INTRODUCTION

Aggregation of the high affinity IgE receptors (FcRI)¹ on basophils and mast cells initiates a cascade of events that results in the release of inflammatory mediators. This pathway includes the activation of several protein-tyrosine kinases such as Lyn, Syk, Btk, and Fak that induce the tyrosine phosphorylation of various proteins (1, 2, 3). There is also stimulation of phospholipase A₂, C, and D, the mobilization of Ca²⁺ from intracellular and extracellular sources and activation of serine and threonine kinases (4, 5).

The high affinity IgE receptor on mast cells and basophils is a tetrameric structure composed of the IgE binding  chain, a  subunit, and a homodimer of disulfide-linked  chains (6). The COOH-terminal cytoplasmic domain of FcRI and the cytoplasmic domain of FcRI appear to be important in receptor-mediated signal transduction (7, 8). These transducing subunits contain a cytoplasmic motif with the amino acid sequence (D/E)X₂YX₂LX_6-7YX₂(L/I) that is critical for cell activation (9, 10, 11). This motif called ITAM (for mmunoreceptor yrosine-based ctivation otif) (12) is present in the  and  subunits of FcRI, in the  subunit of the T-cell receptor complex, and in Ig and Ig of the B-cell receptor. This motif contains all the structural information essential for signal transduction and is rapidly tyrosine-phosphorylated after receptor aggregation (11, 13, 14, 15). Phosphorylation of the tyrosine residues appears to be necessary as mutants where the tyrosines are replaced with phenylalanine are inactive (16, 17). This phosphorylation is probably due to src family tyrosine kinases that associate with these receptors (11). After receptor aggregation, a Syk/ZAP-70 family protein tyrosine kinase associates with these receptors and is critical for the downstream activation signals (10, 11, 18, 19, 20, 21).

Fusion proteins containing the two Src homology 2 (SH2) domains of Syk bind preferentially to the tyrosine-phosphorylated  subunit of FcRI, whereas a fusion protein containing the single Lyn SH2 domain binds to tyrosine-phosphorylated FcRI (22, 23). Therefore, to better understand signal transduction from the FcRI, we used synthetic peptides based on the ITAM sequences of the  and  subunits of FcRI to investigate the interaction of molecules that may be involved in downstream propagation. We observed an association of Syk predominantly with the tyrosine-phosphorylated  ITAM, whereas Lyn, Shc, Grb2, and PLC1 interacted with the ITAM of FcRI. The binding of distinct downstream effector molecules to the different ITAMs may be critical in the activation of distinct signaling pathways from the different receptor subunits.

EXPERIMENTAL PROCEDURES

Streptavidin coupled to agarose beads were from Pierce (Rockford, IL). Protein A, aprotinin, and Triton X were obtained from Sigma. The materials for electrophoresis were purchased from Novex (San Diego, CA), and the source of other materials was as described previously (20).

Mouse monoclonal anti-PLC1 antibody was from Upstate Biotechnology Inc. (Lake Placid, NY), rabbit polyclonal anti-PLC2 antibody was from Santa Cruz Biotechnology (Santa Cruz, CA), mouse monoclonal anti-Grb2 antibody, mouse monoclonal, and rabbit polyclonal anti-Shc antibodies were from Transduction Laboratories (Lexington, KY). All other antibodies have been described previously (20, 24, 25).

Synthesis of ITAM Peptide of FcRI

Peptides based on the sequences of the  and  subunits of FcRI (6) were synthesized by solid-phase Fmoc chemistry on an Applied Biosystems 430A synthesizer. Phosphotyrosine residues were incorporated by using Fmoc-Tyr(PO₃H₂)OH (26). Peptides were also biotinylated on the resins. The purification and characterization of the peptides was as described previously (25).

RBL-2H3 cells were cultured and activated with anti-FcRI antibodies (mAb BC4) as described previously (1, 27). After 10 min of stimulation, the monolayers were rinsed twice with ice-cold phosphate-buffered saline and solubilized in lysis buffer (10 mM Tris, pH 7.5, containing 1.0% Triton X-100, 1 mM Na₃VO₄, 150 mM NaCl, 50 µg/ml leupeptin, 0.5 unit/ml aprotinin, 2 mM pepstatin A, 1 mM phenylmethylsulfonyl fluoride). After leaving the plates on ice for 10 min, the cells were scraped and supernatants were collected after centrifugation for 30 min at 16,000 × g, 4 °C.

