Protein Phosphatase 2C Inactivates F-actin Binding of Human Platelet Moesin*

During activation of platelets by thrombin phosphorylation of Thr558 in the C-terminal domain of the membrane-F-actin linking protein moesin increases transiently, and this correlates with protrusion of filopodial structures. Calyculin A enhances phosphorylation of moesin by inhibition of phosphatases. To measure this moesin-specific activity, a nonradioactive enzyme-linked immunosorbent assay method was developed with the synthetic peptideCys-Lys555-Tyr-Lys-Thr(P)-Leu-Arg560coupled to bovine serum albumin as the substrate and moesin phosphorylation state-specific polyclonal antibodies for the detection and quantitation of dephosphorylation. Calyculin A-sensitive and -insensitive protein-threonine phosphatase activities were detected in platelet lysates and separated by DEAE-cellulose chromatography. The calyculin A-sensitive enzyme was identified as a type 1 protein phosphatase. The calyculin A-insensitive enzyme activity was purified to homogeneity by phenyl- Sepharose, protamine-, and phosphonic acid peptide-agarose chromatography and characterized biochemically and immunologically as a 53-kDa protein(s) and a type 2C protein phosphatase (PP2C). Phosphorylation of Thr558 is necessary for F-actin binding of moesin in vitro. The purified enzyme, as well as bacterially made PP2Cα and PP2Cβ, efficiently dephosphorylate(s) highly purified platelet phospho-moesin. This reverses the activating effect of phosphorylation, and moesin no longer co-sediments with actin filaments. In vivo, regulation of these phosphatase activities are likely to influence dynamic interactions between the actin cytoskeleton and membrane constituents linked to moesin.

Members of the moesin protein family are localized membrane structures rich in actin filaments such as filopodia, membrane ruffles, microvilli, or the cleavage furrow in a wide range of cell types (1)(2)(3). They are proposed to act as linkers between the plasma membrane and the actin cytoskeleton. The N-terminal domain binds phosphoinositides (4), CD43 (5), CD44 (6), intercellular adhesion molecules (ICAM-1, -2, and -3 (7-9), Rho GDI (10), EBP-50 (11), type II protein kinase A regulatory subunit (12), myosin-binding subunit (13), Nhe-Rf (14), and Dbl (15), whereas the C-terminal domain interacts with F-actin (16). These interactions are thought to be dynamically regulated by signaling molecules. One of the candidate molecules is phosphatidylinositol 4,5-bisphosphate, which appears to be necessary for stabilizing moesin-CD44 interaction (4,6). Moesin is phosphorylated in human platelets by thrombin activation at a single site, Thr 558 , that is located within or near the F-actin binding domain in the C-terminal region of the moesin sequence (17). This region is nearly identical for all moesin-like proteins, and experiments in other cell types identified Thr 558 as a common phosphorylation site as well (18,19).
Thrombin stimulation of human platelets induces a rapid but transient increase in phosphorylation that correlates with F-actin binding activity of moesin (18). Although several phosphokinase and phosphatase inhibitors, including some acting on enzymes specific for modifying tyrosine residues, modulated phosphorylation at the Thr 558 site, regulatory mechanisms of moesin phosphorylation are not well understood. Rho-dependent kinase phosphorylates two threonine residues of a C-terminal fragment of the homologous protein radixin in vitro, but the enzyme did not phosphorylate the full-length protein or did so rather poorly (19). Protein kinase C also phosphorylates Thr 558 as one site of moesin in vitro, but it remains to be established how phosphorylation is regulated in blood cells, in which this enzyme is predominantly expressed (20). With regard to phosphatases, in vitro phosphorylated moesin has been dephosphorylated with myosin phosphatase, a calyculin A-sensitive enzyme (13).
Matsui et al. (19) have shown that phosphorylation of Thr 558 in the C-terminal fragment of radixin inhibited the interaction of the fragment with the N-terminal fragment. This modification had no effect on the constitutive F-actin binding activity of the C-terminal fragment. More recent evidence strongly suggests that phosphorylation and other factors in the full-length protein disrupt structural features in vitro under certain conditions, causing a substantial conformational change that exposes the high affinity binding site for F-actin (20,21). Dephosphorylation of the modified protein would be expected to reverse this allosteric change and to lead to inactivation of this binding site.
