gp49B1 Inhibits IgE-initiated Mast Cell Activation through Both Immunoreceptor Tyrosine-based Inhibitory Motifs, Recruitment ofsrc Homology 2 Domain-containing Phosphatase-1, and Suppression of Early and Late Calcium Mobilization*

We define by molecular, pharmacologic, and physiologic approaches the proximal mechanism by which the immunoglobulin superfamily member gp49B1 inhibits mast cell activation mediated by the high affinity Fc receptor for IgE (FcεRI). In rat basophilic leukemia-2H3 cells expressing transfected mouse gp49B1, mutation of tyrosine to phenylalanine in either of the two immunoreceptor tyrosine-based inhibitory motifs of the gp49B1 cytoplasmic domain partially suppressed gp49B1-mediated inhibition of exocytosis, whereas mutation of both abolished inhibitory capacity. Sodium pervanadate elicited tyrosine phosphorylation of native gp49B1 and association of the tyrosine phosphatases src homology 2 domain-containing phosphatase-1 (SHP-1) and SHP-2 in mouse bone marrow-derived mast cells (mBMMCs). SHP-1 associated transiently with gp49B1 within 1 min after coligation of gp49B1 with cross-linked FcεRI in mBMMCs. SHP-1-deficient mBMMCs exhibited a partial loss of gp49B1-mediated inhibition of FcεRI-induced exocytosis at concentrations of IgE providing optimal exocytosis, revealing a central, but not exclusive, SHP-1 requirement in the counter-regulatory pathway. Coligation of gp49B1 with cross-linked FcεRI on mBMMCs inhibited early release of calcium from intracellular stores and subsequent influx of extracellular calcium, consistent with SHP-1 participation. Because exocytosis is complete within 2 min in mBMMCs, our studies establish a role for SHP-1 in the initial counter-regulatory cellular responses whereby gp49B1 immunoreceptor tyrosine-based inhibition motifs rapidly transmit inhibition of FcεRI-mediated exocytosis.

The cross-linking of the high affinity Fc receptor for IgE (Fc⑀RI) 1 on the surface of mast cells activates intracellular signaling pathways that lead to exocytosis of preformed secretory granule components, generation of membrane-derived lipid mediators, and elaboration of immunoregulatory and proinflammatory cytokines. Through the release of their array of bioactive mediators, mast cells have an unusually high proinflammatory potential and contribute to host defense in animal models (1)(2)(3) and to the pathogenesis of allergic reactions and bronchial asthma in humans.
In addition to Fc⑀RI, mouse mast cells express on their surface gp49B1 (4,5), an immunoglobulin (Ig) superfamily member with two Ig-like domains (6) and two cytoplasmic immunoreceptor tyrosine-based inhibition motifs (ITIMs), which have the amino acid sequence (I/V/S)XYXX(L/V) (7)(8)(9)(10)(11)(12). We have shown that when rat IgE bound to Fc⑀RI on mouse interleukin-3-dependent, bone marrow-derived mast cells (mBMMCs) is both cross-linked to itself and coligated with rat anti-gp49B1 mAb bound to gp49B1, by means of F(abЈ) 2 mouse anti-rat IgG (heavy and light chain reactive), the release of secretory granule-and lipid-derived mediators is inhibited (9). As defined by ALIGN amino acid homology analysis, gp49B1 belongs to a family of receptors that has immunoregulatory potential (9,13). In the mouse, the sole other member of this homology-based family (14) that bears an ITIM is PIR-B (also termed p91), which has six Ig-like domains and is expressed as mRNA transcripts in mouse mast cells (15,16). When chimeric receptors possessing the cytoplasmic domain of PIR-B are expressed by transfection in the rat mucosal mast cell-like cell line RBL-2H3 (17) and coligated with cross-linked Fc⑀RI, the chimeric molecule becomes tyrosine-phosphorylated and associates with the tyrosine phosphatases SHP-1 and SHP-2. Early and late calcium fluxes, as well as exocytosis, as assessed by the release of incorporated [ 3 H]serotonin, are inhibited (10,11).
