Accumulating evidence indicates that the Cbl protein plays a negative role in immune receptor signaling; however, the mode of Cbl action in B cell receptor (BCR) signaling still remains unclear. DT40 B cells deficient in Cbl showed enhanced BCR-mediated phospholipase C (PLC)-γ2 activation, thereby leading to increased apoptosis. A possible explanation for the involvement of Cbl in PLC-γ2 activation was provided by findings that Cbl interacts via its Src homology 2 (SH2) domain with B cell linker protein (BLNK) after BCR ligation. BLNK is a critical adaptor molecule for PLC-γ2 tyrosine phosphorylation through its binding to the PLC-γ2 SH2 domains. As a consequence of the interaction between Cbl and BLNK, the BCR-induced recruitment of PLC-γ2 to BLNK and the subsequent PLC-γ2 tyrosine phosphorylation were inhibited. Thus, our data suggest that Cbl negatively regulates the PLC-γ2 pathway by inhibiting the association of PLC-γ2 with BLNK.
Cbl is the cellular homologue of the v-Cbl oncoprotein 1, which was originally identified in the murine Cas NS-1 retrovirus 2. v-Cbl induces pre-B cell and myeloid tumors in mice 2, and transforms rodent fibroblasts 3. Recently, Cbl has been identified as a prominent substrate of protein tyrosine kinases (PTKs) that is rapidly phosphorylated after engagement of several different receptors on hematopoietic cells. Cbl has no known catalytic function, but it contains a RING finger domain, an extensive proline-rich region, and a COOH-terminal leucine zipper (for reviews, see references 4,5,6,7,8). Although it has been thought that Cbl contains a novel NH2-terminal phosphotyrosine-binding domain 9, recent crystal structure analyses indicate that the Cbl NH2-terminal region (Cbl-N) is not similar to a phosphotyrosine-binding domain 10,11. Instead, this region comprises three domains: an NH2-terminal four-helix bundle, a calcium-binding domain, and an unusual Src homology 2 (SH2) domain. Moreover, this study clarifies that G306E mutation in Cbl is located inside its unusual SH2 domain, and that this mutation disrupts binding of Cbl to phosphotyrosine-containing proteins 10.
The functional importance of Cbl in hematopoietic cell signaling was initially suggested by the finding that Cbl binds to several critical signaling molecules, e.g., PTKs such as Src family PTK 12,13,14, Syk/Zap-70 family PTK 8,15,16,17,18, and Btk 19; adaptor molecules, including Grb2 13,20, Crk 21,22,23, and Nck 24; and effector molecules, including phosphatidylinositol 3-kinase (PI3-K [16, 25]) and Vav 26. The most revealing clue about the function of Cbl came from genetic studies in Caenorhabditis elegans, where SLI-1, a C. elegans Cbl homologue, negatively regulates signaling downstream of LET-23, a C. elegans epidermal growth factor receptor homologue 27,28. Recent studies have provided evidence that Cbl is involved in regulating PTKs rather than downstream effectors. Indeed, overexpression of Cbl in mast cells suppresses Syk function upon FcεRI engagement 29. This is presumably because of the interaction of the Cbl SH2 domain with the phosphorylation site in human Syk Tyr-323 (equivalent to Tyr-316 in porcine Syk), since either inactivation of the Cbl SH2 domain (G306E mutation) or Syk-Y323F mutation leads to abrogation of Cbl's negative influence on Syk in a heterologous COS cell system 30. Moreover, this study demonstrates the possibility that Cbl targets the autophosphorylated Syk for degradation, leading to downregulation of the expression level of Syk.
Consistent with the negative role of Cbl in Syk function, Cbl-deficient mouse thymocytes exhibited hyperactivation of Zap-70 after CD3 cross-linking 31,32,33. However, contrary to the expectation from the hyperactivated Zap-70 in the Cbl-deficient thymocytes, neither PLC-γ1 nor PI3-K activities were enhanced in these thymocytes 32. Thus, it is still unclear whether the target molecule of Cbl in antigen receptor signaling is solely a PTK such as Syk/Zap-70.
