The low-molecular-weight phosphotyrosine phosphatase is a negative regulator of FcγRIIA-mediated cell activation

The transient phosphorylation of proteins on tyrosine residues represents a very common and versatile signaling mechanism regulating several cellular functions. Changes in the tyrosine phosphorylation level of multiple intracellular substrates also occur in circulating blood platelets upon stimulation with virtually all the physiological agonists.^1,2  These include soluble molecules acting on 7 transmembrane G-protein–coupled receptors, such as thrombin or thromboxane A₂; adhesive proteins, such as von Willebrand factor (VWF) and collagen, which bind to the peculiar platelet receptors GPIb-IX-V and GPVI, respectively; and integrin ligands, such as fibrinogen, which initiates an outside-in signaling process regulating platelet aggregation and clot retraction.^1–5 

The overall level of protein tyrosine phosphorylation and the regulation of the strength and duration of a phosphorylation-based signaling pathway result from the balanced action of protein kinases and phosphatases. While a large repertoire of soluble cytosolic tyrosine kinases has been described in platelets, and has been extensively studied in the past, relatively little information is available on the expression, substrate specificity, and function of protein-tyrosine phosphatases.⁶  Only 3 enzymes, of a superfamily including more than 90 members, have been described in platelets: PTP-1B, SHP-1, and SHP-2.

PTP-1B is the most abundant phosphatase in platelets, and is activated by calpain-dependent proteolysis or by interaction with the cytoskeleton during integrin αIIbβ3-dependent aggregation.^7,8  Only 2 substrates for PTP-1B have been unambiguously identified in platelets: the adaptor protein LAT and the tyrosine kinase c-Src.^9,10  While the late dephosphorylation of LAT is associated with the negative regulation of platelet activation induced by clustering of the low-affinity receptor for immunoglobulin G, FcγRIIA,⁹  activation of c-Src by dephosphorylation of Tyr⁵²⁹ occurs upon fibrinogen binding to platelets, and implicates PTP-1B in the positive regulation of integrin α_IIbβ₃ outside-in signaling.¹⁰ 

SHP-1 and SHP-2 are SH2 domain–containing tyrosine phosphatases.¹¹  In thrombin- or collagen-stimulated platelets, SHP-1 interacts with the cytoskeleton and associates with c-Src.^12,13  SHP-1 is activated by tyrosine phosphorylation, and negatively regulated by PKC-directed serine phosphorylation.^14,15  During platelet aggregation, both SHP-1 and SHP-2 can be activated by interaction with the tyrosine-phosphorylated intracellular domain of the membrane receptor PECAM-1.¹⁶  Although many substrates for SH2-containing phosphatases have been identified in nucleated cells, only Vav and α-actinin have been demonstrated to be dephosphorylated by SHP-1 in platelets.^15,17  In addition, a reduced, rather than increased, protein phosphorylation has been observed in collagen-stimulated platelets from mice expressing a catalytic inactive form of SHP-1, suggesting a positive role in platelet activation.¹⁸ 

The low-molecular-weight protein tyrosine phosphatase (LMW-PTP) is a cytosolic 18-kDa enzyme that has so far attracted relatively little attention.¹⁹  This protein is widely expressed in multiple isoforms derived by alternative splicing, and shows little sequence similarities to other known phosphatases, apart from a common CX₅R motive in the active site and an identical catalytic mechanism.¹⁹  Although generally referred to as a cytosolic enzyme, LMW-PTP dynamically localizes in the cytoskeleton of activated cells.²⁰  In transfected fibroblasts, LMW-PTP has been implicated in the regulation of growth factor–induced cell stimulation. Activated membrane receptors for PDGF, FGF, MCSF, insulin, and ephrins are, indeed, substrates for LMW-PTP.^20–24  In addition, the enzyme can efficiently dephosphorylate p190RhoGAP and p120FAK, suggesting a role in the control of cell adhesion and spreading.^25,26  Regulation of LMW-PTP is complex, and involves c-Src–directed tyrosine phosphorylation, interaction with the cytoskeleton, and redox modification.^20,27–29 

In this work, we have investigated the expression and function of LMW-PTP in circulating platelets and in DAMI megakaryocytic cells. We report that LMW-PTP is actually expressed in these cells and plays a crucial role in the regulation of cell activation downstream of immunoreceptor tyrosine–based activation motif (ITAM)–containing receptors.

