Identification of a binding sequence for the 14-3-3 protein within the cytoplasmic domain of the adhesion receptor, platelet glycoprotein Ib alpha.

The zeta-form 14-3-3 protein (14-3-3zeta) regulates protein kinases and interacts with several signaling molecules. We reported previously that a platelet adhesion receptor, glycoprotein (GP) Ib-IX, was associated with a 29-kDa protein with partial sequences identical to 14-3-3zeta. In this study, the interaction between GPIb-IX and recombinant 14-3-3zeta is reconstituted. Further, we show that the 14-3-3zeta binding site in GPIb is within a 15 residue sequence at the C terminus of GPIb-alpha, as indicated by antibody inhibition and direct binding of 14-3-3zeta to synthetic GPIb-alpha cytoplasmic domain peptides. The 14-3-3zeta binds to recombinant wild type GPIb-IX but not to the GPIb-alpha mutants lacking C-terminal 5 or more residues, suggesting that the C-terminal 5 residues of GPIb-alpha are critical. Similarity between the GPIb-alpha C-terminal sequence and the serine-rich regions of Raf and Bcr kinases suggests a possible serine-rich recognition motif for the 14-3-3 protein.

The platelet membrane glycoprotein Ib (GPIb) 1 -glycoprotein IX (GPIX) complex (GPIb-IX) plays an important role in the initial platelet adhesion to injured vascular wall under high shear flow conditions such as in arteries and capillaries (1,2). By binding to the subendothelium-bound von Willebrand factor, GPIb-IX not only mediates the physical adherence of platelets to the site of vascular injury but also initiates an activation signal that is transduced across the membrane resulting in a series of biochemical events. GPIb-induced intracellular biochemical changes include synthesis of thromboxane A 2 , hydrolysis of phosphoinositide, activation of protein kinase C and tyrosine kinases, elevation of cytoplasmic calcium level, cytoskeleton reorganization, and exposure of ligand-binding function of other adhesion receptors such as the integrin ␣ IIb ␤ 3 (3)(4)(5)(6)(7)(8). In addition, GPIb␣ binds thrombin and is involved in the signaling process of thrombin-induced platelet activation (9 -11). The mechanism of signal transduction via GPIb-IX has been unclear. In search for a possible intermediate between GPIb-IX and the intracellular signaling pathways, we have recently found that GPIb-IX is associated with a 29-kDa intracellular protein that is identical in partial amino acid sequence to the -form 14-3-3 protein (14-3-3) (12). The 14-3-3 proteins are a family of highly conserved eukaryotic proteins, which are distributed in a variety of cells (13,14). Several functions have been attributed to the 14-3-3 proteins. At the cellular level, the 14-3-3 proteins have been implicated in the regulation of cell cycle and stimulation of exocytosis (15,16). At the molecular level, the reported functions of 14-3-3 proteins include the activation of Pseudomonas aeruginosa exoenzyme S (14), phospholipase A 2 activity (17), and regulation of protein kinase C and the tyrosine and tryptophan hydroxylases (18 -22). The 14-3-3 proteins, including the -form, interact with and activate Raf protein kinase in both yeast and mammalian cells (23)(24)(25)(26)(27). Raf phosphorylates and activates mitogen-activated protein kinase kinase, which subsequently activates mitogen-activated protein kinase. Mitogen-activated protein kinase may phosphorylate and activate several signaling proteins, including a cytosolic phospholipase A 2 (28). Thus, it is possible that association of a 14-3-3 protein with GPIb-IX may serve as a signaling mechanism that transduces adhesioninitiated platelet activation signals.
The functional mechanism of the 14-3-3 protein is unclear. Evidence of direct binding of the 14-3-3 proteins has been shown in Raf, Bcr kinase, and protein kinase C, as well as middle T antigen and tryptophan hydroxylase (19 -27, 29 -31). This suggests that binding to target proteins may be required for its regulatory functions. However, it is not understood how the 14-3-3 proteins interact with different types of proteins. In this study, we have identified the 14-3-3 binding site in GPIb␣. This is the first identified short peptide sequence that binds the 14-3-3 protein. Furthermore, we report a similarity between this 14-3-3 protein binding sequence and a short segment from each of the serine-rich 14-3-3 protein binding regions of Raf and Bcr kinases, suggesting a possible serine-rich recognition motif.

