The Cytoplasmic Domain of the Platelet Glycoprotein Ibα Is Phosphorylated at Serine 609*

The α chain of the platelet von Willebrand factor receptor, glycoprotein (GP) Ib, is not known to be phosphorylated. Here, we report that the cytoplasmic domain of GPIbα is phosphorylated at Ser609; this was detected by immunoblotting with an anti-phosphopeptide antibody, anti-pS609, that specifically recognizes the GPIbα C-terminal sequence S606GHSL610 only when Ser609 is phosphorylated. Immunoabsorption with anti-pS609 removed almost all of the GPIbα from platelet lysates, indicating a high proportion of GPIbα phosphorylation. Anti-pS609 inhibited GPIb-IX binding to the intracellular signaling molecule, 14-3-3ζ. Dephosphorylation of GPIb-IX with potato acid phosphatase inhibited anti-pS609 binding and also 14-3-3ζ binding. A synthetic phosphopeptide corresponding to the GPIbα C-terminal sequence (SIRYSGHpSL), but not a nonphosphorylated identical peptide, abolished GPIb-IX binding to 14-3-3ζ. Thus, phosphorylation at Ser609 of GPIbα is important for 14-3-3ζ binding to GPIb-IX. In certain regions of spreading platelets, particularly at the periphery, there was a reduction in GPIbα staining by anti-pS609 as observed under a confocal microscope, indicating that a subpopulation of GPIbα molecules in these regions is dephosphorylated. These data suggest that phosphorylation and dephosphorylation at Ser609 of GPIbα regulates GPIb-IX interaction with 14-3-3 and may play important roles in the process of platelet adhesion and spreading.

munizing New Zealand White rabbits with this peptide conjugate. To verify the specificity of the antibody (anti-pS609), phosphorylated GPIb␣ C-terminal peptides (CSGHpSL and SIRYSGHpSL) and nonphosphorylated versions of the same peptides were dissolved in 0.1 M NaHCO 3 , pH 9.2, and coated onto the microtiter plates by incubation at 4°C overnight. Comparable amounts of the phosphorylated and nonphosphorylated peptides were coated as indicated by the binding of an anti-peptide antibody, anti-Ib␣C, against the nonphosphorylated Cterminal sequence of GPIb␣. The wells were blocked with 5% bovine serum albumin in phosphate-buffered saline and then incubated with anti-pS609 antiserum or control preimmune serum from the same rabbits at 22°C for 2 h. After three washes, microtiter wells were further incubated with horseradish peroxidase-conjugated goat antirabbit IgG and then washed six times. Bound antibody was quantitated by incubation with peroxidase substrate (0.04% O-phenylenediamine, 0.012% H 2 O 2 in 0.1 M citrate-phosphate buffer, pH 5.0) at 22°C for 30 min. The reaction was stopped by the addition of 50 l/well 2 M H 2 SO 4 and then visualized by determining optical density at 490 nm wavelength.
Confocal Microscopy-Washed platelets in modified Tyrode's buffer (35) were allowed to adhere and spread on vWF-coated glass chamber slides (Nunc) for various lengths of time. Nonadherent cells were removed by three washes. Adherent platelets were fixed by adding 4% paraformaldehyde in phosphate-buffered saline and then permeabilized by incubation for 30 min at 22°C in 0.1 M Tris, 0.01 M EGTA, 0.15 M NaCl, and 5 mM MgCl 2 , pH 7.4, containing 0.1% Triton X-100, 0.5 mM leupeptin, 1 mM phenylmethylsulfonyl fluoride, and 0.1 mM E64. Platelets were then incubated with 10 g/ml of various antibodies at 22°C for 1 h. After three washes, platelets were further incubated with fluorescein-or rhodamine-labeled secondary antibodies at 22°C for 30 min. After additional washes, cells were scanned under a Zeiss LSM510 confocal microscope (ϫ2520).

