RI Ε 2 Requires Syk to Transmit Signals from Fc Scaffolding Adapter Grb2-Associated Binder

Scaffolding Adapter Grb2-Associated Binder 2 Requires Syk to Transmit Signals from Fc ε RI

This information is current as Min Yu, Cliff A. Lowell, Benjamin G. Neel and Haihua Gu of October 2, 2021. J Immunol 2006; 176:2421-2429; ; doi: 10.4049/jimmunol.176.4.2421 http://www.jimmunol.org/content/176/4/2421 Downloaded from References This article cites 49 articles, 26 of which you can access for free at: http://www.jimmunol.org/content/176/4/2421.full#ref-list-1

Why The JI? Submit online. http://www.jimmunol.org/

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

*average by guest on October 2, 2021 Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2006 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

Scaffolding Adapter Grb2-Associated Binder 2 Requires Syk to Transmit Signals from Fc⑀RI1

Min Yu,* Cliff A. Lowell,† Benjamin G. Neel,* and Haihua Gu2*

Scaffolding adapter Grb2-associated binder 2 (Gab2) is a key component of Fc⑀RI signaling in mast cells, required for the activation of PI3K. To understand how Gab2 is activated in Fc⑀RI signaling, we asked which protein tyrosine kinase is required for Gab2 phosphorylation. We found that Gab2 tyrosyl phosphorylation requires Lyn and Syk. In agreement with published results, we found that Fyn also contributes to Gab2 tyrosyl phosphorylation. However, Syk activation is defective in Fyn؊/؊ mast cells, suggesting that Syk is the proximal kinase responsible for Gab2 tyrosyl phosphorylation. Then, we asked which domains in Gab2 are required for Gab2 tyrosyl phosphorylation. We found that the Grb2-Src homology 3 (SH3) binding sites are required for, whereas the pleckstrin homology (PH) domain contributes to, Gab2 tyrosyl phosphorylation. Using a protein/lipid overlay assay, we determined that the Gab2 PH domain preferentially binds the PI3K lipid products, PI3, 4,5P3 and PI3, 4P2. Further- more, the Grb2-SH3 binding sites and PH domain binding to PI3K lipid products are required for Gab2 function in Fc⑀RI-evoked Downloaded from degranulation and Akt activation. Our data strongly suggest a model for Gab2 action in Fc⑀RI signaling. The Grb2 SH3 binding sites play a critical role in bringing Gab2 to Fc⑀RI, whereupon Gab2 becomes tyrosyl-phosphorylated in a Syk-dependent fashion. Phosphorylated Gab2 results in recruitment and activation of PI3K, whose lipid products bind the PH domain of Gab2 and acts in positive feedback loop for sustained PI3K recruitment and phosphatidylinositol-3,4,5-trisphosphate production, required for Fc⑀RI-evoked degranulation of mast cells. The Journal of Immunology, 2006, 176: 2421–2429. http://www.jimmunol.org/ ast cells are the major effector cells for the allergic whereas phosphorylated ITAMs in the ␥-chain recruit and activate response. Fc⑀RI, the high-affinity receptor for IgE pre- another key protein tyrosine kinase (PTK) Syk, which ultimately M sents on mast cells and basophils, binds the Fc portion triggers various mast cell responses (1). A recent report showed of IgE. Fc⑀RI becomes activated when IgE prebound to Fc⑀RI is that another SFK, Fyn, also is important for Fc⑀RI-initiated sig- cross-linked by the binding of multivalent Ag. Activated Fc⑀RI naling and biological responses (2). Subsequent to these initial triggers the release of performed granules (degranulation) and lip- signaling events, critical downstream signaling molecules, includ- id-derived mediators that evoke the immediate hypersensitivity re- ing PI3K, become activated, leading to various mast cell re- action. Fc⑀RI cross-linking also results in activation of gene ex- sponses. Although much is known about initial PTK activation by guest on October 2, 2021 pression for various cytokines and chemokines that can further following Fc⑀RI cross-linking, it is still not well-understood how enhance the initial hypersensitivity response. this results in the activation of various mast cell responses. Fc⑀RI consists of one ligand-binding ␣-chain and three associ- Recent studies indicate that adapter proteins including linker for ated subunits (one ␤-chain and two ␥-chains). The ␤- and ␥-chains activation of T cells (LAT) (3), Src homology 2 domain-containing contain ITAMs without any intrinsic enzymatic activity. It is gen- leukocyte protein of 76 kDa (4), Grb2-associated binder (Gab) 2 erally accepted that upon Fc⑀RI cross-linking, Fc⑀RI-associated (5), and non-T cell activation linker/linker for activation of B cells Src family protein tyrosine kinase (SFK)3 Lyn becomes activated (6, 7) play key roles in mediating Fc⑀RI-evoked responses in mast and phosphorylates the ␤- and ␥-chain ITAMs. Phosphorylated cells. Gab2 belongs to the Gab2/daughter of sevenless family of ITAMs in the ␤-chain recruits and activates additional Lyn scaffolding adapters that include mammalian Gab1, Gab2, and Gab3, Drosophila daughter of sevenless, and Caenorhabditis el- egans Soc-1 (8, 9). Like its relatives, Gab2 contains an N-terminal *Cancer Biology Program, Division of Hematology/Oncology, Department of Med- pleckstrin homology (PH) domain, several proline-rich motifs, and icine, Beth Israel Deaconess Medical Center and Harvard Medical School, Boston, MA 02115; and †Department of Laboratory Medicine, University of California, San multiple tyrosyl phosphorylation sites. The PH domain is a mod- Francisco, CA 94143 ular domain known to bind inositol phospholipids (10). The Gab1 Received for publication September 29, 2005. Accepted for publication November PH domain preferentially binds phosphatidylinositol-3,4,5- 28, 2005. trisphosphate (PIP3) (11, 12), and is essential for Gab1 function in The costs of publication of this article were defrayed in part by the payment of page branching morphogenesis (12). The Gab2 PH domain is critical for charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. recruiting Gab2 to the phagocytic cup in macrophages (13), but 1 This work was supported by National Institutes of Health Grants R01-AI51612 (to dispensable for Gab2 function in IL-3/GM-CSF signaling (14). H.G) and R01-DK50693 (to B.G.N). H.G. is a recipient of the Junior Faculty Scholar Although the lipid-binding specificity of Gab2 PH has not been Award from the American Association of Hematology. determined, Gab2 PH domain recruitment to the phagocytic cup is 2 Address correspondence and reprint requests to Dr. Haihua Gu, Harvard Medical sensitive to pretreatment with wortmannin, a specific inhibitor of School, New Research Building, NRB 1030N, 77 Avenue Louis Pasteur, Boston, MA 02115. E-mail address: [email protected] PI3K, suggesting that Gab2 PH may also bind PI3K lipid products (13). Two of the proline-rich motifs in Gab2 are binding sites for 3 Abbreviations used in this paper: SFK, Src family protein tyrosine kinase; PTK, protein tyrosine kinase; LAT, linker for activation of T cells; Gab, Grb2-associated the SH3 of Grb2 (15, 16). Grb2 binding to Gab2 plays a key role binder; PH, pleckstrin homology; PIP3, phosphatidylinositol-3,4,5-trisphosphate; SH, in recruiting Gab2 to its upstream receptors. In IL-3/GM-CSF sig- Src homology; PTB, phosphotyrosine binding; ␤c, ␤ common; BMMC, bone marrow-derived mast cell; HA, hemagglutinin; WT, wild type; NEAA, nonessential naling, the Gab2/Grb2 complex via Grb2 Src homology 2 (SH2) amino acid; SCF, stem cell factor; PI, phosphatidylinositol. domain interacts with tyrosyl-phosphorylated Shc. Shc via its