A number of human members of the family are encoded in a gene complex on chromosome 19. Ig-like transcript 3 (ILT3, also termed HM18 and leukocyte Ig-like receptor 5), was cloned based on its homology with mouse gp49B1 (18) and by a separate strategy (19). Solid phase anti-ILT3 induces the association of SHP-1 with ILT3, and antibody-mediated coligation of ILT3 to certain cross-linked activating receptors on monocytes inhibits early and late calcium fluxes (19). The extracellular domain of another human relative of gp49B1, ILT2 (also referred to as leukocyte Ig-like receptor 1 and monocyte-macrophage Ig-like receptor 7) (20 -22), interacts broadly with major histocompatibility complex class I molecules derived from multiple human leukocyte antigen loci (23), as does another gp49B1 relative, ILT4/leukocyte Ig-like receptor 2 (24). Treat-* This work was supported by Grants AI-22531, AI-31599, AI-41144, and HL-36110 from the National Institutes of Health. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
¶ Recipient of a postdoctoral fellowship from the Arthritis Foundation.
Human killer cell inhibitory receptors (KIRs), which are primarily expressed on natural killer (NK) cells and certain T cells, are also ITIM-bearing members of the homology-defined family that includes gp49B1 (9,13). In contrast to ILT2, individual human KIRs bind major histocompatibility complex class I molecules expressed by a particular human leukocyte antigen locus. When KIRs bind major histocompatibility complex class I on a target cell or are bound by anti-KIR Ig, they become tyrosine-phosphorylated, recruit SHP-1, and impair both the release of calcium from intracellular stores and the influx of extracellular calcium (7,26). Furthermore, active SHP-1 is required to transduce KIR-mediated inhibition of antibody-dependent cytotoxicity in NK cells (27). As predicted from the studies of NK cells, ITIM-bearing cytoplasmic domains of KIRs transfected into B cell or mast cell lines attenuate both phases of calcium mobilization when coligated with the B cell receptor or Fc⑀RI, respectively, (28,29).
We now demonstrate with site-directed mutagenesis that the ITIMs of gp49B1 are critical to inhibition of Fc⑀RI-mediated mast cell activation and that both tyrosines in the ITIMs of gp49B1 participate. We also show that gp49B1, when tyrosinephosphorylated, binds SHP-1 and SHP-2 recruited from the cytoplasm of mBMMCs and that gp49B1-mediated inhibition of Fc⑀RI-driven activation is reduced in SHP-1-deficient mBMMCs. That inhibition of mast cell activation by gp49B1 is accompanied by decreases in both the release of calcium from intracellular stores and the subsequent influx of extracellular calcium supports the role of SHP-1 in the proximal transmission of the counter-regulatory signal.
Preparation of Stable Transfectants of RBL-2H3 Cells-Mutated gp49B1 cDNAs were generated by polymerase chain reaction mutagenesis (32). The upstream and downstream primers corresponded to nucleotides 1-21 and 1237-1257 of gp49B1, respectively (33). The mutagenic primers possessed adenosine to thymidine substitutions at nucleotides 920 and/or 986 to change tyrosines to phenylalanines in the ITIMs. The wild-type and mutated cDNAs were ligated into the pCR3 mammalian expression vector (Invitrogen, San Diego, CA), and the sequences of all plasmid constructions were determined by dideoxy sequencing to ensure that no other mutations were introduced into the nucleotide sequences during polymerase chain reaction. The cDNA constructs were released from pCR3 with PmeI and XbaI and cloned into the blunted SalI and native XbaI sites of pSR␣neo (34).
The expression of gp49B1 on the surface of each kind of transfectant was assessed by flow cytometry. Cells (5 ϫ 10 5 ) were incubated in 50 l of calcium-and magnesium-free Hanks' balanced salt solution supplemented with 0.1% bovine serum albumin and 0.02% sodium azide (HBA) alone or with a saturating amount of anti-gp49B1 for 30 min at 4°C. Cells were washed with HBA, incubated for 30 min at 4°C with fluorescein isothiocyanate-labeled F(abЈ) 2 goat IgG anti-rat IgM, and washed again with HBA. For detection of Fc⑀RI, cells were incubated with 5 g/ml of mouse IgE and 5 g/ml of fluorescein isothiocyanatelabeled rat anti-mouse IgE mAb (PharMingen, San Diego, CA) in the first and second steps, respectively. After staining, cells were resuspended in 2% paraformaldehyde in Hanks' balanced salt solution and analyzed on a Becton Dickinson FACSORT using logarithmic fluorescence amplification.
