14-3-3 and distinct binding proteins exist in T
(and possibly other) cells. To determine whether PMA increases the
interaction between 14-3-3 and Cbl by indirectly inducing tyrosine
phosphorylation of Cbl, the membrane containing the resolved samples
was probed with an anti-Tyr(P) antibody. Although Cbl was readily
tyrosine-phosphorylated in OKT3-stimulated cells, no Tyr(P) could be
detected in immunoprecipitated Cbl from PMA-stimulated cells (Fig.
1A, bottom panel). Thus, tyrosine phosphorylation
of Cbl appears to be dispensable for its interaction with 14-3-3.
We then ascertained whether PMA could also enhance the interaction
between 14-3-3 and Cbl in intact T cells by probing Cbl IPs from
resting or OKT3- or PMA-stimulated cells, with an anti-14-3-3 mAb.
As shown in Fig. 1B, a very small amount of 14-3-3
could be detected in the Cbl IPs from unstimulated cells (Fig. 1B,
top panel); both anti-CD3 and PMA treatment increased the
interaction between 14-3-3
and Cbl to a similar degree. However, Cbl
was tyrosine-phosphorylated only in OKT3-stimulated cells as revealed by anti-Tyr(P) immunoblotting of the same IPs. We further assessed the
effect of the same activating stimuli on the association of Cbl with
two other proteins, i.e. Grb2 and PI3-K. When the membrane was probed with an antibody specific for Grb2, which is known to
associate constitutively with Cbl (2, 9-11), similar amounts of Grb2
coimmunoprecipitated with Cbl from resting and stimulated cells. In
contrast, probing the membrane with an anti-p85 antibody demonstrated
that p85 was only present in Cbl IPs from OKT3-stimulated cells,
consistent with the finding that the regulatory subunit (p85) of PI3-K
associates with tyrosine-phosphorylated Cbl, possibly via the
Y731EAM motif of the latter (2, 34). Anti-Cbl
immunoblotting confirmed that similar amounts of Cbl were present in
all the IPs (Fig. 1A, bottom panel). These
results demonstrate that PMA induces increased association of Cbl with
14-3-3
, but not with Grb2 or PI3-K, in intact T cells and that this
association is independent of tyrosine phosphorylation of Cbl.
Physiological activation of T lymphocytes via their
TCR·CD3 complex activates both PTKs and Ser/Thr kinases. To determine the effects of Cbl tyrosine phosphorylation in isolation on its association with 14-3-3, we cotransfected COS-1 cells with a Cbl expression vector plus plasmids encoding PTKs potentially involved in
regulating Cbl, i.e. Lck and/or Syk family kinases (Zap-70 or Syk). As shown in Fig. 2A, transfection of
the Cbl expression plasmid resulted in overexpression of a 120-kDa
protein which was recognized by the anti-Cbl antibody (top
panel). Similarly, anti-PTK immunoblotting confirmed the
expression of Syk, Zap-70, and/or Lck in cells transfected with the
corresponding vectors (three bottom panels). When the
membrane containing the resolved cell lysates was immunoblotted with
anti-Tyr(P) antibody, it was found that the Tyr(P) content of Cbl was
very low in cells transfected with the Cbl cDNA alone.
Cotransfection with Zap-70 or, to a larger extent, with Syk alone
increased the tyrosine phosphorylation of Cbl, and this level was
further increased when the cells were triple-transfected with Cbl, Lck,
and Zap-70 or Syk. These results agree with the previous observation
that Cbl is probably a downstream target of Zap-70 and Src family
kinases such as Lck or Fyn in T cells (8), as well as with the recent
report demonstrating that Syk is more active than Zap-70 in terms of
its ability to phosphorylate Cbl in COS-1 cells (39).
Cell lysates from the Cbl- plus PTK-transfected COS-1 cells, which were
either unstimulated or treated with PMA, were then precipitated with
GST-14-3-3, and the binding of Cbl or Raf was determined by
immunoblotting with the corresponding antibodies (Fig. 2B).
GST-14-3-3
precipitated a small amount of Cbl from unstimulated
cells, and this level was markedly increased when the cells were
stimulated with PMA. However, consistent with the results obtained in T
cells (Fig. 1), cotransfection with PTKs and, consequentially, the
level of Cbl tyrosine phosphorylation, did not have a significant
effect on its association with GST-14-3-3
(Fig. 2B,
top panel). These results reinforce the conclusion that tyrosine
phosphorylation of Cbl is not required for its interaction with 14-3-3. As noted before, PMA stimulation did not increase the amount of Raf
associated with GST-14-3-3
(Fig. 2B, bottom panel), although it caused a mobility shift (Fig.
2B, bottom panel) consistent with its
phosphorylation.
