(Received for publication, July 9, 1996, and in revised form, October 2, 1996)
From the Division of Cell Biology, La Jolla Institute for Allergy
and Immunology, San Diego, California 92121 and
Department of Pathology, University of Western Australia,
Nedlands, West Australia 6009, Australia
T cell receptor (TCR) stimulation induces rapid tyrosine phosphorylation of cellular proteins, including Cbl, a protooncogene product whose function remains unclear. As a first step toward elucidating the function of Cbl in TCR-initiated signaling, we evaluated the ability of wild-type Cbl or a transforming Cbl mutant (70Z/3) to induce transcriptional activation of a nuclear factor of activated T cells (NFAT) element derived from the interleukin 2 (IL2) promoter in transiently cotransfected Jurkat-TAg T cells. 70Z/3, but not Cbl, caused NFAT activation which was significantly enhanced by stimulation with calcium ionophore, and was drastically reduced by cyclosporin A pretreatment. A point mutation of a potential phosphatidylinositol 3-kinase (PI3-K) binding site (Y731EAM to Y731EAC) in 70Z/3 disrupted the association of PI3-K with 70Z/3, but did not reduce the induction of NFAT activity, suggesting that the interaction between Cbl and PI3-K is not required in the 70Z/3-mediated induction of NFAT. Additional mapping studies indicated that defined deletions of C-terminal 70Z/3 sequences affected to a variable degree its ability to stimulate NFAT activity. Strikingly, deletion of 346 C-terminal residues augmented this activity, whereas removal of 20 additional residues abolished it. Coexpression of dominant negative Ras abrogated the basal or ionomycin-stimulated, 70Z/3-mediated NFAT activation, suggesting a functional Ras is required for this activation. These results implicate Cbl in Ras-dependent signaling pathways which lead to NFAT activation.
The c-cbl protooncogene is the cellular homologue of v-cbl, a transforming gene of the Cas NS-1 retrovirus. The v-cbl oncogene was generated as a result of a rearrangement between Cas-Br-M virus and a sequence encoding the 355 N-terminal residues of the c-cbl product. This oncogene induces pro-B cell lymphomas and myeloid leukemias in mice (1). The 120-kDa product of c-cbl consists of a highly basic N-terminal region, a zinc Ring finger, multiple proline-rich sequences, and several potential tyrosine phosphorylation sites (2). While the transforming v-cbl product can be detected both in the cytoplasm and the nucleus, c-Cbl is exclusively cytosolic and lacks transforming activity (3). Another c-cbl mutant isolated from the murine 70Z/3 pre-B cell lymphoma encodes a protein with an internal deletion of 17 amino acids near the zinc Ring finger. This mutation (70Z/3) also activates the transforming potential of c-cbl, and the corresponding protein displays enhanced tyrosine phosphorylation (4).
Several recent findings support the notion that Cbl is involved in signal transduction pathways: first, the proline-rich domain of Cbl mediates constitutive associations with SH3 domains of adaptor signaling proteins, i.e., Grb2 (5, 6, 7, 8) and Nck (9); second, engagement of different receptors, including the epidermal growth factor receptor (10, 11, 12, 13, 14) and cytokine receptors (7, 14, 15, 16) causes an increase in the phosphotyrosine (Tyr(P))1 content of Cbl, and tyrosine-phosphorylated Cbl associates with SH2 domains of the regulatory subunit (p85) of phosphatidylinositol 3-kinase (PI3-K) (5, 6, 7, 13, 17, 18) and Crk (19, 20, 21) in an activation-dependent manner. Third, we found recently that in T cells, Cbl interacts with 14-3-3 proteins, and that this association is increased upon TCR stimulation (22). 14-3-3 proteins are thought to participate in signal transduction pathways via their association with different oncogene and protooncogene products (23, 24). Despite all of these findings, however, the physiological function of mammalian Cbl remains unclear.
Cbl is also likely to participate in signaling pathways initiated by the T cell antigen receptor (TCR)-CD3 complex and other immune recognition receptors, i.e. the B cell antigen receptor (18, 25), and the IgG-Fc receptor (14, 26). TCR·CD3 ligation by processed antigenic peptides or anti-receptor antibodies leads to activation of protein tyrosine kinases of the Src and Syk families (27, 28) which, in turn, phosphorylate multiple cellular proteins. These early events trigger signal transduction cascades which lead to T cell activation, lymphokine production, and proliferation. Cbl was recently identified as a prominent protein tyrosine kinase substrate in TCR-stimulated T cells (5, 6, 21, 29), and was found to associate with two protein tyrosine kinases, i.e. Fyn (5, 29, 30) and ZAP-70 (30), in activated T cells. Furthermore, ZAP-70 causes increased tyrosine phosphorylation of Cbl in an Lck- and Fyn-dependent manner (30), indicating that Cbl may couple ZAP-70 to downstream signaling events during T cell activation.