EXPERIMENTAL PROCEDURES
Cloning and Expression of the Recombinant -Form 14-3-3 Protein-A cDNA fragment encoding the 14-3-3 was amplified by polymerase chain reaction from a cDNA library isolated from a human erythrocytic leukemia cell line (kindly provided by Dr. Jerry Ware and Dr. Z. M. Ruggeri, The Scripps Research Institute, La Jolla, CA). The primers used for the amplification were as follows: TGATGAATTCTTCACCAT-GGATAAAAATGAGCTGGTTC and ATGGTCTAGAAATGGTCTACT-GTGTA. The cDNA fragments were first cloned into the pBluescript vector (Promega, Madison, WI) and sequenced. The pBluescript plasmid carrying the cDNA fragment was digested with EcoRI and XbaI, and subcloned into the pmal C2 vector (New England Biolabs, Boston, MA) digested with the same enzymes. The construct was named pmal1433, and its insert matched the published sequence of 14-3-3 (17). The pmal1433 encodes a fusion protein with the N-terminal region corresponding to the Escherichia coli maltose-binding protein (MBP) and C-terminal region corresponding to the entire 14-3-3. Expression of the fusion protein in E. coli (DH5␣) cells was performed as described previously (32). The expressed fusion protein has an molecular mass of ϳ70 kDa as analyzed by SDS-polyacrylamide gel electrophoresis (SDS-PAGE), and specifically reacted with antiserum against the purified human platelet 14-3-3 (12), and antibodies against synthetic peptides corresponding to the central and C-terminal regions of 14-3-3. The fusion protein was purified by affinity chromatography using crosslinked amylose-Sepharose column as described previously.
Cell Lines Expressing Recombinant Wild Type and Mutant GPIb-IX Complex-The stable transfected Chinese hamster ovary cell lines (CHO) expressing recombinant GPIb-IX were established as described previously (33). These include cells expressing the wild type GPIb-IX complex, and cells expressing GPIb-IX with mutant GPIb␣ cytoplasmic domain. 2 The GPIb␣ DNA mutants incorporate stop codons C-terminal to residues 559 (GPIb␣⌬559), 591 (GPIb␣⌬591), and 605 (GPIb␣⌬605) respectively. The actin-binding protein (ABP)-deficient melanoma cell line (34) and expression of GPIb-IX complex in this cell line have been described previously (35).
Antibodies and Peptides-The monoclonal antibody WM23, against the central region of GPIb␣, was kindly provided by Dr. Michael C. Berndt, Baker Institute, Melbourne, Australia (36). Monoclonal antibody P3, specific for the von Willebrand factor binding domain of GPIb␣, was kindly provided by Dr. Z. M. Ruggeri, The Scripps Research Institute, La Jolla, CA. Monoclonal antibody against the integrin ␣ IIb subunit (GPIIb), PMI-1 was kindly provided by Dr. Mark Ginsberg, The Scripps Research Institute, La Jolla, CA (37). Anti-peptide antibodies were raised by immunizing New Zealand White rabbits with peptides conjugated to keyhole limpet hemocyanin (Sigma). Peptides were synthesized using a model 430A automated peptide synthesizer (Applied Biosystems). These include peptides DLLSTVSIRYSGHSL, corresponding to C-terminal 15 residues of GPIb␣; TDPLVAERAGTDES, corresponding to C-terminal 14 residues of GPIb␤; DTQGDEAEAGEGGEN, corresponding to C-terminal 15 residues of 14-3-3; KFLIPNASQAE, corresponding to 11 residues in the central region of 14-3-3; and KSAVTTVVNPKYEGK, corresponding to C-terminal 15 residues of the integrin ␤ 1 subunit.
Binding Assays and Immunoprecipitations-Fresh human platelets (10 9 /ml) were solubilized as described above. CHO cells or melanoma cells were detached from flasks by incubating with 3.5 mM EDTA, 0.01 M Na 2 HPO 4 , 0.15 M NaCl, pH 7.4, washed twice and also solubilized in the solubilization buffer as described above. After centrifuging at 10,000 ϫ g for 10 min, the lysates (200 l) were incubated with 25 l (50%; v/v) of MBP-conjugated control beads or the 14-3-3 protein-conjugated Sepharose 4B beads at 4°C for 2 h. The beads were then washed three times in solubilization buffer by centrifugation. Bound proteins were extracted by adding SDS-PAGE sample buffer (0.125 M Tris, pH 6.8, 20% (v/v) glycerol, 0.004% (w/v) bromphenol blue, 4% (w/v) SDS), and analyzed by SDS-PAGE followed by Western blot with antibodies.