The present work demonstrates that calyculin A-sensitive and -insensitive phosphatases are detectable in human platelet lysates with a newly developed assay with a moesin-specific substrate. The calyculin A-insensitive enzyme was purified and identified as a type 2C protein phosphatase. The purified enzyme efficiently dephosphorylates highly purified in vivo phosphorylated platelet moesin and inactivates F-actin binding. This result lends support to previous speculation that phosphorylation and dephosphorylation regulate F-actin binding of fulllength endogenous moesin by an allosteric mechanism (19,22).

Preparation of Affinity Columns
The KYKcpTLR-agarose (where cpT indicates ␥-phosphono-valine, nonhydrolyzable phosphothreonyl mimetic, the so-called C-P compound) affinity column was prepared as follows. The peptide, Cys-Lys 555 -Tyr-Lys-cpThr-Leu-Arg 560 ) was chemically synthesized by Peptide Company (Osaka, Japan). The synthetic peptide (1 mg/ml of gel) was coupled to SulfoLink Coupling Gel (6% cross-linked beaded agarose; Pierce) according to the procedure recommended by the manufacturer.

Isolation of Subcellular Fractionation of Human Platelets
Outdated platelets were provided as platelet-rich plasma by the Miyagi Red Cross Blood Center (Japan). Gel-filtered platelets (17) were centrifuged at 800 ϫ g for 15 min at 30°C and resuspended in extraction buffer (50 mM Tris-HCl, 0.5 mM EGTA, 50 mM benzamidine, 10 g/ml aprotinin, 10 M E-64, 100 M p-amidinophenylmethanesulfonyl fluoride, and 100 M leupeptin, pH 7.4). The platelets were quickly homogenized using a Dounce homogenizer and subsequently subjected to an ultracentrifugation at 100,000 ϫ g for 1 h at 4°C. This first supernatant is referred to as the "cytosol." The pellet was rehomogenized using 1 ϫ Triton X-100 lysis buffer (1% Triton X-100, 50 mM Tris-HCl, 0.1 mM 2-mercaptoethanol, 10 g/ml aprotinin, 10 M E-64, 100 M p-amidinophenylmethanesulfonyl fluoride, and 100 M leupeptin, pH 7.4) and again centrifuged at 15,600 ϫ g at 4°C for 4 min. This second supernatant is referred to as the "membrane fraction." For other experiments, gel-filtered platelets were resuspended in 100 l of Tyrode's buffer (136 mM NaCl, 2.9 mM KCl, 12 mM NaHCO 3 , 0.36 mM NaH 2 PO 4 , 1.8 mM CaCl 2 , 0.4 mM MgCl 2 , 5.5 mM glucose, pH 7.4) at 1 ϫ 10 9 platelets/ml. Platelets were activated by the addition of 1.0 National Institutes of Health unit of thrombin/ml at 37°C or incubated for 10 min with 100 nM calyculin A or 1 M staurosporine at 37°C. Platelets were lysed by addition of an equal volume of 2 ϫ Triton X-100 lysis buffer. The lysates were immediately centrifuged at 15,600 ϫ g at 4°C for 4 min to sediment cytoplasmic actin filaments. This supernatant is referred to as the "Triton X-100 extract." The pellet was resuspended in 200 l of 1ϫ Triton X-100 lysis buffer containing 500 mM NaCl, homogenized, and again centrifuged at 15,600 ϫ g at 4°C for 4 min. The supernatant of this step is referred to as the "NaCl extract."

Phosphatase Assay
To monitor purification of phosphatase, BSA-KYKpTLR was used as a substrate in the enzyme assays. Microtiter plate wells were coated overnight at room temperature with 100 l of peptide solution (10 mg of peptide, KYKpTLR, as BSA conjugate/ml in 10 mM Tris-HCl, pH 8.5, 100 mM NaCl, 0.02% sodium azide). After coating, wells were washed three times with TTBS (20 mM Tris-HCl, pH 7.6, 136 mM NaCl, 0.05% Tween 20), blocked for 1 h with TTBS containing 5% nonfat milk, and washed again with TTBS. After addition of 90 l of buffer A (20 mM Tris-HCl, pH 7.2, 0.1 mM EDTA, 0.1% ␤-mercaptoethanol, 2 mM MnCl 2 ) the phosphatase assay was started by adding 10 l of enzyme preparation and incubation for 30 min at 30°C. Wells were washed three times with TTBS, incubated with pAbKYKpTLR in TTBS containing 3% BSA for 1 h, washed three times with TTBS, incubated with antirabbit IgG-horseradish peroxidase conjugate in TTBS containing 3% BSA for 1 h, and washed again three times with TTBS. Finally, peroxidase was assayed with 0.01 mg/ml 3Ј,3Ј,5Ј,5Ј-tetramethylbenzidine and 0.01% H 2 O 2 in a buffer of 0.1 M sodium acetate (pH 6.0). After addition of an equal volume of 1 M H 2 SO 4 , the optical density 450 nm was determined. In some experiments, pAbKYKTLR was used instead of pAbKYKpTLR.