In order to begin to address the role of Cbl in TCR·CD3-mediated signaling, we evaluated the effects of wild-type or a mutated form (70Z/3) of Cbl on the transcriptional activation of nuclear factor of activated T cells (NFAT), using transient transfection assays in Jurkat T cells. NFAT is a transcription factor complex that plays a critical role in the induction of interleukin 2 (IL2) and other cytokine genes, and the two signal requirement for its activation mimicks the requirement for optimal proliferation and IL2 production in TCR·CD3-activated T cells (31, 32). NFAT-reporter gene activation in transiently transfected T cells has, therefore, been extensively used as a physiologically relevant assay for analyzing molecular events associated with T cell activation. We report that transient overexpression of 70Z/3 (but not wild-type Cbl) induces an increase in the basal activity of NFAT, which is further enhanced synergistically by calcium ionophore. This NFAT induction is dependent on both calcineurin and functional Ras. These results provide evidence that Cbl is involved in Ras-dependent T cell signaling pathways that lead to the transcriptional activation of cytokine genes.
The anti-CD3 monoclonal antibody (mAb), OKT3,
was purified from culture supernatants of the corresponding hybridoma
by protein A-Sepharose affinity chromatography. Polyclonal rabbit
anti-Cbl antibody was from Santa Cruz Biotechnology (Santa Cruz, CA).
The 4G10 anti-Tyr(P) mAb and a rabbit anti-p85 antibody were purchased from Upstate Biotechnology (Lake Placid, NY). Anti-hemagglutinin (HA;
12CA5) or anti-Grb2 mAbs were from Boehringer Mannheim and Transduction
Laboratories (Lexington, KY), respectively. Horseradish peroxidase-conjugated donkey anti-rabbit or sheep anti-mouse IgG F(ab
)2 fragments were from Amersham Corp.
The cDNAs encoding human c-Cbl, v-Cbl, or the
homologue of murine 70Z/3 (representing an internal 17-amino acid
deletion of Cbl) were subcloned from the pGEM-2z vector into a
mammalian expression vector, pEFneo (33) at the EcoRI and
XbaI sites to generate pEFneo-Cbl, pEFneo-v-Cbl, and
pEFneo-70Z/3, respectively. A sequence encoding an HA tag epitope has
been added to the 5 end of the 70Z/3 cDNA by a standard polymerase
chain reaction. A methionine to cysteine mutation at a potential
PI3-K-binding motif in Cbl (Y731EAM) was introduced by
polymerase chain reaction amplification of a cDNA fragment encoding
the Y731EAC mutation and ligation of the product into the
TA cloning vector (InVitrogen). The sequence was verified by nucleotide
sequencing, and the cDNA fragment digested with SnaBI
and XbaI was ligated into pEFneo-70Z/3 which has been
digested with the corresponding enzymes to create the
pEFneo-Y731EAC expression vector. To construct expression
plasmids encoding 70Z/3 proteins with successive deletions of their
114, 176, 346, 366, or 456 C-terminal residues, internal restriction
sites or a point mutation were used to generate plasmids
1 (deletion
of an EcoRV-XbaI fragment),
2 (deletion of a
SnaBI-XbaI fragment),
3 (point mutation of the
proline-560 codon to a stop codon by polymerase chain reaction),
4
(deletion of a BglII-XbaI fragment), and
5
(deletion of a SacI-XbaI fragment), respectively
(see Fig. 5A). These deletions removed distinct proline-rich
sequences of the protein. The deletion mutants were ligated into the
pEFneo vector.
The luciferase reporter plasmids, NFAT-Luc and IL2-Luc, were generous
gifts from Dr. G. Grabtree; AP-1-Luc and NFB-Luc were kindly
provided by Dr. M. Karin. The dominant negative H-RasN17 plasmid in the
pcDNA3 vector was obtained from Dr. H. Wang.