Copyright © 2006 by The American Association of Immunologists, Inc. 0022-1767/06/$02.00 2422 Gab2 TRANSMIT SIGNALS FROM Fc⑀RI IN MAST CELLS phosphotyrosine-binding (PTB) domain interacts with the tyrosyl- purchased from The Jackson Laboratory. All the mice work was followed phosphorylated IL-3 ␤ common (␤c). Therefore, Gab2 is recruited to a protocol approved by the Harvard Medical Area Standing Committee to the ␤c via the Shc/Grb2 complex (14). Gab2/Grb2 complex via on Animals. BMMC were cultured as described previously (5). Briefly, bone marrow cells from 2- to 4-mo-old mice were incubated in IMDM with the Grb2 SH2 domain is recruited to tyrosyl-phosphorylated BCR- 10% heat-inactivated FBS (HyClone), 2 mM L-glutamine, 0.1 mM nones- Abelson tyrosine kinase and plays key role in mediating BCR- sential amino acids (NEAA), 1 mM sodium pyruvate, 1000 U/ml penicillin, Abelson tyrosine kinase transformation (17). 1000 ␮g/ml streptomycin, 50 ␮M 2-ME, 4 ng/ml recombinant murine IL-3 Ϫ/Ϫ Ϫ/Ϫ Upon receptor activation, Gab2 becomes tyrosyl-phosphory- (BioSource International). For Fyn mice or Lyn mice and WT lit- termates, 4 ng/ml IL-3 and 20 ng/ml stem cell factor (SCF) were used to lated and associated with SHP-2, a tyrosine phosphatase, required ϩ ϩ Ϫ Ϫ Ϫ Ϫ obtain BMMC from littermate WT / and Fyn / or Lyn / mice as for optimal activation of Erk (18, 19) and immediate early gene described (2). After 5 wk, these cultures consist of ϳ95% mast cells re- transcription (20). Tyrosyl-phosphorylated Gab2 also binds p85, flected by surface expression of Fc⑀RI and c-Kit. Five- to 8-wk-old BMMC Ϫ Ϫ the regulatory subunit of PI3K, resulting in activation of the PI3K cultures were used for biochemical and degranulation assays. Syk / BMMC were generated from bone marrow chimeric mice (C57BL/6J) pathway (14, 21). Ϫ/Ϫ ⑀ transplanted with Syk fetal liver cells as described (28). Three months Previously, we found that Gab2 is required for Fc RI-triggered after transplant, chimeric mice were killed and the bone marrows were used responses including degranulation and cytokine gene expression in to obtain SykϪ/Ϫ BMMC as described above. Three-month-old WT bone marrow-derived mast cells (BMMC) (5). Gab2 is also re- C57BL/6J mice were used to generate Sykϩ/ϩ BMMC. The MC/9 murine quired for mast cell development in specific tissues mostly likely mast cell line, purchased from the American Type Culture Collection, was via signaling through c-Kit (5, 22). One key roles of Gab2 in the passaged in RPMI 1640 supplemented with 10% FBS, 5% T-Stim supple- ⑀ ment (BD Biosciences), 0.1 mM NEAA, 2 mM L-glutamine, 0.1 mM Fc RI-initiated response is to activate the PI3K pathway (5), which NEAA, 1 mM sodium pyruvate, 1000 U/ml penicillin, 1000 ␮g/ml strep- is required for mast cell degranulation (23). Interestingly, Gab2 tomycin, and 50 ␮M 2-ME. The anti-DNP IgE hybridoma cell line (106/ has also been found to be more important for Fc⑀RI-evoked che- ml), a gift from Dr. F.-T. Liu (University of California–Davis, Sacramento, Downloaded from mokine gene expression (24). A recent report also suggests that CA) (29), was grown in PFHM-II protein-free hybridoma medium (In- vitrogen Life Technologies) for 4–5 days. The resultant supernatant that Gab2 regulates degranulation by activating small GTPase RhoA typically contains 10–20 ␮g/ml anti-DNP-IgE was used to sensitize mast and microtubule polymerization (25). Consistent with the impor- cells at a 1/10–20 dilution overnight. For cell stimulation, sensitized tant role of Gab2 in Fc⑀RI-signaling responses, Gab2 is found to BMMC were washed, starved in IMDM ϩ 1% BSA for 3–4 h at 37°C, and be in the same specialized membrane domain (osmiophilic resuspended in modified Tyrode’s buffer (135 mM NaCl, 5 mM KCl, 1 mM patches) with Fc⑀RI␤ upon Fc⑀RI cross-linking (26). However, it MgCl, 1.8 mM CaCl2, 10 mM HEPES (pH 7.4), 5.6 mM glucose, and 0.1% http://www.jimmunol.org/ ⑀ BSA) before stimulation with indicated concentrations of DNP is still not clear how Gab2 is recruited to the activated Fc RI and (Sigma-Aldrich). becomes activated to participate in the Fc⑀RI-initiated degranulation response. Furthermore, the PTK required for Gab2 tyrosyl phosphorylation remains controversial. One group reported that Abs, reagents, immunoprecipitation, and Western blots Ϫ/Ϫ Gab2 tyrosyl phosphorylation is dramatically increased in Lyn Anti-Gab2 Abs were generated as described (20). Anti-HA mAb (12CA5) BMMC (2); another group claimed that Gab2 phosphorylation re- was described previously (14). Anti-murine Fc⑀RI ␤-chain mAb was pro- mains normal in LynϪ/Ϫ BMMC (27). Although Fyn is reportedly vided by Dr. J. Rivera (National Institutes of Health, Bethesda, MD). Rab- required for Gab2 phosphorylation (2), it is not clear whether Gab2 bit anti-Syk Abs were provided by Dr. J. Cambier (National Jewish Med- by guest on October 2, 2021 is a direct substrate of Fyn. In this study, we investigated how ical and Research Center, Denver, CO) or purchased from Cell Signaling Technology. Anti-Grb2, Akt1/2, Lyn, SHP-2 rabbit Abs, and anti-GST Gab2 becomes tyrosyl-phosphorylated and which PTKs are re- mAb were purchased from Santa Cruz Biotechnology. Anti-phospho-Akt sponsible for Gab2 tyrosyl phosphorylation. (Ser473)- and -Syk Abs were purchased from Cell Signaling Technology. Monoclonal anti-phosphotyrosine Ab (4G10) was obtained from Upstate Materials and Methods Biotechnology. Anti-Shc rabbit Abs were purchased from BD Transduc- tion Laboratories. The Syk inhibitor piceatannol was obtained from Cal- Constructs biochem. Cells were lysed in 1% Nonidet-P40 lysis buffer as described The Gab2R32C mutant cDNA was generated by PCR using the pBluescript previously (5). Total cell lysates or immunoprecipitates were resolved by (Stratagene) HA-Gab2 as template and the following primers: 1) 5Ј-A CCA SDS-PAGE, immunoblotted with indicated primary Abs, followed by GCA TTT CTT CCA CGC GT A GCG-3Ј is the primer encoding the HRP-conjugated anti-rabbit or anti-mouse IgG (Amersham Biosciences), R3C 32 mutation; 2) 5Ј-ATC GGA TCC CCG AAT ATG AGC GGC-3Ј and developed by ECL (Amersham Biosciences). Bands in Western blots encoding the 5Ј end of Gab2 cDNA; and 3) 5Ј-GCT CTA GAA CTA GTG were quantified by densitometry analysis using NIH Image 1.63F software. GAT CCA-3Ј: encoding the 3Ј end of Gab2 and the hemagglutinin (HA) The intensities of phospho-Akt, phospho-Gab2, and phospho-Syk were epitope Tag. The PCR product was ligated into pBluescript and verified by normalized to the corresponding total Akt, Gab2, and Syk signal, DNA sequencing. The HA-Gab2 wild-type (WT) and R32C fragments re- respectively. leased by restriction enzyme digestion from pBluescript and the HA- Gab2⌬Grb2 fragment released from a pEBB plasmid (17) were cloned into the retroviral vector pMXs-puro (a gift from T. Kitomura, Tokyo Univer- Flow cytometry analysis (FACS) sity, Tokyo, Japan) or MSCV-IRES-GFP (13). To generate the pGex-4T-3 Gab2PH WT and Gab2 PH R32C constructs, Gab2 PH WT (1–120 aa; BMMCs were prebound with anti-DNP IgE, washed with PBS containing WT) and PH R32C cDNAs were amplified by PCR with the following 3% FBS, and stained with FITC-anti-mouse IgE rat mAb R35-72 (IgG1) or primers from pBluescript HA-Gab2WT and Gab2R32C, respectively: 5Ј- FITC-rat monoclonal IgG1 as control. Stained cells were analyzed using a ATC GGA TCC CCG AAT ATG AGC GGC-3Ј and 5Ј-TCA CTC GAG FACScan (BD Biosciences). BMMC were stained with rat mAb FITC- TTA GAA GCC GCA GAT CTG GCA G-3Ј. The PCR fragments were anti-CD117 or FITC-isotype control Ab to measure surface expression of cloned into BamHI- and XhoI-digested pGex-4T-3. The GST-Shc-SH2 and c-Kit. All these FITC-conjugated Abs were purchased from BD GST-Shc-PTB domain constructs were provided by Dr. K. S. Ravichand- Pharmingen. ran (University of Virginia Medical School, Charlottesville, VA). The GST-Grp1 PH plasmid is a gift from Dr. S Field (Beth Israel Deaconess Medical Center, Boston, MA). Mast cell degranulation assays ϫ 5 Mice and cell culture BMMC (5 10 ) were sensitized with anti-DNP IgE supernatants overnight, washed, resuspended in modified Tyrode’s buffer, stimulated with 10 Gab2Ϫ/Ϫ mice (129sv/J ϫ C57BL/6J mixed background) were generated ng/ml DNP for 7 min, and centrifuged (4°C; 300 ϫ g, 5 min). Supernatants as described previously (5). LynϪ/Ϫ mice (129sv/J ϫ C57BL/6J mixed were assayed for ␤-hexosaminidase activity as described (30). Degranula- background) were provided by Dr. T. Yamamoto (University of Tokyo, tion is quantified as the percent of ␤-hexosaminidase released into the Tokyo, Japan). FynϪ/Ϫ mice (C57BL/6J ϫ 129S mixed background) were medium as a fraction of the total cellular ␤-hexosaminidase. The Journal of Immunology 2423