To address this issue, we have used the DT40 B cell system. First, by generating DT40 B cells deficient in Cbl, we have shown that Cbl negatively regulates the B cell receptor (BCR)-induced PLC-γ2 pathway. Given the previous evidence that BCR-mediated PLC-γ2 activation requires its association with BLNK (alternatively named SH2 domain–containing leukocyte protein of 65 kD [SLP-65] or B cell adaptor containing the SH2 domain [BASH]) 34,35,36,37,38, as well as two types of PTKs (Syk and Btk; for reviews, see references 39,40,41,42), two models could be put forward to account for this phenotype. In the first model, Cbl acts on PTKs, particularly Syk, which in turn inhibits PLC-γ2 tyrosine phosphorylation. In the second model, Cbl binds to the phosphorylated BLNK through its SH2 domain, possibly to competitively inhibit the association of the PLC-γ2 SH2 domain with BLNK, leading to inhibition of PLC-γ2 phosphorylation. Our findings suggest that Cbl dominantly utilizes the second mechanism in the BCR signaling context.
Materials And Methods
Cells and Abs.
Wild-type and various mutant chicken DT40 B cells were maintained in RPMI 1640 supplemented with 10% FCS, 1% chicken serum, 50 μM β-mercaptoethanol, 2 mM l-glutamine, penicillin, and streptomycin. Anti–PLC-γ2 Ab 34, anti-Syk Ab 43, anti-BLNK Ab 34, and anti–chicken IgM mAb M4 44 were described previously. Anti-Cbl Ab and anti–glutathione S-transferase (GST) mAb were purchased from Santa Cruz Biotechnology. Anti-T7 mAb and antiphosphotyrosine mAb 4G10 were purchased from Novagen and from Upstate Biotechnology, respectively.
Generation of Cbl-deficient DT40 Cells.
Chicken Cbl cDNA that corresponds to human c-Cbl amino acid residues 77–426 was isolated by reverse transcriptase PCR using RNA from chicken DT40 B cells. Cbl genomic DNA fragments were obtained by screening a chicken genomic library (Clontech Laboratories) using the isolated chicken Cbl cDNA. The targeting vectors pCbl-Bsr, pCbl-HisD, pCbl-Puro, and pCbl-Neo were constructed by replacing the genomic fragment–containing exons that correspond to human c-Cbl amino acids 250–336 with bsr, hisD, puro, or neo cassettes. These cassettes were flanked by 3.4 and 6.6 kb of chicken Cbl genomic sequence on the 5′ and 3′ sides, respectively. The targeting vector pCbl-Bsr was linearized and introduced into wild-type DT40 cells by electroporation at 550 V, 25 μF. Transfectants were selected in the presence of 50 μg/ml blasticidin S, and resistant clones were screened by Southern blot analysis. pCbl-HisD was transfected into the bsr-targeted clone, and was selected with both blasticidin S (50 μg/ml) and histidinol (1 mg/ml). pCbl-Puro was transfected into the bsr/his-targeted clone, and was selected with blasticidin S (50 μg/ml), histidinol (0.5 mg/ml), and puromycin (0.5 μg/ml). pCbl-Neo was transfected into the bsr/his/puro-targeted clone, and was selected with blasticidin S (50 μg/ml), histidinol (0.5 mg/ml), puromycin (0.5 μg/ml), and G418 (2 mg/ml).
DT40 B cells display a stable karyotype with a modal chromosome number of 80, which comprises 11 autosomal macrochromosomes, the ZW sex chromosomes, and 67 microchromosomes. The karyotype does not show obvious abnormalities, except for a trisomy of chromosome 2 and one additional microchromosome 45. The requirement for four rounds of targeting in generating a null mutant suggested that the DT40 B cell has two alleles with a duplicated cbl gene, although the possibility that four copies of a cbl gene are integrated in a single chromosome, such as sex chromosomes, could not be excluded.
