Negative regulation of FcϵRI-mediated mast cell activation by a ubiquitin-protein ligase Cbl-b

Mast cells play a central role in inflammatory and immediate allergic reactions. The aggregation of FcϵRI triggers the activation of Lyn protein-tyrosine kinase (PTK), resulting in the rapid tyrosine phosphorylation of its β and γ subunits.¹  We isolated nonreceptor type PTK Syk.^2,3  Tyrosine phosphorylated γ subunit of FcϵRI then recruits and activates Syk, the essential PTK for antigen-induced Ca⁺⁺ influx, degranulation, and cytokine production in mast cells.^3-9  Syk phosphorylates an adaptor protein linker for activation of T cells (LAT) which is located in the lipid raft (also referred to as the glycolipid-enriched microdomains [GEMs]) to accumulate the signaling molecules and is essential for FcϵRImediated mast cell activation.^6,7,10  In contrast to this Lyn-Syk-LAT pathway, recent findings by using the genetic approach have revealed the existence of another complementary signaling pathway. Expression of the adaptor/scaffold protein Gab2 is necessary for FcϵRI-mediated activation of phosphatidylinositol 3-kinase (PI3-kinase) and downstream mast cell activation.¹¹  Src family PTK Fyn is required for tyrosine phosphorylation of Gab2, independent of Lyn and LAT.¹²  Both of these 2 pathways are important for the FcϵRI-induced degranulation and cytokine production.

Cbl-b, a second member of the Cbl-family of E3 ubiquitin-protein ligase, was originally isolated from the human breast cancer cells and is expressed in a variety of healthy tissues and cancer cells.¹³  It consists of an amino-terminal variant Src homology domain 2 (SH2) domain, a RING finger domain, a proline-rich region, a carboxyl-terminal leucine-zipper domain, and potential tyrosine phosphorylation sites. Genetic studies revealed that c-Cbl and Cbl-b have the distinct and overlapped functions in T-cell regulations.¹⁴  Although c-Cbl regulates thymocyte development and cell surface T-cell receptor (TCR) expression, Cbl-b^–/– mice display a normal development of thymocytes and impaired peripheral T-cell activation mediated by CD28.^15,16  Cbl-b targets PI3-kinase by ubiquitination in T cells.^17,18  In addition, lack of Cbl-b results in the enhanced tyrosine phosphorylation and guanosine diphosphate/guanosine triphosphate (GDP/GTP) exchanging activity of Vav1.^15,16  Therefore, Cbl-b might indirectly regulate the activation of Vav1 through phosphatidylinositol 3-phosphate, the product of PI3-kinase.¹⁹  Analysis of B cells from Cbl-b–deficient mice showed that lack of Cbl-b results in the sustained phosphorylation of Igα, Syk, phospholipase C-γ2 (PLC-γ2), and prolonged Ca⁺⁺ mobilization.²⁰  However, Cbl-b–deficient avian pro-B cells display reduced PLC-γ2 activation and Ca⁺⁺ mobilization, suggesting that Cbl-b may play a different role in immature and mature B cells.²¹  In mast cells, c-Cbl catalyzes FcϵRI-mediated ubiquitination of FcϵRI and Syk, resulting in an inhibition of the downstream inflammatory mediator release.^22-24  However, the function of Cbl-b in mast cells has not been demonstrated yet.

The lipid raft constitutes a membrane domain that concentrates sphingolipids, cholesterol, glycophosphatidylinositol-linked proteins, and numerous signaling molecules.²⁵  Engagement of the immune receptors induces the accumulation of the signaling molecules into the lipid raft for both the effective receptor signals to the downstream and the degradation of the activated/tyrosine phosphorylated signaling molecules by ubiquitination.²⁶  Engagement of FcϵRI induces the translocation of a ubiquitin-protein ligase c-Cbl into the lipid raft.²⁷  Our present experiment shows that Cbl-b translocates into the lipid raft after the FcϵRI engagement. Targeting of the ubiquitin-protein ligase Cbl-b by the localization signal to the lipid raft might emphasize the activity of Cbl-b to have a dominant-active function.