Retroviral infection of BMMC and MC9 tyrosyl phosphorylation. Interestingly, Fc⑀RI cross-linking still in- ϩ/ϩ Ϫ/Ϫ PMXs-puro and MSCV-IRES-GFP retroviral plasmids were transfected duced similar mobility shift of Gab2 in Lyn and Lyn into the ecotrophic packaging cell line PlateE (31) (provided by T. Kita- BMMC (Fig. 1B). In response to various stimuli, Gab2 also shows mura, Tokyo University, Tokyo, Japan) using Fugene reagent (Roche). a decreased in mobility shift in SDS-PAGE, which is mainly due Virus-containing culture supernatants were collected 2 days later. Bone to serine and threonine phosphorylation (14, 34, 35). This result marrow cells cultured in IL-3-containing IMDM medium for 12 days were suggests that loss of Lyn does not affect serine and threonine phos- spin-infected (2500 rpm, 90 min) with pMXs-puro virus supernatants in the presence of 4 ␮g/ml polybrene, and then incubated at 37°C for 18–24 h. phorylation (Gab2). Infected cells were selected in the presence of 0.8 ␮g/ml puromycin for 10–14 days, and then cultured in the absence of puromycin for 4–5 wk. For Syk is required for Gab2 phosphorylation infection of MC9 with MSCV-IRES-GFP viruses, MC9 cells were resus- Because Lyn is required for Syk activation in response to Fc⑀RI pended in fresh medium and incubated with an equal volume of virus supernatant in the presence of 4 ␮g/ml polybrene at 37°C overnight. One activation, we asked whether Gab2 tyrosyl phosphorylation also week later, GFP-positive infected cells were purified by FACS. depends on Syk. We first examined Gab2 tyrosyl phosphorylation in WT BMMC pretreated with piceatannol, a pharmacological in- Protein/lipid overlay assay hibitor of Syk. Piceatannol pretreatment completely eliminated Bacterial-expressing GST fusion proteins were purified as described (32). Fc⑀RI-evoked Syk and Gab2 tyrosyl phosphorylation, and Gab2 GST-fusion proteins purified by glutathione agarose beads (Sigma-Al- mobility shift, respectively (Fig. 2A). To further support a role of drich) were washed, eluted with an equal volume of 20 mM glutathione Syk in Gab2 tyrosyl phosphorylation, we assessed Gab2 tyrosyl (Sigma-Aldrich), resolved by SDS-PAGE, and quantified by staining with ϩ/ϩ Ϫ/Ϫ Coomassie brilliant blue. phosphorylation in Syk and Syk BMMC. FACS analysis ϩ/ϩ Ϫ/Ϫ Phosphatidylinositol (PI) lipids were purchased from Echelon Bio- indicated that Syk and Syk BMMC express similar levels Downloaded from sciences, reconstituted in chloroform:methanol:water (1:2:0.8), and spotted of cell surface IgE receptor and c-Kit (Fig. 2B), suggesting that onto HyBond C extra membranes (Amersham Biosciences) at 100, 50, 25, loss of Syk has no effect on mast cell differentiation (data not and 12.5 pmol/spot, respectively. Membranes were air-dried, blocked in TBST (150 mM NaCl, 10 mM Tris-Cl (pH 8.0), 0.05% Tween 20) ϩ 3% shown). However, in agreement with the effect of piceatannol, fat-free BSA (Sigma-Aldrich)) for 1 h, incubated with 1 ␮g/ml purified Fc⑀RI-evoked Gab2 tyrosyl phosphorylation and mobility shift Ϫ Ϫ GST-fusion proteins in TBST ϩ 3% fat-free BSA overnight, and then was completely blocked in Syk / BMMC (Fig. 2C). We also blotted with anti-GST mAb. found that Syk becomes coimmunoprecipitated with Gab2 in WT BMMC upon Fc⑀RI cross-linking (Fig. 2D). Gab2 association with http://www.jimmunol.org/ Results Syk has also been reported when BaF3 cells were stimulated Lyn is required for Gab2 phosphorylation through integrin cross-linking (36). In contrast, SCF-stimulated Ϫ Ϫ We first asked which PTKs are required for Gab2 tyrosyl phos- Gab2 tyrosyl phosphorylation occurred normally in Syk / ϩ ϩ phorylation. It is generally believed that Lyn is the initial tyrosine BMMC compared with Syk / BMMC (data not shown). Collec- kinase activated upon Fc⑀RI cross-linking (1). However, whether tively, these data suggest that Syk, rather a SFK, is the proximal Lyn is required for Gab2 tyrosyl phosphorylation remains contro- PTK required specifically for Fc⑀RI-evoked Gab2 tyrosyl and versial. Flow cytometry analysis showed that Lynϩ/ϩ and LynϪ/Ϫ serine/threonine phosphorylation.

BMMC have similar surface expression of IgE receptor and c-Kit by guest on October 2, 2021 Decreased Fc⑀RI-evoked Gab2 tyrosyl phosphorylation (Fig. 1A). However, while Gab2 was robustly tyrosyl-phosphory- Ϫ/Ϫ lated in Lynϩ/ϩ BMMC, it was hardly tyrosyl-phosphorylated correlates with decreased Syk activation in Fyn BMMC upon Fc⑀RI cross-linking in LynϪ/Ϫ BMMC (Fig. 1B). We also A recent report suggested that Fyn is the kinase that responsible for found, in agreement with published reports (33), that Syk activa- Fc⑀RI-evoked Gab2 tyrosyl phosphorylation and activation of tion is severely inhibited in LynϪ/Ϫ BMMC (data not shown). PI3K (2). Therefore, we examined Fc⑀RI evoked Gab2 tyrosyl These data suggest that Lyn and/or Syk are required for Gab2 phosphorylation and Akt activation in Fynϩ/ϩ and FynϪ/Ϫ

FIGURE 1. Fc⑀RI-evoked tyrosyl phosphorylation of Gab2 is impaired in LynϪ/Ϫ BMMC. A, Lynϩ/ϩ and LynϪ/Ϫ BMMC express similar surface IgE receptor and c-Kit. BMMC from Lynϩ/ϩ and LynϪ/Ϫ mice were analyzed for cell surface IgE binding (top) and c-Kit (bottom) by FACS as described in Materials and Meth- ods. B,Fc⑀RI-evoked Gab2 tyrosyl phosphorylation is impaired in LynϪ/Ϫ BMMC. Lynϩ/ϩ and LynϪ/Ϫ BMMC were sensitized by anti-DNP-IgE, starved, and stimulated with 10 ng/ml DNP for the indicated times. Lysates were immunoprecipitated with anti-Gab2 serum, immunoblotted with anti-phosphorylation tyrosine Ab (pTyr), and reprobed with anti-Gab2 Abs. Data shown here represent one of the three independent experiments with similar results. 2424 Gab2 TRANSMIT SIGNALS FROM Fc⑀RI IN MAST CELLS