Serine 609 in the Cytoplasmic Domain of GPIb␣ Is
Phosphorylated-We previously showed that the C-terminal 15-residue serine-rich region of GPIb␣ contains a 14-3-3 binding site in which 5 amino acid residues (S 606 GHSL 610 ) at the C terminus are important (35). To investigate the possibility that 14-3-3 binding is regulated by phosphorylation, a phosphorylated peptide, CS 606 GHpSL 610 , was synthesized. This peptide incorporates a phosphoserine (pS) at the residue corresponding to Ser 609 of GPIb␣. An antibody against this peptide was developed. This antibody (anti-pS609) reacted specifically with the phosphorylated peptides CSGHpSL and S 602 IRYSGHpSL 610 corresponding to C-terminal 5-and 9-residue sequences of GPIb␣ but failed to react with the identical nonphosphorylated peptides CSGHSL or SIRYSGHSL (Fig. 1, A and B). Anti-pS609 also failed to react with the SIRYpSGHSL peptide with phosphorylation at Ser 606 (Fig. 1B). As a positive control, anti-Ib␣C antibody against the nonphophorylated GPIb␣ C-terminal sequences was shown to interact with both the nonphosphorylated and phosphorylated peptides coated on the A, synthetic phosphopeptide CSGHpSL or nonphosphorylated identical peptide CSGHSL was coated onto microtiter plates. Anti-pS609 antiserum or preimmune serum (negative result not shown) was incubated in the microtiter wells for 2 h at 22°C. After further incubation with horseradiash peroxidase-conjugated goat anti-rabbit IgG, bound antibody was detemined by measuring A 490 nm as described under "Experimental Procedures." B, synthetic phosphopeptide SIRYSGHpSL with phosphorylated Ser 609 , an identical nonphosphorylated peptide (SIRYSGHSL), or an identical peptide with phosphorylation at Ser 606 but not Ser 609 (SIRYpSGHSL) were coated onto the microtiter well, and binding of the anti-pS609 serum to each of these peptides was measured as described in A. Note that anti-pS609 reacts only with the Ser 609 -phosphorylated peptide. C, microtiter plates were coated with synthetic phosphopeptides SIRYSGHpSL, an identical nonphosphorylated peptide SIRYSGHSL, a phosphopeptide SIRYpSGHSL (phosphorylation at Ser 606 ), or a negative control peptide corresponding to GPIb␤ C-terminal 14-amino acid sequence (Ib␤C). The microtiter wells were incubated with an antiserum, anti-Ib␣C, directed against the nonphosphorylated GPIb␣ C-terminal sequence. The comparable amounts of binding of this antibody to various GPIb␣ peptides indicates that comparable amounts of these peptides were coated on the microtiter wells. Shown in the figure are the results from three samples (mean Ϯ S.D.). microtiter plates (Fig. 1C). These data indicate that anti-pS609 specifically binds to GPIb␣ C-terminal sequences only when Ser 609 is phosphorylated (Fig. 1).
To examine whether Ser 609 of the platelet GPIb␣ is phosphorylated, washed resting platelets were solubilized directly into SDS-containing sample buffer and immunoblotted with anti-pS609. Fig. 2A shows that anti-pS609 specifically reacted with a band with its molecular weight identical to that of GPIb␣. To verify that anti-pS609 indeed reacted with GPIb␣, platelets were solubilzed and GPIb␣ was immunoprecipitated with a monoclonal antibody, SZ2, directed against GPIb␣. The immunoprecipitates were then immunoblotted with anti-pS609. Fig.  2B shows that anti-pS609 indeed reacted with immunoprecipitated GPIb␣. As the binding of anti-pS609 requires phosphorylation at Ser 609 , these results indicate that the Ser 609 in the cytoplasmic domain of GPIb␣ is phosphorylated. To further verify that the reactivity of anti-pS609 with GPIb␣ requires phosphorylation, platelets were solubilized and treated with PAP to dephosphorylate proteins. As shown in Fig. 2C, treatment of platelet lysates with PAP dramatically inhibited the binding of anti-pS609 to GPIb␣. Inhibition in anti-pS609 binding did not result from the loss of GPIb␣ because such treatment did not affect the recognition of GPIb␣ by the antibody (anti-Ib␣C) that reacts with both the phosphorylated and nonphosphorylated GPIb␣ C-terminal sequence. Thus, binding of anti-pS609 indeed requires phosphorylation of GPIb␣. Taken together, the above results indicate that Ser 609 in the cytoplasmic domain of platelet GPIb␣ is phosphorylated in resting platelets.
