Interaction of Phosphorylated FcϵRIγ Immunoglobulin Receptor Tyrosine Activation Motif-based Peptides with Dual and Single SH2 Domains of p72syk

To examine the characteristics of the interaction of the FcεRIγ ITAM with the SH2 domains of p72syk, the binding of an 125I-labeled dual phosphorylated FcεRIγ ITAM-based peptide to the p72syk SH2 domains was monitored utilizing a novel scintillation proximity based assay. The Kd for this interaction, determined from the saturation binding isotherm, was 1.4 nM. This high affinity binding was reflected in the rapid rate of association for the peptide binding to the SH2 domains. Competition studies utilizing a soluble C-terminal SH2 domain knockout and N-terminal SH2 domain knockouts revealed that both domains contribute cooperatively to the high affinity binding. Unlabeled dual phosphorylated peptide competed with the 125I-labeled peptide for binding to the dual p72syk SH2 domains with an IC50 value of 4.8 nM. Monophosphorylated 24-mer FcεRIγ ITAM peptides, and phosphotyrosine also competed for binding, but with substantially higher IC50 values. This, and other data discussed, suggest that high affinity binding requires both tyrosine residues to be phosphorylated and that the preferred binding orientation of the ITAM is such that the N-terminal phosphotyrosine occupies the C-terminal SH2 domain and the C-terminal phosphotyrosine occupies the N-terminal SH2 domain.

Src homology 2 (SH2) 1 domains are regions of approximately 100 -120 amino acid residues present in a variety of proteins including tyrosine kinases, tyrosine phosphatases, phospholipases, and other signal transducing proteins (1)(2)(3)(4)(5)(6). These domains bind with high affinity to tyrosine containing motifs in associating proteins, such as specific cytokine and immunoglobulin receptor subunits, adapter proteins, tyrosine kinases, and other signaling molecules such as STATs, when these motifs are phosphorylated by the action of tyrosine kinases (1-5, 7, 8). This allows recruitment of SH2 domain-containing signaling molecules and phosphotyrosine-containing signaling molecules into receptor-linked signal transduction assemblies (9 -13).
The tyrosine-containing motifs are composed of phosphorylated tyrosine residues followed by 3-4 amino acids (e.g. pYXX(L/I)) which carry the sequence-specific information for SH2 recognition (14 -23). These motifs can occur in isolation or in tandem, thus can bind single SH2 domains (e.g. of src related tyrosine kinases) (4, 24 -26) or dual SH2 domains (e.g. of p70 zap and p72 syk ) (27)(28)(29)(30)(31)(32)(33)(34). Certain antigen receptor subunits, such as the subunit of the T cell receptor (TCR), the Ig␣ and Ig␤ subunits of the B cell receptor and the ␤ and ␥ subunits of the high affinity IgE receptor (Fc⑀RI), contain tyrosine motifs in tandem and these have been termed immunoglobulin receptor tyrosine activation motifs (ITAMs) (24, 27, 32, 34 -36). In hematopoietic cell signaling, tyrosine-phosphorylated ITAMs have been shown to be critical for signaling interactions via their association with the SH2 domains of tyrosine kinases. For example, p70 zap and the src-related kinases p59 fyn and p56 lck , appear to play a role in TCR-mediated T cell activation (32, 36 -40), whereas p72 syk and p56 lyn appear to play a role in B cell receptor-mediated B cell activation (24,41,42) and Fc⑀RImediated mast cell activation (33,34,(43)(44)(45)(46). Data suggest that, in mast cells, Fc⑀RI aggregation results in enhanced p56 lyn catalytic activity leading to phosphorylation of the ITAM tyrosine residues contained within the Fc⑀RI␤ and the Fc⑀RI␥ C termini (33,34). This results in further recruitment of p56 lyn to the Fc⑀RI␤ ITAM and recruitment of p72 syk to the Fc⑀RI␥ ITAM (33,34). The resulting increase in the catalytic activity of p72 syk leads to activation of downstream signaling events and ultimately, degranulation.
The objective of this study was to determine the binding properties of the Fc⑀RI␥ ITAM to the p72 syk dual SH2 domains. We have done this by examining the affinities and kinetics of * 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. the binding of Fc⑀RI␥ ITAM-based peptides to the dual and single SH2 domains of p72 syk . Previously, binding of ITAM peptides to a variety of SH2 domains has been assessed by a number of different approaches including ELISA-based assays (15), surface plasmon resonance determination (Biacore) (8,10,17,21,26), and isothermal titration calorimetry (17,25). The ELISA-based assays are generally formatted with ITAM-based peptides immobilized to ELISA plates and the SH2 domains in solution. Surface plasmon resonance determination involves immobilizing the ITAM peptides on sensor chips. An increase in binding of the SH2 domains to the immobilized peptides and the resulting increase in mass is detected by a change in the refractive index of the sensor chip surface (17). Isothermal titration calorimetry involves determining the extent of binding based on the entropy of the binding reaction (25). Unfortunately, published studies utilizing these techniques have shown considerable variation in the apparent affinities of binding. This variation is a consequence of experimental limitations associated with the assay parameters (17). For example, as GST fusion proteins dimerize in solution, unless the ligand density on the solid phase is reduced, avidity effects could result in inaccurate determination of K d values (17). Avidity effects may also account for the apparently long off-rates described in the literature for certain phosphotyrosine-SH2 domain interactions (8). Furthermore, as competition-based assays utilizing both the ELISA and surface plasmon resonance formats are configured with the ITAM peptides immobilized on solid phase and the competing peptides in solution, a true competition equilibrium may not be achieved. Finally ELISAbased binding assays require several washing steps which may compromise the accuracy of K d determinations. To, therefore, accurately determine the affinities and the binding kinetics of Fc⑀RI␥ ITAM-based peptides to the dual and single SH2 domains of p72 syk , a novel scintillation proximity (SPA) based assay system was developed which circumvents the experimental limitations described above.