Simian virus 40 T antigen
(TAg)-transfected human leukemic Jurkat T cells (Jurkat-TAg) were grown
in RPMI 1640 medium (Life Technologies, Inc.) supplemented with 10%
fetal bovine serum (Harlan, Indianapolis, IN), L-glutamine,
and antibiotics. Cells in a logarithmic growth phase were transfected
with the indicated amount (usually 10 µg) of Cbl plasmid DNA and 5 µg of the reporter plasmid by electroporation as described previously
(22, 34). The cells were transferred to 24-well plate 24 h later
and cultured in 1 ml of RPMI 1640 medium containing 0.2% fetal calf
serum for 4-6 h. The transfected cells were then left unstimulated or
stimulated with an anti-CD3 mAb (OKT3; 10 µg/ml), ionomycin (100 ng/ml), and/or phorbol myristate acetate (PMA; 10 ng/ml) for another
8-10 h. In some experiments, the cells were treated in triplicate with cyclosporin A (50 ng/ml) for 10 min prior to stimulation. In order to
assess the effect of dominant negative Ras, the cells were transfected
with a mixture of 5 µg of pEFneo-70Z/3, 10 µg of RasN17, and 5 µg
of reporter plasmid DNAs. Similar amounts of the corresponding empty
vectors were used as controls in all transfection experiments.
Transfected Jurkat-TAg cells were
harvested and lysed in 100 µl of lysis buffer (100 mM
KPO4, pH 7.8, 10 mM dithiothreitol, 0.2%
Triton X-100) for 10 min at room temperature. The lysates were then
centrifuged (15,000 × g, 5 min), and their protein
content was determined in an ELISA plate reader (Bio-Rad) using a
Bio-Rad microassay. Aliquots of the lysates (30-50 µl) normalized to
the same protein concentration were mixed with 100 µl of assay buffer (17.5 mM glycyl-glycine, pH 7.8, 10 mM
MgCl2, 5 mM ATP, 0.135 mM coenzyme
A, 0.235 mM luciferin). Luciferase activity was determined in triplicate and expressed as arbitrary units (AU). The standard deviation among triplicates was 10%, and each experiment was repeated at least three times.
Immunoprecipitation, immunoblotting, and PI3-K enzymatic assays were performed as described previously (22, 34).
To explore whether Cbl or 70Z/3 can
transduce signals leading to IL2 gene transcription, we evaluated the
effects of cotransfected Cbl or 70Z/3 cDNA on the transcriptional
activity of an IL2 promoter-reporter plasmid (IL2-Luc), or reporters
corresponding to defined response elements in the IL2 promoter,
i.e., NFAT-Luc, NFB-Luc or AP-1-Luc, in Jurkat-TAg cells.
Cell lysates were assayed for luciferase activity following culture in
low serum-containing medium and stimulation with ionomycin and/or PMA.
As shown in Fig. 1A, when the full-length Cbl
was transfected into unstimulated Jurkat T cells, no increase (or a
slight decrease) in NFAT activity was observed by comparison with the
empty control vector. Similarly, Cbl did not activate NFAT in
ionomycin- or PMA-stimulated cells. In contrast, transient
overexpression of 70Z/3 induced a ~3-4-fold increase of NFAT
activity in unstimulated cells, and stimulation of the
70Z/3-transfected cells with ionomycin which was minimally active by
itself (~2-fold activation), but not with PMA, resulted in a
synergistic effect (~70-fold NFAT activation). Cbl or 70Z/3 did not
cause detectable activation of AP-1, and 70Z/3 overexpression led to
only a slight induction of the IL2 promoter and NF
B under unstimulated or stimulated conditions (data not shown).
To rule out the possibility that the observed difference between Cbl and 70Z/3 was due to different expression levels of the two proteins, we assessed protein expression by immunoblotting cell lysates with an anti-Cbl antibody. As shown in Fig. 1B, transient transfection of Cbl or 70Z/3 increased the intracellular level of a 120-kDa protein which was recognized by anti-Cbl antibody to a similar degree. The 70Z/3 protein could also be detected with an anti-HA antibody, since an HA tag was added to its N terminus (Fig. 1B). Collectively, these results indicate that 70Z/3 is constitutively active in T cells in terms of its ability to induce NFAT-dependent transcription, and synergizes with a Ca2+-dependent signal in this regard. The former observation is consistent with the finding that 70Z/3 can transform fibroblasts (4).