Stoichiometry of Ser 609 Phosphorylation-To examine the stoichiometry of Ser 609 phosphorylation, washed platelets were solubilized and lysates were immunoprecipitated with anti-pS609 to deplete the GPIb␣ population containing phosphorylated Ser 609 . The GPIb␣ that remained in platelet lysates was then detected by immunoblotting with the antibody anti-Ib␣C. Fig. 3A shows that preabsorption by anti-pS609, but not by preimmune serum, removed most of the GPIb␣ molecules (Ͼ95%) from platelet lysates. In contrast, anti-pS609 failed to remove GPIb␣ from the PAP-dephosphorylated platelet lysates (Fig. 3B). Thus, the majority of the GPIb␣ molecules in Triton X-100-soluble platelet lysates are phosphorylated at Ser 609 . As a population of GPIb-IX is associated with the Triton X-100insoluble cytoskeleton of platelets, we also examined whether phosphorylated GPIb␣ is present in the Triton X-100-insoluble fractions corresponding to the cytoskeleton (precipitated by centrifugation at 15,000 ϫ g) and the membrane skeleton (precipitated at 100,000 ϫ g) using the method reported by Fox (9). Fig. 3C shows that anti-pS609 reacted with GPIb␣ in the cytoskeleton and the membrane skeleton, and the distribution pattern of anti-pS609-binding GPIb␣ in these different fractions is similar to that reactive with anti-Ib␣C. This finding suggests that a majority of the GPIb-IX population in both the Triton X-100 soluble and insoluble fractions of platelet lysates is phosphorylated at Ser 609 .
Anti-pS609 Inhibits 14-3-3 Binding to Platelet GPIb-IX-To examine whether the phosphoserine-dependent epitope of anti-pS609 is involved in 14-3-3 binding, platelet lysates were preincubated with anti-pS609 and then with 14-3-3-coated Sepharose beads. As a control, platelet lysates were preincubated with preimmune serum from the same rabbit. Preincubation with anti-pS609 but not the control serum inhibited the binding of GPIb-IX to 14-3-3-coated beads, suggesting that the SGHpSL sequence recognized by anti-pS609 is proximate to the 14-3-3 binding site (Fig. 4).
Phosphorylation at Ser 609 of GPIb␣ Is Important for the Binding of Platelet GPIb-IX to 14-3-3 Protein-To investigate whether 14-3-3 binding to GPIb-IX is regulated by phosphorylation, the platelet lysates were pretreated with PAP to dephosphorylate proteins. This treatment inhibited the binding of anti-pS609 to GPIb␣ (Fig. 2B), suggesting that Ser 609 is dephosphorylated. PAP-treated platelet lysates were then allowed to interact with recombinant 14-3-3-conjugated beads. As shown in Fig. 5, GPIb-IX from platelet lysates bound to 14-3-3-coated beads, and this binding was dramatically reduced by PAP treatment. Thus, phosphorylation of GPIb-IX is required for high affinity binding between GPIb-IX and 14-3-3.
To further investigate whether phosphorylation at Ser 609 of GPIb␣ is important for the interaction between the GPIb␣ C-terminal sequence and 14-3-3, a nonphosphorylated peptide with a sequence corresponding to the C-terminal region of FIG. 2. Phosphorylation-dependent binding of anti-pS609 to the platelet GPIb␣. A, washed platelets were directly solubilized into SDS-sample buffer, analyzed by SDS-PAGE, and then immunoblotted with a control preimmune serum (Control) or anti-pS609 serum. B, platelets were solubilized in Triton X-100-containing buffer as described under "Experimental Procedures." GPIb-IX were immunoprecipitated with control IgG or an anti-GPIb␣ monoclonal antibody, SZ2 (SZ-2). Control IgG and SZ-2 immunoprecipitates as well as platelet lysates were immunoblotted with anti-pS609. C, the lysates were treated with (PAP) or without (No PAP) 6 units/ml potato acid phosphatase as described under "Experimental Procedures" and then analyzed by SDS-PAGE and immunoblotting with anti-pS609 or anti-Ib␣C (reactive with both phosphorylated and nonphosphorylated GPIb␣ Cterminal sequence). Note that potato acid phosphatase treatment inhibited anti-pS609 binding to GPIb␣ but did not affect the binding of anti-Ib␣C.
FIG. 3. Stoichiometry and distribution of GPIb␣ Ser 609 phosphorylation. A, washed platelets were solubilized, immunoabsorbed with anti-pS609 to remove Ser 609 -phosphorylated GPIb␣, and then analyzed by SDS-PAGE and immunoblotting with the monoclonal anti-GPIb␣ antibody, WM23. B, platelets were first treated with (ϩPAP) or without (No PAP) PAP and then immunoabsorbed with anti-pS609 or control preimmune serum. Lysates were then analyzed by SDS-PAGE and immunoblotting with anti-GPIb␣ monoclonal antibody SZ2. C, washed platelets were solubilized as described previously (9). The platelet lysates were centrifuged at 14,000 ϫ g for 5 min (low speed). and the supernatant was again centrifuged at 100,000 ϫ g for 3 h (high speed). The pellets from low speed and high speed centrifugations as well as the final supernants were solubilized in identical final volumes of SDSsample buffer and Western blotted with anti-pS609 (phosphorylationspecific) or anti-Ib␣C (reactive with both phosphorylated or nonphosphorylated GPIb␣). Note the similar distribution patterns of GPIb␣ as detected with anti-pS609 or anti-Ib␣C.