Although some critical experiments (see Fig. 3) were carried out using another independent Cbl-deficient clone (C8-12), a single Cbl-deficient clone (C8-10) was analyzed extensively and transfected with c-Cbl cDNAs.
The hemagglutinin-tagged human c-Cbl cDNA (provided by Dr. W.Y. Langdon, University of Western Australia, Nedlands, Australia ) was subcloned into the pAzeo expression vector, which was generated by replacing the CMV promoter on pcDNA3.1/Zeo (Invitrogen) with the chicken actin promoter on pApuro 43. Mutation of Gly-306 to Glu, designated G306E, was introduced into the hemagglutinin-tagged human c-Cbl cDNA by PCR-directed mutagenesis. Transfection into Cbl-deficient DT40 cells was done by electroporation at 550 V, 25 μF. Selection was performed in the presence of zeocin (1 mg/ml). T7-tagged Btk cDNA subcloned into both pApuro and pAhygro was transfected into wild-type or Cbl-deficient DT40 cells, respectively. Mutation of porcine Syk Tyr-316 to Phe (Y316F), which corresponds to human Syk Tyr-323, was produced by PCR-directed mutagenesis. Wild-type and Syk-Y316F cDNAs were subcloned into pApuro, and were transfected into Syk-deficient DT40 cells. Expression of each transfectant was confirmed by Western blot analysis.
Northern Blot Analysis.
RNA was prepared from wild-type and Cbl-deficient DT40 cells using the guanidium thiocyanate method. Total RNA (20 μg) was separated in 1.2% formaldehyde gel, transferred to Hybond™-N+ membrane (Amersham Pharmacia Biotech), and probed with 32P-labeled chicken Cbl cDNA.
Immunoprecipitation, Western Blot Analysis, and In Vitro Kinase Assay.
For immunoprecipitation, cells were solubilized in NP-40 lysis buffer supplemented with protease and phosphatase inhibitors 43, and precleared lysates were sequentially incubated with proper Abs and protein A-agarose. The immunoprecipitates were washed four times with lysis buffer. Whole cell lysates were prepared from unstimulated or M4-stimulated cells using SDS sample buffer. Whole cell lysates or immunoprecipitates were separated by SDS-PAGE, transferred to nitrocellulose membranes, and detected by appropriate Abs and the enhanced chemiluminescence (ECL) system (Amersham Pharmacia Biotech). To examine in vitro kinase activity of Syk, immunoprecipitated Syk was suspended in kinase buffer (20 mM Tris, pH 7.5, 10 mM MgCl2, 10 mM MnCl2) in the presence of 10 μCi of [γ-32P]ATP (>3,000 Ci/mmol). Recombinant GST fusion protein containing a cytoplasmic domain of mouse Igα (5 μg) was used as an exogenous substrate 46. The reactions were allowed to incubate at 30°C for 10 min. For in vitro kinase assay of Btk, immunoprecipitated Btk was suspended in kinase buffer (20 mM Pipes, pH 7.5, 20 mM MnCl2) containing 10 μCi of [γ-32P]ATP (>3,000 Ci/mmol) and 5 μg of enolase (Sigma Chemical Co.), and was incubated at 25°C for 3 min. The kinase reactions were terminated by the addition of SDS sample buffer. The mixtures were separated by SDS-PAGE gel, followed by autoradiography.
Inositol 1,4,5-Trisphosphate Generation Assay.
Cells (2 × 106) were stimulated with mAb M4 (2 μg) at 37°C for the indicated time. Determination of inositol 1,4,5-trisphosphate (IP3) production was performed using the Biotrak IP3 assay system (Amersham Pharmacia Biotech) according to the manufacturer's protocol. Results are shown as the mean from three independent experiments. Error bars represent the SD from the mean.
Cells (5 × 106) were suspended in PBS containing 20 mM Hepes (pH 7.2), 5 mM glucose, 0.025% BSA, and 1 mM CaCl2, and were loaded with 3 μM acetoxymethyl ester of fura-2 (fura-2 AM) at 37°C for 45 min. Cells were washed twice, and were adjusted to 106 cells/ml. Continuous monitoring of fluorescence from the cell suspension was performed using a fluorescence spectrophotometer (model F-2000; Hitachi Instruments) at an excitation wavelength of 340 nm and an emission wavelength of 510 nm. Calibration and calculation of calcium levels were done as described 47.