The present study demonstrated that the overexpression of Cbl-b in the lipid raft inhibits FcϵRI-mediated tyrosine phosphorylation of FcϵRI, Syk, Gab2, and, therefore, the downstream degranulation and cytokine gene transcription, suggesting that Cbl-b functions as a negative regulator of both Lyn-Syk-LAT and complementary signaling pathways from FcϵRI. In particular, overexpression of Cbl-b in the lipid raft dramatically down-regulates the protein amount of Gab2. This result gives us a new insight to develop gene therapy methods to treat immune and allergic diseases by using Cbl-b.

Protein A-agarose beads, mouse monoclonal anti-dinitrophenyl IgE monoclonal antibody (mAb; anti-DNP IgE, clone SPE-7), anti-Flag mAb (M2), Ca⁺⁺ ionophore A23187, thapsigargin, and PMA (phorbol 12-myristate 13-acetate) were purchased from Sigma (St Louis, MO). Protein-G–Sepharose beads were from Amersham (Piscataway, NJ). Antiphosphotyrosine (pTyr) mAb (4G10), anti-JNK/SAPK1 (c-Jun N-terminal kinase/stress-activated protein kinase 1) antibody, anti–PLC-γ1 mAb, and anti-Gab2 antibody were purchased from Upstate Biotechnology (Lake Placid, NY). Anti–Cbl-b, anti-Lyn, anti–PLC-γ2, anti-Syk, anti-p38, and anti-IKK (inhibitor of nuclear factor κB kinase) antibodies were from Santa Cruz Biotechnology (Santa Cruz, CA). Antihemagglutinin epitope (HA) mAb was from Covance (Princeton, NJ). Antiphospho-p44/42 ERK (extracellular signal regulated kinase) (Thr202/Tyr204), anti-p44/42 ERK, antiphospho-Akt (Ser473), and anti-Akt antibodies were from New England Biolabs (Beverly, MA). Antiphospho-SAPK/JNK (Thr183/Tyr185), antiphospho-specific p38 MAP (mitogen-activated protein) kinase (Thr180/Tyr182), and antiphospho-IKKα (Ser180)/IKKβ (Ser181) antibodies were obtained from Cell Signaling Technology (Beverly, MA). Antigen 2, 4-dinitrophenylated bovine serum albumin (DNP-BSA; 30 mol DNP/1 mol BSA) was from LSL (Tokyo, Japan). Anti-FcϵRIβ mAbs were kindly provided by Dr Juan Rivera²⁸  and Dr Reuben P. Siraganian (National Institutes of Health, Bethesda, MD).

The human Cbl-b cDNA was kindly provided by Dr Stanley Lipkowitz (National Naval Medical Center, Bethesda, MD). The cDNA encoding Cbl-b with the membrane localization signal was constructed by the polymerase chain reaction (PCR)–based method.^9,29  We generated a chimeric molecule of Cbl-b with the myristoylation and palmitoylation signals. The 16 amino acids of c-Src N-terminus region with a Ser3 to Cys mutation [GEM-tag, MGCNKSKPKDASQRRR] was fused to the second amino acid in the N-terminal of human Cbl-b (GEM–Cbl-b). The cDNA of GEM–Cbl-b was then used to generate its mutant form which could not function as ubiquitin-protein ligase by a point mutation with a Cys373 to Ala (C373A) using the Site-Directed Mutagenesis kit (Stratagene, La Jolla, CA).³⁰  All mutations were confirmed by DNA sequencing. The cDNA of Flag-tagged ubiquitin (Flag-Ub) was a gift from Dr Keiji Tanaka (Tokyo Metropolitan Institute of Medical Science, Japan). All cDNAs were then subcloned into the pSVL expression vector (Amersham). The schematic diagram of Cbl-b mutants used in this study was shown in Figure 2A.