MATERIALS AND METHODS
Plasmid Expression Vectors-Human p72 syk cDNA was initially cloned by reverse transcriptase-PCR from the KU812 human mast cell line. Three oligonucleotide primers were then prepared on an Applied Biosystems DNA synthesizer: A, 5Ј-TCAGTTCTCGAGGAACCACCT-GCCCTTCTTTTTCG-3Ј; B, 5Ј-TACGTTCTCGAGGTCCCCTATACTA-GGTTATTGGAAA-3Ј; C, 5Ј-CAGGTAAGATCTCCTTATTTTTGACAT-GGGACAGTAAGA-3Ј. Primers A and C were used to amplify a segment of human p72 syk cDNA (47) corresponding to amino acid residues 10 -261, using standard PCR methodology (48). The PCR fragment was cleaved with XhoI and BglII, and then ligated with the XhoI/BamHI vector fragment of pDW363 (49) to construct p(bio-dual Syk SH2) (pTC1). p(bio-dual GST-Syk SH2) (pTC2) was constructed by amplifying a portion of a previously constructed dual GST-Syk SH2 fusion vector with primers B and C containing the same SH2 domain amino acid residues as stated above, cleaving the PCR product with XhoI and BglII, and then ligating it with the XhoI/BamHI vector backbone of pDW363 (49). Standard recombinant DNA techniques were employed (48), and the structures of the recombinant plasmids were confirmed by dideoxy nucleotide sequencing (50).
For the N-terminal Arg-42 mutation, PCR reactions were performed using the dual SH2 domain of p72 syk as the template with primers 1 and 2 (product A) and primers 3 and 4 (product B). Aliquots of products A and B were combined in a second PCR reaction with primers 1 and 4. For the N-terminal Arg-45 mutation, PCR reactions were performed using the dual SH2 domains of p72 syk as the template with primers 6 and 7 (product A) and primers 5 and 8 (product B). Aliquots of products A and B were combined in a second PCR reaction with primers 5 and 6. For the C-terminal mutation, PCR reactions were performed with primers 6 and 9 (product C) and primers 5 and 10 (product D) using the p72 syk dual SH2 domains as the template. Aliquots of products C and D were combined in a second PCR reaction with primers 5 and 6. The second PCR products (mutated p72 syk SH2 domain constructs) were purified (Wizard PCR preps, Promega), cleaved with BamHI and EcoRI, purified using low melt agarose followed by Wizard PCR preps, and subcloned into the pGex-3X vector (Pharmacia). The subclones were sequenced to verify inclusion of the desired mutations.
Expression of Biotinylated and Non-biotinylated Forms of Syk SH2 Proteins in Escherichia coli-Cultures of MC1061 cells (51) containing pTC1 or pTC2 were grown overnight from single colonies at 37°C in LB (48) supplemented with 50 M biotin and 100 g/ml ampicillin. The cultures were diluted 10-fold in the same medium, grown to OD 600 ϭ 1.0 at 37°C and, after reducing the temperature to 30°C, expression was induced with 1 mM isopropyl-1-thio-␤-D-galactopyranoside for 3 h. Cells were collected by centrifugation and the cell pellet was resuspended in ice-cold lysis buffer (50 mM Tris-HCl, pH 8.0, 200 mM NaCl, 1 mM EDTA, 1 mM dithiothreitol, 5% glycerol, 100 g/ml phenylmethylsulfonyl fluoride, 25 g/ml leupeptin, 25 g/ml pepstatin A, and 1% Triton X-100) (20 ml/liter of original culture media). Lysozyme (20 g/ml) was added to the cell suspension and the cells were gently shaken for 1 h at 4°C. MgCl 2 (final concentration: 20 mM), DNase, and RNase (final concentration: 20 g/ml) were then added and the cells were incubated for an additional 1 h at 4°C with shaking. The cell debris was removed by centrifugation for 45 min at 10,000 ϫ g and the proteins were purified from the supernatant as described below.