Flow Cytometric Analysis.
For DNA content analysis, stimulated or unstimulated cells were pelleted and resuspended in hypotonic DNA staining solution (50 μg/ml propidium iodide, 0.1% sodium citrate, 0.1% Triton X-100). Samples were kept at 4°C overnight, and were subjected to analysis by FACScan™ (Becton Dickinson). Debris and doublets were excluded by appropriate gating 48. For cell surface expression of BCR, DT40 cells were washed, incubated with FITC-conjugated anti–chicken IgM (Bethyl Laboratories), and analyzed using FACScan™. The x and y axes for the histograms indicate fluorescence intensity (four-decade log scales) and relative cell number, respectively.
Binding Assay with GST Fusion Proteins.
Human Cbl sequences encoding amino acids 25–351 were amplified by PCR from wild-type or G306E Cbl cDNAs, and were cloned as BamHI-XhoI fragments into the pGEX-4T-1 vector (Amersham Pharmacia Biotech). GST fusion proteins were affinity purified by glutathione-Sepharose beads (Amersham Pharmacia Biotech). For in vitro binding assays, stimulated or unstimulated DT40 cell lysates (1 × 107) were incubated with 20 μg of purified GST fusion proteins for 1 h at 4°C. The binding mixtures were washed extensively in NP-40 lysis buffer. Bound proteins or whole cell lysates were resolved by SDS-PAGE, transferred to nitrocellulose membranes, and detected by anti-BLNK Ab and ECL. For far-Western blotting, anti-BLNK immunoprecipitates were separated on a 7% SDS-PAGE gel and then transferred to polyvinylidene difluoride membranes. The blotted membranes were incubated with 2.5 μg/ml eluted GST fusion proteins, and were detected using anti-GST mAb and ECL.
Cbl Negatively Regulates the PLC-γ2 Pathway.
To disrupt the cbl locus in DT40 B cells, we transfected a targeting construct pCbl-Bsr into parental DT40 B cells. Southern blot analysis indicated that 10 out of 24 blasticidin-resistant clones had sustained a targeting event (data not shown). The hybridization signal of a rearranged 5.4-kb band was two- or threefold weaker than the wild-type 4.9-kb band, suggesting that DT40 cells contain more than two alleles of the cbl locus (Fig. 1 A). Disruption of the remaining alleles of the cbl locus was accompanied by the sequential transfection of three other targeting constructs, pCbl-HisD, pCbl-Puro, and pCbl-Neo (Fig. 1 A). Hybridization with a bsr, hisD, puro, or neo probe indicated that the targeted clone had incorporated a single copy of each construct (data not shown). Lack of Cbl was verified by Northern and Western analyses (Fig. 1B and Fig. C). The level of cell surface expression of BCR on the targeted DT40 cells was the same as that of parental cells (Fig. 2).
One of the hallmarks of BCR-induced signaling is calcium mobilization. Fura-2–loaded wild-type and Cbl-deficient (C8-10) DT40 cells were stimulated with anti-BCR mAb M4, and the rise of intracellular Ca2+ ([Ca2+]i) was measured. As shown in Fig. 3 A, the peak height of the [Ca2+]i increase upon receptor stimulation was enhanced by ∼1.3-fold by loss of Cbl. Because IP3 generation mediated by PLC-γ2 is primarily important for BCR-induced calcium mobilization through its binding to IP3 receptors, BCR-induced IP3 generation in wild-type and Cbl-deficient B cells was examined. Cbl-deficient cells exhibited increased IP3 generation compared with wild-type cells; the peak stimulation after BCR ligation was threefold more than wild-type DT40 cells (Fig. 3 B). We also examined another independent Cbl-deficient clone (C8-12), which indicated essentially the same abnormalities as clone C8-10. Furthermore, transfection of wild-type Cbl cDNA into clone C8-10 restored normal BCR-mediated calcium mobilization (Fig. 3 A), as well as IP3 generation (see Fig. 8 B). Based on these data, we conclude that Cbl negatively regulates PLC-γ2 activation after BCR stimulation.