Lysates from 1.5 × 10⁷ cells in 1.0 ml were precleared by mixing for 90 min at 4 °C with streptavidin coupled to agarose beads. The lysates were then incubated with 1 nmol of biotinylated ITAM peptides that had been preincubated with 20 µl of streptavidin beads. After gentle rotation at 4 °C for 90 min, the beads were washed three times with wash buffer (lysis buffer with Triton concentration decreased to 0.5%), once with 150 mM NaCl, 50 mM Tris, pH 7.4, and the proteins eluted by boiling for 5 min with Laemmli's sample buffer as described previously (22).

For reimmunoprecipitation experiments, proteins were first precipitated with ITAM peptides bound to beads and then eluted by boiling for 10 min in 1% SDS. The proteins were reimmunoprecipitated with antibody prebound to protein A and analyzed by immunoblotting.

Precipitation was performed as above with the biotinylated ITAM peptides bound to the streptavidin beads. After the four washes, the beads were further washed with kinase buffer (30 mM HEPES, pH 7.5, 10 mM MgCl₂, and 2 mM MnCl₂), and resuspended in 20 µl of kinase buffer. The kinase reaction was started by the addition of 1 µCi of [-³²P]ATP (ICN) and incubated for 10 min at room temperature. The reaction was stopped by washing once with 1 ml of ice-cold 50 mM Tris, 150 mM NaCl, then by adding sample buffer. After boiling for 5 min, the proteins were separated by SDS-PAGE electrophoresis on 10% Tris glycine gels, transferred to PVDF (polyvinylidene difluoride membranes, Millipore, Bedford, MA) membranes, and ³²P was detected by autoradiography at 70 °C.

Samples from the precipitates were separated by SDS-PAGE under reducing conditions and electrotransferred to PVDF membranes. The membranes were incubated overnight in blocking buffer. Tyrosine-phosphorylated proteins were detected with monoclonal antibody PY-20 conjugated to horseradish peroxidase as described previously (28). Lyn, Syk, Shc, PLC2, and FcRI were detected by rabbit antibodies using peroxidase-conjugated donkey anti-rabbit as a secondary antibody. PLC1, Grb2, and FcRI were detected by mouse antibodies and the secondary antibody was peroxidase-conjugated donkey anti-mouse antibody. In all blots, proteins were visualized using the enhanced chemiluminescence kit from DuPont and Kodak X-Omat radiographic film (Eastman Kodak Co.). In some experiments antibodies were stripped from the membranes according to the protocol of the manufacturer and then membranes reprobed with other antibodies. Scanning densitometry on the film was with a Pharmacia LKB Imagemaster.

The SH2 domains of Lyn and Syk expressed as GST fusion proteins have been described previously (22). The plasmids containing the SH2 domains of PLC1 and Shc as fusion proteins were kindly provided by Dr. T. Pawson, Mount Sinai Hospital Research Institute, Toronto, Canada. The proteins were expressed in Escherichia coli, affinity purified on glutathione-Sepharose beads, and characterized as described previously (22).

PVDF membranes were used to determine the interaction of GST fusion proteins with the ITAM peptides. Ten pmol and 2 pmol of the GST fusion proteins containing the SH2 domains of PLC1, Shc, Syk, and Lyn were spotted onto PVDF membranes. The membranes were then blocked by overnight incubation with 4% bovine serum albumin and for 1 h with 100 µg/ml donkey IgG. After extensive washing, membranes were incubated with 1 µM biotinylated ITAM peptides for 1 h. The membranes were then washed and incubated with horseradish peroxidase-conjugated streptavidin and the signal was detected by enhanced chemiluminescence.