Purification of Biotinylated and Non-biotinylated Syk SH2 Proteins-The bio-dual Syk SH2 and bio-dual GST-Syk SH2 proteins were purified by affinity chromatography using Soft-link monomeric avidin resin (Promega) or glutathione-Sepharose (Pharmacia), respectively, in accordance with the instructions provided by the manufacturers. After elution from the resin, the proteins were dialyzed exhaustively against Tris buffer (50 mM Tris-HCl, pH 8.0, 100 mM NaCl) then concentrated by diafiltration using an Amicon membrane with a 10-kDa molecular mass cut off. Protein concentrations were determined spectrophotometrically (A 280 reading) or by a BCA protein determination kit (Pierce).
Non-biotinylated p72 syk SH2 domain protein was expressed in E. coli as a GST fusion protein, with a Factor Xa cleavage site directly preceeding the p72 syk SH2 domains. The fusion protein was purified (Ͼ90% pure) by affinity chromatography on a glutathione-Sepharose (Pharmacia) column and a fraction was then digested with Factor Xa to release the SH2 domain fragment. The fragment was purified to homogeneity from the Xa-digest. On SDS gels, this protein appeared as a 29-kDa band. The fusion protein was purified (Ͼ90% pure) by affinity chromatography on a glutathione-Sepharose (Pharmacia) column. N-terminal sequencing and mass spectrometric analysis indicated that the protein was Ͼ95% pure. For some assays, the purified GST fusion protein was biotinylated chemically using either a 5-fold molar excess of sulfosuccinimidyl-6-(biotinamido)-hexanoate (Immunopure-NHS-Lc-biotin, Pierce) or a 10-fold molar excess of Biotin-xx-NHS (Calbiochem) in the presence of phosphate-buffered saline, pH 7.4. The excess biotin was removed by extensive dialysis against phosphate-buffered saline containing 2 mM dithiothreitol at 4°C.
SDS-PAGE and Western Blot-Samples were removed from the cultures before and after induction with isopropyl-1-thio-␤-D-galactopyranoside, and the cells were pelleted at 10,000 ϫ g. The pellets were resuspended in 200 l of sample buffer (52) and heated at 100°C for 4 min. The proteins in the samples were resolved by SDS-PAGE on a 12% pre-cast mini-gel (Bio-Rad). After electrophoresis, the gel was either stained with Coomassie Brilliant Blue or the proteins were electroblotted onto a nitrocellulose membrane using a Bio-Rad Trans-Blot SD device at 100 volts for 1 h. Following transfer, the membrane was treated overnight at 4°C with blocking buffer (25 mM Tris-HCl, pH 7.5, 0.15 M NaCl, 0.2% Tween 20 and 5% BSA), washed three times with blocking buffer (without BSA), then incubated with streptavidin-horseradish peroxidase conjugate (Boehringer Mannheim; 1/2000 dilution) for 1 h at room temperature. After washing the membrane three times with blocking buffer (without BSA), the streptavidin-reactive proteins were visualized using the ECL detection system (Amersham).

Synthesis and Iodination of Phosphorylated Fc⑀RI-␥ ITAM-based
Peptide-A peptide (YKSDGVY 64 (p)TGLSTRNQETY 75 (p)ETLKHEK-NH 2 ; bold type denotes critical phosphotyrosine residues) corresponding to amino acid residues 59 -82 of the human Fc⑀RI␥ chain, which includes the ITAM region with an additional tyrosine residue, was synthesized on a 131 Applied Biosystems instrument. The side chain protections were as follows: Arg(Pmc), Asn(Trt), Asp(OtBu), Gln(Trt), Glu(OtBu), His(Trt), Lys(Boc), Ser(OtBu), and Thr(OtBu). The synthesis was performed on benzhydrylamine resin functionalized with (53) as starting material and Fmoc protection was used for all amino acids. The ␣-amino Fmoc protection was removed with 20% piperidine in dimethylformamide and acylations were achieved by a preformed hydroxybenzotriazole active ester. After the amino acid synthesis was complete, the resin was transferred to a reaction vessel on a Glass Col bench top shaker and the tyrosines at positions 64 and 75 were phosphorylated using the procedure described by Kitas et al. (54) and Bannworth et al. (55). The di-O-benzyl-phosphotyrosine containing peptide was cleaved from the resin and the protecting groups were removed by the action of trifluoroacetic acid (0.05 M) containing ethanedithiol (8 equivalents), thioanisole (10 equivalents), and anisole (4 equivalents) at 25°C for 2 h. The crude peptide was precipitated in ether, collected, purified by preparative HPLC (1-20% acetonitrile, H 2 0, 0.1% trifluoroacetic acid), and lyophilized. The purity of the peptide was assessed by analytical HPLC and mass spectrometry. Analytical HPLC was performed on a Milton Roy CM 4000 system using a Milton Roy Spectromonitor #100 system and a Vydac C 18 analytical column. Preparative HPLC was performed on a Waters 4000 system using a Waters 486 detector and a Vydac C 18 preparative column.