To establish the specificity of the interaction between
Cbl and 14-3-3, we extended our analysis to examine another 14-3-3 isoform (), as well as two additional Cbl-related proteins: 70Z/3, a
transforming mutant of Cbl with a 17-amino acid deletion near the Ring
zinc finger motif (32); and Cbl-b, a recently identified Cbl homolog
expressed in different tissues and cell lines, including hematopoietic
cells (35). Lysates from Jurkat-TAg cells transfected with HA-tagged
70Z/3, Cbl, or Cbl-b expression vectors were incubated with
GST-14-3-3
or 14-3-3
, and the binding of Cbl was analyzed by
immunoblotting with an anti-HA mAb. Following PMA stimulation, both
70Z/3 and Cbl displayed an increased association with GST-14-3-3
; however, no binding of Cbl-b to GST-14-3-3
was detectable in either
resting or PMA-stimulated cells (Fig. 3A),
despite the fact that similar amounts of HA-tagged Cbl-related proteins
were present in the lysates from all the transfected samples (Fig. 3B). Similar results were obtained when the interaction with
GST-14-3-3
was evaluated (Fig. 3C). These results
indicate that both 14-3-3 isoforms can bind Cbl or 70Z/3, but not
Cbl-b.
PMA-induced Serine Phosphorylation of Cbl
Since 14-3-3 proteins were recently found to recognize phosphoserine-containing
motifs (30), it was likely that the increased association of Cbl with
14-3-3 following PMA stimulation results from its phosphorylation. To
address this question, lysates from resting or PMA-stimulated
Jurkat-TAg cells metabolically labeled with
32Pi were immunoprecipitated with an anti-Cbl
antibody, and the phosphorylation status of Cbl was determined. PMA
stimulation dramatically increased the phosphorylation of a 120-kDa
protein (Fig. 4A), which was recognized by an
anti-Cbl antibody (data not shown). Immunoprecipitated 14-3-3 did
not display increased phosphorylation under the same conditions (data
not shown). Phosphoamino acid analysis of the 120-kDa band revealed a
clear increase in the phosphoserine content following PMA stimulation
(Fig. 4B). No phosphothreonine or Tyr(P) was detectable.
These results strongly suggest that the PMA-enhanced interaction of Cbl
with 14-3-3 results from serine phosphorylation of the former.
Dephosphorylation of Cbl Disrupts Its Interaction with 14-3-3
To further confirm the importance of phosphoserine
residues in Cbl for the interaction with 14-3-3, Cbl IPs from
unstimulated or PMA-stimulated Jurkat-TAg cells were subjected to
treatment with a recombinant Ser/Thr phosphatase, PP1. The PP1-treated
and untreated samples were then analyzed for binding of GST-14-3-3 by Far Western blotting. Consistent with our previous observations in
anti-CD3-stimulated T cells (18), GST-14-3-3
bound directly to a
120-kDa protein that comigrated with Cbl in the untreated samples, and
binding was markedly enhanced by PMA stimulation (Fig.
5, top panel). PP1 treatment completely
abolished the direct interaction of 14-3-3
with Cbl. Stripping and
reprobing the membrane with an anti-Cbl antibody demonstrated that
similar amounts of Cbl were present in all the samples (Fig. 5,
bottom panel). The results indicated that phosphoserine
residues in Cbl were indispensable for its direct interaction with
14-3-3.
Mapping of the 14-3-3-Binding Site in Cbl
It was recently
reported that the binding of proteins to 14-3-3 is mediated by the
phosphorylated consensus motif RSXXP
(30). Inspection of the sequence of Cbl did not reveal such a motif. Therefore, experiments were conducted to map the 14-3-3-binding domain
in Cbl using HA-tagged 70Z/3 proteins containing successive truncations
at their carboxyl terminus (34). As shown earlier (Fig. 3A),
70Z/3 binds 14-3-3 to the same extent as wild-type Cbl. Jurkat-TAg
cells transiently transfected with 70Z/3 expression plasmids, and the
binding of the transfected gene products from untreated or
PMA-stimulated cells to GST-14-3-3 in vitro was analyzed
with an anti-HA antibody. GST-14-3-3
bound the full-length 70Z/3
protein and two deletion mutants,
1-792, and
1-730, from which
114 and 176 amino acids were removed, respectively (Fig. 6, top panel). Additional carboxyl-terminal
deletions that removed 346, 366, and 456 carboxyl-terminal residues
(
1-560,
1-540, and
1-450, respectively) eliminated the
binding to GST-14-3-3
. The product of v-cbl, which
encodes the 361 amino-terminal residues of Cbl, did not associate with
GST-14-3-3
. As a control, the membrane was stripped and reprobed
with an anti-Raf antibody. Comparable amounts of Raf could be detected
in all the samples (Fig. 6, middle panel). To confirm that
the differential binding to GST-14-3-3
does not reflect different
expression levels of the 70Z/3 proteins, cell lysates of the
transfected cells were immunoblotted with the anti-HA antibody. Similar
amounts of HA-tagged proteins were present in the different groups
(Fig. 6, bottom panel). These results demonstrate that a
region of 70Z/3 between amino acid residues 560 and 730 most likely
contains the 14-3-3
-binding motif.