Binding of the 125 I-labeled 14-3-3 to Synthetic Peptides-The recombinant 14-3-3 protein was 125 I-labeled using IODOBEADS (Pierce). The synthetic peptides were solubilized in distilled water and then diluted to 50 g/ml in 0.1 M NaHCO 3 , pH 9.2. The peptides were coated onto the microtiter wells (Immunolon II Removawells, Dynatech Laboratories, Chantilly, VA) by incubation at 22°C for 6 h and then 4°C overnight. The peptide coated microtiter wells were incubated at 22°C with 5% bovine serum albumin for 2 h and then with 125 I-labeled 14-3-3 protein for 2 h. After three washes, microtiter well-bound radioactivity were estimated in a ␥-counter. In some experiments, the peptide-coated microtiter wells were preincubated with control serum or the anti-Ib␣C antiserum before incubating with the 125 I-labeled 14-3-3.

Binding of the Recombinant 14-3-3 to the Platelet GPIb-IX
Complex-In order to verify that 14-3-3 binds GPIb-IX, a cDNA fragment encoding the entire -form 14-3-3 protein was amplified by polymerase chain reaction from a cDNA library from a human erythrocytic leukemia (HEL) cell line. The cDNA fragment was expressed as a fusion protein with E. coli MBP. The fusion protein was purified and then conjugated to CNBractivated Sepharose 4B. As a control, the equivalent amount of MBP was also conjugated to CNBr-activated Sepharose 4B beads. Equal volumes of platelet lysates were allowed to pass through the recombinant 14-3-3 column or the control column simultaneously. The column-bound proteins were eluted with a 0.15-1 M NaCl gradient. Fig. 1 shows the SDS-PAGE analysis of the proteins that were eluted at 1 M NaCl. Silver staining of the gel showed that two major bands with molecular masses of 250 and 135 kDa, corresponding to ABP and GPIb␣ (that are known to exist as a complex in the platelet lysate), were present in the 14-3-3 column eluates but not in the eluates from the control maltose-binding protein-conjugated column. Western blots of the eluates showed that the 135-kDa band reacted with a monoclonal antibody against the ␣ chain of GPIb. In contrast, a control monoclonal antibody against platelet glycoprotein IIb (GPIIb) was not reactive with the eluates, indicating that GPIIb-IIIa, which is the most abundant platelet membrane protein, did not bind to the 14-3-3 column. These data indicate that the GPIb-IX-ABP complex specifically bound to the recombinant 14-3-3.
The 14-3-3 Protein Binds to the C-terminal Domain of GPIb-IX-GPIb-IX consists of a smaller globular domain in the N terminus of GPIb␣ that contains the ligand binding sites, a rodlike central domain, and a larger C-terminal globular domain that contains the transmembrane and cytoplasmic domains of GPIb␣, GPIb␤, and GPIX (39) (cf. Fig. 2B). Cleavage by proteases, such as calpain, releases the 135-kDa fragment corresponding to the N-terminal and central region of GPIb␣, called glycocalicin, from the C-terminal globular domain (39). Calpain also cleaves ABP, releasing it from the C-terminal globular domain (40). To determine which component of GPIb-IX-ABP complex binds to 14-3-3, the platelets were solubilized in the presence of calcium to allow the cleavage of GPIb␣ by endogenous calpain (41). The platelet lysates were then passed 2 J. Cunningham, personal communication. Washed platelets (2 ϫ 10 9 /ml) were solubilized and lysates were allowed to pass through a Sepharose 4B column conjugated with the recombinant MBP-14-3-3 fusion protein (14-3-3), or a control column (MBP). After extensive washing, the bound proteins were eluted with a 0.15 M to 1 M NaCl gradient. The proteins eluted at 1 M NaCl were analyzed by SDS-PAGE, followed by silver staining and immunoblotting with an anti-GPIb␣ antibody (WM23) or a control antibody against the platelet glycoprotein IIb (PMI-1). through the 14-3-3-conjugated column or the control column. Bound proteins were eluted with 1 M NaCl and analyzed by SDS-PAGE, followed by Western blot with anti-peptide antibodies against C-terminal 15 residues of GPIb␣ (anti-Ib␣C). As shown in Fig. 2, the majority of GPIb␣ in the platelet lysate is hydrolyzed to produce a 20-kDa fragment containing the epitope for anti-Ib␣C. The 20-kDa C-terminal fragment of GPIb␣ was detected in the eluates from the 14-3-3 column but not from the control column (Fig. 2). This indicates that the C-terminal globular domain of GPIb-IX binds to 14-3-3. Since the C-terminal domain of GPIb-IX does not bind to ABP in lysates in which calpain is active (40), this result also indicates that 14-3-3 interaction with GPIb is not mediated by ABP. This conclusion was further supported in the experiment in which recombinant GPIb-IX expressed in an ABP-deficient cell line (34) was solubilized and incubated with the 14-3-3-conjugated Sepharose beads or control MBP-conjugated beads. GPIb-IX bound to the 14-3-3 beads even though ABP was absent (data not shown).