BCR-mediated Apoptosis Is Enhanced in Cbl-deficient DT40 Cells.
As PLC-γ2 activation is required for BCR-induced apoptosis in DT40 B cells 49, we reasoned that the receptor-induced apoptosis might be augmented by the enhanced PLC-γ2 activation in Cbl-deficient DT40 cells. As shown in Fig. 4, the BCR-induced apoptosis was clearly enhanced by loss of Cbl, demonstrating that Cbl negatively regulates BCR-mediated apoptosis through the PLC-γ2 pathway.
Enhancement of BCR-induced Tyrosine Phosphorylation of PLC-γ2 by Loss of Cbl.
To explore the mechanism by which Cbl mediates its inhibitory effect on PLC-γ2 activation, the BCR-mediated tyrosine phosphorylation status of PLC-γ2 was determined in wild-type and Cbl-deficient cells. Compared with wild-type cells, PLC-γ2 tyrosine phosphorylation in Cbl-deficient DT40 B cells was increased by about three- to fivefold, as assessed by antiphosphotyrosine mAb blotting analysis (Fig. 3 C). These observations indicate that PLC-γ2 hyperactivation in Cbl-deficient DT40 cells is most likely due to the enhanced tyrosine phosphorylation of PLC-γ2 in the BCR signaling context.
As Syk and Btk are the PTKs responsible for BCR-induced PLC-γ2 tyrosine phosphorylation 43,50, one possible explanation for the enhanced PLC-γ2 phosphorylation in Cbl-deficient cells is that Syk and/or Btk are hyperactivated in Cbl-deficient DT40 cells, leading to hyperphosphorylation of PLC-γ2. To evaluate this possibility, the BCR-induced whole tyrosine phosphorylation pattern was compared in wild-type and mutant cells. As shown in Fig. 5 A, there were no significant changes between wild-type and Cbl-deficient cells, except that the band corresponding to Cbl itself was absent in the mutant cells. Consistent with the whole tyrosine phosphorylation data, both Syk and Btk were inducibly tyrosine phosphorylated to an almost similar extent in wild-type and Cbl-deficient DT40 cells. In addition, the BCR-induced in vitro kinase activities of Syk and Btk were almost the same between wild-type and mutant DT40 cells (Fig. 5B and Fig. C).
Previous overexpression experiments using the COS cell system have indicated that a potential target of the Cbl SH2 domain is phosphorylated tyrosine 316 of porcine Syk 30. To formally demonstrate that PLC-γ2 hyperactivation by loss of Cbl cannot be accounted for by the interaction between Cbl and Syk, we determined the effect of the Syk-Y316F mutation on BCR-mediated PLC-γ2 activation. Wild-type Syk and its mutant (Y316F) were transfected into Syk-deficient DT40 cells, and the resulting DT40 clones expressing similar levels of Syk were characterized (Fig. 6 A). As shown in Fig. 6 (B and C), BCR-mediated tyrosine phosphorylation of PLC-γ2 and subsequent IP3 production were only marginally enhanced by this mutation. Together, these results demonstrate that activation of Syk and Btk is not significantly perturbed by loss of Cbl.
Recruitment of PLC-γ2 to BLNK Is Significantly Enhanced in Cbl-deficient B Cells.