The peptide was iodinated by the IODO-GEN procedure (Pierce Chemical Co.). The peptide was dissolved (2 mg/ml) in distilled water and a 10-g fraction was diluted to a final volume of 50 l with Tris-iodination buffer (25 mM Tris-HCl, pH 7.5, 0.4 M NaCl). A borosilicate glass tube (12 ϫ 75 mm) was coated with IODO-GEN (50 g) according to the manufacturer's protocol. Tris-iodination buffer (100 l) and 1 mCi of Na 125 I (Amersham Corp.) were added to the IODO-GENcoated tube and allowed to react for 6 min. The activated 125 I solution was added to the peptide solution and the reaction continued for 10 min. The reaction solution was diluted with 1 ml of Tris-iodination buffer and applied to a C-18 Sep-Pak cartridge (Waters). The C-18 cartridge was washed with 10 ml of Tris-iodination buffer to remove unreacted 125 I and eluted with 70% acetonitrile in 0.01% trifluoroacetic acid. The radioiodinated peptide was dried and redissolved in Tris-BSA buffer (25 mM Tris-HCl, pH 7.5, 0.4 M NaCl, 5 mM EDTA with 0.25% BSA) to a concentration of 1-2 ϫ 10 8 cpm/ml (approximately 1 M). The specific activity of the labeled peptide was 250 -300 cpm/fmol. SPA Assay-The biotinylated p72 syk SH2 domain proteins (3 g, unless otherwise noted) were incubated with 1.0 ml of SPA beads (Amersham). 50 mg of beads were resuspended in 5 ml of Tris buffer (TBS; 50 mM Tris-HCl, pH 8.0, 200 mM NaCl) and rotated overnight at 4°C. The SPA beads were washed twice with TBS, pH 8.0, 0.1 mg/ml BSA, 1 mM dithiothreitol, then resuspended to the original volume in the same buffer. 5-l aliquots of the suspended beads were then added to the individual wells of a 96-well plate which had previously been blocked with 1 mg/ml BSA in TBS for 30 min at room temperature. 25 l of various concentrations of 125 I-Fc⑀RI␥-based peptides were added to each well then TBS buffer was added to bring the final volume to 200 l. The plates were incubated at room temperature and the binding of the peptides to the SH2 domain proteins was determined by counting the plates 2 h later. Nonspecific binding was determined by running an identical set of beads in the presence of 100 ϫ concentration of unlabeled Fc⑀RI␥ ITAM-based peptide.
ELISA Assay-Biotinylated Fc⑀RI␥ ITAM-based peptide (bio-KS-DGVYpTGLSTRNQETYpETLKHEK) was bound to streptavidincoated 96-well ELISA plates by incubating for 1 h at room temperature. After washing three times with TBS buffer containing Tween 20 (0.05%), dual GST-Syk SH2 domain protein was added to each well and binding was allowed to proceed at room temperature for 1 h. After a further washing step, a 1:5000 dilution of anti-GST antibody was added to each well, the incubation continued for 1 h, then the plate was rinsed again. A 1:8000 dilution of goat anti-rabbit horseradish peroxidaseconjugated antibody (Zymed) was then added to each well and the plate was then incubated for an additional hour. After a final rinse, the plate was developed using TMB (Pierce) (1:1) for 5 min at room temperature then the reaction terminated by the addition of 1 M phosphoric acid. The binding was quantified by the OD 450 determination in a 96-well plate reader (Molecular Devices).

Binding of dual GST-Syk SH2 Fusion Proteins to Dual Phosphorylated Fc⑀RI ITAM-based Peptide as Determined by ELISA
Assay-The binding of the dual phosphorylated Fc⑀RI ITAMbased peptide to the dual p72 syk SH2 domain protein was initially confirmed by an ELISA-based assay. Binding was dependent on the concentration of both the peptide and the protein (data not shown). The non-biotinylated dual phosphorylated Fc⑀RI ITAM-based peptide competed for binding to the dual SH2 domains of p72 syk with a mean IC 50 value of 22 nM (n ϭ 8). In addition to the limitations of the ELISA assay described earlier, the IC 50 results observed for competition with the non-biotinylated dual phosphorylated Fc⑀RI ITAMbased peptide were variable (range 2-40 nM). A novel SPAbased assay was, therefore, developed for further studies.
Expression and Characterization of Bio-dual Syk SH2 and Bio-dual GST-Syk SH2 Domain Proteins-The binding of the p72 syk SH2 domain proteins to the streptavidin-coated SPA beads required that the proteins be biotinylated. The wide range of chemical reagents designed to biotinylate different functional groups on macromolecules (56) are not inherently site-selective, therefore chemical biotinylation will result in some molecules tagged with a single biotin, some with more than one, and others with none at all. In addition, these molecules may be targeted randomly on the surface of a protein. As some of the reactive groups on the surface of the protein may be critical for protein-protein interactions, chemical biotinylation may sharply reduce binding capacity of the protein. In addition, when the biotinylated proteins are immobilized on avidin or streptavidin-coated surfaces, at least some of the many different orientations will be sterically unfavorable. For these reasons, the dual SH2 domains of p72 syk was expressed in a vector that produces a protein in E. coli with a 23-residue peptide leader sequence at the N terminus which is a substrate for in vivo biotinylation by E. coli biotin holoenzyme synthetase (49). Unlike chemical reagents, enzymatic biotinylation at a single specific site ensures that all molecules will be immobilized in the same orientation, thereby resulting in the highest possible specific activity.