Fc⑀RI Cross-linking and Coligation with gp49B1 in RBL-2H3 Transfectants-Duplicate samples of cells were suspended at 1 ϫ 10 5 /ml in enriched EMEM containing 0.5 g/ml of rat IgE and incubated overnight at 37°C in 16-mm-diameter culture wells (1 ml/well). The supernatant media were aspirated, and 0.3 ml of fresh enriched EMEM containing the same concentrations of IgE, with a saturating concentration of anti-gp49B1, was added to appropriate wells. Cells were incubated on ice for 20 min, washed twice with PGC buffer (0.025 M PIPES, 0.12 M NaCl, 0.005 M KCl, 0.04 M NaOH, 5.6 mM glucose, 0.1% bovine serum albumin, and 1 mM CaCl 2 ), and incubated with 0.3 ml of F(abЈ) 2 mouse anti-rat IgG (25 g/ml) for 40 min at 37°C. The percentage of release of ␤-hexosaminidase (9, 35) elicited by the cross-linking of Fc⑀RI alone and combined with coligation to gp49B1 was calculated for each transfectant according to the formula (S/(S ϩ P)) ϫ 100, where S and P are the mediator contents of the samples of each supernatant and cell pellet, respectively. Cysteinyl leukotrienes were measured by reverse phase high performance liquid chromatography as described (36).
Immunochemical Detection of gp49B1 Tyrosine Phosphorylation and Association of Cytosolic Phosphatases-Samples of mBMMCs (1 ϫ 10 7 ) were pelleted by centrifugation and resuspended in 1 ml of enriched medium alone or containing 2.5 mM sodium pervanadate (10 l of a 1:1 mixture of 500 mM sodium orthovanadate and 3% hydrogen peroxide per ml). After a 10-min incubation at 37°C, the cells were lysed in an extraction buffer containing 1% Nonidet P-40, 10 mM Tris, pH 7.4, 150 mM NaCl, 2 mM sodium orthovanadate, 10 mM sodium fluoride, 10 mM pyrophosphate tetrasodium decahydrate, 5 mM EDTA, 4 g/ml leupeptin, 10 g/ml pepstatin, 10 g/ml aprotinin, 1 mM phenylmethylsulfonyl fluoride, and 20 g/ml N-acetyl-Leu-Leu-methioninal. Whole cell lysates were centrifuged at 14,000 rpm for 20 min at 4°C in a microcentrifuge to sediment cell nuclei and debris, and the supernatants were incubated at 4°C overnight with 70 l of a 25% (v/v) solution of anti-gp49B1 coupled to Sepharose. The immunoprecipitates were washed four times with extraction buffer, eluted from the Sepharose beads by boiling for 5 min in 150 l of 1% SDS, and precipitated with 750 l of cold acetone. Precipitates were resuspended in 50 mM potassium phosphate buffer, pH 7.0, containing 0.5% Nonidet P-40, 30 mM EDTA, 1 M leupeptin, 1 M pepstatin, 0.2% SDS, and 2 units of N-glycosidase F (Boehringer Mannheim) and were incubated overnight at 37°C. Reaction products were precipitated with cold acetone, and the pellets were dried, resuspended in 25 l of 1ϫ SDS-polyacrylamide gel electrophoresis (PAGE) sample buffer containing 5% 2-mecaptoethanol, boiled for 5 min, and electrophoresed in 10% acrylamide gels under reducing conditions. The resolved proteins were transferred to Immobilon P membranes (Millipore, Bedford, MA) and immunoblotted sequentially with anti-phosphotyrosine mAb 4G10 and anti-phospho-gp49B1 315-326 , and immunoreactive proteins were visualized with HRP-labeled goat antimouse IgG and HRP-labeled goat anti-rabbit IgG, respectively, using chemiluminescence.
To assess the association of phosphatases with gp49B1 with and without exposure of mBMMCs to pervanadate, gp49B1 was immunoprecipitated from detergent lysates with anti-gp49B1 coupled to Sepharose as described above for tyrosine phosphorylation of gp49B1, except that the N-glycosidase F step was omitted. The non-deglycosylated immunoprecipitates were resuspended in SDS-PAGE sample buffer, boiled, resolved on a 10% SDS-PAGE gel, transferred to Immobilon P membranes, and immunoblotted sequentially with rabbit anti-SHP-1, anti-SHP-2, and anti-SHIP. Immunoreactive bands were visualized with HRP-conjugated goat anti-rabbit IgG and chemiluminescence.