The specific phosphatase activity was calculated with purified phospho-moesin as the substrate as follows. The standard reaction mixture of phosphatase assay contained buffer B (10 mM Tris-HCl, pH 7.2, 0.1 mM EDTA, 0.1% ␤-mercaptoethanol, 5 mM MgCl 2 ) and enzyme. Reaction at 30°C was initiated by the addition of phospho-moesin (final concentration, 50 g/ml ϭ 750 nM) and terminated with equal volume of 2ϫ SDS sample buffer (125 mM Tris-HCl, 4% SDS, 20% glycerol, 10% 2-mercaptoethanol, pH 6.8) followed by boiling for 5 min. A control incubation was performed without enzyme. The sample (500 ng of moesin) was separated by SDS-PAGE on a 9.0% polyacrylamide gel run under reducing conditions. The stoichiometry of moesin dephosphorylation was determined from densitometric data of Western blots obtained with affinity purified pAbKYKpTLR antibodies and the enhanced chemiluminescence detection system (Amersham Pharmacia Biotech). The phosphatase activity was calculated from the decrease in the phosphorylation level of moesin. One unit of phosphatase activity is the amount of enzyme that catalyzes the release of 1 nmol phosphate/ min from phospho-moesin in the standard assay. In some assays 20 mM sodium acetate/acetic acid, 20 mM MES, 20 mM Tris-HCl, or 20 mM glycine/NaOH was used as the reaction buffer to obtain appropriate pH conditions.

Protein Assay
Protein concentrations were determined with the bicinchoninic acid protein assay reagent (Pierce) with BSA as a standard. Because of limits of sensitivity and interference by ␤-mercaptoethanol in this assays, protein concentrations were also determined by densitometric analysis of silver-stained SDS-polyacrylamide gels.

Purification of Moesin Phosphatase
Outdated platelets were washed as previously described (26) and lysed with 1ϫ Triton X-100 lysis buffer. After centrifugation at 25,000 ϫ g for 30 min at 4°C, the supernatant was subjected to successive chromatographic steps of purification either directly or after ethanol treatment (27). All purification procedures were performed at 4°C. At each step, phosphatase activity was measured by the ELISA procedure.
Phenyl-Sepharose HP Chromatography-Phenyl-Sepharose HP column (15 ϫ 57 mm) was washed with ethanol, followed by buffer C containing 1 M NaCl. After loading the sample from step 1, the column was eluted at a flow rate of 1 ml/min with a linear salt gradient (150 ml) from 1 to 0 M NaCl in the equilibrating buffer. The phosphatase fractions were pooled.
Protamine-Agarose Chromatography-The pooled fractions from the previous step were loaded onto a protamine-agarose column (10 ϫ 13 mm), equilibrated with Buffer C. The phosphatase was eluted with a linear salt gradient (40 ml) from 0 to 1 M NaCl in the equilibrating buffer at a flow rate of 0.5 ml/min. Fractions containing phosphatase were pooled and diluted 1:4 with 10 mM Tris-HCl, pH 7.2.
KYKcpTLR-Agarose Affinity Chromatography-The diluted sample 1 The abbreviations used are: PP1, protein phosphatase type 1; PP2C, protein phosphatase type 2C; BSA, bovine serum albumin; phosphomoesin, phosphorylated moesin; ELISA, enzyme-linked immosorbent assay; rr, recombinant rat; PAGE, polyacrylamide gel electrophoresis; from the previous step were loaded onto a KYKcpTLR affinity column (10 ϫ 13 mm) equilibrated with 10 mM Tris-HCl, pH 7.2. Phosphatase was eluted with a linear salt gradient (24 ml) from 0 to 0.5 M NaCl in the equilibrating buffer at a flow rate of 0.3 ml/min. Fractions containing phosphatase were pooled.