Three constructs were therefore initially utilized to produce the dual p72 syk SH2 domain proteins: (i) one encoded residues 10 -261 of p72 syk with a 23-residue peptide leader sequence at the N terminus which was a substrate for in vivo biotinylation by E. coli biotin holoenzyme synthetase (49) (bio-dual Syk SH2); (ii) one encoded the same protein but with a N terminus GST coding sequence (bio-dual GST-Syk SH2), and (iii) one encoded residues 10 -261 of p72 syk with the N-terminal GST but not the biotinylation site leader sequence. This protein was chemically biotinylated (lc-bio-dual GST-Syk SH2). When analyzed by SDS-PAGE (Fig. 1), the purified bio-dual Syk SH2 protein, bio-dual GST-Syk SH2 protein, and dual Syk SH2 domain protein that was subsequently chemically biotinylated migrated as single bands with apparent molecular masses of 30, 55-60, and 55 kDa, respectively, as would be predicted by their amino acid sequences. Scanning densitometry indicated that the proteins were greater than 95% pure. The bio-dual Syk SH2 and bio-dual GST-Syk SH2 proteins were likely to be maximally biotinylated with a single biotin at the N terminus of each molecule, as the enzymatically biotinylated proteins were purified and eluted from Soft-link monomeric avidin resin (Promega). The biotinylation of these proteins was confirmed by probing the proteins on an SDS-PAGE gel with streptavidinhorseradish peroxidase conjugate. Biotinylation of the chemically biotinylated protein was confirmed by ELISA determination. The chemically biotinylated GST-Syk SH2 protein, however, likely contains a mixture of biotinylated and non-biotinylated molecules.
Binding of the 125 I-Dual Phosphorylated Fc⑀RI ITAM-based Peptide to Bio-dual Syk SH2 Proteins-The ability of the biodual p72 syk SH2 domain protein to bind the Yp 64 Yp 75 peptide was initially confirmed utilizing the ELISA based format (data not shown). To establish parameters for subsequent SPA binding studies, streptavidin-coated SPA beads were incubated with increasing concentrations of bio-dual Syk SH2 domain protein. After rinsing, the beads were incubated to equilibrium (2 h) with increasing concentrations of 125 I-Yp 64 Yp 75 . As shown in Fig. 2, the binding of the 125 I-dual phosphorylated Fc⑀RI ITAM-based peptide to the dual p72 syk SH2 protein was specific and saturable, especially for the lower concentrations of the bio-dual Syk SH2 domain protein. Utilizing a protein concentration where there was saturation of binding (3 g/ml), the binding isotherm revealed that the K d for this binding was 1.4 nM (Fig. 3). Scatchard analysis and best fit comparison utilizing "Radlig" software indicated that there was only one binding site (Fig. 3).

Binding of the 125 I-Dual Phosphorylated Fc⑀RI ITAM-based Peptide to Enzymatically and Chemically Biotinylated Dual GST-p72 syk SH2 Domain Fusion Proteins-Previous studies
utilizing other binding assays, such as surface plasmon resonance (Biacore), indicate that GST fusion proteins can dimerize in solution resulting in inaccurate determination of the K d values (17). As it may not always be possible to utilize the biotinylation vector to express SH2 domain proteins, e.g. where the proteins need to be expressed in insect cells or the protein yield necessitates a higher binding capacity to solid phase for purification purposes, it may be necessary to express the SH2 domain proteins with a GST tag to purify the proteins on a glutathione column. The ability of bio-dual GST-Syk SH2 protein to bind to the 125 I-Yp 64 Yp 75 peptide was, therefore, compared to the ability of bio-dual Syk SH2 to bind to the 125 I-Yp 64 Yp 75 peptide. From the equilibrium binding studies (Fig. 3,  Fig. 4A), it can be seen that the K d for the binding of the 125 I-Yp 64 Yp 75 to the GST fusion protein (1.3 nM) was identical to that observed for the non-GST fusion protein (1.4 nM). A Scatchard plot of the data and best fit comparison utilizing Radlig software revealed only one binding site. Taken together, these data indicate that there is no difference in the binding characteristics of GST and non-GST SH2 domain fusion proteins utilizing the SPA format.
To determine whether chemically biotinylating the p72 syk dual SH2 domain protein interfered with its ability to recognize the 125 I-Yp 64 Yp 75 peptide, dual GST-Syk SH2 domain protein Bio-dual Syk SH2, bio-dual GST-Syk SH2, and dual GST-Syk SH2 proteins were expressed in E. coli and purified as described under "Materials and Methods." The dual GST-Syk protein was chemically biotinylated to produce the lc-bio-dual Syk SH2 protein as described under "Materials and Methods." Approximately 0.5-1 g of the proteins were loaded into the wells of a 12% precast gel and, after running, the proteins were visualized by staining with Coomassie Blue.