RESULTS

Previously we observed that Syk coprecipitated with FcRI in activated cells (20) and that fusion proteins containing the SH2 domains of Syk bound to the  subunit of FcRI (22). To further investigate these interactions we used synthetic peptides that encompass the ITAM motif of both the  and  subunits of FcRI (Fig. 1). The different ITAM synthetic peptides (both non-phosphorylated and tyrosine-phosphorylated) were used to affinity precipitate proteins from both non-stimulated and stimulated RBL-2H3 cells (Fig. 2). The tyrosine-diphosphorylated  and  ITAM peptides precipitated several proteins that were tyrosine-phosphorylated. The most prominent was a 72-kDa protein that was only precipitated from lysates of stimulated RBL-2H3 cells by tyrosine-phosphorylated  ITAM peptide. In contrast,  phosphorylated ITAM peptide precipitated 145- and 52-kDa tyrosine-phosphorylated proteins, both of which were present in both non-stimulated and activated cells. However, no tyrosine-phosphorylated proteins were precipitated by the non-phosphorylated ITAM peptides. These results indicate that proteins were selectively precipitated by ITAM peptides and that the ITAM peptides had to be tyrosine-phosphorylated for them to bind these proteins.

Synthetic peptides corresponding to the ITAMs of the  and  subunits of FcRI.

Tyrosine-phosphorylated ITAM peptides precipitated Syk, Shc, and PLC1.

Several antibodies were used to examine whether these ITAM-associated proteins were previously identified tyrosine-phosphorylated proteins. Immunoblotting identified Syk as the 72-kDa phosphorylated protein (Fig. 2C). The immunoblots also demonstrated that Syk was precipitated by the PP and the monophosphorylated YP peptides. As Syk was approximately equally precipitated from both non-stimulated and stimulated cells, tyrosine phosphorylation of Syk is not important for its interaction with the ITAM peptides. On longer exposures we could detect some Syk precipitation by the other monophosphorylated PY peptide, whereas no Syk was precipitated by the non-phosphorylated peptides. Combining the two different monophosphorylated  peptides did not result in increased precipitation of Syk (Fig. 3). Similarly addition of the mono-phosphorylated  peptides (PY and YP) to PP did not affect the binding of Syk. The efficiency of precipitating Syk from the cell lysates was PPPP>YP (by densitometric analysis the ratios were 100:13:0.5). These results support the reports of the preferential interaction of Syk with the tyrosine-phosphorylated  subunit of FcRI and demonstrate that this binding requires that both tyrosines in the ITAM be phosphorylated. Previously we observed that fusion proteins containing the two SH2 domains of Syk were more efficient than either one of the two SH2 domains alone in interacting with the  subunit of FcRI. Therefore, the binding of Syk to  requires both SH2 domains of Syk and phosphorylation of both tyrosines of the ITAM.

Activation of RBL-2H3 cells results in tyrosine phosphorylation of several 72-kDa proteins that are not Syk (20, 29). We therefore investigated whether Syk was the only 72-kDa tyrosine-phosphorylated protein precipitated with the PP peptide. Syk was depleted from cell lysates by incubation with protein A beads preincubated with anti-Syk antibodies. This removed in parallel the 72-kDa tyrosine-phosphorylated protein precipitated by PP (data not shown). To further confirm that the 72-kDa tyrosine-phosphorylated protein was Syk, the proteins precipitated from activated cell lysates with peptide-streptavidin beads were eluted by boiling in SDS. These were then immunoprecipitated with control or anti-Syk antibodies. Analysis by anti-phosphotyrosine and anti-Syk antibodies confirmed that this 72-kDa tyrosine-phosphorylated protein was Syk (data not shown). Therefore, Syk was the only tyrosine-phosphorylated 72-kDa protein precipitated with PP.

In vitro kinase reactions were used to further define proteins that were precipitated with the ITAM peptides (Fig. 4). The major proteins phosphorylated in vitro were 72 kDa in size and were seen after precipitation with the PP peptide. The long exposure of this radioautograph is used here to show that similar bands were detected with PP and YP. Depletion of Syk from lysates with antibodies confirmed that this in vitro kinase activity was due to Syk. As was observed in the immunoblotting experiments, there was good association of the kinase with the ITAM peptides in both stimulated and non-stimulated cells.