GPIb␣ Cytoplasmic Domain Contains the 14-3-3 Binding Site-To further characterize the binding site for 14-3-3 on GPIb-IX, the platelet lysates were preincubated with either a control rabbit serum, an anti-peptide antibody directed against the C-terminal 15 residues of GPIb␣ (anti-Ib␣C), or an antipeptide antibody against the cytoplasmic domain of GPIb␤ (anti-Ib␤C). The platelet lysates were then incubated with Sepharose beads conjugated with 14-3-3. As shown in Fig. 3, GPIb-IX binds to the 14-3-3but not MBP-conjugated beads. Pretreatment with control serum had no effect on the binding of GPIb-IX. However, when the platelet lysates were pretreated with anti-Ib␣C, the binding of GPIb-IX to 14-3-3-conjugated beads was dramatically inhibited. In contrast, preincubation with anti-GPIb␤C antibody had no inhibitory effect. The different effects between the anti-GPIb␣ and anti-GPIb␤ cytoplasmic domain antibodies in inhibiting GPIb-IX binding to 14-3-3 is not due to a difference in the antibody binding to GPIb-IX, because equivalent amounts of GPIb-IX were immunoprecipitated by these two antibodies (Fig. 3B). These data suggest that the C-terminal 15 residues of GPIb␣ is critical for the 14-3-3 binding.
The C-terminal 15 Residues of GPIb␣ Contains the 14-3-3 Binding Site-To verify whether the C-terminal 15 residues of GPIb␣ recognized by the inhibitory antibody contains the 14-3-3 binding site, 125 I-labeled 14-3-3 was incubated in microtiter wells coated with the synthetic peptides corresponding to the C-terminal 15 residues of GPIb␣ (Ib␣C) or C-terminal 15 residues of GPIb␤ (Ib␤C). The 14-3-3 bound to the Ib␣C peptide in a saturable manner, with an estimated K d of 850 nM (Fig. 4). In contrast, there was no specific binding to the Ib␤C peptide (Fig. 4) or a control peptide corresponding to C-terminal 15 residues of the integrin ␤ 1 subunit (not shown). To exclude the possibility of nonspecific binding, a truncated form of the 14-3-3 (1433T3, truncated at residue 36) fused with E. coli MBP was shown not to bind to the Ib␣C-coated wells (Fig.  4). In addition, binding of the 14-3-3 to Ib␣C peptide was inhibitable by the unlabeled 14-3-3 protein and by the anti-Ib␣C antibody (data not shown). Thus, the C-terminal 15residue sequence of GPIb␣ (Asp 596 -Leu 610 ) contains the 14-3-3 binding site.
The C-terminal 5 Residues of GPIb␣ Cytoplasmic Domain Are Critical for the Binding of 14-3-3-To further examine the region of GPIb␣ cytoplasmic domain involved in the binding of 14-3-3, recombinant wild type and mutant GPIb-IX expressed in the CHO cells were used. The mutants incorporated stop codons at the C-terminal side of the residues 559 (⌬559), 591 (⌬591), and 605 (⌬605) of GPIb␣ (Fig. 5C). GPIb-IX from ⌬559 cell line does not bind ABP, while GPIb␣⌬605 only lacks 5 residues from the C terminus of the wild type GPIb␣ and is still able to bind ABP. 2 The CHO cells expressing recombinant wild type GPIb-IX or GPIb-IX mutants were solubilized and then incubated with the 14-3-3-conjugated Sepharose beads or the control beads. Similar to platelet GPIb-IX, the recombinant wild type GPIb-IX bound to 14-3-3-conjugated beads but not to the control beads. Pretreatment of the CHO lysates with the anti-Ib␣C antibody inhibited the binding of the 14-3-3 protein (Fig. 5A). However, the GPIb-IX mutants with truncations at residues 559, 591 or 605 of GPIb␣ all lost their ability to bind 14-3-3 (Fig. 5). This suggests that the C-terminal 5 residues of the GPIb␣ cytoplasmic domain (Ser 606-Gly-His-Ser-Leu 610 ) is critical for the binding of the 14-3-3 protein.