Rat basophilic leukemia RBL-2H3 cells were maintained as monolayer cultures in Dulbecco modified Eagle medium (DMEM) (Sigma), 100 U/mL penicillin, and 10% heat-inactivated fetal calf serum. For stable transfection, each 20 μg linearized expression construct and 2 μg pSV2-neo vector were cotransfected into 5 × 10⁶ RBL-2H3 cells by electroporation (950 μF, 310 V).³¹  Stably transfected cell lines were selected by 0.4 mg/mL active G418 (Invitrogen, Carlsbad, CA). Cell lines were screened by the level of protein expression by the immunoblotting of total lysates with anti-HA and anti–Cbl-b antibodies. The immunoblotting with anti-FcϵRIβ mAb was used as an internal control. RBL-2H3 cells stably expressing Flag-Ub were used for immunoprecipitation study to demonstrate ubiquitination of Gab2.

The cell monolayers cultured overnight with anti-DNP IgE (1:5000) were washed once with Tyrode-HEPES (2-(4-(2-hydroxyethyl)-1-piperazinyl)ethanesulfonic acid) buffer (10 mM HEPES, pH 7.4, 127 mM NaCl, 4 mM KCl, 0.5 mM KH₂PO₄, 1 mM CaCl₂, 0.6 mM MgCl₂, 10 mM LiCl₂, 5.6 mM glucose, and 0.1% BSA) and then stimulated with 30 ng/mL antigen DNP-BSA in the same buffer for the indicated times.³²  For the immunoprecipitation studies, cells were washed twice with ice-cold phosphate-buffered saline (PBS) and solubilized in 1% Triton lysis buffer (1% Triton X-100, 50 mM Tris (tris(hydroxymethyl)aminomethane), pH 7.4, 150 mM NaCl, 10 mM EDTA (ethylenediaminetetraacetic acid), 100 mM NaF, 1 mM Na₃VO₄, 1 mM phenylmethylsulfonyl fluoride, and 90 mU/mL aprotinin) on ice. For immunoprecipitation with anti–Cbl-b, anti-pTyr or anti-Gab2 antibody, cells were solubilized in the denature buffer (1% Triton X-100 lysis buffer containing 0.1% sodium dodecylsulfate [SDS] and 0.5% deoxycholic acid) to dissociate protein complexes. Cell lysates were precleared by centrifugation, and then resultant supernatants were incubated with the indicated antibodies prebound to protein A-agarose or protein G-Sepharose beads. After rotation for 1 hour at 4°C, the beads were washed 4 times with lysis buffer, and the immunoprecipitated proteins were eluted by heat treatment at 100°C for 5 minutes with 2 × sampling buffer. For the preparation of total cell lysates, monolayers were rinsed as described earlier with PBS and lysed by the direct addition of 2 × sampling buffer.

Total cell lysates, immunoprecipitated proteins, and subcellular fractions were separated by SDS-polyacrylamide gel electrophoresis (SDS-PAGE) and electronically transferred to polyvinylidene diflouride transfer membrane (Millipore, Bedford, MA). After blocking with 5% skim milk in TBST (10 mM Tris, pH 7.4, 150 mM NaCl, and 0.1% Tween20), the blots were probed with the indicated primary antibodies for 1 hour at 4°C. After washing with TBST, the membrane was reacted with the appropriate horseradish peroxidase–conjugated secondary antibodies for 30 minutes at room temperature. After extensive washing with TBST, proteins were visualized by the enhanced chemiluminescence reagent (Western Lightning; PerkinElmer Life Sciences, Boston, MA). Densitometric analysis of immunoblotting was performed by using National Institutes of Health (NIH) Image software. Relative amounts of protein-tyrosine phosphorylation which compared to the maximum phosphorylation in the control cells were normalized by protein expression and indicated at the bottom of the figures. In some experiments, relative amounts of protein expression were analyzed.

The low-density detergent-insoluble fractions were prepared by sucrose density gradient centrifugation essentially as described.^33,34  The proteins were concentrated with 0.02% deoxycholic acid and 10% trichloroacetic acid. For the immunoprecipitation study, 1 to 3 fractions, 4 to 6 fractions, and 7 to 10 fractions were collected together.