Sepharose CL-2B and protein G–Sepharose were from GE Healthcare (Cologno Monzese, Italy). FURA-2-AM was from Calbiochem (Milan, Italy). Thrombin, protein A–Sepharose, and anti–mouse IgG F(ab′)₂ fragments were from Sigma (Milan, Italy). Convulxin was provided by Dr K. J. Clemetson (Theodor Kocher Institute, University of Berne, Switzerland). Rabbit polyclonal anti–LMW-PTP antibody was prepared as previously described.³⁰  Human recombinant LMW-PTP was expressed in E coli, and purified as previously described³¹ ; 1U defines the amount of enzyme able to hydrolyze 1 μmol p-nitrophenyl phosphate in 1 minute at 37°C and pH 5.5. Phosphotyrosine monoclonal antibody 4G10, anti-LAT, and anti–FcR γ-chain polyclonal antibodies were from Upstate Biotechnology (Lake Placid, NY). Polyclonal antibodies against FcγRIIA and phospholipase C γ2 (PLCγ2), and monoclonal antibody against Syk were from Santa Cruz Biotechnology (purchased through Tebu-Bio, Magenta, Italy). Monoclonal antibody IV.3 was from Medarex (Annandale, NJ). The rabbit polyclonal antibody against phospho-(Ser)-PKC substrates was from Cell Signaling Technology (purchased through Celbio, Pero, Italy). Appropriate peroxidase-conjugated anti-IgG antibodies were from Bio-Rad (Milan, Italy). Lipofectamine 2000 was from Invitrogen (Milan, Italy). Hygromycin was from Serva (purchased through Celbio).

Samples were obtained from the Immunohematology and Transfusion Service, IRCCS Policlinico San Matteo, University of Pavia (Pavia, Italy), where blood was withdrawn from healthy volunteers, using 10% citric acid—citrate–dextrose as anticoagulant. Samples were centrifuged at 120g for 10 minutes at room temperature to obtain the platelet-rich plasma. Upon addition of 0.02 U/mL apyrase and 1 μM PGE₁, the platelet-rich plasma was centrifuged at 720g for 15 minutes. The platelet pellet was washed with PIPES buffer (20 mM PIPES, 136 mM NaCl, pH 6.5), centrifuged at 600g for 15 minutes, and finally resuspended in HEPES buffer (10 mM HEPES, 137 mM NaCl, 2.9 mM KCl, 12 mM NaHCO₃, pH 7.4). The cell count was adjusted to 1 × 10⁹ platelets/mL with HEPES buffer. Platelet stimulation was performed at 37°C under gentle stirring with thrombin (1 U/mL, 1 minute) or convulxin (100 ng/mL, 1 minute), or by clustering of FcγRIIA, obtained by addition of 2 μg/mL IV.3 monoclonal antibody for 2 minutes followed by 30 μg/mL sheep antimouse F(ab′)₂ fragments for 1 minute.

DAMI human megakaryocytic cells were routinely cultured in RPMI 1640 supplemented with 10% heat-inactivated fetal calf serum, l-glutamine, and antibiotics (penicillin/streptomycin). The PTP13 DAMI cell clone overexpressing LMW-PTP has been previously described.³²  A DAMI cell clone with a reduced expression of LMW-PTP was generated by siRNA interference. The pRNAT-U6.1/Hygro vector containing the siRNA 5′-CAGCGGTGCTGTTTCTGACTG-3′ (sense sequence) under the control of the U6 promoter was obtained from GenScript (Piscataway, NJ). As a negative control, a plasmid containing the scrambled sequence 5′-GTCGGCTGGTATCTTACGTGC-3′ was used. Cells (0.5 × 10⁵) were plated in 0.5 mL RPMI without serum and transfected with 1 μg DNA plus 4 μL Lipofectamine 2000. After 24 hours, cells were given complete medium and, after 48 hours, selection was started by adding 100 μg/mL hygromycin. DAMI S2C cell clone displaying a 50% reduction in LMW-PTP content (as evaluated from immunoblotting with anti–LMW-PTP antibody) was so obtained. Stable suppression was maintained by periodically growing the cells in selective medium. v-src–transformed NIH3T3 murine fibroblasts expressing LMW-PTP were obtained and cultured as previously described.²⁷ 