To measure the association of enzymes with gp49B1 after its coligation to Fc⑀RI, mBMMCs were incubated for 1 h at 4°C with 3 g/ml of rat IgE. A saturating amount of anti-gp49B1 was added for the last 3 min before the cross-linking of Fc⑀RI and coligation to gp49B1 with F(abЈ) 2 mouse anti-rat IgG. Before and at several times after the crosslinking and coligating step, samples of cells (1.5 ϫ 10 7 ) were lysed in extraction buffer, nuclei and debris were removed by centrifugation as described above, and gp49B1 bound to anti-gp49B1 was immunoprecipitated from the lysates with goat anti-rat IgM coupled to Sepharose. The immunoprecipitates were processed as described above to resolve and visualize gp49B1 and associated enzymes.
Calcium Measurements-mBMMCs (1 ϫ 10 7 /ml) were sensitized with 3 g/ml rat IgE for 1 h at 4°C and incubated with 2.5 M Fura-2 (Molecular Probes, Eugene, OR) at 37°C for 30 min. Medium alone or with a saturating amount of anti-gp49B1 was added, and the cells were incubated for an additional 15 min at 4°C. Cells were washed, resuspended in Hanks' balanced salt solution containing 1 mM each CaCl 2 and MgCl 2 , and placed in a fluorometer, and 25 g/ml of F(abЈ) 2 mouse anti-rat IgG was added. In some experiments, EGTA at a final concentration of 18 mM was added to replicate samples before the addition of F(abЈ) 2 mouse anti-rat IgG to chelate extracellular calcium. Relative free cytosolic calcium in the cells was measured by the ratio of fluorescence emission intensities at 510 nm when the samples were exposed to excitation wavelengths of 340 and 380 nm, respectively.

Analysis of the Requirement of ITIM Tyrosines in gp49B1mediated Inhibition of Mast Cell
Activation-RBL-2H3 cells were stably transfected with cDNAs encoding either wild-type gp49B1 (YY) or gp49B1 with tyrosine-to-phenylalanine mutations in the upstream (FY), downstream (YF), or both ITIMs (FF) (Fig. 1A) to establish the requirement for ITIMs in the inhibitory function of gp49B1. The expression of gp49B1 and Fc⑀RI on transfected clones was assessed by flow cytometry, and YY, FY, and YF clones were selected, for which the net mean channel numbers of fluorescence for gp49B1 and Fc⑀RI ranged from 43-48 and 78 -81, respectively; the channel numbers were 84 and 147, respectively, for the closest available FF clone (n ϭ 2).
Cells were incubated with rat IgE (0.5 g/ml), without or with a saturating concentration of anti-gp49B1, and were washed. F(abЈ) 2 mouse anti-rat IgG (25 g/ml) was added both to cross-link Fc⑀RI molecules sensitized with rat IgE to each other and to coligate transfected gp49B1 molecules bearing rat IgM anti-gp49B1 with Fc⑀RI. The coligation of gp49B1 possessing wild-type ITIMs with cross-linked Fc⑀RI resulted in a 38 Ϯ 10% (mean Ϯ S.E., n ϭ 3) inhibition of ␤-hexosaminidase release compared with cells activated by Fc⑀RI cross-linking alone (Fig. 1B). The FY and YF mutations each partially suppressed gp49B1-mediated inhibition of ␤-hexosaminidase release. In contrast, mutation of both ITIM tyrosines abolished the ability of gp49B1 to inhibit mediator release. Similar results were obtained with additional clones expressing each type of gp49B1 (data not shown). Activation of the cells with a 10-fold lower concentration of IgE afforded a greater percentage of inhibition of exocytosis by the YY, FY, and YF clones, but the FF clone remained fully noninhibitory (data not shown).
Tyrosine Phosphorylation of Native gp49B1 in mBMMCs-To demonstrate the tyrosine phosphorylation of endogenous gp49B1 in mast cells, mBMMCs were incubated for 10 min at 37°C in enriched medium alone or containing sodium pervanadate, an inhibitor of tyrosine phosphatases. Cell lysates were subjected to immunoprecipitation with either anti-gp49B1 or a rat IgM negative control mAb, and precipitates were deglycosylated with N-glycosidase F, resolved by SDS-PAGE, transferred to an Immobilon P membrane, and immunoblotted sequentially with anti-phosphotyrosine mouse mAb 4G10 and antibody from a rabbit immunized with a tyrosinephosphorylated peptide encompassing the downstream ITIM of gp49B1 (anti-phospho-gp49B1 315-326 ). In the absence of pervanadate, a 37-kDa molecule immunoreactive with antiphospho-gp49B1 315-326 (Fig. 2, lane 3, upper panel) but not with anti-phosphotyrosine (Fig. 2, lane 3, lower panel) was present; 37 kDa is the size of the protein core of gp49B1 (5). When cells had been treated with pervanadate, the apparent M r of the band that was reactive with anti-phospho-gp49B1 315-326 increased to 40 kDa (Fig. 2, lane 4, upper panel), concomitant with the acquisition of immunoreactivity with anti-phosphotyrosine (Fig. 2, lane 4, lower panel). Inasmuch as the 37-and 40-kDa bands were not detected in negative control immunoprecipitates (Fig. 2, lanes 1 and 2, upper and lower  panels), they represent nonphosphorylated and tyrosinephosphorylated gp49B1, respectively.