The adaptor molecule BLNK is shown to be essential for BCR-mediated tyrosine phosphorylation of PLC-γ2, in that phosphorylated BLNK provides docking sites for PLC-γ2 SH2 domains, leading to recruitment of PLC-γ2 into the close proximity of Btk and Syk 36. Thus, the effect of disruption of Cbl on association of PLC-γ2 to BLNK after BCR ligation was examined. As this association is SH2–phosphotyrosine dependent, we first checked the tyrosine phosphorylation status of BLNK (Fig. 7 A), which indicated the same level between wild-type and Cbl-deficient DT40 cells. Despite the same level of BLNK phosphorylation, BCR-induced association of PLC-γ2 with BLNK was significantly augmented by loss of Cbl (Fig. 7 B). Given that the extent of PLC-γ2 association to BLNK correlates well with the phosphorylation extent of PLC-γ2 36, the enhanced association between PLC-γ2 and BLNK in Cbl-deficient cells is likely to cause hyperphosphorylation of PLC-γ2.
Association of Cbl with BLNK Is Dependent on the Cbl SH2 Domain.
The above observations suggest that Cbl acts through its SH2 domain to competitively inhibit the association between PLC-γ2 and BLNK. To verify this possibility, the effect of the Cbl SH2 mutant (G306E) on BCR-mediated PLC-γ2 activation was determined. Wild-type Cbl and its SH2 mutant (G306E) were transfected into the Cbl-deficient DT40 clone (C8-10), and DT40 clones expressing similar expression levels of Cbl, as assessed by immunoblotting analysis (Fig. 8 A), were selected and further characterized. Wild-type Cbl reverted both PLC-γ2 tyrosine phosphorylation and IP3 generation upon BCR cross-linking, whereas the SH2 mutant could not, demonstrating the importance of the SH2 domain in Cbl's negative influence on PLC-γ2 activation (Fig. 8B and Fig. C). Moreover, recruitment of Cbl to phosphorylated BLNK was dependent on the Cbl SH2 domain (Fig. 8 D).
To further examine the ability of the Cbl SH2 domain to bind to phosphorylated BLNK in vitro, we used GST fusion protein containing the Cbl-N. Binding of GST–Cbl-N to BLNK in lysates of unstimulated DT40 cells was low, but it increased substantially upon BCR stimulation. Importantly, G306E mutation in Cbl-N completely abrogated binding to phosphorylated BLNK (Fig. 9 A). These results establish that the Cbl SH2 domain interacts with phosphorylated BLNK in a lysate binding assay. However, it is possible that this interaction is indirectly mediated by an adaptor molecule. To determine if the Cbl SH2 domain directly binds to phosphorylated BLNK, a far-Western blotting was performed. As seen in Fig. 9 B, GST–Cbl-N directly bound to phosphorylated BLNK; in contrast, no binding was observed with GST or GST–Cbl-N (G306E). These data demonstrate that the Cbl SH2 domain can directly bind to phosphorylated BLNK.
In this study, we have addressed three questions regarding the function of Cbl: (a) does Cbl function as a negative regulator in BCR signaling, as well as in the case of TCR or FcεRI 29,31,33; (b) if so, which signaling pathway(s) is a physiological target of Cbl in the BCR signaling context; and (c) what is the molecular mechanism by which Cbl influences the BCR signaling pathway?
DT40 B cells deficient in Cbl exhibited hyperactivation of PLC-γ2 (Fig. 3) and hyperapoptosis (Fig. 4) in response to BCR ligation, supporting the previous contention that Cbl functions as a negative regulator in immune receptor signaling. Since the previous evidence has shown that BCR-mediated apoptosis in DT40 B cells requires the PLC-γ2 pathway 49, BCR-mediated hyperapoptosis in Cbl-deficient DT40 cells is most likely accounted for by hyperactivation of the PLC-γ2 pathway. In contrast to threefold enhancement of IP3 production in Cbl-deficient cells compared with wild-type DT40, the BCR-mediated calcium mobilization was increased by ∼1.3-fold by loss of Cbl, suggesting that the molecule(s) lying between PLC-γ2 activation and calcium mobilization, such as IP3 receptors, may have a limiting factor for transmission from IP3 generation to calcium mobilization in the DT40 B cell system. Indeed, we observed previously that the expression level of PLC-γ2 in DT40 cells correlated well with BCR-mediated IP3 production, whereas calcium mobilization was not linearly correlated 49.