Comparison between GPIb␣ C-terminal Domain and the 14-3-3 Protein Binding Region of Bcr and Raf Kinases-To characterize possible common structure responsible for the 14-3-3 protein recognition, we examined similarity between the Cterminal region of the GPIb␣ cytoplasmic domain and the reported 14-3-3 protein binding regions of Raf and Bcr kinases using BESTFIT. There was an alignment between the GPIb␣ C-terminal region (residues 582-610) (42) and the Bcr 14-3-3 protein-binding region (residue 298 -412) (29) with a 44.4% similarity and 29.6% identity in a segment of 30 residues. Alignment with the 14-3-3 protein binding region of Raf kinase (residues 148 -256) (26,27) showed that the C-terminal region of GPIb␣ had a similarity of 39.3% (28.6% identity) also within a segment of 30 residues (Fig. 6A). A significant feature manifested in alignment of these three sequences is that serine (or threonine) residues are the major factor responsible for their similarity (Fig. 6). These serines and threonines often reside 3 or 4 residues apart. Thus, in a helical conformation, they would be clustered in one face of the ␣-helix (Fig. 6B).

DISCUSSION
In this study, we have reconstituted the binding of the 14-3-3 protein to an important platelet adhesion receptor, GPIb-IX, and identified the 14-3-3 binding sequence at the C terminus of GPIb␣ cytoplasmic domain. The identification of a short peptide sequence that binds the 14-3-3 protein provides insight into the structural basis required for the 14-3-3 protein recognition. The similarity between the 14-3-3 protein binding sequence in GPIb␣ and the segments from the 14-3-3 protein binding region of the Bcr and Raf kinases suggests a possible serine-rich recognition motif in the ligands of the 14-3-3 protein. In addition, characterization of the interaction of the 14-3-3 with the cytoplasmic domain of a membrane receptor may help to understand the roles of the 14-3-3 protein in the receptor-mediated signaling pathways.
Reconstitution of the binding between the recombinant 14-3-3 and GPIb-IX (Fig. 1) confirmed the identity of the previously reported GPIb-IX-associated 29-kDa protein (12) as 14-3-3. In platelets, 14-3-3 was first purified and cloned as a platelet intracellular phospholipase A 2 (17,43), although this function of 14-3-3 was recently disputed by other groups (44). The 14-3-3 is relatively abundant in platelets and has been shown to be both associated with the plasma membrane and present in the cytosol (43). Purified 14-3-3, however, does not bind to phospholipid vesicles (45). Thus, association with the cytoplasmic domain of GPIb-IX may account for at least part of the membrane-associated fraction of this protein. The location of the 14-3-3 binding site on GPIb-IX is indicated by following data. 1) The proteolytically generated Cterminal domain of GPIb-IX bound to the 14-3-3 column; 2) an anti-peptide antibody against the C-terminal 15 residues of GPIb␣ inhibited the binding of 14-3-3; 3) 125 I-labeled 14-3-3 directly bound to this C-terminal 15 residue peptide; and 4) mutagenesis that truncated 5 or more residues from C terminus of GPIb␣ abolished 14-3-3 binding. These data suggest that 14-3-3 binds to the cytoplasmic domain GPIb␣ in the region between Asp 596 and Leu 610 , in which the C-terminal 5 residues (Ser 606 -Gly-His-Ser-Leu 610 ) are critical. A significant feature of this 14-3-3 binding sequence is the presence of a serine every 3-4 residues (XXXSXXSXXXSXXSX) (Fig. 6). Thus, if this sequence were to form an ␣-helix in the intact protein, the serine residues may form a cluster in one face of the helix (Fig. 6). Indeed, from the recently resolved structure of the 14-3-3 protein, a ligand that fits into the ligand binding groove of the 14-3-3 protein was predicted to be an amphipathic helix (46,47). Furthermore, the serine-rich regions in Bcr and Raf kinases appear to be critical for the 14-3-3 protein binding, and serine residues from a 30-residue segment from each of the serine-rich regions of Bcr and Raf kinases are well aligned with serine residues from the 14-3-3 binding sequence of GPIb␣ (Fig.  6). This suggests that clusters of serine residues in a helical structure may be a common recognition motif important for the binding of the 14-3-3 protein.