Binding of Phosphatases to gp49B1 ITIMs-Phosphorylated peptides encompassing each of the gp49B1 ITIM sequences bind the tyrosine phosphatases SHP-1 and SHP-2, as well as

FIG. 1. Effects of tyrosine-to-phenylalanine ITIM mutations on gp49B1-mediated inhibition of Fc⑀RI exocytosis.
Clones of stable RBL-2H3 transfectants (1 ϫ 10 7 /ml) expressing gp49B1 with either wild-type ITIMs (YY) or with tyrosine-to-phenylalanine substitutions in the upstream (FY), downstream (YF), or both (FF) ITIMs (A) were activated by cross-linking 0.5 g/ml rat IgE bound to Fc⑀RI with 25 g/ml of F(abЈ) 2 mouse anti-rat IgG (heavy and light chain reactive) to establish a reference value for uninhibited activation for each transfectant. Replicate cell samples were coincubated with rat IgE and rat anti-mouse gp49B1 so that added F(abЈ) 2 mouse anti-rat IgG would both cross-link and coligate receptors. B, the percentage of release of secretory granule ␤-hexosaminidase from each transfectant was measured 40 min after Fc⑀RI cross-linking with and without coligation to gp49B1. Data are expressed as percentage of inhibition Ϯ S.E. of ␤-hexosaminidase release obtained with coligation (n ϭ 3). The percentage of ␤-hexosaminidase release values for YY, FY, YF, and FF transfectants activated by cross-linking IgE without gp49B1 coligation were 36 Ϯ 6, 15 Ϯ 2, 17 Ϯ 6, and 25 Ϯ 1%, respectively (mean Ϯ S.E., n ϭ 3). the inositol polyphosphate 5-phosphatase SHIP in detergent extracts of RBL-2H3 (37) and mBMMCs (data not shown). To identify the preferred native interaction(s), the ability of each enzyme to associate with full-length gp49B1 expressed at the plasma membrane of mBMMCs was assessed by stimulating tyrosine phosphorylation with pervanadate, immunoprecipitating gp49B1 with anti-gp49B1 coupled to Sepharose, resolving immunoprecipitates with SDS-PAGE, and immunoblotting with each of the IgG anti-phosphatases. Pervanadate-mediated phosphorylation resulted in the coprecipitation of the tyrosine phosphatases SHP-1 (Fig. 3, panel A, lane 1) and SHP-2 (Fig. 3,  panel B, lane 1) with gp49B1. Despite the fact that mBMMCs contain SHIP (38), the enzyme did not coimmunoprecipitate detectably with native gp49B1 after exposure of mBMMCs to pervanadate (Fig. 3, panel C, lane 1).
To extend these findings to the association of phosphatases with native gp49B1 during physiologic inhibition of Fc⑀RImediated activation, Fc⑀RI on mBMMCs were sensitized with rat IgE, and gp49B1 was bound with rat anti-gp49B1; the cells were washed and extracted in detergent before and at several times within 60 s of cross-linking and coligation with F(abЈ) 2 mouse anti-rat IgG. gp49B1 bound by anti-gp49B1 was immunoprecipitated with goat anti-rat IgM-Sepharose, eluted and resolved by SDS-PAGE, and analyzed for the presence of coprecipitated phosphatases by immunoblotting. SHP-1 transiently associated with gp49B1, with maximal association occurring at 30 s, after which there was a decrease at 45 and 60 s (Fig. 4). In four additional experiments, maximal association of SHP-1 with gp49B1 occurred 15-45 s after coligation. In all cases, the decrease in SHP-1 association at the later time points was not the result of failure to immunoprecipitate gp49B1, as judged by sequential immunoblotting with anti-gp49B1 72-87 Ig (9)(data not shown). Several studies with time points beyond 60 s were entirely negative for coimmunoprecipitated SHP-1, and SHP-2 and SHIP were not detected in association with gp49B1 at any of the time points within the first 60 s (data not shown).