Degranulation of different cell lines was determined by the measurement of β-hexosaminidase release. Cells (10⁵) were seeded in 24-well plates and cultured overnight with or without anti-DNP IgE. The cell monolayers were washed once with Tyrode-HEPES buffer and then stimulated with different concentrations of the antigen DNP-BSA or Ca⁺⁺ ionophore A23187 in the same buffer. After incubation for 1 hour at 37°C, the medium was recovered for the analysis of β-hexosaminidase activity. Total cell lysates were obtained by the addition of 1% NP-40 in the same medium. The medium and total cell lysates were incubated with 1.3 mg/mL P-nitrophenyl-N-acetyl-β-d-glucopyranoside (Nacalai, Osaka, Japan) in 0.1 M sodium citrate buffer (pH 4.5) for 30 minutes at 37°C. The reaction was terminated by the addition of 0.2 M glycine buffer (pH 10.7). The release of the product 4-P-nitrophenol was monitored by the absorbance at 405 nm by using a Microplate reader (Model 550; Bio-Rad, Hercules, CA). The antigen-induced β-hexosaminidase release was expressed as a percentage of the maximal release induced by A23187.³¹ 

Intracellular free Ca⁺⁺ concentration ([Ca⁺⁺]_i) was measured by means of fluorescent indicator fura-2 (Dijindo, Kumamoto, Japan) as described previously.³¹  The cell monolayers of the different cell lines sensitized with anti-DNP IgE were detached from the tissue-culture dish and then loaded with 3 μM fura-2 for 30 minutes at 37°C. Cells were stimulated by either 30 ng/mL antigen DNP-BSA or 1 μM thapsigargin. Fura-2 fluorescence was monitored with the fluorescence spectrophotometer (Hitachi F-4500, Tokyo, Japan) by the excitation wavelength at 340, 380 nm and the emission wavelength at 510 nm.

The total RNA was purified by RNeasy Mini Kit (Qiagen, Hilden, Germany). The production of the cytokine mRNA was quantitatively measured by using Multi-probe RNase Protection Assay Kit (BD Biosciences, San Jose, CA).³⁵  Briefly, by Riboquant In Vitro Transcription Kit (BD Biosciences), ³²P-labeled RNA probes were synthesized by using Rat Cytokine Multi-Probe Template Sets (BD Biosciences). The synthesized probes were purified by using the Spin+STE 10 Column (BD Biosciences) and then hybridized overnight at 56°C with 20 μg total RNA purified from cells with or without either the antigen or PMA plus calcium ionophore stimulation. After the RNase treatment, the protected double-stranded RNAs were separated by the urea gel and visualized by autoradiography.

Either unstimulated or antigen-stimulated RBL-2H3 cells or cells transfected with the different cDNAs were solubilized in Triton lysis buffer. Anti-Lyn immunoprecipitates were resuspended in the kinase buffer without adenosine triphosphate (ATP), then incubated with 40 μL kinase buffer (40 mM HEPES, pH 7.5, 10 mM MgCl₂, 4 μM ATP, 4 μCi [0.148 MBq] γ-³²P] ATP) and 2.5 μg acid-treated enolase (Sigma) at 30°C for 5 minutes. Reactions were terminated by heat treatment at 100°C for 5 minutes in the sample buffer, and eluted proteins were separated by 10% SDS-PAGE. The gels were incubated with 1 N KOH for 1 hour at 56°C to remove phosphoserine and most of the phosphothreonine. After the gel fixing and drying, radiolabeled proteins were visualized by autoradiography. Immunoprecipitation of Lyn was confirmed by immunoblotting.^36,37 