Calcineurin, a calcium-calmodulin dependent phosphatase which plays a critical role in NFAT induction, is a major target for the immunosuppressive effect of cyclosporin A and FK506 (35, 36). To determine whether 70Z/3-induced NFAT activation is dependent on calcineurin, the cells were pretreated with cyclosporin A. As shown in Fig. 1C, cyclosporin A blocked the ionomycin-stimulated increase in NFAT activity in cells transfected with the empty vector, as well as the synergistic effect of 70Z/3 plus ionomycin. This suggests that functional calcineurin is required for the 70Z/3-mediated signal which leads to the transactivation of NFAT.
v-Cbl Fails to Activate NFATCbl was originally isolated as a
product of a transforming viral oncogene (v-cbl). Since both
v-Cbl (1) and 70Z/3 (4) possess transforming activity, we wished to
determine whether v-Cbl shares with 70Z/3 the ability to activate NFAT.
When cells expressing the NFAT-Luc reporter plasmid were cotransfected
with 70Z/3, an increase of 3- and 60-fold in NFAT activity was detected in the absence or presence of ionomycin stimulation, respectively, consistent with the results shown in Fig. 1. In contrast,
cotransfection with the v-Cbl expression vector failed to stimulate the
basal or ionomycin-induced NFAT activity (Fig.
2A). This difference was not due to different
expression levels of 70Z/3 and v-Cbl, since immunoblotting with an
anti-HA antibody revealed similar expression levels of the two
transfected proteins (Fig. 2B). Therefore, despite the fact
that both v-Cbl and 70Z/3 display transforming activity, only 70Z/3 is
capable of activating NFAT, suggesting that transformation by the two
proteins involves different mechanisms.
70Z/3-mediated NFAT Activation Is Independent of PI3-K Association with 70Z/3
TCR·CD3 cross-linking causes rapid tyrosine
phophorylation of Cbl which, in turn, leads to an association of Cbl
with PI3-K in an activation-dependent manner (5, 6, 17,
29). PI3-K has been implicated in TCR·CD3- and or CD28-mediated
signaling events leading to IL2 transcription (37, 38). Thus, we wished to address the possibility that an activation-induced Cbl/PI3-K association plays a role in the observed NFAT activation mediated by
70Z/3. Cbl contains two potential PI3-K-binding motifs (39), i.e., Y371CEM and Y731EAM which may
mediate binding to the SH2 domain of p85 (17). The first of these
motifs is deleted in 70Z/3 (4). To examine whether the
Y731EAM is essential for the interaction between Cbl and
PI3-K and, if so, whether this association is required for the
70Z/3-mediated NFAT activation, we mutated the methionine residue in
this motif of 70Z/3 to cysteine (Y731EAC). A similar
methionine mutation was found to disrupt the association of the CD28
YXXM motif with PI3-K (37). The relative specificity of this
point mutation is indicated by the findings that: first, it did not
have a global effect on the basal or anti-CD3-induced tyrosine
phosphorylation of 70Z/3, as revealed by anti-Tyr(P) immunoblotting of
the corresponding immunoprecipitates (Fig.
3C); second, the mutation did not appear to
have a gross effect on the integrity and/or conformation of the 70Z/3
protein since it could still associate in an activation-independent
manner with Grb2 (Fig. 3D), which is known to constitutively
associate with Cbl (5, 6, 7, 8).
Jurkat-TAg cells were transiently transfected with the 70Z/3 expression vectors (or empty vector), and their association with endogenous PI3-K was assessed by coimmunoprecipitation using the anti-HA tag mAb. The results (Fig. 3A) demonstrate that 70Z/3 coimmunoprecipitated with p85 in unstimulated cells, and that the amount of coprecipitated p85 was markedly increased following anti-CD3 stimulation. The lower association in unstimulated cells most likely reflects the basal tyrosine phosphorylation of Cbl. In contrast, very little association of Y731EAC with PI3-K was observed under the same conditions (panel A), although the expression levels of the two Cbl proteins were equivalent as revealed by anti-HA immunoblotting (Fig. 3B). This result was confirmed by performing an in vitro PI3-K enzymatic assay on the same immunoprecipitates. Thus, anti-HA immunoprecipitates from stimulated, Y731EAC-transfected cells contained a markedly lower PI3-K activity than their counterparts from 70Z/3-transfected cells (Fig. 3E). Together, these results suggest that the Y731EAM motif in 70Z/3 is largely responsible for the interaction with p85.