FIGURE 2. Syk is required for Fc⑀RI- evoked tyrosyl phosphorylation of Gab2. A, Syk inhibitor blocks Fc⑀RI-evoked Gab2 tyrosyl phosphorylation. DNP-IgE-sensitized WT BMMC were pretreated with DMSO (Ϫ)or100 ␮g/ml Syk inhibitor piceatannol (ϩ)for1h,and stimulated with 10 ng/ml DNP for 5 min. Ly- sates were immunoprecipitated with anti-Gab2 or Syk serum, immunoblotted with anti-pTyr, and reprobed with anti-Gab2 and Syk Abs, respectively. B, Sykϩ/ϩ and SykϪ/Ϫ BMMC express similar IgE receptor and c-Kit. Sykϩ/ϩ and SykϪ/Ϫ BMMC were analyzed for cell surface IgE binding (top) and c-Kit (bottom)by FACS as described in Materials and Methods. C,Fc⑀RI-evoked Gab2 tyrosyl phosphorylation is completely inhibited in SykϪ/Ϫ BMMC. BMMC were sensitized by anti-DNP-IgE, starved, and stimulated with 10 ng/ml DNP for the indicated times. Gab2 tyrosyl phosphorylation was analyzed as in Fig. 1B. D, Syk becomes Downloaded from associated with Gab2 upon Fc⑀RI cross-linking. WT BMMC were sensitized with anti-DNP-IgE, starved, and stimulated with 10 ng/ml DNP for 1 min. Lysates were immunoprecipitated with anti- Gab2 serum, immunoblotted with anti-pTyr, and reprobed with anti-Gab2 or Syk Abs. Data shown here represent one of the three independent exper- http://www.jimmunol.org/ iments with similar results.

BMMC. In agreement with the previous report, loss of Fyn did not whether the Shc SH2 domain or the PTB domain interacts with the affect surface expression of IgE receptor and c-Kit in BMMC (Fig. tyrosyl-phosphorylated Fc⑀RI ␤-chain. To address this question, 3A). In addition, Fc⑀RI-evoked degranulation (data not shown) and we incubated purified GST-fusion proteins containing either the Akt phosphorylation (Fig. 3B) were decreased in FynϪ/Ϫ BMMC Shc SH2 domain or PTB with lysates from unstimulated or Fc⑀RI- compared with Fynϩ/ϩ BMMC as reported (2). Gab2 tyrosyl phos- activated mast cells. Only the GST-Shc SH2 domain was able to by guest on October 2, 2021 phorylation was also reduced in FynϪ/Ϫ BMMC (Fig. 3C), but not immunoprecipitate the Fc⑀RI ␤-chain, as validated by anti- completely inhibited as we found in SykϪ/Ϫ BMMC. Importantly, phosphotyrosine and anti-Fc⑀RI ␤ immunoblotting (Fig. 4B). Fc⑀RI-evoked total Syk tyrosyl phosphorylation (Fig. 3D) was de- These data suggest that Shc, via its SH2 domain, is recruited to creased in FynϪ/Ϫ BMMC compared with Fynϩ/ϩ BMMC. These Fc⑀RI ␤-chain. Shc then becomes tyrosyl-phosphorylated and re- data suggest that the Fc⑀RI-evoked Syk activation is impaired in cruits the Grb/Gab2 complex. FynϪ/Ϫ BMMC. Collectively, these data indicate that Fyn is re- ⑀ quired for the optimal activation of Syk in response to Fc⑀RI en- Gab2 PH domain also contributes to Fc RI-evoked Gab2 tyrosyl gagement. Since we found that Syk is absolutely required for Gab2 phosphorylation tyrosyl phosphorylation (Fig. 2, A and B), our data strongly support Our previously published data suggest that the Gab2 PH domain a model that Fyn, via Syk, indirectly contributes to the Fc⑀RI- may bind PI3K lipid products, and is critical for recruiting Gab2 to evoked Gab2 tyrosyl phosphorylation. the phagocytic cup in macrophages (13). However, the phospholipid-binding specificity of Gab2 PH has not been determined. To Grb2-SH3 domain binding sites are required for Gab2 tyrosyl address this question, we performed protein/lipid overlay assays by phosphorylation incubating the GST-Gab2 PH domain fusion protein with mem- Next, we asked which domain(s) in Gab2 is required for Fc⑀RI- branes spotted with different amounts of PI lipids (Fig. 5A). The evoked Gab2 tyrosyl phosphorylation. To address the role of Grb2 GST-Gab2 PH domain preferentially interacted PIP3 and P3,4P2, SH3 binding sites, we expressed HA-tagged Gab2 WT or a mutant and to a lesser extent PI5P and PI4,5P2. As a control for the lipids of Gab2 (Gab2⌬Grb2), in which the Grb2-SH3 domain binding spotted on the membrane, the GST-Grp1PH domain interacted sites are mutated, in the MC/9 mast cell line. Upon Fc⑀RI cross- only with PIP3, consistent with published reports about this PH linking, HA-tagged Gab2 WT or Gab2⌬Grb2 were immunopre- domain’s specificity (38). To verify the PI lipid-binding specificity cipitated by anti-HA Ab followed by anti-phosphorylation tyrosine of Gab2 PH, we generated a mutant (R32C) Gab2 PH domain in immunoblotting (Fig. 4A). HA-Gab2 WT became robustly tyrosyl- which R32 is mutant to C. R32, a residue found conserved in other phosphorylated and associated with Shc and Grb2. By contrast, PH domains, is required for PI lipid binding (39). Importantly, the HA-Gab2⌬Grb2 showed only weak tyrosyl phosphorylation GST-Gab2 PH R32C fusion protein failed to bind PIP3 and (ϳ75% decrease). HA-Gab2⌬Grb2 also failed to associate with PI3,4P2 in the lipid overlay assay (Fig. 5A). These data indicate Shc and Grb2. Thus, the Grb2-SH3 domain binding sites are re- that the Gab2 PH domain binds preferentially to PI3,4,5P3 and quired for Gab2 association with Shc and the majority of the Gab2 PI3,4P2, the lipid products of class IA PI3K, similar to the Gab1 tyrosyl phosphorylation in response to Fc⑀RI cross-linking. PH domain as reported (11, 12). Shc was reported to bind a phosphorylated ITAM peptide de- Next, we assessed the role of the Gab2 PH domain in Fc⑀RI- rived from the Fc⑀RI ␤-chain (37). However, it is not clear triggered tyrosyl phosphorylation. We expressed HA-tagged The Journal of Immunology 2425 Downloaded from http://www.jimmunol.org/ FIGURE 3. Decreased Fc⑀RI-evoked Gab2 tyrosyl phosphorylation is correlated to reduced Syk phosphorylation in FynϪ/Ϫ BMMC. A, Fynϩ/ϩ and FynϪ/Ϫ BMMC express similar IgE receptor and c-Kit. BMMC derived from Fynϩ/ϩ and FynϪ/Ϫ mice were analyzed for cell surface IgE binding (top) and c-Kit (bottom) by FACS as described in Materials and Methods. B,Fc⑀RI-evoked Akt activation is reduced in FynϪ/Ϫ BMMC. Anti-DNP IgE- sensitized BMMC were starved and stimulated with 10 ng/ml for the indicated time. Total cell lysates were immunoblotted with anti-phospho-Akt (Ser473) and reprobed with anti-Akt Abs. C,Fc⑀RI-evoked Gab2 tyrosyl phosphorylation is reduced in FynϪ/Ϫ BMMC. BMMC were sensitized by anti-DNP-IgE, starved, and stimulated with 10 ng/ml DNP for the indicated times. Gab2 tyrosyl phosphorylation was analyzed as in Fig. 1B (top). Gab2 tyrosyl phosphorylation quantified by densitometry was shown (bottom). D,Fc⑀RI-evoked-Syk tyrosyl phosphorylation is reduced in FynϪ/Ϫ BMMC. BMMC were sensitized by anti-DNP-IgE, starved, and stimulated with 10 ng/ml DNP for the indicated times. Total Syk tyrosyl phosphorylation was analyzed as in Fig. 2C (top). Syk tyrosyl phosphorylation quantified by densitometry was shown (bottom). Data shown here represent one of the three independent experiments with similar results. by guest on October 2, 2021