Platelet samples (0.3 mL) were stimulated at 37°C with convulxin or by cross-linking of FcγRIIA. DAMI cells were serum starved for 24 hours in RPMI containing 1% BSA, washed with PBS, and resuspended at the final concentration of 2 × 10⁷ cells/mL. Cells in 0.2-mL samples were stimulated by cross-linking of FcγRIIA for increasing time periods. Platelet or cell stimulation was stopped by addition of an equal volume of ice-cold 2 × lysis buffer (100 mM Tris, pH 7.4, 300 mM NaCl, 2 mM EGTA, 2% nonidet P-40, 0.5% sodium deoxycholate, 2 mM PMSF, 2 mM NaF, 5 μg/mL leupeptin, 5 μg/mL aprotinin, 2 mM Na₃VO₄), and samples were placed on ice for 10 minutes. Immunoprecipitation of Syk, PLCγ2, LAT, FcγRIIA, or FcR γ-chain was then performed essentially as previously described.³³  Immunoprecipitated proteins were then separated by sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE) on 7.5% linear or 10% to 20% gradient acrylamide gels, transferred to nitrocellulose, and analyzed by immunoblotting with antiphosphotyrosine antibody as previously described.³³  Upon stripping, membranes were reprobed with the same antibody used for the immunoprecipitation. For analysis of LMW-PTP expression in platelets, whole lysates were obtained by addition of 2% SDS to platelet samples. Upon determination of protein concentration by the bicinchoninic acid assay, aliquots of the lysates were analyzed by immunoblotting with anti–LMW-PTP antibody. This was performed through the same procedure for immunoblotting as previously described,³³  using the anti–LMW-PTP antibody at a 1:1000 final dilution and appropriate peroxidase-conjugated anti–rabbit IgG antibody. All the reported figures are representative of results obtained at least in 3 different experiments.

Platelet samples (0.1 mL) were stimulated with thrombin or convulxin, or by clustering of FcγRIIA, as indicated above, and reactions were stopped by addition of an equal volume of 2 × lysis buffer. Identical aliquots of the lysates were incubated in the absence or in the presence of 5 U/mL purified recombinant LMW-PTP³¹  at 30°C for 5 minutes. At the end of the incubation, an equal volume of SDS-sample buffer was added, and samples were heated at 96°C for 3 minutes. The level of protein tyrosine phosphorylation in identical aliquots from each sample was evaluated by immunoblotting with an antiphosphotyrosine antibody.

Syk, PLCγ2, LAT, FcγRIIA, and FcR γ-chain were immunoprecipitated from platelets stimulated by cross-linking of FcγRIIA or with convulxin, as described above. Immunoprecipitated Syk, PLCγ2, and FcγRIIA, immobilized on protein A-Sepharose, were resuspended in 50 μL 0.1-M sodium acetate, pH 5.5, 1 mM EDTA, and 1 mM DTT. LAT and FcR γ-chain were resuspended in 50 μL 62-mM HEPES, pH 7.0, 0.2 M NaCl, 1 mM EDTA, and 5 mM DTT. This different buffer was adopted, because in preliminary experiments it was found more suitable in order to minimize a nonspecific loss of the phosphate group from these immunoprecipitated proteins during the following incubation at 30°C. Proteins were then incubated in the absence or presence of 5 U/mL purified recombinant LMW-PTP at 30°C for 30 minutes. The reaction was terminated by centrifugation at 17 900g for 30 seconds, and the supernatants were removed. The pellets were dissociated with SDS-sample buffer and heated at 96°C for 3 minutes. The tyrosine phosphorylation of the immunoprecipitated proteins was then determined by immunoblotting with an antiphosphotyrosine antibody.

Alternatively, immunoprecipitated proteins were immediately separated by SDS-PAGE and transferred to nitrocellulose. Membranes were subsequently incubated with or without 1 U/mL LMW-PTP (in 0.2 mL 0.1-M sodium acetate, pH 5.5, 1 mM EDTA, 1 mM DTT for Syk, PLCγ2, and FcγRIIA or 62 mM HEPES, pH 7.0, 0.2 M NaCl, 1 mM EDTA, 5 mM DTT for LAT and FcRγ-chain), for 1 hour at room temperature. After extensive washing, nitrocellulose membranes were incubated with a specific antibody in order to determine the level of tyrosine phosphorylation of immunoprecipitated proteins.