Engagement of FcϵRI results in the translocation and concentration of many signaling molecules, including c-Cbl, into the lipid raft.²⁷  In RBL-2H3 mast cell line, 2 kinds of Cbl family of ubiquitin ligase, c-Cbl and Cbl-b, are expressed.³⁷  Therefore, we first examined whether there is a similar change in Cbl-b association with the lipid raft or not (Figure 1). The result shows that the majority of Cbl-b localized in detergent-soluble fractions (Figure 1, upper panels). After FcϵRI engagement, Cbl-b protein was shifted into the lower-density fractions and was detectable in the lipid raft fractions (Figure 1, fractions 2 and 3). Densitometric analysis revealed that 3% of total Cbl-b was translocated into the lipid raft (Figure 1A). Recently, we have demonstrated that antigen stimulation causes the inducible association of Cbl-b with Lyn, which is known to be localized in the lipid raft.³⁷  Both Lyn and FcϵRIβ were colocalized in the lipid raft fraction prior to antigen stimulation (Figure 1, middle and bottom panels). These results demonstrated that antigen stimulation induces the translocation of Cbl-b into the lipid raft in RBL-2H3 mast cells (Figure 1).

To investigate the role of Cbl-b in the lipid raft during mast cell activation, we generated stable cell lines overexpressing the mutant forms of Cbl-b which constitutively localize in the lipid raft. The myristoylation and palmitoylation of cytoplasmic proteins by addition of 16 amino acids from the N-terminus of c-Src in which Ser3 is substituted to Cys (GEM-tag) can direct molecules into the lipid raft.^9,29  This GEM-tag was fused to the N-terminus of wild-type Cbl-b (Figure 2A).³⁰  Expression constructs of GEM-Cbl-b or GEM-Cbl-b (C373A) which loses the ability as a ubiquitin-protein ligase were stably transfected into RBL-2H3 cells and cloned lines were screened by immunoblotting (Figure 2B). For further analysis, 2 or 3 cloned lines transfected with each kind of cDNA were selected in which the level of expression of the mutant form of Cbl-b was highest among the clones. The molecular weight of GEM-Cbl-b (C373A) seems to be higher than that of GEM-Cbl-b (Figure 2B). The intracellular localization of chimeric protein was further analyzed by the subcellular fractionation study. The GEM-tagged Cbl-b was mainly localized in the lipid raft (Figure 2C). Densitometric analysis demonstrated that 55% of GEM-Cbl-b was constitutively localized in the fraction of the lipid raft (Figure 2C).

To examine the function of Cbl-b in the lipid raft, we examined the FcϵRI-induced β-hexosaminidase release by using the stable cloned lines overexpressing GEM-Cbl-b and GEM-Cbl-b (C373A) (Figure 2B). The cell lines were stimulated by antigen or Ca⁺⁺ ionophore A23187. Antigen-induced β-hexosaminidase release was suppressed in GEM-Cbl-b–overexpressing cells compared with that in the control cells (Figure 3A). However, overexpression of GEM-Cbl-b (C373A) relatively enhanced the antigen-induced β-hexosaminidase release, suggesting that GEM-Cbl-b (C373A) has a dominant-negative function for the endogenous Cbl-b (Figure 3A). The A23187-induced release as a percentage of the total β-hexosaminidase activity was comparable among all these cell lines (Figure 3A, legend). In addition, thapsigargin-induced release was similar among these clones (data not shown). Similar results were obtained when the other cloned lines overexpressing the mutant form of Cbl-b were tested. These results demonstrate that overexpression of Cbl-b in the lipid raft results in a suppression of FcϵRI-mediated degranulation, depending on its ubiquitin-protein ligase activity.

Because antigen-induced increase in [Ca⁺⁺]_i is critical for mast cell degranulation, we compared Ca⁺⁺ mobilization. The parental control cells and cells overexpressing GEM-Cbl-b were stimulated by either antigen or thapsigargin (Figure 3B). As shown, the response to the antigen stimulation was suppressed in the cells overexpressing GEM-Cbl-b compared with that in control cells. Overexpression of GEM-Cbl-b (C373A) resulted in an increase in Ca²⁺ mobilization (Figure 3B). In contrast, there was similar response to thapsigargin among the clones, indicating that overexpression of GEM-Cbl-b or GEM-Cbl-b (C373A) has no effects on Ca⁺⁺ release from the intracellular storage. Thus, the overexpression of Cbl-b in the lipid raft suppresses the signals upstream of the elevation of [Ca⁺⁺]_i. Then, we focused on PLC-γ which hydrolyzes phosphatidylinositol to generate diacylglycerol and inositol 1,4,5-trisphosphate (IP₃) to release Ca⁺⁺ ion from the internal stores, resulting in a Ca⁺⁺ influx.