FIG. 2. The binding of 125 I-dual tyrosine-phosphorylated Fc⑀RI␥ ITAM-based peptide to the bio-dual Syk SH2 domain protein as a function of peptide and protein concentrations.
The concentrations of the 125 I-peptide are shown on the x axis and the concentrations of the bio-dual Syk SH2 domains are represented by the following symbols: Q, 0.1 g/ml; ࡗ, 0.3 g/ml; µ, 1 g/ml; Ç, 3 g/ml;ˆ, 10 g/ml; छ, 30 g/ml; Ⅺ, 100 g/ml. The data represents the mean values of duplicate determinations. Bio-dual Syk SH2 domain protein was incubated overnight with SPA beads at the concentrations indicated in the figure. The beads were rinsed and distributed into the individual wells of a 96-well microtiter plate. The ITAM peptide was then added to the wells and binding was allowed to proceed for 120 min at room temperature. The extent of binding was then determined by scintillation counting as described under "Materials and Methods. "   FIG. 3. Binding of the 125 I-dual tyrosine phosphorylated Fc⑀RI␥ ITAM-based peptide to bio-dual Syk SH2 domain protein. The insert is the Scatchard plot based on the data presented in the larger graph. The data represents the mean values of duplicate determinations. The bio-dual Syk SH2 protein (3 g/ml) was bound to the SPA beads overnight. After rinsing the beads, the binding of the ITAM peptide (3 nM) was determined as described in the legend for Fig. 2. The K d determined from these data was 1.37 nM. that had no biotinylation leader sequence was expressed in E. coli, purified, then chemically biotinylated utilizing two different systems (see "Materials and Methods"). Lc-bio-dual GST-Syk SH2 fusion protein and the GST fusion protein enzymatically biotinylated in E. coli showed similar binding properties (Fig. 4). Although the actual counts were lower, the K d values (1.4 nM) were virtually identical. In a similar strategy, the activity of the lc-bio-dual GST-Syk SH2 protein was determined by comparing the affinities of 125 I-peptide binding to the protein bound to glutathione-Sepharose versus bound to streptavidin-agarose. Whether bound to glutathione-Sepharose or streptavidin-agarose, the protein exhibited apparent K d values of ϳ0.6 and ϳ0.3 nM, respectively, indicating that the biotinylation procedure did not inactivate the protein. Again, a Scatchard plot of the data and best fit comparison utilizing Radlig software revealed only one binding site. When the protein was biotinylated with the biotin-xx-NHS reagent, however, there was substantially weaker binding of the 125 I-Yp 64 Yp 75 peptide as determined by SPA assay (data not shown).
Although these data demonstrate that chemical biotinylation of the Syk SH2 domains using the lc-reagent did not affect the ability of the SH2 domains to recognize the ITAM peptides, this was likely dependent on the particular SH2 domain in question and the particular biotinylation protocol used. Furthermore, preliminary studies from our laboratory suggest that binding of phosphotyrosine peptides to other SH2 domains may be affected by chemical biotinylation (data not shown). Thus, for development of other SPA-based SH2-domain binding assays, the method of choice for biotinylating the protein would be to use the E. coli biotinylation vector. If chemical biotinylation was the only option, it would be necessary to determine whether the biotinylation procedure was interfering with phosphotyrosine recognition.
Kinetics of Binding of 125 I-Dual Phosphorylated Fc⑀RI ITAM-based Peptide to Dual p72 syk SH2 Domains-The high affinity binding of the dual phosphorylated Fc⑀RI ITAM-based peptide to the dual p72 syk SH2 domains was reflected in the fast kinetics of association (Fig. 5A) of the peptide with the Syk protein. The t 1 ⁄2 was less than 1 min and was independent of the protein concentration over the range of protein examined (1-30 g/ml). The dissociation rate was determined by displacing the 125 I-Yp 64 Yp 75 peptide bound to the bio-dual Syk SH2 domain protein at equilibrium with Ͼ100-fold excess unlabeled Yp 64 Yp 75 . This produced a rapid rate of dissociation which again was independent of concentration (Fig. 5B). The K d for The beads were rinsed and distributed into the individual wells of a 96-well microtiter plate. The ITAM peptide (3 nM) was then added to the wells and binding was allowed to proceed for 120 min at room temperature. The extent of binding was then determined as before.  Fig. 2. The concentrations of the bio-dual Syk SH2 domain protein was 1 g/ml (ç), 3 g/ml (å), 10 g/ml (f), and 30 g/ml (F). The concentration of the ITAM peptide was 3 nM. The dissociation rate experiment was conducted by allowing the ITAM peptide to bind for 120 min then adding unlabeled dual phosphorylated Fc⑀RI␥ ITAMbased peptide (10 M).
the binding of the doubly phosphorylated Fc⑀RI ITAM-based peptide to the dual p72 syk SH2 domains was calculated utilizing the relative rates of association (4.17 ϫ 10 6 M Ϫ1 s Ϫ1 ) and dissociation (0.98 ϫ 10 Ϫ3 s Ϫ1 ) by "Radlig Kinetic" software. The association rate constant was determined utilizing the data from 4 different concentrations of peptide used in the "on rate" studies. The dissociation constant determined from the 1 and 3 nM peptide concentrations and the association rate data was determined to be 2.34 nM. This value is consistent with the K d determined from the saturation binding isotherm, indicating that the observed rates of association and dissociation reflected real time kinetics.