Quantitative analysis of LMW-PTP–mediated dephosphorylation of the immunoprecipitated substrates was performed by densitometric analysis of 3 to 5 different experiments for each protein.

Membrane and cytosolic fractions were rapidly separated essentially as described.³⁴  Briefly, platelets (0.2 mL) were lysed with 0.05% digitonin in PBS, pH 7.4, and centrifuged at 17 900g for 1 minute. The supernatants (cytosolic fractions) were removed and collected. The pellets (membrane-rich fractions, also containing membrane skeleton and membrane-associated cytoskeleton) were resuspended in an equal volume of PBS containing 0.05% digitonin. Cytosolic and membrane fractions were finally dissociated with 2% SDS, and heated at 96°C for 3 minutes. Cytoskeleton was extracted from resting and stimulated platelets (0.4 mL) upon lysis with 1% Triton X-100, and was isolated by high-speed centrifugation, as previously described.³⁵  Membrane, cytosolic, and cytoskeletal proteins from aliquots of each sample corresponding to the same number of platelets were separated on a 10% to 20% SDS-PAGE and analyzed by immunoblotting with anti–LMW-PTP antibody.

Serum-starved cells (5 × 10⁶ cells/mL) were loaded with 3 μM FURA-2-AM in RPMI containing 1% BSA for 1 hour at 37°C. Cells were then washed and resuspended in HEPES buffer containing 0.5% BSA and 5.5 mM glucose at the final concentration of 2 × 10⁶ cells/mL. Measurement of cytosolic Ca²⁺ was performed on 0.4-mL samples prewarmed at 37°C under gentle stirring in a PerkinElmer Life Sciences LS3 spectrofluorometer (Shelton, CT), essentially as previously described.³⁶ 

Platelet lysates from 4 different donors were analyzed by immunoblotting with a polyclonal antibody against LMW-PTP. Figure 1A shows that a major, single, and strong band with a molecular mass of about 20 kDa was detected in all the samples with a comparable intensity. This band had the same electrophoretic mobility as the purified phosphatase (Figure 1B). In order to quantify the expression of LMW-PTP, increasing amounts of the purified, recombinant enzyme and total platelet lysates were loaded on the same gel, and analyzed by immunoblotting (Figure 1B). By densitometric analysis, we calculated that 10⁷ platelets contained 8.2 ± 1.7 ng (n = 4) LMW-PTP, which represents about 0.05% of the total cell proteins.

We next investigated whether endogenous substrates for LMW-PTP are present in human platelets. In order to analyze different subsets of tyrosine-phosphorylated proteins, platelets were treated with 3 different agonists: thrombin, convulxin, and antibodies able to cross-link the FcγRIIA. Aliquots of cell lysates were incubated in the absence or presence of purified recombinant LMW-PTP for 5 minutes. Immunoblotting analysis with antiphosphotyrosine antibody revealed a single band of about 60 kDa in nonstimulated platelets. This protein presumably corresponded to pp60^src, which is known to be abundant and constitutively phosphorylated in platelets, and was not affected by incubation with LMW-PTP (Figure 2). In stimulated platelets, multiple substrates were tyrosine phosphorylated, and many of them underwent evident dephosphorylation upon incubation with LMW-PTP. As documented by a number of previous publications,^4,33,37  we have been interested for a long time in the role of FcγRIIA in platelet activation, a process that is critically dependent on tyrosine phosphorylation of many intracellular effectors, including the receptor itself. Therefore, although LMW-PTP appeared to selectively recognize multiple substrates in cells stimulated with all the agonists analyzed, we decided to focus our further investigations on platelets activated by clustering of FcγRIIA, where many of the substrates promptly dephosphorylated by LMW-PTP showed a molecular mass ranging between 30 and 50 kDa.