Effects of SHP-1 Deficiency on gp49B1-mediated Inhibition of Fc⑀RI-triggered Exocytosis in mBMMCs-mBMMCs were cultured from the bone marrow of motheaten-viable and ϩ/ϩ mice and were activated by the addition of F(abЈ) 2 mouse anti-rat IgG to cross-link and coligate the occupied Fc⑀RI and gp49B1 molecules. Coligation suppressed the release of ␤-hexosaminidase from ϩ/ϩ mBMMCs by 83-56% with increasing amounts of IgE, even though the percentage of exocytosis was near or at a plateau level at all IgE concentrations examined. For motheaten-viable mBMMCs, the gp49B1-mediated inhibition fell from 60 to 0% in the presence of increasing IgE input, even though the percentage of exocytosis in the absence of anti-gp49B1 was near or at plateau levels throughout and indistinguishable from the ϩ/ϩ mBMMCs (Fig. 5). Similar results were obtained for the generation and release of the cysteinyl leukotriene, LTC 4 , as measured by high performance liquid chromatography (n ϭ 3; data not shown).

Effects of gp49B1/Fc⑀RI Coligation on the Mobilization of Intracellular Calcium-The cross-linking of rat IgE bound to
Fc⑀RI on the surface of mBMMCs with F(abЈ) 2 mouse anti-rat IgG elicited an early, steep rise in the intracellular calcium level, beginning at 15 s and peaking at 90 s, followed by a sustained period of elevated intracellular calcium (Fig. 6A). There was no response to cross-linking gp49B1 in the absence of occupancy of Fc⑀RI by rat IgE (data not shown). When Fc⑀RI was cross-linked and coligated with gp49B1, there was no peak of calcium mobilization, and the level was ϳ30% of that achieved at 90 s with Fc⑀RI cross-linking alone (Fig. 6A). In three experiments, including that depicted in Fig. 6A, the peak early calcium flux after Fc⑀RI cross-linking occurred at 97 Ϯ 15 s (mean Ϯ S.E.), and cross-linking of Fc⑀RI with coligation to gp49B1 ablated the initial peak and reduced the calcium flux at 97 s to 29 Ϯ 2.7% (mean Ϯ S.E.) of that observed with crosslinking Fc⑀RI alone. FIG. 2. Effects of pervanadate on tyrosine phosphorylation of gp49B1. mBMMCs (1 ϫ 10 7 /ml) were incubated in medium with or without 2.5 mM pervanadate for 10 min at 37°C and were extracted in detergent buffer. gp49B1 was immunoprecipitated from extracts with anti-gp49B1 coupled to Sepharose (lanes 3 and 4), and parallel extracts were exposed to rat IgM mAb coupled to Sepharose as a negative control (lanes 1 and 2). Precipitates were deglycosylated with N-glycosidase F, resolved by SDS-PAGE, and transferred to an Immobilon P membrane. The membrane was immunoblotted with mouse anti-phosphotyrosine (anti-pY) mAb 4G10 and developed with HRP-labeled goat anti-mouse IgG (lower panel). The membrane was stripped and probed with rabbit anti-phospho-gp49B1 315-326 and developed with HRP-labeled goat antirabbit IgG (upper panel). IP, immunoprecipitation; IB, immunoblot.

FIG. 4. Association of SHP-1 with gp49B1 after coligation with
Fc⑀RI. Five samples of mBMMCs (1 ϫ 10 7 /ml) were incubated for 1 h with 3 g/ml of rat IgE at 4°C, and a saturating amount of anti-gp49B1 was added for the last 3 min. Cells were washed by centrifugation and subjected to Fc⑀RI cross-linking and coligation with F(abЈ) 2 mouse anti-rat IgG at 37°C for 0 -60 s. Samples were extracted in detergent, and gp49B1/anti-gp49B1 complexes were immunoprecipitated with goat anti-rat IgM coupled to Sepharose. Immunoprecipitates were resolved by SDS-PAGE and transferred to an Immobilon P membrane, which was immunoblotted with rabbit anti-SHP-1 and HRP-goat antirabbit IgG.