The highly purified preparations were stored at 4°C until they were used. For longer storage, the enzyme was stored in buffer C containing 20% glycerol at Ϫ85°C.

Fast Protein Liquid Chromatography Gel Filtration Chromatography
An aliquot of the purified phosphatase was concentrated in a ultrafree centrifugal filter device (Millipore) and chromatographed on a Superose 12 column (HR 10/30) equilibrated at 4°C with 10 mM Tris, pH 7.4, 150 mM NaCl, 0.2 mM EDTA, 1 mM dithiothreitol, 5 mM MgCl 2 , and 10% glycerol, at a flow rate of 0.4 ml/min. The volume at which standard proteins eluted from the column was determined in separate runs.

F-actin Co-sedimentation Assay
Phosphorylated moesin (0.3 M) was incubated with phosphatase in buffer B. After various periods of time, the mixture was incubated with phalloidin-stabilized F-actin in buffer F (10 mM Tris-HCl, pH 7.2, 0.5 mM Na 2 ATP, 5 mM MgCl 2 , 140 mM KCl, 0.2 mM dithiothreitol, 0.2 mM CaCl 2 , 0.1% dodecyl trimethyl ammonium chloride) for 30 min at 25°C (20). The filaments were then sedimented by centrifugation at 100,000 ϫ g for 20 min at 25°C. Proteins in the supernatants were precipitated with trichloroacetic acid containing 2 mg/ml sodium deoxycholate, and the precipitates were washed with ice-cold acetone. Samples were then solubilized in SDS gel sample buffer and subjected to SDS-PAGE. Moesin and phosphorylated moesin were detected by immunoblotting.

Alkaline or Acid Phosphatase Treatment
Purified phospho-moesin was treated with alkaline phosphatase from calf intestine or Escherichia coli (Takara, Japan) (19) or wheat germ acid phosphatase (Nacalai Tesque, Kyoto, Japan) (28) as described previously. The dephosphorylation was detected by Western blotting probed with pAbKYKpTLR.

RESULTS
Detection of Protein-threonine Phosphatase Activity in Human Platelets-Originally, dephosphorylation of moesin was observed after lysis of 32 P-labeled platelets in the absence of phosphatase inhibitors, suggesting the presence of a moesin phosphatase (17). To study this further, we developed two nonradioactive methods, an ELISA and an in vitro dephosphorylation assay. Both assays are based on an antibody reagent that specifically recognizes the phosphorylation state of moesin. The ELISA was performed with BSA-KYKpTLR as the substrate. The substrate constitutes a hexapeptide centered around the phosphorylation site of moesin. This substrate can be dephosphorylated by addition of platelet lysates, and dephosphorylation is detected with specific antibodies (Fig. 1A). To confirm that this assay is measuring phosphatase and not protease activity, which might have removed the phosphopeptide, we tested with antibodies that specifically recognize the dephosphorylated peptide (18). As shown in Fig. 1B, pAbKYK-TLR reacted with BSA-KYKpTLR after this substrate was incubated with cell lysate, indicating dephosphorylation rather than proteolytic modification. This was also substantiated by the unchanged reactivity of the antibodies with BSA-KYKpTLR when incubated with the cell lysate in the presence of the protein phosphatase inhibitor calyculin A (Fig. 1A, lane  4). Because ELISA is simple and time-saving, we used this method to monitor phosphatase activity during purification. Both antibodies can be used for ELISA; however, the pAbKYKpTLR reagent is more sensitive and specific than pAbKYKTLR ( Fig. 1 and Ref. 18). We therefore primarily used pAbKYKpTLR. To make sure that KYKpTLR phosphatase dephosphorylates full-length phospho-moesin, an in vitro dephosphorylation assay was also carried out using highly puri-fied platelet phospho-moesin as the substrate, followed by Western blotting with pAbKYKpTLR. This method allows to determine specific phosphatase activity, because the amount of moesin and phospho-moesin can be quantitated by Western blotting using pAbMo (95/2) or mAbMo (38/87) and pAbKYKpTLR, respectively.