It has been shown recently that Cbl regulates ubiquitination of receptor-type PTKs, such as platelet-derived growth factor receptor, in a manner dependent on its SH2 and RING finger domains 51,52,53,54,55. Indeed, Cbl has the capacity to act as a component of ubiquitin–protein ligases (E3s ). The expression levels of TCR components in Cbl-deficient thymocytes are higher than in wild-type mice 31,32,33; in contrast, the expression level of BCR in Cbl-deficient DT40 cells is not significantly upregulated (Fig. 2). Thus, one explanation for the functional difference of Cbl in TCR versus BCR signaling may be that targeting signals for degradation by Cbl are comprised in CD3 ζ chain components, but not BCR Igα/Igβ subunits. Cbl-b, another Cbl-related molecule, is expressed in hematopoietic cells, and the NH2-terminal half of Cbl-b, including SH2 and RING finger domains, is highly homologous to Cbl 56. These structural characteristics of Cbl-b suggest that Cbl-b plays a redundant role in ubiquitin-mediated degradation. Thus, it is also possible that relatively high expression of Cbl-b in B cells compared with T cells 57 may compensate for the degradation function of Cbl in B cells.
Given that phosphatidylinositol 3,4,5-trisphosphate, a product of PI3-K, participates in positive modulation of PLC-γ activation 58,59,60, one possible explanation for hyperactivation of PLC-γ2 in Cbl-deficient DT40 cells is that the PI3-K pathway is hyperactivated, leading to enhancement of PLC-γ2 activation. However, the observations that BCR-mediated Akt activation, a readout of the PI3-K pathway, was not affected in Cbl-deficient DT40 cells (data not shown) suggest that this hyperactivation of PLC-γ2 is mediated through a PI3-K–independent mechanism.
Recent experiments have demonstrated that the adaptor molecule BLNK, in addition to Syk and Btk, is required for tyrosine phosphorylation of PLC-γ2 and its subsequent activation 34. Indeed, the phosphorylated BLNK by Syk provides docking sites for PLC-γ2 SH2 domains and its subsequent phosphorylation 36,61. Thus, one possible explanation for the hyperphosphorylation of PLC-γ2 by loss of Cbl is that Cbl negatively regulates Syk function, as proposed previously in mast cell and COS cell systems 29,30. However, this possibility is unlikely in the DT40 B cell system based on the following data: (a) overall tyrosine phosphorylation was almost the same between wild-type and Cbl-deficient DT40 cells; (b) a Syk substrate, BLNK, was normally phosphorylated in Cbl-deficient DT40 cells; and (c) although the Cbl SH2 domain was essential for exerting Cbl's negative effect on PLC-γ2 phosphorylation, Syk-Y316F (disruption mutant for binding to the Cbl SH2 domain) did not exhibit significant enhancement of PLC-γ2 phosphorylation, as observed in Cbl-deficient cells. BCR-mediated Btk activation, as in the case of Syk, occurred normally in Cbl-deficient DT40 B cells. The present results differ from previous studies using a transfected rat basophilic leukemia mast cell system, which suggested that Cbl directly inhibits the enzymatic activity of Syk 29. It is possible that when Syk and Cbl proteins are overexpressed using the vaccinia virus system in rat basophilic leukemia cells, SH2-independent binding of Cbl to Syk could be promoted, thereby leading to a direct negative impact on Syk enzymatic activity. Another potential mechanism, enhanced degradation of Syk by Cbl, was proposed based on overexpression experiments in COS cells 30. The degradation effect of Cbl on Syk may be dependent on the expression level of the activated Syk; overexpressed Syk could be more susceptible to degradation. Thus, it is possible that the effect of Cbl on the endogenous expression level of Syk in DT40 cells may be too subtle to be detected by our assay system.