In the presence of EGTA, the peak time of calcium increase with cross-linking of Fc⑀RI alone occurred earlier (Fig. 6B) as a result of the inhibition of calcium influx (57 Ϯ 3 s, mean Ϯ S.E., n ϭ 3). Furthermore, cross-linking of Fc⑀RI with coligation to gp49B1 in the presence of EGTA virtually eliminated the increase in intracellular calcium that occurred with the crosslinking of Fc⑀RI alone. DISCUSSION We have established that gp49B1 initiates a counter-regulatory series of cellular responses within 60 s of coligation to cross-linked Fc⑀RI on the surface of mast cells. Inasmuch as exocytosis and lipid mediator generation are complete 120 -150 s after Fc⑀RI cross-linking in mBMMCs (39), the rapid recruitment of the tyrosine phosphatase SHP-1 to gp49B1 is consistent with the enzyme initiating an effective silencing pathway for the immediate-onset response of the mast cell to Fc⑀RI cross-linking. The known functions of SHP-1 (26) are compatible with the capacity of gp49B1 to inhibit the release of calcium from intracellular stores and the influx of extracellular calcium. Taken together, our findings demonstrate that an endogenous mast cell membrane protein can down-regulate IgE-mediated mast cell activation in nontransformed cells via the actions of recruited SHP-1.
The participation of gp49B1 ITIMs in the inhibition of mast cell activation was established by mutagenesis of their tyrosine residues in the context of native gp49B1. Inasmuch as the rat anti-gp49B1 does not bind to RBL-2H3 cells, this rat mucosal mast cell line (17) was optimal for transfection with wild-type and mutated mouse gp49B1 molecules. In these transfectants, the native extracellular domain of wild-type and mutant gp49B1 bound by anti-gp49B1 was coligated to Fc⑀RI sensitized with rat IgE by means of F(abЈ) 2 mouse anti-rat IgG, as described previously for mBMMCs (9) and utilized here for studies of calcium flux and SHP-1 deficiency in mBMMCs (Figs. 5 and 6). The coligation of wild-type, transfected gp49B1 to cross-linked Fc⑀RI inhibited the exocytosis achieved by the cross-linking of Fc⑀RI alone (Fig. 1). Mutation of the tyrosines in both ITIMs abolished the ability of coligated gp49B1 to inhibit ␤-hexosaminidase release, whereas mutation of the tyrosines individually caused a partial loss of inhibitory capacity (Fig. 1). When the PIR-B cytoplasmic domain was expressed as a chimeric receptor by transfection of RBL-2H3 cells, single tyrosine mutations in either of the two most downstream ITIMs (Tyr-794 and Tyr-824) resulted in either no appreciable effect (10) or partial loss of inhibitory function with each mutation (11); in yet another profile of effects, there was a partial loss of inhibition of calcium flux with mutation of the Tyr-794 ITIM and no loss for mutation of the Tyr-824 ITIM in a B cell line (12). The differences in these findings, each of which resulted from studies that used a single concentration of activating agonist, may relate to our findings that both the inhibitory capacity of gp49B1 ITIMs bearing single ITIM mutations and the SHP-1 independence of the native gp49B1 counter-regulatory mechanism decrease as the amount of IgE utilized for activation via Fc⑀RI on the cell surface increases.
The ability of endogenous mast cell tyrosine kinases to phosphorylate native gp49B1 in mBMMCs was established by inhibiting tyrosine phosphatases pharmacologically with pervanadate (Fig. 2). When native gp49B1 was phosphorylated by exposing mBMMCs to pervanadate, SHP-1 and SHP-2 were coprecipitated detectably with tyrosine-phosphorylated gp49B1, as assessed by SDS-PAGE/immunoblotting (Fig. 3). Only SHP-1 coprecipitated detectably with gp49B1 after coligation with Fc⑀RI under conditions that inhibit mast cell activation (Fig. 4), although it is possible that SHP-2 and/or SHIP may associate in amounts below their respective levels of immunochemical detection. The association of SHP-1 with native gp49B1 was transient, being detected only within 60 s of crosslinking of Fc⑀RI to itself with coligation to gp49B1. Although time course data have not been reported for the binding of SHP-1 to any of the human or mouse relatives of gp49B1, their cross-linking to themselves, coligation to an activating receptor, or binding to a natural ligand leads to their association with SHP-1 when assessed 1-2 min later (7, 10, 19, 26 -28). That the association of SHP-1 with gp49B1 is substantially reduced from the maximum by 60 s after coligation with Fc⑀RI ( Fig. 4) may indicate that tyrosine-phosphorylated gp49B1 is a substrate for SHP-1, inasmuch as certain phosphorylated ITIM peptides can be substrates for SHP-1 in broken cell assays (40). The ability of gp49B1 to inhibit Fc⑀RI-directed exocytosis (Fig.   FIG. 5. Effects of SHP-1 deficiency on gp49B1-mediated inhibition of exocytosis. mBMMCs generated from motheaten-viable (me-v) (open symbols and dotted lines) and wild-type (ϩ/ϩ) (closed symbols and solid lines) mice were sensitized at 1 ϫ 10 7 /ml with incremental amounts of rat IgE alone or in the presence of anti-gp49B1 and activated by adding F(abЈ) 2 mouse anti-rat IgG to both cross-link and coligate the occupied receptors, as described in Fig. 1. The data are expressed as the percentage of ␤-hexosaminidase released (mean Ϯ S.E.; n ϭ 3 for 10 and 35 g/ml and n ϭ 4 for 15 and 25 g/ml IgE).