The analysis was refined by creating additional but more limited
truncations in the region encompassing residues 560-730 of 70Z/3. The
respective proteins from transfected Jurkat-TAg cells were analyzed for
binding to GST-14-3-3 as before. GST-14-3-3 bound the
1-698,
1-655, and
1-645 70Z/3 mutants as well as the control
1-730, and PMA treatment increased the interaction of all three
mutants with GST-14-3-3
(Fig. 7A, top panel). The
1-629 mutant, in which 16 amino acids were further deleted relative to
1-645, exhibited a markedly reduced binding, and the deletion of
another 15 residues (
1-614) abolished the binding to GST-14-3-3
. Similarly, the
1-597 and
1-571 mutants, as well as the negative control,
1-560, did not bind to GST-14-3-3
. As before,
endogenous Raf from all groups of the transfected COS-1 cells bound
equally well to GST-14-3-3
(Fig. 7A,
middle panel). Anti-HA immunoblotting of the cell lysates
showed the presence of comparable amounts of the mutant proteins in all
the transfected samples (Fig. 7A, bottom panel).
Identical results were obtained when binding to 14-3-3
was assessed
(Fig. 7B). Based on these results, we conclude that residues
615-644 of Cbl contain the binding site for 14-3-3 proteins.
Identification of Serine Residues Critical for 14-3-3 Binding
Although Cbl does not contain an exact
RSXSXP which has been implicated in 14-3-3 binding (30), residues 615-644 of Cbl contain five serine
residues. We focused our attention on two di-serine-containing
stretches in this region, i.e.
R617HLPF, and
R626LGTF as plausible sites
involved in the interaction with 14-3-3 (Fig.
8A). To determine the importance of these
residues, we replaced the two tandem serine residues in the first,
second, or both motif(s) with alanine by site-directed mutagenesis to
generate the constructs A2S2, S2A2, or A4, respectively. These, as well
as the non-mutated Cbl construct (S4), were fused in-frame to a pEF
vector encoding the Fc fragment of human IgG1 (33), and the
plasmids were then transfected into Jurkat-TAg cells to determine their
ability to interact with GST-14-3-3 in vitro.
The protein products of the four expression vectors were expressed at a
similar level as revealed by immunoblotting with an anti-human IgG
antibody (Fig. 8B). When lysates from the transfected, untreated, or PMA-stimulated cells were incubated with GST-14-3-3, it was found that only the wild-type Cbl fragment (S4) bound to the
fusion protein; mutation in either or both of the di-serine motifs
eliminated the binding (Fig. 8C). Consistent with the
earlier findings, the association with 14-3-3
was markedly increased by PMA stimulation. These results are also consistent with the finding
that relative to the
1-645 construct, which interacted well with
14-3-3, the additional deletion of residues 630-644 or 615-629
(constructs
1-629 and
1-614, respectively), which encompass the
relevant serine residues, reduced considerably or abolished the binding
to GST-14-3-3
(Fig. 7).
The striking difference in 14-3-3 binding between Cbl and Cbl-b (Fig. 3) prompted us to compare the sequences of the two proteins in the 14-3-3-binding domain. Sequence alignment showed that Cbl-b contains the first consensus motif present in Cbl (R617HLPF), but lacks the second motif (Fig. 8D). This sequence difference most likely explains the failure of Cbl-b to interact with 14-3-3 and, moreover, strongly suggests that both di-serine-containing consensus motifs are essential for 14-3-3 binding, a notion also supported by the analysis of the Cbl mutants (Fig. 8C).
Previous studies have shown that TCR/CD3 cross-linking induces rapid tyrosine phosphorylation of Cbl (2, 3). We reported recently that the same activating conditions enhance the direct interaction between Cbl and 14-3-3 (18). However, the contribution of Tyr(P) residues in Cbl to this increased association was unclear, and therefore, we have sought to define the nature of the activation-associated events that induce a greater interaction between Cbl and 14-3-3. Evidence is presented that tyrosine phosphorylation of Cbl is not required for the interaction with 14-3-3. Rather, based on the effects of PMA or phosphatase treatment and the analysis of Cbl mutants, serine phosphorylation is responsible for this enhanced association. Our conclusion that phosphoserine residues in Cbl are critical for its direct physical contact with 14-3-3 proteins is consistent with the findings that: first, Ser259 in Raf, which is a known phosphorylation site (40), is critical for 14-3-3 binding (41); second, 14-3-3 proteins associate with phosphorylated, but not unphosphorylated, tyrosine hydroxylase (42) or keratin (43); and third, 14-3-3 interacts with its ligands via a conserved phosphoserine-containing consensus motif, RSXXP (30).