Despite the same levels of BLNK tyrosine phosphorylation in wild-type and Cbl-deficient DT40 cells, association of PLC-γ2 with BLNK was increased by about threefold in Cbl-deficient cells, indicating that Cbl negatively modulates the PLC-γ2–BLNK association. Since the Cbl SH2 domain was also recruited to phosphorylated BLNK after BCR ligation (Fig. 8 D and Fig. 9), it would be reasonable to anticipate that the Cbl SH2 domain and the PLC-γ2 SH2 domains are competitively recruited to the phosphorylated BLNK. This idea is further supported by the reported preferred sequences of these SH2 domains. A phosphopeptide library screen with the Cbl NH2-terminal region demonstrated that the binding motif for this region could be D(N/D)XpY 9. Indeed, chicken BLNK comprises this binding motif (DDSY115), and this motif is well conserved in both human and mouse BLNK 34. Interestingly, the sequences that match well the motif selected by the COOH-terminal SH2 domain of PLC-γ2 (Y103VVP and Y194IVP) are found near this Cbl binding motif 62. Thus, our biochemical data, together with previous evidence that the extent of BLNK–PLC-γ2 association correlates with the level of BCR-mediated PLC-γ2 tyrosine phosphorylation 36, suggest that enhanced BLNK–PLC-γ2 association by loss of Cbl competition causes increased PLC-γ2 phosphorylation, leading to its hyperactivation. Additional studies are underway to define exactly the in vivo phosphorylation sites on BLNK for binding to the PLC-γ2 SH2 domains and the Cbl SH2 domain, and to examine the functional consequences of phosphorylation on these sites. The importance of the Cbl SH2 domain in TCR signaling has been suggested by recent results that TCR-mediated nuclear factor of activated T cells modulation by oncogenic Cbl mutant 70Z was abolished by its SH2 mutation 63,64. As nuclear factor of activated T cells activation is dependent on both PKC activation and calcium mobilization, this notion is further supported by our findings that the Cbl SH2 domain is essential for its influence on BCR-mediated PLC-γ2 activation. Although a D(N/D) XpY motif in Zap-70 (Y292) and Syk (Y316) binds to the Cbl SH2 domain 9,10,30,65, recent studies indicate that this interaction cannot fully account for the effect of the Cbl SH2 domain in immune cells 64,66,67. Thus, our identification here of BLNK as a target of the Cbl SH2 domain provides another potential mode of Cbl action.
In contrast to the thymic T cell changes in Cbl-deficient mice, the B cell development in bone marrow appears to be normal, suggesting differential requirements for Cbl in T and B cell development 31. Developmental changes in T cells might reflect the fact that Cbl downregulates the expression levels of TCR components. In support of this idea, the BCR expression level in Cbl-deficient mice seems to be normal 31. Despite no apparent B cell developmental changes, the peripheral lymph nodes of Cbl-deficient mice demonstrated increases in both B and T cells 31. This observation suggests that Cbl functions as a negative regulator in peripheral B and T cells. Thus, our finding that Cbl is a negative regulator in the BCR-mediated PLC-γ2 pathway may reflect the status of murine peripheral B cells rather than developing B cells. Comparative studies using Cbl-deficient mice will be needed to test this hypothesis.
We would like to thank Dr. W.Y. Langdon for providing human c-Cbl cDNA, Dr. T. Adachi for construction of Syk mutant, Dr. M. Ishiai for critical discussions and for reading our manuscript, Dr. H. Umemori for discussions, and P. Winding for his helpful reading of our manuscript.
This work was supported by grants to T. Yamamoto and T. Kurosaki from the Ministry of Education, Science, Sports, and Culture of Japan; to T. Kurosaki from the Toray Science Foundation, Takeda Science Foundation, Hoh-ansha Foundation, and the Human Frontier Science Program; and to A. Maeda from the Novartis Foundation (Japan) for the Promotion of Science.
Abbreviations used in this paper: BCR, B cell receptor; BLNK, B cell linker protein; [Ca2+]i, intracellular Ca2+ concentration; Cbl-N, NH2-terminal region of Cbl; ECL, enhanced chemiluminescence; GST, glutathione S-transferase; IP3, inositol 1,4,5-trisphosphate; PI3-K, phosphatidylinositol 3-kinase; PLC, phospholipase C; PTK, protein tyrosine kinase; SH2, Src homology 2.