We next investigated the ability of LMW-PTP to dephosphorylate selected elements in the signal transduction pathway initiated by FcγRIIA: the receptor itself, the tyrosine kinase Syk, the adaptor protein LAT, and the enzyme PLCγ2.^38,39  These proteins were immunoprecipitated from stimulated platelets, and immobilized on nitrocellulose by Western blotting. Membranes were incubated with or without recombinant LMW-PTP for 30 minutes, and the level of tyrosine phosphorylation of the proteins was evaluated by immunoblotting. Figure 3A shows that immobilized FcγRIIA and LAT underwent evident dephosphorylation when incubated with LMW-PTP, while Syk and PLCγ2 did not. In a different approach, the immunoprecipitated proteins were directly incubated with LMW-PTP in solution, and then subjected to SDS-PAGE and immunoblotting with the antiphosphotyrosine antibody. In this way, we confirmed that FcγRIIA and LAT, but not Syk or PLCγ2, were efficiently dephosphorylated by LMW-PTP (Figure 3B). It is well known that tyrosine phosphorylation of FcγRIIA occurs within the ITAM.^36,38  In addition to FcγRIIA, platelets express another ITAM-bearing receptor, the FcR γ-chain, which is associated with GPVI and is phosphorylated upon platelet stimulation with collagen or convulxin.^3,40,41  Although some dephosphorylation of FcR γ-chain was observed upon incubation with LMW-PTP both in solution and upon immobilization on nitrocellulose (Figure 3A,B), this effect was marginal compared with that seen on FcγRIIA or LAT. Quantification of LMW-PTP–mediated dephosphorylation was evaluated by densitometric analysis, and is reported in Figure 3C. It is confirmed that LAT, but not Syk or PLCγ2, was dephosphorylated by LMW-PTP, with a greater efficiency when the protein and the enzyme were incubated in solution. In addition, it is evident that LMW-PTP had no significant effects on FcR γ-chain. These results, however, indicate that LMW-PTP can act as negative regulator of FcγRIIA.

Since LMW-PTP dephosphorylates critical players in the signal transduction pathway activated by FcγRIIA, we investigated whether this phosphatase could be directly targeted by engagement of this receptor. It has been previously shown that enzymatic activity of LMW-PTP is stimulated by Src-mediated tyrosine phosphorylation.^27,28  By immunoprecipitation and immunoblotting experiments, however, we found that, in platelets, LMW-PTP was not tyrosine phosphorylated upon clustering of FcγRIIA (Figure 4Ai) or upon stimulation with thrombin or convulxin (data not shown). Concomitant evaluation of platelet aggregation confirmed the effective activation of platelets stimulated by clustering of FcγRIIA (Figure 4Ai). Although we have been unable to observe tyrosine phosphorylation of LMW-PTP even in platelets treated with pervanadate (data not shown), previous works have documented that the antiphosphotyrosine antibody used in this study is actually suitable to detect the phosphorylated form of LMW-PTP.⁴²  Moreover, we also documented phosphorylation of LMW-PTP immunoprecipitated from v-src–transformed NIH3T3 cells by immunoblotting with antiphosphotyrosine antibody (Figure 4Aii). These results, therefore, demonstrate that neither platelet activation nor aggregation can trigger LMW-PTP tyrosine phosphorylation. In fibroblasts, LMW-PTP has been reported to be regulated through interaction with the cytoskeleton,²⁵  and, in platelets, the cytoskeleton is known to coordinate several signaling molecules.⁴³  Upon platelet stimulation by clustering of FcγRIIA, LMW-PTP underwent a transient and time-dependent translocation to the cytoskeleton, which was maximal within 60 seconds, as aggregation reached about 90%, and decreased afterward (Figure 4B). A similar time-dependent translocation of LMW-PTP to the cytoskeleton was also seen in platelets stimulated with thrombin or convulxin (data not shown).

We next analyzed the distribution of LMW-PTP between the cytosol and the membrane-rich fraction. In resting platelets, LMW-PTP was almost exclusively present in the cytosol, but an evident, time-dependent translocation to the membrane-rich fraction was seen upon platelet stimulation by clustering of FcγRIIA (Figure 4C). Altogether, these results indicate that the subcellular distribution of LMW-PTP is dynamically regulated upon platelet activation. In particular, the different kinetics reported in Figure 4B,C indicate that upon stimulation of FcγRIIA, interaction of LMW-PTP with the intracellular actin-based cytoskeleton precedes its association with the membrane fraction.