Mast cells express 2 kinds of PLC-γ, PLC-γ1 and PLC-γ2. Compared with the control cells, FcϵRI-induced tyrosine phosphorylation of PLC-γ1 and PLC-γ2 was reduced in the cells overexpressing GEM-Cbl-b (Figure 3C-D). Overexpression of GEM-Cbl-b (C373A) lightly enhanced the antigen-induced tyrosine phosphorylation of PLC-γ1 and PLC-γ2 (Figure 3C-D). Therefore, these results indicate that overexpression of Cbl-b in the lipid raft down-regulates FcϵRI-induced activation of PLC-γ1 and PLC-γ2, and thereby the subsequent Ca⁺⁺ mobilization and degranulation. The ubiquitin-protein ligase activity is necessary to regulate this pathway.

Because overexpression of Cbl-b in the lipid raft suppresses the antigen-induced immediate degranulation, we next examined the transcription of cytokine genes by the RNase protection assay (Figure 4A). Compared with the control cells, there was significant suppression of the antigen-induced increase in the mRNA of interleukin-3 (IL-3), IL-4, IL-6, IL-10, and tumor necrosis factor α (TNFα) in the cells overexpressing GEM-Cbl-b (Figure 4A). Unlike degranulation, expression of GEM-Cbl-b (C373A) could not restore the defect of cytokine gene transcription. The production of control housekeeping genes (L32 and GAPDH) was comparable among the control cells and cells overexpressing GEM-Cbl-b or GEM-Cbl-b (C373A) (Figure 4A). Moreover, PMA plus calcium ionophore–induced production of cytokine genes (0 minute versus 60 minutes induced by PMA plus calcium ionophore) were identical among the clones (Figure 4A). These results suggest that the endogenous Cbl-b plays a negative regulatory role in FcϵRI-mediated transcriptional activation of cytokine genes, independent of its ubiquitin-protein ligase activity. Perhaps the other domains in Cbl-b contribute to this suppression through the passive adaptor function of Cbl-b.

Then we tested the effects on the activation of the upstream MAP kinases. Overexpression of Cbl-b in the lipid raft suppressed the antigen-induced phosphorylation of JNK (Figure 4B) and extracellular signal regulated kinase (ERK) (Figure 4C). Suppression of JNK was more remarkable than ERK. However, overexpression of GEM-Cbl-b (C373A) could result in the suppression of neither JNK nor ERK, suggesting that Cbl-b–mediated negative regulation of JNK and ERK is RING finger domain dependent. In contrast, phosphorylation of p38 MAP kinase displayed similar patterns among the different cell lines (Figure 4D). In addition to MAK kinases, we tested IKK which is known to regulate nuclear factor κB (NF-κB) (Figure 4E). Similar to MAP kinases, activation of IKK was down-regulated by the overexpression of GEM-Cbl-b but not by GEM-Cbl-b (C373A). These results suggest that the suppression of multiple cytokine gene transcriptions by Cbl-b is regulated by both RING finger–dependent and –independent specific mechanisms.