Next, we determined the effect of the 70Z/3 Y731EAC
mutation on the 70Z/3-mediated activation of NFAT. The mutant plasmid
or 70Z/3 was cotransfected with the NFAT-Luc reporter plasmid into Jurkat-TAg T cells. As showed in Fig. 4A, cell lysates
from Y731EAC-transfected cells showed 1.5 to 2 times higher
luciferase activity both in resting and ionomycin-stimulated cells than
that from 70Z/3-transfected cells. Similar results were consistently observed in a number of repeated experiments. The expression levels of
the transfected plasmids were determined by immunoblotting and no
apparent difference was detected between 70Z/3 and Y731EAC
(Fig. 4B). These data strongly suggest that an optimal
interaction between PI3-K and 70Z/3 is not required for NFAT activity
induced by the latter.
The Effects of C-terminal Deletions on 70Z/3-mediated NFAT Activation-In order to further define the region in 70Z/3 which is responsible for NFAT activation, we focused on the C-terminal region of the protein, since v-Cbl, which corresponds to the 361 N-terminal residues of human c-Cbl, lacked any NFAT-stimulating activity (Fig. 2). The C-terminal region of c-Cbl contains a proline-rich region and several potential tyrosine phosphorylation sites (2). Expression vectors encoding 70Z/3 proteins with successive C-terminal deletions were generated (Fig. 5A) and assessed for their ability to activate NFAT in Jurkat-TAg cells by transient cotransfection assays (Fig. 5B).
As shown earlier (Figs. 1 and 2), 70Z/3 caused a ~3-fold increase in
the basal NFAT activity and stimulated it by ~70-fold in the presence
of ionomycin, whereas v-Cbl was inactive. Similarly, mutation of the
major PI3-K-binding motif (Y731EAC) caused some increase in
the corresponding activity. Deletion of 114 or 176 C-terminal residues
(1 and
2, respectively) reduced the activity by ~50% relative
to the untruncated 70Z/3. These results suggest that the C-terminal 114 amino acids of 70Z/3 contribute to maximal NFAT activation.
Surprisingly, the
3 construct, in which the 346 C-terminal 70Z/3
residues have been deleted, induced higher basal or
ionomycin-stimulated NFAT activity than the untruncated 70Z/3
construct, and deletion of additional 20 (
4) or 110 (
5) residues
essentially abolished this activity (Fig. 5B). The
differences in biological activity among the distinct 70Z/3 constructs
did not reflect differences in their expression levels since
immunoblotting lysates of the transiently transfected cells with an
anti-HA tag antibody revealed similar expression levels of the
corresponding proteins (Fig. 5C).
The
finding that 70Z/3 synergizes with ionomycin to induce NFAT activation
was reminiscent of the effect of constitutively active Ras which can
also cooperate with ionomycin to induce NFAT or IL2 promoter activity
(40, 41, 42, 43). To examine whether NFAT activation by 70Z/3 is dependent on
functional Ras, we cotransfected Jurkat-TAg cells with combinations of
70Z/3 and/or a dominant negative Ras expression vector (or with the
corresponding empty vectors), and assessed the degree of NFAT
activation. The 70Z/3-induced NFAT activation was nearly completely
inhibited by RasN17 in either resting or ionomycin-stimulated cells
(Fig. 6A). The expression level of 70Z/3 was,
however, similar in cells cotransfected with the control (pcDNA3)
or RasN17 expression vector (Fig. 6B). These results suggest
that functional Ras is required for the 70Z/3-mediated activation of
NFAT.
In the present study, we have demonstrated that 70Z/3, a mutated version of the Cbl protooncogene product, but not wild-type Cbl or v-Cbl, functions in a constitutively active manner to induce NFAT activation in Jurkat T cells and, furthermore, that 70Z/3 acts in synergy with calcium ionophore in this regard. Cyclosporin A, which selectively inhibits the function of calcineurin (35, 36), as well as dominant negative Ras, blocked both the basal and the ionomycin-stimulated induction of NFAT by 70Z/3. On the other hand, a mutation that markedly reduced the association of 70Z/3 with PI3-K did not impair the activation of NFAT and, in fact, seemed to increase it. These results suggest that Cbl is involved in the TCR·CD3-mediated signaling pathway(s) which ultimately lead to the activation of NFAT, one of the critical transcription factor complexes involved in induction of IL2 and other cytokine genes (31, 32).