FIGURE 4. Grb2-SH3 binding sites are required for Fc⑀RI-evoked Gab2 tyrosyl phosphorylation and association with Shc. A, MC9 cells were infected MSCV-HA Gab2WT and MSCV-HA-Gab2 ⌬Grb2. Infected cells were sensitized with anti- DNP-IgE, starved, and stimulated with DNP for 5 and 10 min. Lysates were immunoprecipitated with anti-HA Ab, immunoblotted with anti-pTyr, and reprobed with anti-HA, Shc, and Grb2 Abs. Gab2 tyrosyl phosphorylation quantiﬁed by densitometry was shown (middle). B, Shc SH2 domain can bind to Fc⑀RI ␤ upon Fc⑀RI cross-linking. BMMC were sensitized, starved, and stimulated with DNP. Lysates were incubated with GST alone, and GST- ShcPTB or GST-ShcSH2 fusion protein, immunoblotted with anti-pTyr, and reprobed with anti-Fc⑀RI ␤ Ab. Data shown here represent one of the three independent experiments with similar results. 2426 Gab2 TRANSMIT SIGNALS FROM Fc⑀RI IN MAST CELLS

ever, compared with Gab2WT expressing cells, phosphorylation of Akt was decreased by ϳ60% in Gab2⌬Grb2 cells and Gab2R32C cells (at 10–15 min). As Akt is a downstream effector of PI3K, the results suggest that both Gab2 mutants of Gab2 cannot fully activate PI3K. Thus, both Grb2 SH3 binding sites and the Gab2 PH domain binding to PI3K lipid products are critical for Fc⑀RI- evoked responses such as activation of PI3K and degranulation.

Discussion Gab2 is one key tyrosyl-phosphorylated scaffolding adapter protein that plays a critical role in Fc⑀RI-evoked mast cell responses including degranulation, cytokine, and chemokine gene transcription (5, 24, 25). In our present study, we have shown that Fc⑀RI- evoked Gab2 tyrosyl phosphorylation requires Syk by using the pharmacologic inhibitor of Syk and Syk-deﬁcient mast cells. Al- though Lyn and Fyn are also critical for Gab2 tyrosyl phosphorylation, they act likely through activation of Syk. By expressing the mutant of the Gab2 protein in mast cells, we found that the

Grb2-SH3 domain-binding sties and the PH domain are required Downloaded from for Fc⑀RI-evoked Gab2 tyrosyl phosphorylation, Akt activation, and degranulation. We showed that the Grb2-SH3 binding sites are required for Gab2 interaction with Shc while the Gab2 PH domain can preferentially bind the PIP3 and PI3,4P2.

Gab2 recruitment to Fc⑀RI http://www.jimmunol.org/ A previous study using immunogold labeling suggested that Gab2 FIGURE 5. Gab2 PH domain binding to PI3K lipid products contrib- becomes associated with a pool of the membrane domain contain- ⑀ utes to Fc RI-evoked Gab2 tyrosyl phosphorylation. A, Gab2 PH domain ing Fc⑀RI ␤ and p85, the regulatory submit of PI3K, upon Fc⑀RI preferentially binds to PIP3 and PI3,4P2 in a protein/lipid overlay assay in cross-linking (26). Our study provides a biochemical explanation vitro. Membranes were spotted with PI lipids at 2-fold less in mass se- ⑀ ␤ ⑀ quentially, incubated with GST-Gab2 PH, GST-Gab2-PH R32C, and GST- for how Gab2 is recruited to the Fc RI complex. Upon Fc RI ␤ Grp1PH domain fusion proteins, and probed with anti-GST Abs. B, Gab2 cross-linking, the ITAM in the cytoplasmic tail of the -chain R32C mutant showed transient Fc⑀RI-evoked tyrosyl phosphorylation. becomes tyrosyl-phosphorylated. Shc is recruited to the Fc⑀RI ␤ via its SH2 domain and becomes tyrosyl-phosphorylated. Phos-

BMMC were infected with vector alone, Gab2WT, and Gab2R32C pMXs- by guest on October 2, 2021 puro retrovirus. Infected cells were starved, sensitized with IgE, and stim- phorylated Shc provides a docking site for the Grb2 SH2 domain ulated with anti-DNP for the indicated times. Gab2 immunoprecipitates or and recruits Gab2 that is in a constitutive complex with Grb2 via total cell lysates from infected BMMC were immunoblotted with anti- the Grb2 SH3 domain (16, 20). Subsequently, Gab2 becomes ty- pTyr, and probed with anti-Gab2 Abs. Gab2 tyrosyl phosphorylation quan- rosyl-phosphorylated in an Syk-dependent manner and recruits tiﬁed by densitometry was shown (bottom). Data shown here represent one SH2 domain-containing signaling molecules including PI3K. The of the three independent experiments with similar results. Gab2-activated PI3K lipid products bind to the Gab2 PH domain around the Fc⑀RI ␤, further potentiating recruitment. In this positive feedback loop, Gab2 contributes to the sustained activation of Gab2WT or Gab2 R32C in BMMC using retroviral transduction. PI3K, required for Fc⑀RI-evoked mast cell responses (Fig. 7). This Although HA-Gab2 R32C tyrosyl phosphorylation occurred nor- model is similar to the role of Shc in recruiting Gab2 to IL-3/GM- mally initially (1 min), its phosphorylation was decreased by CSF ␤c although Shc binds ␤c via its PTB domain (14). Likewise, ϳ40% at later time points compared with Gab2 WT (Fig. 5B). This the function of Gab2 PH binding to PIP3 is analogous to the pro- result indicates that Gab2 binding to PIP3 is critical for the sus- posed model for Gab1 action in epidermal growth factor receptor tained tyrosyl phosphorylation of Gab2 in Fc⑀RI signaling. signaling (11). Previously, phosphorylated Fc⑀RI ␤ was reported to interact Gab2 Grb2 SH3 binding sites and PH domain-binding PI lipids with adapter protein Shc (37). Shc is also reported to become ty- ⑀ are required for Fc RI-evoked mast cell responses rosyl-phosphorylated and associated with Grb2 upon Fc⑀RI cross- To assess the role of the Grb2-SH3 domain binding sites and PH linking (40). However, it is not clear how Shc interacts with Fc⑀RI domain binding to PIP3 in Fc⑀RI-evoked mast cell responses, we ␤ and what role Shc plays in Fc⑀RI signaling. We showed here that re-expressed Gab2WT, Gab2⌬Grb2, and Gab2R32C in Gab2Ϫ/Ϫ Shc can interact with the phosphorylated Fc⑀RI ␤ speciﬁcally via BMMC using retroviral transduction, and measured Fc⑀RI-evoked its SH2 domain, not the PTB domain (Fig. 4B). This result is mast cell degranulation by analyzing the release of ␤-hexosamini- consistent with the predication by Scansite (41) that the cytoplas- dase. Re-expression of Gab2 WT increased degranulation by mic tail of Fc⑀RI ␤ only contains tyrosine-containing motifs ϳ2.5-fold compared with vector control (Fig. 6A). This difference (Tyr217, Tyr223, and Tyr227) that may bind the Shc SH2 domain. in degranulation shows comparable difference in degranulation be- Consistent with Shc recruiting Gab2 to Fc⑀RI ␤ (Fig. 7), we found tween Gab2ϩ/ϩ and Gab2Ϫ/Ϫ BMMC reported earlier (5). Re- that Gab2 increases its association with Shc upon Fc⑀RI cross- expression of either Gab2⌬Grb2 or Gab2R32C only slightly in- linking. A mutant of Gab2 (⌬Grb2), that cannot bind Grb2, lost its creased degranulation (ϳ50%) compared with vector control (Fig. association with Shc and tyrosyl phosphorylation (Fig. 4A). There- 6A). Furthermore, re-expression of Gab2WT also increased Fc⑀RI- fore, the Fc⑀RI-evoked Gab2 tyrosyl phosphorylation requires evoked Akt phosphorylation dramatically (Fig. 6, B and C). How- Gab2 association with the Shc/Grb2 complex. The Journal of Immunology 2427 Downloaded from