In order to get further insights into the physiological role of LMW-PTP in the regulation of FcγRIIA-mediated cell activation, we analyzed 2 different clones of the megakaryocytic cell line DAMI: the PTP13 clone, overexpressing the phosphatase, and the S2C clone, in which expression of LMW-PTP was reduced by siRNA. By immunoblotting analysis, we confirmed that in comparison with wild-type cells, the expression of LMW-PTP was reduced in the S2C and increased in the PTP13 clones, and we verified that all these cells expressed comparable levels of FcγRIIA (Figure 5A).

Figure 5B,C show the kinetics of antibody-induced FcγRIIA tyrosine phosphorylation in DAMI cells. Figure 5B reports some representative immunoblots, whereas Figure 5C summarizes data from 3 to 5 different experiments. In wild-type cells, clustering of FcγRIIA caused a rapid and transient tyrosine phosphorylation of the receptor, which persisted up to 45 seconds, and was followed by a complete dephosphorylation within 2 minutes. In our experiments with DAMI cells, the antiphosphotyrosine antibody detected an additional band of slightly higher molecular mass in immunoprecipitated FcγRIIA. The identity of this band is currently unknown, but it may represent either a differently glycosylated form of FcγRIIA itself or a FcγRIIA-associated protein. In the S2C clone with reduced expression of LMW-PTP, the initial rapid antibody-induced tyrosine phosphorylation of the receptor was not altered, but the subsequent dephosphorylation was impaired and incomplete even after 5 minutes. By contrast, in the PTP13 clone, overexpression of LMW-PTP was associated with a reduced initial phosphorylation of FcγRIIA and to a more rapid dephosphorylation, which was already complete within 45 seconds (Figure 5B).

We next investigated the effect of LMW-PTP suppression or overexpression on intracellular signaling events elicited by FcγRIIA engagement. It is well known that tyrosine-phosphorylated FcγRIIA binds the kinase Syk, which is activated by autophosphorylation, and contributes to the phosphorylation and activation of PLCγ2.^38,44,45 Figure 6 shows that dephosphorylation of Syk on tyrosine residues was accelerated in the DAMI cells overexpressing LMW-PTP (clone PTP13), and delayed when the expression of the phosphatase was reduced (clone S2C). Interestingly, the kinetics of Syk phosphorylation in the 3 different clones of DAMI megakaryocytic cells was superimposable to that of FcγRIIA (Figure 5B). Since Syk is not a direct substrate for LMW-PTP (Figure 3A,B), these results reflect the altered upstream regulation of the FcγRIIA, rather than a direct action on Syk itself.

We next investigated the activation of PLC triggered by clustering of FcγRIIA in DAMI cells, by measuring the release of Ca²⁺ from intracellular stores. The initial rapid and transient increase of cytosolic Ca²⁺ in FcγRIIA-stimulated cells was not altered when the expression of LMW-PTP was reduced, but was clearly inhibited in the PTP13 clone, overexpressing LMW-PTP (Figure 7A). As expected, all 3 clones underwent comparable increase of the intracellular Ca²⁺ concentration upon stimulation with thrombin (Figure 7A). Although signaling through activated FcγRIIA has been previously found to potentiate Ca²⁺ signaling induced by low doses of thrombin,³³  at the high concentration used in the present experiments (1 U/mL), thrombin-induced intracellular Ca²⁺ mobilization was recognized to occur mainly through G-protein–mediated stimulation of PLCβ isoforms rather than through stimulation of PLCγ2 by ITAM-dependent signaling.³³  These results indicate that overexpression of LMW-PTP results in an impaired PLC activity downstream of FcγRIIA engagement.

In this work, we have identified LMW-PTP as a cytosolic protein tyrosine phosphatase expressed in human platelets, and we have found that this enzyme can efficiently dephosphorylate the ITAM-bearing FcγRIIA, both in vitro and in vivo. This action has relevant consequences on the final outcome of cell response to FcγRIIA engagement, as LMW-PTP–mediated dephosphorylation of the receptor was found to affect the correct organization of the intracellular signaling pathway for the propagation of the activating signal, eventually resulting in a defective PLC activation and altered second messenger generation. In addition, we have found that LMW-PTP may also promote, at least in vitro, the dephosphorylation of FcR γ-chain, the ITAM-containing subunit of the collagen receptor GPVI. Therefore, LMW-PTP may represent a new versatile and important regulator of platelet activation through immunoreceptors. We calculated that, in platelets, LMW-PTP accounts for about 0.05% of the total platelet proteins, and, thus, it appears less abundant than other phosphatases, such as PTP-1B, which was reported to represent 0.2% of the platelet proteins.⁷  However, LMW-PTP represents, to our knowledge, the first described enzyme able to dephosphorylate FcγRIIA, and to negatively regulate cell activation triggered by this immunoreceptor.