GPIb␣, S 602 IRYSGHSL 610 , and identical peptides phosphorylated at Ser 609 or Ser 606 were synthesized. Sepharose beads conjugated with recombinant 14-3-3 were preincubated with these peptides (1 mM) and then allowed to interact with GPIb-IX. As shown in Fig. 6, preincubation with nonphosphorylated or Ser 606 -phosphorylated peptides did not significantly affect the binding of GPIb-IX to 14-3-3-conjugated beads. In contrast, preincubation with the Ser 609 -phosphorylated peptide almost completely abolished GPIb-IX binding. Inhibition by the Ser 609 -phosphorylated GPIb␣ cytoplasmic domain peptide was concentration-dependent, with the half-maximal inhibition at ϳ50 M (Fig. 6B). These data suggest that phosphorylation at Ser 609 of GPIb␣ is required for the high affinity binding of platelet GPIb-IX to 14-3-3.
In Situ Distribution of Ser 609 -phosphorylated and Dephosphorylated GPIb␣ in Spreading Platelets-To examine whether phosphorylation of GPIb␣ is regulated in intact platelets, freshly washed platelets were allowed to adhere to a vWF-or fibrinogen-coated surface and were then fixed and permeabilized. These platelets were double-stained with anti-pS609 and a monoclonal antibody, WM23, against the extracellular region of GPIb␣ and were then scanned under a confocal microscope. Platelets adherent on fibrinogen and vWF were similar in staining patterns (Fig. 7). Although anti-pS609 stain (red) and WM23 stain (green) were colocalized in most parts of the platelet as indicated by the orange and yellow colors (depending on the relative intensity of each color), WM23 stain (green) was also observed in regions where there was no or very weak staining of anti-pS609 (red), suggesting that the GPIb␣ population in these regions was mostly dephosphorylated. In particular, the lamellipodium-like edge of spreading platelets was strongly stained by WM23 only. This WM23-only staining pattern was also seen at the tips of pseudopodia. Thus, it appears that dephosphorylated GPIb␣ is distributed at the leading edge of spreading platelets. These data suggest that the phosphorylation state of GPIb␣ at Ser 609 is dynamically regulated in FIG. 4. Inhibition of 14-3-3

binding to GPIb-IX by anti-pS609.
Platelet lysates were first incubated with anti-pS609 serum or preimmune negative control serum and then further incubated with control MBP-conjugated beads or 14-3-3 conjugated beads using the methods described previously (29). After three washes, the 14-3-3-bound GPIb␣ were detected by Western blotting with monoclonal antibody WM23.

FIG. 5. Phosphorylation of GPIb-IX regulates 14-3-3 binding.
Platelet lysates were first incubated with (ϩPAP) or without (No PAP) potato acid phosphatase as described under "Experimental Procedures" and then further incubated with 14-3-3-or MBP-conjugated beads. Bead-bound GPIb-IX was detected by Western blot with anti-GPIb␣ monoclonal antibody, WM23. GPIb␣ in platelet lysates (Lysate) treated with (ϩPAP) or without PAP (No PAP) were also immunoblotted with WM23 to show that amounts of GPIb-IX were not significantly changed following PAP incubation.

FIG. 6. A GPIb␣ C-terminal peptide phosphorylated at Ser 609 inhibits 14-3-3 binding to GPIb-IX.
A. synthetic peptides (1 mM) corresponding to the GPIb␣ C-terminal sequence SIRYSGHSL, a Ser 606 -phosphorylated identical peptide, SIRYpSGHSL, a Ser 609 -phosphorylated form of the same peptide, SIRYSGHpSL, or a negative control peptide corresponding to the GPIb␤ C-terminal 14 amino acid residues (Ib␤C) were incubated with 14-3-3-conjugated Sepharose beads (14-3-3) or MBP-conjugated beads (Control) at 4°C for 1 h. Platelet lysates (150 l) were then added and incubated for additional 1 h at 4°C, and after washing, the bead-bound GPIb-IX was detected by immunoblotting with anti-GPIb␣ antibody, anti-Ib␣C. B, control MBPconjugated beads or 14-3-3-conjugated beads were preincubated with increasing concentrations of nonphosphorylated GPIb␣ C-terminal peptide (SIRYSGHSL) or the Ser 609 -phosphorylated peptide (SIRYS-GHpSL) and then allowed to bind to GPIb-IX as described in A. The relative quantity of bead-bound GPIb-IX was estimated by scanning the GPIb␣ bands and then analyzing them by NIH Image for optical density. Percentages of inhibition by the peptides were calculated by the formula: Inhibition % ϭ intact platelets, and regulation of GPIb␣ phosphorylation may be involved in regulating GPIb-IX function during platelet adhesion and spreading.