Phosphorylated ITAM Peptides of FcRI Precipitated Lyn, PLC1, Shc, and Grb2 from Lysates of RBL-2H3 Cells

Several studies suggest that Lyn associates with FcRI in both non-stimulated and stimulated cells (30, 31, 32, 33) and that this interaction is preferentially with the  subunit of the receptor (22, 34). However, we could not detect Lyn precipitation with the ITAM peptides by immunoblotting or by the more sensitive in vitro kinase reaction (Fig. 2 and 3). There was still no Lyn when ITAM peptides based on the  and  subunits of FcRI were used together for precipitation (data not shown). Under these conditions Lyn is normally tyrosine-phosphorylated which could result in masking of the SH2 domain. Interestingly, when cell lysates were prepared in the absence of vanadate to allow for dephosphorylation, there was binding of Lyn to the diphosphorylated  ITAM peptide (Fig. 5). In contrast, binding to the -diphosphorylated peptide was weaker. These results suggest that tyrosine phosphorylation of the regulatory site of Lyn controls its binding to FcRI after cell activation.

Tyrosine-phosphorylated proteins of 145, 52, and 46 kDa were selectively precipitated by the diphosphorylated  ITAM peptide (Fig. 2). Using a panel of antibodies, the 145-kDa protein was identified as PLC1. This identification of PLC1 was confirmed by reimmunoprecipitation and depletion experiments as was described above for Syk (data not shown). Although PLC1 and PLC2 are structurally very similar, only PLC1 was precipitated by PP (data not shown). PLC1 was not precipitated by mono- or diphosphorylated  peptides (Fig. 2). Furthermore, addition to PP of the monophosphorylated PY peptide, but not of the diphosphorylated PP peptide, resulted in a decrease in the precipitation of PLC1 (Fig. 3). Altogether this data suggests that PLC1 can interact with the tyrosine-phosphorylated  subunit of FcRI.

The 52- and 46-kDa proteins were identified as Shc by blotting with both monoclonal and polyclonal anti-Shc antibodies (Fig. 2). In longer exposures, Shc could be visualized in the precipitates with PP from both non-stimulated and stimulated cells. The PP peptides still precipitated PLC1 from cell lysates depleted of Shc. Therefore, the interaction of PLC1 with PP is independent of Shc. There was increased precipitation of Shc from non-stimulated cells, but not from stimulated cells with the mixture of YP and PP peptides (Fig. 3). These results suggest that Shc can preferentially bind to the tyrosine-phosphorylated  ITAM and that this binding may be modulated by the ITAM of the  subunit.

When Shc protein is tyrosine-phosphorylated there is strong SH2-mediated binding of Grb2, an adaptor protein (35). Grb2 was precipitated by the diphosphorylated  subunit of the ITAM peptide (Fig. 6). However, none of the other synthetic tyrosine-phosphorylated and non-phosphorylated peptides precipitated Grb2.

Grb2 was precipitated by diphosphorylated FcRI ITAM peptide.

A membrane binding assay was used to examine the potential for direct interaction between the SH2 domains of the different proteins with the synthetic ITAM peptides. GST fusion proteins containing the SH2 domains were bound to membranes and then reacted with two concentrations of the synthetic ITAM peptides (Fig. 7). The Syk fusion proteins contained the amino-terminal, carboxyl-terminal, or both SH2 domains expressed in tandem (Fig. 7A). Diphosphorylated  and  ITAM peptides bound to fusion proteins containing the SH2 domains of Syk. By contrast, there was very little binding of ITAM peptides to the NH₂-terminal SH2 domain of Syk, more to the COOH-terminal SH2 domain, and markedly stronger binding when both SH2 domains were expressed in tandem. Binding by the lower concentrations of the fusion proteins to the diphosphorylated ITAM of the  subunit was better than with ITAM of the  subunit. This suggests that the binding to  ITAM was with higher affinity. The monophosphorylated peptides based on the  ITAM bound much less and this binding was only when both SH2 domains of Syk were expressed as fusion proteins. There was very little selectivity in the binding by the Lyn SH2, with it binding to some extent to all the mono- and diphosphorylated synthetic peptides. Nevertheless, neither the Syk nor the Lyn SH2 fusion proteins bound to the non-phosphorylated ITAM synthetic peptides (not shown).