FcγRIIA is a 40-kDa single-chain transmembrane receptor, with 2 IgG-like motifs in the extracellular domain and a ITAM in the short cytoplasmic sequence.⁴⁶  When cross-linked by immunocomplexes, as it occurs in some autoimmune diseases or during heparin-induced thrombocytopenia, it initiates an intracellular signaling pathway similar to the one elicited by other immunoreceptors in B and T lymphocytes.^38,46  Tyrosine-phosphorylated ITAM of FcγRIIA recruits and activates the tyrosine kinase Syk, which propagates the signal through the adaptor protein LAT and the enzyme PLCγ2. Eventually, PLC-driven intracellular Ca²⁺ mobilization and PKC activation trigger platelet secretion and aggregation. Using a megakaryocytic cell line model, we have demonstrated that dephosphorylation of FcγRIIA by LMW-PTP efficiently dampens the downstream signaling cascade and attenuates activation of both Syk and PLCγ2. Activation of Syk has been evaluated by measuring its tyrosine phosphorylation, whose intensity and duration have been found to be influenced by the altered LMW-PTP expression. It should be noted that LMW-PTP is unable to directly target Syk, as demonstrated by in vitro dephosphorylation assays. Therefore, the alteration of the levels of Syk phosphorylation is most likely a consequence of the effects of LMW-PTP on FcγRIIA phosphorylation. Altogether, our results indicate that by targeting the phosphorylated ITAM, LMW-PTP may regulate the overall cellular response to FcγRIIA engagement. While overexpression of LMW-PTP was clearly associated with inhibition of FcγRIIA-mediated cell activation, the reduced expression of the phosphatase caused a prolonged activation of Syk, but did not result in any evident potentiation of PLC activation. Identical results were obtained from the analysis of a different DAMI clone, S27, with a comparable reduction of LMW-PTP expression, obtained by siRNA, ruling out the possibility that this lack of effect on cell activation was related to the particular clone analyzed (data not shown). It is, however, reasonable to suppose that additional and more selective regulatory mechanisms predominate over the persistence of ITAM phosphorylation in the temporal regulation of IP₃- and DAG-mediated effects. Nevertheless, the intensity of FcγRIIA activation represents the original signal triggering the rapid PLC-dependent response, and, in this context, our findings are consistent with the observation that the initial rapid phosphorylation of FcγRIIA was unaltered in the cells expressing a reduced level of LMW-PTP.

The ability of LMW-PTP to dephosphorylate FcγRIIA may appear particularly relevant in light of a number of recent pieces of evidence implicating this receptor in numerous and diverse processes of platelet activation. In fact, the ITAM-orchestrated signaling pathway involving Syk, LAT, and PLCγ2 represents a signaling cassette widely exploited by many agonists in order to elicit full platelet activation. In particular, many studies have documented the direct involvement of ITAM-bearing receptors in VWF-GPIb-IX-V–dependent platelet activation,^47–49  in integrin outside-in signaling,⁵⁰  as well as in platelet stimulation downstream of G-protein–coupled receptors.^33,51  In this regard, it is interesting to note that LMW-PTP was found to dephosphorylate substrates in thrombin-stimulated platelets with molecular masses comparable with those detected upon clustering of FcγRIIA. This observation may be explained in the light of the previous finding that ITAM-mediated signaling initiated by transactivation of FcγRIIA in thrombin-stimulated platelets significantly contributes to the agonist-induced protein tyrosine phosphorylation, and potentiates platelet response to low doses of the agonist.³³  It is therefore possible that LMW-PTP exerts a more widespread and general negative control on platelet function. This intriguing possibility, however, deserves further investigation.