FIG. 6. Effects of Fc⑀RI/gp49B1 coligation on intracellular calcium mobilization in mBMMCs. A, mBMMCs (1 ϫ 10 7 /ml) were sensitized with rat IgE for 1 h at 37°C in the absence and presence of a saturating amount of anti-gp49B1, labeled with 25 M Fura-2, and stimulated with F(abЈ) 2 mouse anti-rat IgG; the point of stimulation is indicated by the closed triangle. Free cytosolic calcium was detected by fluorometry. B, EGTA was added to replicate cells (open triangles) to chelate extracellular calcium before the addition of F(abЈ) 2 mouse antirat IgG (closed triangle). 5) in SHP-1-deficient mBMMCs was absent at high inputs of IgE, indicating that SHP-1 has a central role in mediating gp49B1-dependent inhibition of mast cell activation. However, there was substantial but incomplete gp49B1-dependent inhibition in the SHP-1-deficient mBMMCs at lower amounts of IgE that were still sufficient for near-maximal or maximal IgE-dependent exocytosis, which may reflect residual levels of SHP-1 activity but more likely reveals that SHP-1 is not the only transmitter of gp49B1-mediated inhibition of mast cell activation.
The intracellular calcium flux associated with the crosslinking of Fc⑀RI to itself was markedly suppressed by coligation of Fc⑀RI to gp49B1. Both the initial rate of release reflecting intracellular stores and the magnitude of calcium increase were markedly attenuated (Fig. 6A). In the presence of EGTA, the time of the peak calcium increase after Fc⑀RI cross-linking decreased by ϳ40 s (Fig. 6B), and the magnitude of total increase in calcium was markedly diminished, indicating that the peak at ϳ100 s in the absence of EGTA reflects the sum of two overlapping curves representing the release of intracellular calcium and the influx of extracellular calcium, respectively. In the presence of EGTA, the cross-linking of Fc⑀RI to itself with coligation to gp49B1 provided virtually no calcium flux, again indicating that gp49B1 mediated inhibition of the release of calcium from intracellular stores. The inhibition of NK cellmediated target cell lysis directed through KIRs and their associated SHP-1 is also characterized by an absence of calcium release from intracellular stores and lack of influx of extracellular calcium (26).
Type IIb1 and IIb2 Fc receptors for IgG (Fc␥RIIb1 and Fc␥RIIb2), which are the products of alternative splicing of mRNA from the Fc␥RII␤ gene (41), are Ig superfamily members that do not exhibit significant amino acid sequence homology with gp49B1 (9) but that nonetheless possess a single ITIM. We and others have shown that these receptors are expressed on mouse mast cells (42,43), and the coligation of the Fc␥RIIb species with cross-linked Fc⑀RI suppresses mast cell mediator release through a process that depends on ITIM phosphorylation (38,44,45). In contrast to gp49B1, the inhibition of mBMMCs activation by Fc␥RIIb1 and Fc␥RIIb2 proceeds through the inositol polyphosphate 5-phosphatase SHIP, resulting primarily in inhibition of Fc⑀RI-induced sustained influx of extracellular calcium, with little attenuation of the early increase in free cytosolic calcium (38). This difference highlights a distinction between the two inhibitory receptors in mast cells and may provide the mast cell with a degree of selectivity in its inhibitory processes. In addition, the fact that mouse mast cells express several ITIM-bearing receptors of the Ig superfamily (gp49B1, Fc␥RIIb1/b2, and PIR-B) is consistent with the hypothesis that these cells are controlled by an intricate network of inhibitory receptors. Inasmuch as gp49B1 and PIR-B have two and six Ig-like domains, respectively, the specificity of their SHP-1-transduced inhibition of exocytosis is likely regulated by their different counterligands.