PMA treatment that induced a mobility shift of Raf consistent with its
PKC-mediated phosphorylation (38) did not increase the interaction
between Raf and 14-3-3. PKC was found to phosphorylate Raf at serine
residues 259 and 499 (38). One of these (Ser259), as well
as Ser621 , is required for 14-3-3 binding (30, 41). The
use of Raf-based phosphoserine-containing synthetic peptides
demonstrated that both R256STTP and
R618SAEP derived from Raf-1 bind to 14-3-3 and, hence, disrupt the interaction between these two proteins (30).
Since PMA treatment did not affect the interaction between Raf and
14-3-3 in our experiments, PKC-dependent phosphorylation of
Raf at Ser259 seems not to be critical for 14-3-3 binding,
in agreement with the recent suggestion that Ser621 is the
primary 14-3-3-binding site (30). The observation that the latter
residue is constitutively phosphorylated in cells (40) is also
compatible with our finding that PMA does not enhance the association
between Raf and 14-3-3.
Interestingly, Cbl does not contain the exact consensus sequence RSXSXP that was recently identified as a 14-3-3-binding motif (30). Instead, our analysis of Cbl deletion or point mutants defined a serine-rich 30-amino acid region (residues 615-644) that contains two putative 14-3-3-binding sites encompassed in the sequences R617HSLPFS and R626LGSTFS. The second site is not present in Cbl-b. This difference in 14-3-3 binding between Cbl and Cbl-b predicts some distinctions in their biological functions. Taken together, our findings define a novel serine-based putative 14-3-3-binding motif represented by the consensus sequence RX1-2SX2-3S. Thus, the previously defined 14-3-3-binding consensus motif, RSXSXP (30), does not account for all interactions between 14-3-3 proteins and their ligands.
Our studies do not identify the Ser/Thr kinase(s) that phosphorylates
Cbl to increase its association with 14-3-3 proteins. The
14-3-3-binding motif defined by our experiments contains a basic
residue (arginine) in position 2 or
3 relative to the first serine
residue. This motif could represent a consensus substrate site for
several Ser/Thr kinases, including cyclic AMP-dependent protein kinase, (protein kinase A) calmodulin-dependent
protein kinase II, or PKC (44). However, treatment of Jurkat T cells with a protein kinase A agonist (forskolin) did not increase the association between 14-3-3 and Cbl (data not shown). On the other hand,
the finding that PMA, a known PKC activator, induced this effect both
in T and in COS-1 cells implicates a direct or indirect ubiquitous role
for member(s) of the PKC family in this modification. In this regard,
the presence of different 14-3-3-binding motifs in Cbl and Raf and the
differential effects of anti-CD3 or PMA stimulation on the binding of
14-3-3 proteins to Cbl versus Raf could mean that distinct
Ser/Thr kinases, which respond differentially to anti-CD3 or PMA
stimulation, phosphorylate these two targets to facilitate their
interaction with 14-3-3.
In summary, our findings strongly suggest that TCR/CD3 ligation
transduces two distinct signals to Cbl: one delivered by
receptor-coupled PTKs, most likely Src and/or Syk family kinases (8),
which induce the tyrosine phosphorylation of Cbl; and another, mediated by a Ser/Thr kinase, which phosphorylates Cbl on serine residues. Each
of these post-translational modifications leads to a distinct outcome.
Whereas tyrosine phosphorylation of Cbl causes it to associate with the
Src homology 2 domain of signaling proteins such as Src or Syk family
kinases (3, 5, 8), PI3-K (2, 9, 11-13), and Crk (14-17), its serine
phosphorylation leads to enhanced recruitment of 14-3-3 proteins as
demonstrated in the present study. Thus, the convergent action of both
tyrosine and Ser/Thr kinases on Cbl may stimulate the formation of
Cbl-associated protein complexes that are necessary for optimal signal
transduction from the TCR·CD3 complex to downstream targets. In
addition, serine phosphorylation of Cbl may regulate its
phosphorylation by tyrosine kinases in a manner similar to the
regulation of phospholipase C (45, 46). The results reported here,
as well as our recent finding that Cbl (70Z/3) interacts with
Ras-dependent signaling pathways leading to the nuclear
factor of activated T cells activation in T cells (34), may open the
way to additional studies aimed at delineating the function of Cbl in
signal transduction pathways.