Two Distinct Phosphatase Activities Are Present in Human Platelet Lysates-We have previously shown that the addition of calyculin A to human platelets caused a substantial increase in the number of phosphorylated moesin molecules (17). This large increase is explained by inhibition of all phosphatase activity in the platelet lysate. However, some phosphatase activity was detectable when 10 mM MgCl 2 was added together with calyculin A (Fig. 2A). This suggested that distinct calyculin A-sensitive and -insensitive phosphatases existed in the human platelets lysate. The contribution of each phosphatase activity for the dephosphorylation of KYKpTLR peptide is about 1:1 in this assay ( Fig. 2A). This result is inconsistent with our previous observations that preincubation of human platelets with calyculin A for 10 min induced phosphorylation of all or nearly all moesin molecules (17). We examined whether addition of calyculin A and/or MgCl 2 before lysis affected the calyculin A-insensitive phosphatase activity, but MgCl 2 made no significant difference (Fig. 2B).
Distribution of Calyculin A-sensitive and -insensitive Phosphatase Activity in Triton X-100 Fractions-Cytosol and membrane fractions were prepared from human platelets (29), and the associated phosphatase activity was measured by the ELISA procedure. About 65% of the total phosphatase activity was recovered in the Triton X-100 soluble membrane fraction, whereas about 35% was in the soluble cytosolic fraction (Fig.  3A). The calyculin A-insensitive phosphatase activity preferentially fractionated with the Triton X-100 soluble membrane material (compare bars 3 and 4 in Fig. 3A), whereas the calyculin A-sensitive activity was almost equally distributed in the two fractions (bars 2 minus 1 versus bars 4 minus 3 in Fig. 3A). Thrombin activation did not change this distribution, and phosphatase activity was not contained in the Triton X-100 insoluble fraction either before or after thrombin activation of human platelets (Fig. 3B).
Purification of Protein-threonine Phosphatase from Human Platelets-The extraction studies established the Triton X-100- solubilized preparation of resting human platelets as the optimal starting material for purification of calyculin A-insensitive phosphatases. The Triton X-100 extract was treated with ethanol, and this treatment was necessary to remove contaminants prior to chromatographic steps. The ethanol-treated sample was solubilized in buffer C and passed through the DE52 column, which resolved two peaks of phosphatase activity. Phosphatase activity was detected in fractions eluted with an NaCl gradient at ionic strengths corresponding to 110 -200 mM NaCl (Fig. 4). Addition of 100 nM calyculin A during ELISA inhibited part of the activity (130 -200 mM NaCl in Fig. 4), indicating the calyculin A-insensitive phosphatase activity to elute at 110 -130 mM NaCl. The calyculin A-sensitive fractions contain PP1 and/or 2A by Western blotting with anti-PP1 (data not shown). The calyculin A-insensitive activity was further purified using phenyl-Sepharose HP column chromatography. Phosphatase activity eluted at ionic strengths corresponding to 750 -550 mM NaCl (Fig. 5A). Sodium chloride was used instead of ammonium sulfate, because ammonium sulfate negatively affected the phosphatase assay. Peak fractions were pooled and chromatographed on a protamine-agarose column (Fig. 5B), and fractions containing phosphatase were pooled again, diluted 1:4 with 10 mM Tris-HCl, pH7.2, and affinity purified on a KYKcpTLR affinity matrix. The phosphonopeptide KYKcpTLR is similar to KYKpTLR but contains a nonhydrolyzable phosphate mimetic instead of the phosphate group (30). The diluted sample was adsorbed and eluted after washing with equilibrating buffer with a linear gradient of 0 -500 mM NaCl. A major peak of activity eluted at 130 -200 mM NaCl (Fig. 5C). Peak fractions from each purification step were analyzed on 12% SDS-polyacrylamide gels. As shown in Fig. 6, this multi-step procedure resulted in the isolation of a 53-kDa protein that was obtained by 1130-fold purification with a yield of 4.7% (Table I).
Characterization of the Purified Protein-threonine Phosphatase-To ascertain that the 53-kDa polypeptide was in fact the protein-threonine phosphatase activity, an aliquot of the purified preparation was subjected to Superose 12 gel permeation chromatography. A single major peak was associated with phosphatase activity, and this peak of activity again yielded the 53-kDa band on silver-stained 12% SDS-polyacrylamide gels (data not shown). Enzyme activity eluted as a single peak near fractions where BSA (67-kDa) eluted in separate runs, indicating that the purified enzyme chromatographically behaved as a monomeric protein (data not shown).