FIGURE 6. ⑀ A Grb2-SH3 domain binding sites and the Gab2 PH domain are required for Fc RI-evoked degranulation and Akt activation in BMMC. , http://www.jimmunol.org/ Expression of Gab2⌬Grb2 or Gab2 R32C cannot rescue the defective Fc⑀RI-evoked degranulation in Gab2Ϫ/Ϫ BMMC. Gab2Ϫ/Ϫ BMMC were infected with vector, Gab2WT, Gab2⌬Grb2, and Gab2R32C pMXs-puro-viruses as described in Materials and Methods. BMMC were sensitized with anti-DNP- IgE, and stimulated with DNP for 10 min. Degranulation was measured as percentage of ␤-hexosaminidase release (top). Cell lysates were immunoblotted with anti-Gab2 Abs (bottom), indicating similar expression of Gab2WT, Gab2⌬Grb2, and Gab2R32C. B, Expression of Gab2⌬Grb2 cannot rescue the defective Akt activation compared with Gab2WT in Gab2Ϫ/Ϫ BMMC. BMMC were sensitized with anti-DNP-IgE, starved, and stimulated with DNP for the indicated times. Lysates were immunoblotted with Abs against phospho-Akt and Gab2, and reprobed with Abs against Akt. Akt phosphorylation was quantiﬁed by densitometry was shown (bottom). C, Expression of Gab2 R32C cannot rescue the defective Akt activation compared with Gab2 WT in Gab2Ϫ/Ϫ BMMC. BMMC were sensitized with anti-DNP-IgE, starved, and stimulated with DNP for the indicated times. Lysates were immunoblotted with Abs against phospho-Akt and Gab2, and reprobed with Abs against Akt. Akt phosphorylation quantiﬁed by densitometry was shown (bottom). Data shown here represent one of the three independent experiments with similar results. by guest on October 2, 2021

A previous study strongly suggested that the Gab2 PH domain PH domain also binds preferentially PIP3 and PI3,4P2 (data not binds PI3K lipid products and is critical for Gab2 function in shown), in agreement with previously published work using a lipid Fc␥RI-mediated phagocytosis (13). Here, we showed that the vesicle-binding assay (11), validating the method of protein/lipid Gab2 PH domain can preferentially bind PIP3, and PI3,4P2, the overlay assay in determining PI lipid-binding speciﬁcity for the lipid products of PI3K, and the mutant Gab2 PH R32C domain lost Gab2 PH domain. We also found consistently, although to a lesser its binding to these lipids (Fig. 5A). Gab1 and Gab2 PH domains extent, that the GST-Gab2 PH domain binds to PI5P (Fig. 5A), a share ϳ70% of the sequence identity. Especially, the regions in PH PI lipid with essentially unknown function (42). Likewise, the domain involved in direct PI lipid binding are conserved (20). Gab1 PH domain binds to PI5P in the similar preference in this in Using the same protein/lipid overlay assay, we found that the Gab1 vitro-binding assay (data not shown). Future study is needed to