Other tyrosine phosphatases may be involved in regulation of platelet activation by immunoreceptors as well, but their precise contribution has not been clearly demonstrated. In particular the SH2 domain–containing phosphatases SHP-1 and SHP-2 are known to be activated in stimulated platelets by binding to the tyrosine-phosphorylated ITIM-containing receptor PECAM-1.^16,52,53  Although in B cells SHP-1 negatively regulates BCR-dependent signaling,⁵⁴  there is no evidence that it can dephosphorylate ITAM-bearing receptors in platelets. Moreover, loss of catalytic activity of SHP-1 in mev/mev mice was found to cause an unexpected platelet hyporeactivity to collagen.¹⁸  PTP-1B has been proposed to regulate FcγRIIA-mediated platelet aggregation by promoting dephosphorylation of LAT.⁹  Interestingly, we have found in this work that LAT is also a good substrate for LMW-PTP. In this context, the reported ability of LMW-PTP remains unique and peculiar, as it may regulate platelet activation at 2 different levels in the signaling pathway initiated by phosphorylated ITAM.

Other tyrosine phosphatases expressed in platelets are typically regulated by means of tyrosine phosphorylation, calpain-mediated proteolysis, or cytoskeleton interaction. LMW-PTP possesses 2 adjacent tyrosine residues, Tyr¹³² and Tyr¹³³, and a strong increase of the enzymatic activity has been found to be associated with phosphorylation of Tyr¹³².^27,28  However, we failed to detect any tyrosine phosphorylation of LMW-PTP upon FcγRIIA clustering. We found, by contrast, that LMW-PTP underwent an evident subcellular redistribution in FcγRIIA-stimulated platelets, as it was found to translocate from the cytosol to the membrane fraction and to interact with the intracellular cytoskeleton. It seems therefore most likely that upon FcγRIIA engagement, LMW-PTP undergoes regulation through cellular redistribution rather than by covalent modification. In this context, interaction with the cytoskeleton may be critical, as it appears to precede membrane translocation. LMW-PTP interaction with the cytoskeleton has also been reported to occur in nucleated cells, and it has been shown to be important in determining substrate accessibility.^19,20  In stimulated platelets, the intracellular cytoskeleton interacts with several membrane glycoproteins and with elements of the membranoskeleton.⁴³  It should be noted that in our study the rapid separation of the membrane-rich fraction from the cytosol was obtained upon platelet treatment with low concentrations of the weak detergent digitonin, according to a previously published procedure.³⁵  Since this treatment is most likely to preserve membrane interaction with the cytoskeleton in stimulated platelets, our membrane-rich fractions also contain the cytoskeletal elements that have interacted with membrane components. It is therefore possible that upon FcγRIIA cross-linking, the reorganized actin-based network recruits LMW-PTP from the cytosol in order to subsequently convey it to its physiological substrate at the membrane level. The possible regulation of LMW-PTP by interaction with the cytoskeleton is consistent with the evidence, that, as a down-regulator of platelet activation, it is recruited in a delayed phase of platelet activation. In this respect, LMW-PTP differs from SHP-1 or SHP-2, which are activated during the first seconds of platelet stimulation, but behaves similarly to PTP-1B whose activation by calpain-mediated proteolysis is temporally associated with platelet aggregation.^7–9 

In conclusion, we have identified a new tyrosine phosphatase expressed in human platelets, able to dephosphorylate the ITAM-bearing receptor FcγRIIA, and we have provided new insights into the mechanism regulating immunoreceptor-triggered platelet activation. Because of the versatility of FcγRIIA as signal transducer in many processes of platelet activation, LMW-PTP may represent an excellent target for new antithrombotic agents.

This work was supported by grants from the University of Pavia, and from the Consorzio Interuniversitario Biotecnologie (CIB).

We thank Dr Ilaria Canobbio and Dr Paolo Lova (Department of Biochemistry, University of Pavia) for technical assistance and critical advice.

Contribution: F.M. designed and performed experiments, analyzed data, and wrote the paper; S.R. provided vital reagents, performed experiments, and edited the paper; A.B. contributed vital new reagents; C.B. analyzed data and edited the paper; M.T. designed research, analyzed data, wrote the paper, and provided overall direction.

Conflict-of-interest disclosure: The authors declare no competing financial interests.

Correspondence: Mauro Torti, Department of Biochemistry, University of Pavia, via Bassi 21, 27100 Pavia, Italy; e-mail: mtorti@unipv.it.