Competition of Binding of 125 I-Dual Phosphorylated Fc⑀RI ITAM-based Peptide to Dual p72 syk SH2 Domains with Soluble p72 syk SH2 Domain Proteins-N-terminal (N 42Ϫ ,C ϩ -Syk SH2) and C-terminal (N ϩ ,C Ϫ -Syk SH2) p72 syk SH2 domain knockouts were prepared by mutating critical arginine residues 42 and 195 (34) in the N-and C-terminal SH2 domains, respectively, to alanines. A second mutation in the C-terminal SH2 domain, Arg 45 to Ala (N 45Ϫ ,C ϩ -Syk SH2) was also made to determine whether other arginines within this SH2 domains contribute to the high affinity binding to the 125 I-Yp 64 Yp 75 peptide. The ability of these proteins and the soluble dual p72 syk protein to bind to the 125 I-Yp 64 Yp 75 peptide was examined in an SPA-based competition assay. Initially, the ability of dual GST-Syk SH2 to compete for binding was compared to the ability of the non-dual GST-Syk SH2 domain protein to compete for binding. As shown in Fig. 6, the GST fusion protein had a similar IC 50 value (6.9 nM; K i ϭ 2.1 nM) to the non-GST fusion protein (4.6 nM; K i ϭ 1.4 nM), again demonstrating, that in the SPA format, the GST tag does not substantially influence the binding characteristics of the SH2 domains. Although these IC 50 values are slightly higher than the K d values calculated from the direct binding studies, the equilibrium dissociation constants (K i values) for these inhibitors were similar to the observed K d values. The soluble dual p72 syk SH2 domain protein was much more efficient at competing for the 125 I-Yp 64 Yp 75 peptide binding to the SPA bead-bound bio-dual Syk SH2 protein than were the soluble N-and C-terminal single SH2 domains. The IC 50 values for the competition with the N 45Ϫ ,C ϩ -Syk SH2, N 42Ϫ ,C ϩ -Syk SH2, and N ϩ ,C Ϫ -Syk SH2 proteins were 228 nM, 4.6 M, and Ͼ6 M, respectively. Utlizing these data as analyzed by the "Graphpad Prizm" software, the equilibrium dissociation constants for the binding of these proteins were determined to be 69 nM, 1.3 M, and Ͼ2.3 M, respectively. The above data demonstrate that both SH2 domains are required for the high affinity binding of p72 syk to the Fc⑀RI␥ ITAM. In addition, a comparison of the IC 50 data for the N 42Ϫ ,C ϩ -Syk SH2, and N ϩ ,C Ϫ -Syk SH2 knockouts suggest that the binding to the C-terminal SH2 domain is of higher affinity than that of the binding to the N-terminal SH2 domain. Finally, the competition data with the N 45Ϫ ,C ϩ -Syk SH2 knockout suggest that other arginines in the C-terminal SH2 domain, apart from Arg 42 , may also contribute to the high affinity binding, but to a lesser extent than the Arg 42 .
Competition TCR, B cell receptor, and the Fc⑀RI is dependent upon the recruitment of specific tyrosine kinases into the receptor-signaling complex (24,(27)(28)(29)(30)(31)(32)(33)(34)(43)(44)(45)(46). In the case of Fc⑀RI-mediated mast cell activation, association of p72 syk with the cytosolic domain of the Fc⑀RI␥ subunit is essential for activation of downsteam signaling and for subsequent mediator release (33,34). The recruitment of p72 syk is a consequence of the phosphorylation of the dual tyrosine residues within the ITAM motif of the Fc⑀RI␥ cytosolic domain (33,34) and the resulting high affinity interaction between the phosphorylated ITAM and the dual SH2 domains of p72 syk . In this paper, we have described studies aimed at investigating the nature of this high affinity interaction. This has been accomplished by examining the binding parameters of the interaction between the dual and single SH2 domains of p72 syk and Fc⑀RI␥ ITAM-based peptides.
To examine the properties of the binding of phosphorylated Fc⑀RI␥ ITAM-based peptides to dual and single SH2 domains of p72 syk , a novel SPA-based binding assay was developed. This assay has allowed us to circumvent the limitations encountered utilizing other approaches described in the introduction. For example, when the p72 syk SH2 domain proteins are immobilized on the SPA beads and the binding ITAM-based peptides are in solution, avidity problems associated with GST fusion proteins are minimized (17). To illustrate this, our studies revealed that there was no difference in the K d values of binding of the Fc⑀RI␥ ITAM-based peptide to the E. coli-expressed biotinylated GST fusion Syk SH2 dual domain protein and the E. coli-expressed biotinylated non-GST fusion protein in the SPA-based binding assay. Second, in the competition studies conducted in the SPA assay, both the binding peptides and competing peptides are in the same phase. Thus, a true competition equilibrium should be achieved. Finally, as the SPA assay does not require washing steps or binding of secondary antibodies, the SPA-based assay has proved to be more rapid and reproducible compared to the ELISA assay.