Optimal activation of NFAT and the IL2 promoter requires two signals which can be mimicked by agents that increase the free intracellular calcium concentration and activate the protein kinase C/Ras pathway in T cells, respectively (40, 41, 42, 43). Constitutively active calcineurin, which can replace the calcium signal, or ionomycin, can each synergize with phorbol ester or constitutively active Ras to induce NFAT activation (41, 42). Thus, calcineurin is a major target for the calcium signal involved in NFAT activation and IL2 gene transcription. Our finding that 70Z/3 synergized with ionomycin (but not with PMA) to induce NFAT activation, and that this effect was blocked by cyclosporin A pretreatment (Fig. 1), strongly suggest, therefore, that this constitutively active Cbl can cooperate with calcineurin to provide the requisite signals for NFAT activation. 70Z/3 mimicks in this respect the effects of constitutively active Ras (41, 42). The similarity with Ras also extends to the finding that, like Ras (41, 42), 70Z/3 synergized with ionomycin (Fig. 1), but not with an anti-CD3 mAb (data not shown), to activate NFAT. This connection between Cbl and the Ras signaling pathway is supported by our finding that dominant negative Ras inhibited almost completely the 70Z/3-induced activation of NFAT, indicating that functional Ras is necessary for NFAT induction in the context of 70Z/3. This finding does not allow us to conclude at present whether Ras acts downstream of 70Z/3, or in a parallel pathway which interacts with a Cbl-dependent pathway, to induce NFAT activation.
Recent studies have indicated that, in hematopoietic (5, 6, 17) and other (13) cells, tyrosine-phosphorylated Cbl recruits PI3-K, an interaction that may be mediated by tyrosine-containing YXXM motifs in Cbl and the SH2 domain of p85 (17). Since PI3-K may be involved in signaling pathways leading to IL2 production (37, 38), it was of interest to examine whether PI3-K association with Cbl is required in order for 70Z/3 to activate NFAT. We found that a point mutation of methionine to cysteine in the Y731EAM motif of 70Z/3 reduced the association of 70Z/3 with PI3-K to a minimum, indicating that this motif is largely responsible for the interaction of 70Z/3 with PI3-K. However, this mutation increased the activation of NFAT by 70Z/3, suggesting that activation-dependent recruitment of PI3-K to Cbl is not required for the induction of NFAT.
A striking observation in the present study is that the 70Z/3 3
mutant, from which 346 C-terminal amino acid residues were deleted,
exhibited the highest NFAT-inducing transcriptional activity. Further
deletion of only 20 additional amino acid residues (
4) completely
abrogated basal or ionomycin-stimulated NFAT induction. One potential
explanation is that the region represented by residues 541-560 in
70Z/3 interacts directly or indirectly with an essential effector
involved in NFAT activation, and that removal of more proximal residues
(i.e. residues 561-906) facilitates this interaction, perhaps by abolishing other interactions with a negative regulator(s). Alternatively, differences in biological activity among the various constructs represent conformational changes in the structure of 70Z/3,
in which case
3 may have a more favorable conformation for mediating
the effector function of Cbl. Since the 70Z/3 deletions remove
distinct proline-rich sequences which have been implicated in binding
to SH3 domains of other proteins (5, 6, 7, 8, 44), it is possible that
interactions with SH3-containing signaling proteins are a major target
for the functional effects of these deletions. This possibility is
currently being analyzed.
What is the molecular mechanism by which the constitutively active Cbl mutant, 70Z/3, affects the transcriptional regulation of NFAT? The finding that 70Z/3, but not wild-type Cbl, can mediate this event suggests that amino acid residues 366-383 (which have been deleted in 70Z/3) regulate negatively the function of Cbl or its association with effector proteins, and that their removal unmasks the ability of Cbl to activate NFAT. This is compatible with the findings that the same deletion activates the transforming potential of Cbl (4). The notion of a negative regulatory activity associated with Cbl is supported by the recent finding that a Cbl homologue, Sli-1, negatively regulates vulval development in Caenorhabditis elegans, a process which is known to be initiated by a homologue of the epidermal growth factor receptor via a tyrosine kinase- and Ras-dependent signaling pathway (45). The internal deletion may activate Cbl directly, or indirectly by abolishing an interaction between Cbl and a putative negative regulator that binds to the deleted region. It is possible that this activity can also be unmasked by an undefined physiological signal which leads to Cbl activation during T cell activation. In any event, recent findings that Cbl functions as an adapter protein and associates with many signaling proteins suggest that Cbl-associated proteins may act as intermediate effectors to transduce signals from Cbl which lead to induction of NFAT, and perhaps other transcriptional elements involved in T cell activation. Our findings open the way for further studies to directly investigate the mechanism by which Cbl acts to regulate signaling pathways initiated by immune recognition receptors.