DISCUSSION
In this study, we provide the first evidence that GPIb␣ is phosphorylated and that a phosphorylation site is at Ser 609 . Previously, phosphorylation of platelet membrane proteins has been studied (30,36). The only serine/threonine-phosphorylated major membrane glycoprotein identified was GPIb␤ (30). GPIb␤ is phosphorylated at Thr 166 when platelets are stimulated with agents that enhance intracellular cAMP level (36,37). However, previous studies used a 32 P labeling technique to detect protein phosphorylation, which is dependent upon incorporation of exogenous 32 P into the phosphorylated proteins and thus is not sensitive to phosphoproteins that are protected from dephosphorylation or rephosphorylation during the procedure. In this study, we used an anti-phosphopeptide antibody, anti-pS609, recognizing the SGHpSL sequence at the C terminus of GPIb␣ but not the nonphosphorylated peptide. The phosphorylation-dependent antibody can be used to detect protein phosphorylation whether or not phosphorylation is metabolically active and thus is capable of detecting GPIb␣ phosphorylation that has not been detected by the 32 P labeling technique. Use of the phosphorylation-dependent antibody also enabled us to detect phosphorylation at a specific residue. Two serine residues (Ser 606 and Ser 609 ) are present at the C-terminal SGHSL region of GPIb␣, which is important for 14-3-3 binding (29). We showed that the phosphoserine 609-specific antibody, anti-pS609, specifically bound to platelet GPIb␣ and that its binding was inhibited by dephosphorylation of GPIb-IX with potato acid phosphatase. In addition, an antibody raised against the phosphorylated Ser 606 -containing sequence (SIRYpSGH) did not react with GPIb␣ from resting platelets (data not shown). These data indicates that the C-terminal domain of GPIb␣ is phosphorylated at Ser 609 .
Phosphorylation states of proteins are balanced by the actions of protein kinases and phosphatases. In platelet lysates, the percentage of phosphorylated GPIb␣ in the whole GPIb-IX population is high, as indicated by the removal of nearly all GPIb-IX by the anti-phosphopeptide antibody (Fig. 3A); this suggests that balance under these conditions is tilted toward phosphorylation of GPIb␣. Thus, it appears that, unlike many other phosphoproteins, the default state of GPIb␣ is a phosphorylated state. One possible mechanism for this default phosphorylation state is that 14-3-3 may play a protective role, since the phosphoserine 609 is located in the 14-3-3 binding site. 14-3-3 has previously been shown to protect 14-3-3 ligands from dephosphorylation (38). The protein kinase that catalyzes phosphorylation of GPIb␣ remains to be identified. Several protein kinase inhibitors had no effect on GPIb␣ phosphorylation, including inhibitors of protein kinase A, protein kinase G, and protein kinase C (data not shown). It is thus possible that these kinases are not involved in GPIb␣ phosphorylation. However, as the default state of GPIb␣ appears to be a phosphorylated form, it is also possible that the ineffectiveness of these protein kinase inhibitors is due to the fact that GPIb␣ is already in a relatively stable phosphorylated state and thus immune to the effects of protein kinase inhibitors.