Interaction of ITAM peptides with GST fusion proteins containing the SH2 domains of PLC1 and Shc. Ten and 2 pmol of the different GST fusion proteins containing the SH2 domains were spotted on PVDF membranes, blocked, and then incubated with 1 µM biotinylated ITAM peptides. The membranes were washed and incubated with horseradish peroxidase-conjugated streptavidin and proteins were detected by enhanced chemiluminescence. A, SH2 containing proteins were: the NH₂-terminal (SykSH2, N), COOH-terminal (SykSH2, C), both NH₂-terminal and COOH-terminal in tandem (SykSH2, NC) of Syk and Lyn (Lyn, SH2). GST alone was used as a control (GST). B, the SH2 containing fusion proteins used were the following: the NH₂-terminal (PLC1SH2, N), COOH-terminal (PLC1SH2, C), the NH₂-terminal and COOH-terminal expressed in tandem (PLC1SH2, NC) of PLC1 and Shc (Shc-SH2).

The SH2 domains of phospholipase C1 and Shc were also tested in membrane binding assays (Fig. 7B). The amino-terminal, carboxyl-terminal, and tandemly expressed SH2 domains of phospholipase C-1 all strongly bound to the diphosphorylated  ITAM. The protein with both SH2 domains of PLC1 bound weakly to the non-phosphorylated  and  ITAM peptides and to the phosphorylated  ITAM peptide. The lack of the binding of the two SH2 domains of PLC1 to the monophosphorylated  ITAM peptides indicates the specificity in the binding with this fusion protein. In this membrane binding assay, the SH2 domain of Shc interacted with only the -phosphorylated ITAM peptide. Therefore, the SH2-mediated binding of Syk is to the tyrosine-phosphorylated ITAM of FcRI, whereas PLC1 and Shc binding is to the tyrosine-phosphorylated ITAM of FcRI.

DISCUSSION

One of the earliest events after aggregation of FcRI is the activation of kinases that result in the rapid phosphorylation on Tyr and Ser of the  subunit of the receptor and on Tyr and Thr of the  subunit (36). The phosphorylation on tyrosines of the ITAM in these subunits then recruits other proteins that are crucial for propagating the intracellular signals. The binding of such proteins is mediated by SH2 domains present on many proteins involved in signal transduction (35). The specificity of SH2 domains for phosphorylated tyrosines (37) would explain the selectivity of the interactions of Syk, Lyn, PLC1, and Shc with the  and  subunits of the FcRI. Binding of these downstream signaling molecules then can result in conformational changes and in their activation (25, 38).

The present results further refine the model for FcRI-induced activation of mast cells. Aggregation of the receptor results in the activation of Lyn which then tyrosine phosphorylates the  and  subunits of the receptor. Syk is then recruited and activated by binding to the  subunit of the receptor. PLC1 is recruited to the  subunit of the receptor, bringing it in close proximity to Syk where it could be efficiently tyrosine phosphorylated. Previous experiments suggest that both the  and  subunits of FcRI contribute to signal transduction. Chimeric proteins have been expressed in RBL-2H3 cells that contain the extracellular and transmembrane domains of CD25 fused with the carboxyl-terminal domain of either the  or  subunits of FcRI (8, 34, 39, 40). Whereas there is some secretion by aggregation of the chimeric proteins that contain the  domain, there is minimal if any by the  containing chimeras. Furthermore, signaling by the chimeric proteins containing the  sequence is less than through the normal FcRI, suggesting that the  subunit is important for signal transduction. The present results suggest an explanation for such observations requiring cooperation between the two receptor subunits in inducing secretion in mast cells.

The much stronger binding of Syk with the  than with the  ITAM-diphosphorylated peptides is probably due to the fine structural differences in the sequence of the ITAM of these two molecules. There is also one less amino acid between the two tyrosines in the  motif. By membrane binding studies, the best binding to ITAM peptides was with fusion proteins containing both SH2 domains of Syk expressed in tandem. These results strongly suggest that the two SH2 domains of Syk and phosphorylation of both tyrosines in the ITAMs are necessary for optimal binding. The recently described crystal structure of ZAP-70 associated with the phosphorylated ITAM of  T cell receptor supports this model for binding (41). The binding of Syk/ZAP-70 tyrosine kinases to the tyrosine-phosphorylated ITAM peptides results in their activation and tyrosine phosphorylation (25, 42, 43). The interaction of Syk with the tyrosine-phosphorylated ITAM results in a conformational change (25) and an increase in its kinase activity (42, 43). These conformational changes in Syk and increased enzymatic activity are probably critical for downstream propagation of intracellular signals in both mast cells and B cells (20, 21, 44, 45, 46).