FIGURE 7. Model of Gab2 action in Fc⑀RI signaling. Upon cross-linking of IgE prebound Fc⑀RI by Ag, Fc⑀RI-associated SFK (Lyn and Fyn) are activated. Then, the ITAM in ␤-chain becomes phosphorylated and recruits Shc via the Shc SH2 domain. Shc becomes phosphorylated and interacts with Grb2 via the Grb2 SH2 domain, and subsequently recruits Gab2, which is in a constitutive complex with Grb2 via the Grb2 SH3 domain. Gab2 becomes tyrosyl-phosphorylated in a Syk-dependent manner and becomes associated with PI3K that produces PIP3. The Gab2 PH domain binds to PIP3, resulting in a sustained Gab2-signaling complex around Fc⑀RI and Fc⑀RI-evoked mast cell responses. 2428 Gab2 TRANSMIT SIGNALS FROM Fc⑀RI IN MAST CELLS look into how PI5P binds to the Gab PH domain and whether this (2) study was in the pure C57BL6 background, which could con- regulates Gab protein function. tribute to the different responses of the mast cells derived. Notably, the Fc⑀RI-evoked degranulation is impaired in LynϪ/Ϫ BMMC in Kinases phosphorylate Gab2 our study (data not shown) while the degranulation was enhanced Our data strongly suggest that Syk is the major tyrosine kinase in LynϪ/Ϫ BMMC used in the Parravicini et al. (2) study. One responsible for Gab2 tyrosyl phosphorylation upon Fc⑀RI engage- possible implication from this discrepancy is the genetic back- ment whereas Lyn and Fyn contribute to Gab2 tyrosyl phosphor- ground that may contribute to different expression and/or activa- ylation by acting upstream of Syk. A previous study suggested that tion of different SFKs that can activate parallel pathways critical Fc⑀RI-triggered Gab2 tyrosyl phosphorylation requires Fyn while for Fc⑀RI-evoked responses in mast cells (48). Syk is dispensable. Furthermore, Gab2 became associated with Besides tyrosyl phosphorylation, our data also indicates that Syk Fyn upon Fc⑀RI engagement (2). However, under the condition is required, while Lyn or Fyn is dispensable, for Fc⑀RI-evoked that we immunoprecipitated almost all of Gab2 or Fyn from mast serine/threonine phosphorylation. Published reports showed that cell lysates using Abs against Gab2 or Fyn, we were unable to activation of the PI3K (14, 34) and/or Erk (35) pathways contrib- detect Gab2 association with Fyn (data not shown). Importantly, utes to serine phosphorylation in Gab2. Serine phosphorylation on we found that Fc⑀RI-triggered Syk tyrosyl phosphorylation was Gab2 has been suggested as a negative feedback mechanism to reduced accordingly in FynϪ/Ϫ BMMC compared with Fynϩ/ϩ turn off the Gab2-initiated positive signaling by tyrosyl phosphor- BMMC (Fig. 3D), suggesting that Fyn is required for optimal ac- ylation in response to growth factor and cytokine stimulation (34, tivation of Syk. 35). It will be important to study in the future how serine/threonine Because our result shows that Syk is absolutely required for phosphorylation affects Gab2 functions in Fc⑀RI-evoked responses Fc⑀RI-evoked Gab2 tyrosyl phosphorylation (Fig. 2C), there are in mast cells. Downloaded from several possibilities regarding how Fyn and Syk control Gab2 ty- The Lyn/Syk/LAT pathway and the alternative pathway involv- rosyl phosphorylation. A simple model is that Fyn regulates the ing Fyn/Gab2/PI3K have been suggested to play critical roles in activation of Syk, which phosphorylates Gab2 directly (Fig. 7). Fc⑀RI-evoked degranulation of mast cells (2). Our result indicates Alternatively, Syk phosphorylates Gab2 at several sites including that Syk, at least a pool of Syk, is critical for the activation of the the site that recruits Fyn. Upon interacting with Gab2, Fyn may Gab2/PI3K in the Fc⑀RI-evoked degranulation response (Fig. 7). phosphorylate other tyrosine residues in Gab2. Lastly, Syk may Furthermore, our data provide a further biochemical basis for de- http://www.jimmunol.org/ contribute to Gab2 tyrosyl phosphorylation by controlling the veloping potential drugs that specifically inhibit Syk activation phosphorylation of the tyrosine residue in Fc⑀RI ␤ that recruits the (49) to treat allergy. Shc/Grb2/Gab2 complex as discussed above (Fig. 7). Consistent with this idea, a previous study revealed that the Fc⑀RI-evoked Acknowledgments Shc/Grb2 complex depends on Syk (40). At present, we could not We thank Naoko Imanaka, Yongping Wang, and Wenkai Yang for excel- determine whether Syk also indirectly contributes to tyrosyl phos- lent technical assistance. phorylation of Gab2 through activation of other protein kinases, such as Tec family kinase member Btk and Itk (1), present in mast Disclosures by guest on October 2, 2021 cells. Syk is required for activation of Btk in mast cells (43). The The authors have no financial conflict of interest. use of Btk and/or Itk-deficient mast cells are required to address the potential role of Tec family kinase Fc⑀RI-evoked Gab2 tyrosyl References phosphorylation in the future. 1. Nadler, M. J., S. A. Matthews, H. Turner, and J. P. Kinet. 2000. Signal transduction by the high-affinity immunoglobulin E receptor Fc⑀RI: coupling form to Our results also indicate that Lyn is required for Gab2 phos- function. Adv. Immunol. 76: 325–355. phorylation. Because Lyn acts upstream of Syk in Fc⑀RI signaling 2. Parravicini, V., M. Gadina, M. Kovarova, S. Odom, C. Gonzalez-Espinosa, (1), it is logical to conclude that Lyn controls Gab2 phosphoryla- Y. Furumoto, S. Saitoh, L. E. Samelson, J. J. O’Shea, and J. Rivera. 2002. Fyn kinase initiates complementary signals required for IgE-dependent mast cell de- tion mainly through activation of Syk. We cannot exclude the pos- granulation. Nat. Immunol. 3: 741–748. sibility that Lyn may phosphorylate Gab2 directly. Consistent with 3. Saitoh, S., R. Arudchandran, T. S. Manetz, W. Zhang, C. L. Sommers, P. E. Love, ⑀ this idea, Gab2 has been suggested to mediate G-CSF-invoked J. Rivera, and L. E. Samelson. 2000. LAT is essential for Fc( )RI-mediated mast cell activation. Immunity 12: 525–535. differentiation in an Lyn-dependent manner (44). However, our 4. Pivniouk, V. I., T. R. Martin, J. M. Lu-Kuo, H. R. Katz, H. C. Oettgen, and result does not agree with published reports from Rivera’s group R. S. Geha. 1999. SLP-76 deficiency impairs signaling via the high-affinity IgE ⑀ receptor in mast cells. J. Clin. Invest. 103: 1737–1743. (2) which showed that Lyn negatively regulates Fc RI-evoked 5. Gu, H., K. Saito, L. D. Klaman, J. Shen, T. Fleming, Y. Wang, J. C. Pratt, G. Lin, Gab2 tyrosyl phosphorylation through the activation of C-terminal B. Lim, J.-P. Kinet, and B. G. Neel. 2001. Essential role for Gab2 in the allergic Src kinase, an inhibitory PTK for SFK, which inhibits Fc⑀RI- response. Nature 412: 186–190. 6. Volna, P., P. Lebduska, L. Draberova, S. Simova, P. Heneberg, M. Boubelik, evoked Fyn activation in mast cells (45). There are several possi- V. Bugajev, B. Malissen, B. S. Wilson, V. Horejsi, et al. 2004. Negative regu- bilities that may account for this discrepancy. The conditions for lation of mast cell signaling and function by the adaptor LAB/NTAL. J. Exp. generating BMMC are different. We typically use only IL-3-con- Med. 200: 1001–1013. 7. Zhu, M., Y. Liu, S. Koonpaew, O. Granillo, and W. Zhang. 2004. Positive and taining medium to generate BMMC from bone marrow. However, negative regulation of Fc⑀RI-mediated signaling by the adaptor protein LAB/ Fc⑀RI-evoked Gab2 tyrosyl phosphorylation was still severely im- NTAL. J. Exp. Med. 200: 991–1000. Ϫ/Ϫ 8. Liu, Y., and L. R. Rohrschneider. 2002. The gift of Gab. FEBS Lett. 515: 1–7. paired in Lyn BMMC derived from the same medium condi- Trends Cell Biol. Ϫ Ϫ 9. Gu, H., and B. G. Neel. 2003. The “Gab” in signal transduction. tion (IL-3 ϩ SCF) as described (2). The Lyn / mice used in our 13: 122–130. study were generated differently from the LynϪ/Ϫ mice used in the 10. Maffucci, T., and M. Falasca. 2001. Specificity in pleckstrin homology (PH) domain membrane targeting: a role for a phosphoinositide-protein co-operative Parravicini et al. (2) study. In the former case, the promoter of mechanism. FEBS Lett. 506: 173–179. region of the Lyn gene was disrupted (46). In the latter case, exon 11. Rodrigues, G. A., M. Falasca, Z. Zhang, S. H. Ong, and J. Schlessinger. 2000. A 3-7 of the Lyn gene was eliminated (47). However, both LynϪ/Ϫ novel positive feedback loop mediated by the docking protein Gab1 and phosphatidylinositol 3-kinase in epidermal growth factor receptor signaling. Mol. mice are null for Lyn protein. The possibility that we favor is the Cell. Biol. 20: 1448–1459. difference of genetic background of these two LynϪ/Ϫ mice. The 12. Maroun, C. R., D. K. Moscatello, M. A. Naujokas, M. Holgado-Madruga, Ϫ/Ϫ A. J. Wong, and M. Park. 1999. A conserved inositol phospholipid binding site Lyn mice used in our study was in the mixed background of within the pleckstrin homology domain of the Gab1 docking protein is required sv129/J and C57BL6 while the mice used in the Parravicini et al. for epithelial morphogenesis. J. Biol. Chem. 274: 31719–31726. The Journal of Immunology 2429

13. Gu, H., R. J. Botelho, M. Yu, S. Grinstein, and B. G. Neel. 2003. Critical role for S. J. Burakoff, et al. 1997. A comparative analysis of the phosphoinositide bind- scaffolding adapter Gab2 in Fc␥R-mediated phagocytosis. J. Cell Biol. 161: ing specificity of pleckstrin homology domains. J. Biol. Chem. 272: 1151–1161. 22059–22066. 14. Gu, H., H. Maeda, J. J. Moon, J. D. Lord, M. Yoakim, B. H. Nelson, and 33. Nishizumi, H., and T. Yamamoto. 1997. Impaired tyrosine phosphorylation and B. G. Neel. 2000. New role for Shc in activation of the phosphatidylinositol Ca2ϩ mobilization, but not degranulation, in lyn-deficient bone marrow-derived 3-kinase/Akt pathway. Mol. Cell. Biol. 20: 7109–7120. mast cells. J. Immunol. 158: 2350–2355. 15. Schaeper, U., N. H. Gehring, K. P. Fuchs, M. Sachs, B. Kempkes, and 34. Lynch, D. K., and R. J. Daly. 2002. PKB-mediated negative feedback tightly W. Birchmeier. 2000. Coupling of Gab1 to c-Met, Grb2, and Shp2 mediates regulates mitogenic signalling via Gab2. EMBO J. 21: 72–82. biological responses. J. Cell Biol. 149: 1419–1432. 35. Arnaud, M., C. Crouin, C. Deon, D. Loyaux, and J. Bertoglio. 2004. Phosphor- 16. Lock, L. S., I. Royal, M. A. Naujokas, and M. Park. 2000. Identification of an ylation of Grb2-associated binder 2 on serine 623 by ERK MAPK regulates its atypical Grb2 carboxyl-terminal SH3 domain binding site in Gab docking pro- association with the phosphatase SHP-2 and decreases STAT5 activation. J. Im- teins reveals Grb2-dependent and -independent recruitment of Gab1 to receptor munol. 173: 3962–3971. tyrosine kinases. J. Biol. Chem. 275: 31536–31545. 36. Yu, W. M., T. S. Hawley, R. G. Hawley, and C. K. Qu. 2002. Role of the docking 17. Sattler, M., M. G. Mohi, Y. B. Pride, L. R. Quinnan, N. A. Malouf, K. Podar, ␤ protein Gab2 in 1-integrin signaling pathway-mediated hematopoietic cell ad- F. Gesbert, H. Iwasaki, S. Li, R. A. Van Etten, et al. 2002. Critical role for Gab2 hesion and migration. Blood 99: 2351–2359. in transformation by BCR/ABL. Cancer Cell 1: 479–492. 37. Kimura, T., H. Kihara, S. Bhattacharyya, H. Sakamoto, E. Appella, and 18. Liu, Y., B. Jenkins, J. L. Shin, and L. R. Rohrschneider. 2001. Scaffolding protein R. P. Siraganian. 1996. Downstream signaling molecules bind to different phos- Gab2 mediates differentiation signaling downstream of Fms receptor tyrosine phorylated immunoreceptor tyrosine-based activation motif (ITAM) peptides of kinase. Mol. Cell. Biol. 21: 3047–3056. the high affinity IgE receptor. J. Biol. Chem. 271: 27962–27968. 19. Meng, S., Z. Chen, T. Munoz-Antonia, and J. Wu. 2005. Participation of both 38. Klarlund, J. K., A. Guilherme, J. J. Holik, J. V. Virbasius, A. Chawla, and Gab1 and Gab2 in the activation of the Erk mitogen-activated protein kinase M. P. Czech. 1997. Signaling by phosphoinositide-3,4,5-trisphosphate through pathway by epidermal growth factor. Biochem. J. 391(Pt. 1): 143–151. proteins containing pleckstrin and Sec7 homology domains. Science 275: 20. Gu, H., J. C. Pratt, S. J. Burakoff, and B. G. Neel. 1998. Cloning of p97/Gab2, 1927–1930. the major SHP-2 binding protein in hematopoietic cells, reveals a novel pathway 39. Salim, K., M. J. Bottomley, E. Querfurth, M. J. Zvelebil, I. Gout, R. Scaife, for cytokine-induced gene activation. Mol. Cell 2: 729–740. R. L. Margolis, R. Gigg, C. I. Smith, P. C. Driscoll, et al. 1996. Distinct speci- 21. Kong, M., C. Mounier, J. Wu, and B. I. Posner. 2000. Epidermal growth factor-