Dephosphorylation of isolated phospho-moesin was catalyzed by the isolated phosphatase and was complete by 30 min (Fig. 7). The purified enzyme thus has a specific activity (V max ) for purified platelet phospho-moesin of 0.83 mol/mg of protein/min and a K m value of 0.74 M (Table II). The pH dependence of the enzymatic activity was relatively broad and ranged from 6.5 to 8.0. p-NPP, a substrate for many serine/threonine and tyrosine protein phosphatases, was not a substrate for the isolated phosphatase when assayed at several pH values (data not shown).
Purified Phosphatase Inactivates F-actin Binding of Moesin-To determine whether the purified phosphatase is able to alter the F-actin binding activity of phosphorylated moesin (20), this was examined by actin co-sedimentation. When A, aliquots of Triton X-100 extract of the insoluble fraction (membrane, white bars), and soluble fraction (cytosol, black bars) from resting platelets were tested and both contained calyculin-sensitive and -insensitive activity. B, Triton X-100 extracts (membrane, white bars) and NaCl extracts (cytoskeleton, black bars) were obtained from human platelets before and after thrombin activation. Thrombin stimulation does not appreciably change phosphatase activity in either fraction. mixed with F-actin, a large fraction of purified platelet phospho-moesin, but only a small fraction of nonphospho-moesin co-pelleted with actin filaments. Partial dephosphorylation with the purified enzyme reduced the fraction of moesin cosedimenting with actin by about 40%, and the dephosphorylated molecules were recovered in the supernatant fraction (Fig. 8).
Identification of the Calyculin A-insensitive Platelet Phosphatase as Type 2C-Serine/threonine protein phosphatases are classified based on their biochemical characteristics, divalent cation requirements, and sensitivity to inhibitors. According to this classification the purified enzyme is a type 2C protein phosphatase (Table II). This was confirmed by Western blotting with specific antibodies to PP2C and PP2C␣ (Fig. 9). We were unable to test whether the enzyme preparation also contains PP2C␤, because human-specific antibodies are not available. Both recombinant PP2C␣ and PP2C␤, but neither alkaline nor acid phosphatase, are able to dephosphorylate phospho-moesin (Fig. 10) and to inactivate F-actin binding (data not shown). DISCUSSION Our initial observations on the Thr 558 phosphorylation of moesin in resting and thrombin-activated human platelets (17) led us to search for kinases and phosphatases that regulate modification at this unique site. To detect and to monitor purification of potential moesin-specific enzymes, we developed nonradioactive assays in which phosphorylation by protein kinases or dephosphorylation by phosphatases of synthetic peptide substrates are measured with phosphorylation state-specific antibodies (18). These methods also proved to be quite useful to study enzyme kinetics and to screen for potential inhibitors.
Synthetic peptides containing amino acid sequences centered around phosphorylation or dephosphorylation sites have been used as immobilized ligands for the purification of kinases or phosphatases (31,32). Hydrolytically stable thiophosphorylated peptides are particularly useful for the isolation of phosphatases (31). Because such peptides are prepared by kinasephosphorylation with ATP␥S, the application depends on whether a particular kinase is available or not. As an alternative, the phosphonic acid mimetic, the so-called C-P compound, can be chemically synthesized and is nonhydrolyzable (33). Moesin kinases were unknown at the time when we started to purify moesin phosphatases from platelets, and we therefore employed the synthetic phosphonic acid peptide, KYKcpTLR as the affinity ligand for purification. cpT is ␥-phosphonovaline or a phosphothreonyl mimetic. We purified and identified a type 2C protein phosphatase as one of at least two moesin phosphatases in human platelets. PP2C is one of four major protein serine/threonine protein phosphatases (PP1, 2A, 2B, and 2C) in eukaryotic cells and is distinct from the other three classes of phosphatases, because it is Mg 2ϩ or Mn 2ϩ -dependent, because it is calyculin A-and okadaic acid-insensitive, because it consists only of a catalytic subunit, and because its amino acid sequence is unrelated to the catalytic subunits of other types of phosphatases (34). The PP2C family consist of multiple isoforms including PP2C␣ (34), PP2C␤ (␤-1, -2, -3, -4, and -5) (35), PP2C␥ (36), Wip1 (37), and FIN13 (38) in mammals. Although little is known of their physiological role, they appear to function in Ca 2ϩ -dependent signal transduction (39), DNA repair systems (40), mitogenactivated protein kinase systems (41), and the dephosphorylation of cofilin (42).