Lyn, a member of the Src family of protein-tyrosine kinases, is associated with FcRI in both non-stimulated and activated cells (30, 31, 32, 33). The present precipitation results suggest that after cell activation the interaction of Lyn is predominantly with the  subunit. There is association of Lyn with the COOH-terminal cytoplasmic domain of the  subunit which contains the ITAM, although mutation of one of the Tyr to Phe does not decrease this association (34). Furthermore, a fusion protein containing only the SH2 domain of Lyn precipitates and binds with the tyrosine-phosphorylated  subunit and with the phosphorylated ITAM. These data suggest that the SH2 domain of Lyn can potentially bind the phosphorylated ITAM, but in the intact Lyn this SH2 site is not available due to intramolecular interaction with its regulatory domain. Stimulation of cells and activation of Lyn could result in dephosphorylation of these regulatory sites and the release of the SH2 domain for binding to the phosphorylated subunits of FcRI that are in close proximity. This can explain the increased association of Lyn with FcRI after cell stimulation (32). However, Lyn is tyrosine phosphorylated to the same extent in non-stimulated compared to stimulated cells, suggesting that there may be rapid rephosphorylation of these regulatory sites. Therefore, Lyn association with FcRI in activated cells is probably mediated by SH2 domain binding.

Ligand binding to receptor protein-tyrosine kinases, such as the epidermal growth factor receptor or the platelet-derived growth factor receptor, induces receptor oligomerization and autophosphorylation, generating docking sites for PLC1 (47). PLC1 can also associate with ITAM-containing receptors such as the T cell receptor (48). PLC1 is tyrosine-phosphorylated after FcRI activation (49, 50) and translocates to the membrane (51). Tyrosine phosphorylation activates PLC1 inducing it to hydrolyze phosphatidylinositol 4,5-bisphosphate to produce two second messengers, inositol trisphosphate and diacylglycerol, these in turn result in the increase in intracellular Ca²⁺ and the activation of protein kinase C (52). PLC1 associates with a tyrosine kinase after FcRI aggregation (53), and a complex containing Syk, PLC1, and a 120-kDa phosphoprotein has been seen in B-cells (54). This 120-kDa protein has been postulated to be a bridge between Syk and PLC1. Similarly, fusion proteins containing the SH2 domains of PLC1 bind tyrosine-phosphorylated Syk (55). However, the present experiments suggest direct interaction of PLC1 with the tyrosine-phosphorylated ITAM of FcRI on the basis of both precipitation and membrane binding studies. Furthermore, the PP peptide precipitated Syk, but there was no detectable PLC1. PLC1 has two SH2 domains, one SH3, and two split pleckstrin homology domains. The interaction between PLC1 and the phosphorylated FcRI ITAM is SH2-mediated, whereas the pleckstrin homology domains may play a role in other membrane interactions.

Tyrosine-phosphorylated  ITAM peptide precipitated Shc and Grb2, two members of a family of adaptor molecules that have SH2 domains. Stimulation of T cell and B cell antigen receptors results in tyrosine phosphorylation of Shc and the formation of a complex that contains Shc, Grb2, and Sos (56, 57). However, in RBL-2H3 cells Shc is constitutively tyrosine- phosphorylated (data not shown and Ref. 58). The single SH2 domain at the carboxyl terminus of Shc probably bound to the tyrosine-phosphorylated  ITAM sequence pYEEL (37). Grb2 is another adaptor molecule with an SH2 domain that binds to tyrosine-phosphorylated Shc. The single SH2 domain of Grb2 is flanked by two SH3 domains. One of the SH3 domains binds Sos1, the Ras guanine nucleotide exchange factor, and thus the complex can activate the Ras pathway (59). FcRI aggregation results in the activation of the ERK1 and ERK2 (60, 61), which are probably important for regulating nuclear events and for the generation of arachidonic acid (62). Therefore, association of Shc with the  subunit of FcRI could be important for initiating the activation of the Ras pathway that leads to nuclear events, such as the synthesis of cytokines and for the generation of arachidonic acid.