The location of the 14-3-3 binding site in the cytoplasmic domain of GPIb␣ suggests that the binding of 14-3-3 to GPIb-IX is likely to occur in intact platelets, and thus may be of relevance to the functions of the cytoplasmic domain of GPIb-IX. The cytoplasmic domain of GPIb␣ is known to bind to the cytoskeletal protein ABP (48). ABP, however, is not required for the binding of the 14-3-3 protein, as co-immunoprecipitation of the 14-3-3 with GPIb-IX was not inhibited by the treatment of cell lysates with DNase I and N-ethylmaleimide, which disrupted the interaction between GPIb and ABP (12). Furthermore, in the present study, we show that lysis of platelets under conditions in which calpain was active and the ABP-GPIb-IX interaction disrupted, did not prevent the interaction of the C-terminal domain of GPIb-IX with 14-3-3 (Fig.  2). Moreover, 14-3-3 bound to GPIb-IX from an ABP-deficient cell line (not shown). Conversely, the 14-3-3 protein is not required for the ABP-GPIb interaction, as the GPIb␣ mutant that was truncated at residue 605 retained its capacity to interact with ABP, 2 yet lost its 14-3-3 binding capacity. The ABP binding site is located in the central region (Thr 736 -Phe 568 ) of the GPIb␣ cytoplasmic domain (48), 2 while the Cterminal region contains the 14-3-3 binding site (Fig. 4). Binding of ABP to the cytoplasmic domain of GPIb␣ links GPIb-IX to the membrane skeleton framework (40). It is interesting to speculate that the adjacent location of the binding sites of the membrane skeleton protein, ABP, and a kinase regulator, 14-3-3, within the GPIb␣ cytoplasmic domain may be important in the shear-dependent signaling transduction through GPIb-IX. For example, it is possible that mechanical force generated by immobilized von Willebrand factor binding to N-terminal domain of GPIb␣ under high shear stress may change the conformation of the C terminus of GPIb␣ by leverage of the membrane skeleton and thus regulate the binding or signaling functions of the 14-3-3.
It has been unclear how the 14-3-3 proteins regulate their target proteins. A recent report suggests that by binding to Raf kinase, the 14-3-3 protein may prevent its inactivation by protein phosphatase (49). It is possible that 14-3-3 protein may prevent the inactivation of Raf by interaction with phosphorylated serine residues in the serine-clustered region. Similarly, it is possible that by binding to the C-terminal serine-rich region of the cytoplasmic domain of GPIb␣, the 14-3-3 may regulate the function of the GPIb␣ cytoplasmic domain by preventing serine (or phosphoserine) residues from being modified. Similarity between the GPIb␣ C-terminal region and a segment from the 14-3-3 protein binding region of both Raf and Bcr also suggests a possibility that GPIb␣ may compete with kinases for the 14-3-3 protein binding sites in the membrane compartment and thus regulate kinase activity. Alternatively, as the membrane translocation is a mechanism of the Raf and protein kinase C activation, it is also possible that the dimeric 14-3-3 may be involved in the translocation of the protein kinases to the membrane (50). Although phospholipid association of some isoforms of the 14-3-3 protein family has been indicated, 14-3-3 does not appear to associate with phospholipid vesicles (45). Thus, a possible mechanism is that by binding to the cytoplasmic domain of GPIb␣, the 14-3-3 may mediate translocation of the protein kinases to cytoplasmic face of the membrane and to the cytoplasmic domain of this membrane receptor in order to relay signals. If this were to be the case, there may also be membrane receptors in other cell types that interact with the 14-3-3 proteins. FIG. 6. A, alignment of the C-terminal region of GPIb␣ (residues 582-610) with segments of Raf (residues 221-252) and Bcr (residues 327-359) kinases from the serine-rich regions reported as critical for the binding of the 14-3-3 protein. Identical residues are shaded. B, helical wheel analysis of the sequences from A. Note that the serine residues are clustered in one side of the helices in all the three sequences.