Phosphorylation at Ser 609 of GPIb␣ is important for GPIb-IX interaction with 14-3-3. This conclusion is supported by our finding that the Ser 609 -phosphorylated GPIb␣ C-terminal domain peptide (SIRYSGHpSL), but not the identical nonphosphorylated or Ser 606 -phosphorylated peptides, inhibited GPIb-IX interaction with 14-3-3 in a concentration-dependent manner (Fig. 6), suggesting that interaction between GPIb-IX and 14-3-3 involves a binding site in 14-3-3 that interacts with the Ser 609 -phosphorylated GPIb␣ C-terminal sequence. This result is consistent with the previous result of Andrews et al. (11) showing that a nonphosphorylated GPIb␣ C-terminal peptide failed to abolish the binding between 14-3-3 and GPIb-IX. Furthermore, dephosphorylation of GPIb-IX by PAP or preincubation with anti-pS609 inhibited 14-3-3 binding (Figs. 4 and 5). Thus, it is likely that high affinity interaction between the intact platelet GPIb-IX and 14-3-3 requires phosphorylation of Ser 609 of GPIb␣. It is interesting to note that the 14-3-3 binding site of GPIb␣ (RYSGHSL) shares similarities with the RSX-pSXP-like motifs of other phosphorylated 14-3-3 ligands; they all contain an arginine and a serine at the N-terminal side of the phosphorylated serine (27,28). Most of the identified RSX-pSXP motifs are present in the middle of the protein sequence, and the proline in the motif may possibly form a turn exposing the phosphoserine. Because the 14-3-3 binding site in GPIb␣ is already exposed at the C terminus, it may not require the presence of a proline residue. However, despite the similarities, there are striking differences between GPIb␣ and the RSX-pSXP-like ligands. The prototype RSXpSXP-like ligand of 14-3-3, c-Raf, requires the helix G region of 14-3-3 (33), and the crystal structure data suggest that phosphoserine in the RSX-pSXP motif may interact with residues in the more N-terminal helix C and E region of 14-3-3 (28,39). In contrast, GPIb␣ binds to the helix I region of 14-3-3 (33), which forms an amphiphilic ligand contact surface (40). Furthermore, synthetic peptides corresponding to C-terminal 15 residues of GPIb␣ bound to 14-3-3 without requiring phosphorylation (11,29), and the recombinant GPIb␣ cytoplasmic domain, which is not phosphorylated at Ser 609 (data not shown), also binds to 14-3-3 but with FIG. 7. In situ distribution of phosphorylated GPIb␣. Freshly washed platelets were allowed to spread on vWFor fibrinogen-coated glass chamber-slides at 37°C for 60 min, fixed, and then permeabilized as described under "Experimental Procedures." The platelets were then double-stained with anti-pS609 (red), and WM23 (green). The slides were scanned under a Zeiss LSM510 confocal microscope (amplification factor ϭ 2520). The orange and yellow colors indicates the colocalization of the red and green stains. Note the green color at the edge of the spreading platelets, which indicates the presence of nonphosphorylated GPIb␣. a much lower affinity than GPIb-IX from platelets (33). This suggests that the interaction of 14-3-3 with GPIb␣ may involve both phosphorylation-dependent and phosphorylation-independent binding mechanisms. However, in intact platelet GPIb-IX, Ser 609 phosphorylation is required for the high affinity binding of 14-3-3.
Phosphorylation of the Ser 609 of GPIb␣ is likely to play important roles in GPIb-IX-mediated platelet adhesion and signaling. First, phosphorylation of Ser 609 of GPIb␣ regulates 14-3-3 binding (Figs. 4 and 5), and we have evidence that 14-3-3 binding to GPIb-IX plays an important role in GPIb-IX signaling. 2 Furthermore, our data indicate that a population of GPIb␣ becomes dephosphorylated at the periphery of platelets during platelet spreading on vWF or fibrinogen (Fig. 7), suggesting that the phosphorylation state of GPIb␣ can be dynamically regulated and that phosphorylation or dephosphorylation of GPIb␣ may have a functional role during platelet spreading. Although further studies are required to understand how phosphorylation of GPIb␣ may play a role in GPIb-IX function, one possibility is that phosphorylation regulates GPIb-IX-associated membrane skeleton organization and thus regulates the movement of GPIb-IX. This possibility is supported by the finding of Dong et al. (41) that a GPIb-IX mutant, lacking the C-terminal 4 amino acid residues including Ser 609 is more likely to move laterally on the membrane. However, we show in Fig. 3 that Ser 609 -phosphorylated GPIb␣ is distributed in both cytoskeleton and non-cytoskeleton fractions, suggesting that Ser 609 phosphorylation does not directly regulate association between GPIb-IX and the membrane skeleton. Consistent with this result, GPIb-IX association with the membrane skeleton has been shown to be mediated by filamin, which binds to the central region of the GPIb␣ cytoplasmic domain distinct from the C terminus (7), and mutant GPIb-IX that lacks the C-terminal domain of GPIb␣ is still associated with filamin and the membrane skeleton (42). However, it is possible that GPIb-IX-associated cytoskeleton organization or movement of GPIb-IX may be indirectly regulated by phosphorylation of GPIb␣ and 14-3-3 binding via intracellular signaling pathways.