ficity in the recognition of phosphoinositides by the pleckstrin homology domains Downloaded from induced phosphatidylinositol 3-kinase activation and DNA synthesis: identifica- of dynamin and Bruton’s tyrosine kinase. EMBO J. 15: 6241–6250. tion of Grb2-associated binder 2 as the major mediator in rat hepatocytes. J. Biol. 40. Jabril-Cuenod, B., C. Zhang, A. M. Scharenberg, R. Paolini, R. Numerof, Chem. 275: 36035–36042. M. A. Beaven, and J. P. Kinet. 1996. Syk-dependent phosphorylation of Shc: a 22. Nishida, K., L. Wang, E. Morii, S. J. Park, M. Narimatsu, S. Itoh, S. Yamasaki, potential link between Fc⑀RI and the Ras/mitogen-activated protein kinase sig- M. Fujishima, K. Ishihara, M. Hibi, et al. 2002. Requirement of Gab2 for mast naling pathway through SOS and Grb2. J. Biol. Chem. 271: 16268–16272. cell development and KitL/c-Kit signaling. Blood 99: 1866–1869. 41. Yaffe, M. B., G. G. Leparc, J. Lai, T. Obata, S. Volinia, and L. C. Cantley. 2001. 23. Nakanishi, S., H. Yano, and Y. Matsuda. 1995. Novel functions of phosphati- A motif-based profile scanning approach for genome-wide prediction of signaling dylinositol 3-kinase in terminally differentiated cells. Cell. Signal. 7: 545–557. pathways. Nat. Biotechnol. 19: 348–353. 24. Gonzalez-Espinosa, C., S. Odom, A. Olivera, J. P. Hobson, M. E. Martinez, http://www.jimmunol.org/ A. Oliveira-Dos-Santos, L. Barra, S. Spiegel, J. M. Penninger, and J. Rivera. 42. De Matteis, M. A., A. Di Campli, and A. Godi. 2005. The role of the phosphoi- 2003. Preferential signaling and induction of allergy-promoting lymphokines nositides at the Golgi complex. Biochim. Biophys. Acta 1744: 396–405. upon weak stimulation of the high affinity IgE receptor on mast cells. J. Exp. Med. 43. Kawakami, Y., J. Kitaura, S. E. Hartman, C. A. Lowell, R. P. Siraganian, and ␤ 197: 1453–1465. T. Kawakami. 2000. Regulation of protein kinase C I by two protein-tyrosine 25. Nishida, K., S. Yamasaki, Y. Ito, K. Kabu, K. Hattori, T. Tezuka, H. Nishizumi, kinases, Btk and Syk. Proc. Natl. Acad. Sci. USA 97: 7423–7428. D. Kitamura, R. Goitsuka, R. S. Geha, et al. 2005. Fc⑀RI-mediated mast cell 44. Zhu, Q. S., L. J. Robinson, V. Roginskaya, and S. J. Corey. 2004. G-CSF-induced degranulation requires calcium-independent microtubule-dependent translocation tyrosine phosphorylation of Gab2 is Lyn kinase dependent and associated with of granules to the plasma membrane. J. Cell Biol. 170: 115–126. enhanced Akt and differentiative, not proliferative, responses. Blood 103: 26. Wilson, B. S., J. R. Pfeiffer, Z. Surviladze, E. A. Gaudet, and J. M. Oliver. 2001. 3305–3312. High resolution mapping of mast cell membranes reveals primary and secondary 45. Odom, S., G. Gomez, M. Kovarova, Y. Furumoto, J. J. Ryan, H. V. Wright, domains of Fc⑀RI and LAT. J. Cell Biol. 154: 645–658. C. Gonzalez-Espinosa, M. L. Hibbs, K. W. Harder, and J. Rivera. 2004. Negative

27. Hernandez-Hansen, V., G. A. Mackay, C. A. Lowell, B. S. Wilson, and regulation of immunoglobulin E-dependent allergic responses by Lyn kinase. by guest on October 2, 2021 J. M. Oliver. 2004. The Src kinase Lyn is a negative regulator of mast cell J. Exp. Med. 199: 1491–1502. proliferation. J. Leukocyte Biol. 75: 143–151. 46. Nishizumi, H., I. Taniuchi, Y. Yamanashi, D. Kitamura, D. Ilic, S. Mori, 28. Mocsai, A., M. Zhou, F. Meng, V. L. Tybulewicz, and C. A. Lowell. 2002. Syk T. Watanabe, and T. Yamamoto. 1995. Impaired proliferation of peripheral B is required for integrin signaling in neutrophils. Immunity 16: 547–558. cells and indication of autoimmune disease in lyn-deficient mice. Immunity 3: 29. Liu, F. T., J. W. Bohn, E. L. Ferry, H. Yamamoto, C. A. Molinaro, 549–560. L. A. Sherman, N. R. Klinman, and D. H. Katz. 1980. Monoclonal dinitrophenyl- 47. Chan, V. W., C. A. Lowell, and A. L. De Franco. 1998. Defective negative specific murine IgE antibody: preparation, isolation, and characterization. J. Im- regulation of antigen receptor signaling in Lyn-deficient B lymphocytes. Curr. munol. 124: 2728–2737. Biol. 8: 545–553. 30. Lu-Kuo, J. M., D. A. Fruman, D. M. Joyal, L. C. Cantley, and H. R. Katz. 2000. 48. Furumoto, Y., C. Gonzalez-Espinosa, G. Gomez, M. Kovarova, S. Odom, Impaired kit- but not Fc⑀RI-initiated mast cell activation in the absence of phos- V. Parravicini, J. J. Ryana, and J. Rivera. 2004. Rethinking the role of Src family phoinositide 3-kinase p85␣ gene products. J. Biol. Chem. 275: 6022–6029. protein tyrosine kinases in the allergic response: new insights on the functional 31. Morita, S., T. Kojima, and T. Kitamura. 2000. Plat-E: an efficient and stable coupling of the high affinity IgE receptor. Immunol. Res. 30: 241–253. system for transient packaging of retroviruses. Gene Ther. 7: 1063–1066. 49. Kovarova, M., and J. Rivera. 2004. A molecular understanding of mast cell ac- 32. Rameh, L. E., A. Arvidsson, K. L. Carraway, A. D. Couvillion, G. Rathbun, tivation and the promise of anti-allergic therapeutics. Curr. Med. Chem. 11: A. Crompton, B. VanRenterghem, M. P. Czech, K. S. Ravichandran, 2083–2091.