To date, several proximal molecules have been identified which possess a critical function in the activation of mast cells induced by the antigen stimulation.³⁸  The sequential activation of 2 types of protein-tyrosine kinases Lyn and Syk is critical for mast cell activation. Therefore, we tested the effect of the overexpression of Cbl-b in the lipid raft on the antigen-induced activation of Lyn and Syk. Antigen-induced tyrosine phosphorylation of FcϵRIβ and γ subunits was dramatically suppressed in the cells overexpressing GEM-Cbl-b, compared with that in the control cells (Figure 5A). In parallel to this result, the enzymatic activity of Lyn was suppressed by the overexpression of GEM-Cbl-b (Figure 5B). Similar results were obtained for tyrosine phosphorylation of Syk (Figure 5C). The inhibition of the activation of Lyn and Syk were abrogated by a point mutation in the RING domain of Cbl-b (GEM-Cbl-b [C373A]). The expressing protein amounts of FcϵRI, Lyn, and Syk were not affected by the overexpression of Cbl-b in the lipid raft (Figure 5, lower panels). Therefore, overexpression of Cbl-b down-regulates the kinase activity of Lyn, tyrosine phosphorylation of FcϵRIβ and γ subunits which cause the suppression of FcϵRI signals.

Earlier findings revealed that the adaptor/scaffold protein Gab2 is critical for the complementary signaling pathway by regulating PI3-kinase in mast cells.¹¹  FcϵRI-mediated tyrosine phosphorylation of Gab2 was dramatically suppressed, and the expressing amount of Gab2 protein was also markedly decreased in the cells overexpressing GEM-Cbl-b (Figure 6A). Overexpression of GEM-Cbl-b (C373A) resulted in an increase in tyrosine phosphorylation of Gab2 (Figure 6A). The mobility shift of Gab2 is lightly enhanced by the expression of GEM-Cbl-b (C373A). This result suggests that overexpression of Cbl-b in the lipid raft may affect the stability of Gab2, depending on the activity of ubiquitin-protein ligase. In fact, antigen stimulation induces the multi-ubiquitination of Gab2 in RBL-2H3 mast cells (Figure 6B). The protein amount of Gab2 was decreased after the antigen stimulation. Therefore, overexpression of Cbl-b in the lipid raft might directly affect the protein amount of Gab2 by ubiquitin-protein ligation and degradation of Gab2.

We have demonstrated that overexpression of Cbl-b in the lipid raft suppresses antigen-induced degranulation and cytokine gene transcription (Figures 3A and 4A). Among the signaling molecules, overexpression of Cbl-b in the lipid raft dramatically down-regulates the protein amount of Gab2 and, therefore, the antigen-induced tyrosine phosphorylation of Gab2 (Figure 6A). Although Cbl-b possesses multiple domains and tyrosine phosphorylation sites, the effect on Gab2 is mediated by the RING finger domain because overexpression of GEM-Cbl-b (C373A) could not suppress the expression level of Gab2, relatively enhanced its antigen-induced tyrosine phosphorylation and downstream degranulation (Figures 3A and 6A). Our results suggest that Cbl-b has a function to down-regulate the adaptor/scaffold protein Gab2 by ubiquitination in the lipid raft (Figure 6). Genetic analysis revealed that Gab2 is essential for FcϵRI-mediated activation of PI3-kinase.¹¹  Therefore, Cbl-b may negatively regulate the complementary signaling pathway from FcϵRI in mast cells.¹²  Furthermore, overexpression of Cbl-b in the lipid raft strongly suppresses tyrosine phosphorylation of FcϵRIβ, γ subunits, and Syk (Figure 5). Previous results have demonstrated that c-Cbl catalyzes ubiquitination of TCR and down-regulates TCR-CD3 complex in thymocytes, whereas Cbl-b promotes ubiquitination of p85 of PI3-kinase and inhibits the signals by CD28 costimulation in peripheral T cells.¹⁴  In B cells, it is reported that Cbl-b negatively regulates B-cell antigen receptor–mediated signaling through ubiquitination of Syk.²⁰  All together, these findings suggest that endogenous Cbl-b could inhibit both Lyn-Syk-LAT and Gab2-mediated complementary signaling pathways from FcϵRI in mast cells. Here we proposed that Gab2 is a possible direct target of Cbl-b in FcϵRI signaling.