Calyculin A treatment of platelets induces complete and okadaic acid induces partial phosphorylation of moesin in micromolar concentrations (17). This pattern is most consistent a Determined by the bicinchoninic acid protein assay. b Estimated by the densitometric analysis of silver-stained SDS-polyacrylamide gels.
FIG. 7. The purified enzyme dephosphorylates purified platelet phospho-moesin. Equal amounts of phosphorylated platelet moesin were incubated with aliquots of the purified protein-threonine phosphatase for the designated time periods. SDS sample buffer was added to stop the reaction. Aliquots of the mixture (500 ng of moesin) was separated by SDS-PAGE on a 9.0% polyacrylamide gel. Phosphorylated moesin was detected by immunoblotting using pAbKYKpTLR and the enhanced chemiluminescence detection system. The antibodies were removed from the membrane according to the manufacturer's protocol. Moesin was then detected by immunoblotting with monoclonal antibodies (mAbMo) and enhanced chemiluminescence detection. Note complete dephosphorylation of full-length moesin in 30 min.  8. Purified platelet phospho-moesin loses F-actin binding activity after treatment with the purified phosphatase. Purified phosphorylated (p-moesin) and nonphosphorylated (np-moesin) platelet moesin (0.3 M) were incubated either each alone or together with purified enzyme for 30 min at 25°C. The mixture was incubated with phalloidin-stabilized F-actin (2 M) in buffer F for 20 min at 25°C prior to centrifugation as described under "Experimental Procedures." Equal volumes of supernatant (S) and pellet (P) fractions were analyzed by SDS-PAGE. Moesin (anti-moesin) and phosphorylated moesin (anti-KYKpTLR) were detected by immunoblotting. A typical result from three independent experiments is shown. Note the decrease in pelleted moesin in the phosphatase-treated sample as compared with untreated p-moesin and increased recovery in the supernatant. with a type 1 protein phosphatase. In fact, PP1 myosin phosphatase is expressed in human platelets and is able to dephosphorylate recombinant C-terminal and full-length moesin in vitro. It is a serine/threonine phosphatase composed of a 38-kDa catalytic subunit PP1␦, a 130-kDa myosin-binding subunit, and a 20-kDa subunit (43). The calyculin A-sensitive phosphatase that we have detected by Western blotting in chromatographic fractions could thus be identical to myosin phosphatase. However, this would require further analysis, because substrate specific properties have not been firmly established. For example, Fukata et al. (13) phosphorylated recombinant C-terminal and full-length moesin with Rho-kinase in vitro but did not determine phosphorylation sites and stoichiometry. Matsui et al. (19) showed that this enzyme phosphorylated ϳ100% of Thr 564 (corresponding to Thr 558 of moesin) of a recombinant radixin C-terminal domain fragment and with ϳ40% efficiency an additional site. These could be dephosphorylated with alkaline phosphatase (19), whereas we could not remove the phosphate from the single physiologically relevant site of highly purified full-length platelet phospho-moesin under similar experimental conditions, as shown here.
We could also demonstrate that dephosphorylation of moesin with PP2C inactivates F-actin binding. It has been proposed that F-actin binding is regulated by an allosteric mechanism that involves, at least in part, phosphorylation of Thr 558 (20,21). PP2C could play a role in regulating phosphorylation at this site, because the physiological concentration of Mg 2ϩ is about 10 mM. It needs to be clarified, however, how addition of calyculin A to platelets induces complete phosphorylation of moesin. A possible candidate is certainly PP1. This enzyme is sensitive to calyculin A, and the myosin-binding subunit interacts with moesin (13). On the other hand, PP2C activity may be modulated by phosphorylation (44) or through association with proteins or lipids via signaling pathways that are affected by calyculin A. This drug also causes major rearrangements of cytoskeletal proteins, including moesin, in the platelet (17,20), which could certainly compromise physiological relationships at the plasma membrane.
The detection and purification methods described here are unique and appear to be specific for the identification of phosphatases. They are applicable to the identification of specific moesin kinases among the 14 distinct threonine/serine kinases that have been found in platelets (45) and for their characterization in other cell systems. The methods can be adapted to other phosphorylated peptides and, together with recent pro-gress in two-dimensional gel electrophoresis and mass spectroscopic analysis, will provide powerful tools for the identification of phosphorylation sites and facilitate future functional studies (46,47).