Utilizing the SPA assay, the binding of the dual phosphorylated peptide, corresponding to amino acids 59 -82 of the Fc⑀RI␥ ITAM, to the p72 syk SH2 domains was observed to be of high affinity with a K d of 1.4 nM. Scatchard analysis and analysis of the rates of association of the dual phosphorylated peptide with the dual Syk SH2 domain protein revealed that there was only one binding site, suggesting that there is a preferred orientation for the binding of the dual phosphorylated ITAM peptide to the p72 syk dual SH2 domains. p72 syk has a 57% sequence homology with the TCR chain-associated tyrosine kinase p70 zap and it has been suggested that the nature of the SH2 domain-ITAM interactions may be very similar for both p72 syk and p70 zap (28). The recently published x-ray crystal structure of p70 zap has revealed that there is a close interaction between the N-and C-terminal SH2 domains (28). Within the interface between the N-and C-terminal SH2 domains, there is a shared contribution to phosphotyrosine recognition. If the N-terminal SH2 domain is expressed in isolation, the structure is not sufficiently complete to recognize phosphotyrosine (28). Thus, the data suggest that the C-terminal of p70 zap alone, but not the N-terminal SH2 domain of p70 zap alone, is capable of binding phosphotyrosine. The data presented in our study suggest that, to a certain extent, this may be similar for the binding of the single SH2 domains of p72 syk . In this context, the relative abilities for the Fc⑀RI␥ ITAM peptide to bind to the dual p72 syk SH2 domains, an N-terminal SH2 domain knockout, and to a C-terminal SH2 domain knockout in a competition assay, revealed that both domains are necessary to confer high affinity binding. This data, also suggest that the binding of the ITAM peptide to the C-terminal SH2 domain is of higher affinity than to the N-terminal SH2 domain of p72 syk . The above findings are supported by the observation that the C-terminal SH2 of p72 syk alone, but not the N-terminal SH2 domain of p72 syk alone, can recognize tyrosine phosphorylated proteins from RBL cell lysates (34). Competition data, with the N 45Ϫ ,C ϩ -Syk SH2 knockout, however, suggest that other arginines in the C-terminal SH2 domain, apart from Arg 42 , may contribute to the high affinity binding but to a lesser extent.
Our studies revealed that both tyrosines of the Fc⑀RI␥ ITAM peptide are required to be phosphorylated to allow the high affinity interaction with the dual SH2 domains of p72 syk . In this respect, the ability of the monophosphorylated peptides to compete for binding was reduced by 3 to 4 orders of magnitude in comparison with the doubly phosphorylated peptide. In addition, the N-terminal monophosphorylated Fc⑀RI␥ ITAM peptide was more efficient than the C-terminal monophosphorylated Fc⑀RI␥ ITAM peptide in competing with the dual phosphorylated Fc⑀RI␥ ITAM peptide for binding to the dual p72 syk SH2 domains. The p70 zap x-ray crystallographic study suggest that the TCR chain ITAM C terminus phosphotyrosine is bound in a pocket formed by both domains and that the TCR chain ITAM N-terminal phosphotyrosine is in the C terminus SH2 domain pocket (28). Our data therefore suggests that the preferred orientation of the binding for the Fc⑀RI␥ ITAM to the dual SH2 domains of p72 syk would be similar to the p70 zap model. In this model, the N-terminal phosphotyrosine occupies the C-terminal SH2 domain pocket and the ITAM C-terminal phosphopeptide occupies the N-terminal SH2 domain pocket. Confirmation of this orientation awaits x-ray crystallographic structural studies.
Finally, the marked differences in the relative affinities of the singly and doubly phosphorylated Fc⑀RI␥ ITAM peptides to bind to the SH2 domains of p72 syk may have important physiological implications. Both the Fc⑀RI␤ and Fc⑀RI␥ ITAMs are targets for Fc⑀RI-associated tyrosine phosphatases (58,59). When these phosphatases are activated, the resulting dephosphorylation of either tyrosine residue within these ITAMs would shift the binding equilibrium resulting in dissociation of p56 lyn and p72 syk from the Fc⑀RI␤ and Fc⑀RI␥ subunits, respectively. Our preliminary studies have demonstrated that tyrosine phosphatases in RBL 2H3 cell lysates are indeed capable of dissociating the Fc⑀RI␥ ITAM peptide from the p72 syk ITAM in the SPA assay (60). The relatively rapid off-rates observed in our study would mean that dephosphorylation of even one of the ITAM tyrosines would result in the termination of ongoing cell activation process.