Although tyrosine phosphorylation of Gab2 was completely abolished, phosphorylation of Akt, which is also the downstream of PI3-kinase is not affected (data not shown).¹¹  This defect might be explained by the fact that RBL-2H3 cells are factor independent because of a point mutation in c-Kit (Asp817 to Tyr).³⁹  A positive regulatory signal from the constitutive active form of c-Kit may be involved in the normal activation of Akt.⁴⁰ 

Mast cells express 2 isoforms of PLC-γ, PLC-γ1 and PLC-γ2. Antigen stimulation induces the translocation of PLC-γ1 into the lipid raft, whereas PLC-γ2 distributes in the plasma membrane and Golgi region and does not redistribute appreciably after FcϵRI engagement.⁴¹  Antigen-induced translocation and activation of PLC-γ1, but not PLC-γ2, is sensitive to the PI3-kinase inhibitor, wortmannin.⁴¹  Moreover, antigen-induced tyrosine phosphorylation of PLC-γ2, not PLC-γ1, requires the activation of Vav1.⁴²  Overexpression of Cbl-b in the lipid raft suppresses tyrosine phosphorylation of Vav1 (data not shown). A genetic study revealed that both Vav1 and Slp-76, which associates with Vav1, play critical roles in FcϵRI-mediated mast cell activation.^42,43  Our results indicate that overexpression of Cbl-b in the lipid raft affects FcϵRI-mediated activation of Vav1. Therefore, although the antigen-induced activation of PLC-γ1 and PLC-γ2 is mediated by the distinct pathways, endogenous Cbl-b could negatively regulate the activation of both PLC-γ1 and PLC-γ2 (Figure 3C-D).

In immune receptor signaling, adaptor and molecular scaffold proteins lack the enzymatic and transcriptional domains, but they have multiple motifs and domains to create protein complexes to integrate signals from cell surface receptors.⁴⁴  In addition to the passive adaptor function, these molecules possess the active regulatory function in immune receptor signaling. We have reported that an adaptor protein 3BP2 is a putative ligand of SH2/SH3 domains of Src family PTK Lyn to unclamp the closed-inactive conformation and stimulate the enzymatic activation of Lyn.^36,45  In this model, an adaptor protein 3BP2 acts as an active regulator of nonreceptor type of PTK. Moreover, point mutations in the 3bp2 gene are correlated with human inherited disease cherubism, suggesting that 3BP2 may have a specific role in the transcriptional regulation in osteoclasts.⁴⁶  In the current study, overexpression of Cbl-b in the lipid raft results in the suppression of the protein amount of Gab2 prior to stimulation, suggesting that Gab2 might be synthesized and degraded in the rapid turnover in resting cells (Figure 6A). This finding leads to the idea that Cbl-b has a potential active function in nonstimulated mast cells. Our result also demonstrates that RING finger–independent, passive adaptor function may be involved in the regulation of antigen-induced cytokine gene transcription (Figure 4A). Because FcϵRI-mediated activation of JNK, ERK, and IKK depends on the function of RING-finger of Cbl-b, the additional pathways for activating the transcriptional factor might be involved in the regulation of cytokine gene transcription (Figure 4). In T cells, it is expected that Ser/Thr kinases might be involved in controlling TCR-mediated transcriptional activation of the NF-κB p65 subunit.^47,48  A possible role of p65 phosphorylation is to serve as a binding site for transcriptional co-activator such as cyclic adenosine monophosphate (cAMP) response element-binding protein/p300.⁴⁷  It is possible that an adaptor function of Cbl-b down-regulates the transcription of cytokine genes through the co-activator of transcriptional factor in mast cells.

This present study suggests that endogenous Cbl-b negatively regulates FcϵRI, Lyn, PLC-γ1, PLC-γ2, Gab2, and subsequent mast cell activation. Overexpression of the novel chimeric protein GEM-Cbl-b has a strong inhibitory role on allergic reaction. Here, we propose that a ubiquitin-protein ligase Cbl-b is a target molecule to develop novel anti-allergic drugs and gene therapy for the immune diseases.

We thank Dr Reuben P. Siraganian, Dr Stanley Lipkowitz, Dr Keiji Tanaka, and Dr Juan Rivera for providing the reagents. We also thank Dr Keiko Kawauchi-Kamata and Dr Zhi-Hui Xie for the instructions.