cAMP- and cGMP-elevating agents inhibit GPIbα-mediated aggregation but not GPIbα-stimulated Syk activation in human platelets
The glycoprotein (GP) Ib-IX-V complex is a unique platelet plasma membrane receptor, which is essential for platelet adhesion and thrombus formation. GPIbα, part of the GPIb-IX-V complex, has several physiological ligands such as von Willebrand factor (vWF), thrombospondin and distinct coagulation factors, which trigger platelet activation. Despite having an important role, intracellular GPIb-IX-V signaling and its regulation by other pathways are not well defined. Our aim was to establish the intracellular signaling response of selective GPIbα activation in human platelets, in particular the role of the tyrosine kinase Syk and its regulation by cAMP/PKA and cGMP/PKG pathways, respectively. We addressed this using echicetin beads (EB), which selectively bind to GPIbα and induce platelet aggregation.
Purified echicetin from snake Echis carinatus venom was validated by mass spectrometry. Washed human platelets were incubated with EB, in the presence or absence of echicetin monomers (EM), Src family kinase (SFK) inhibitors, Syk inhibitors and the cAMP- and cGMP-elevating agents iloprost and riociguat, respectively. Platelet aggregation was analyzed by light transmission aggregometry, protein phosphorylation by immunoblotting. Intracellular messengers inositolmonophosphate (InsP1) and Ca2+i were measured by ELISA and Fluo-3 AM/FACS, respectively.
EB-induced platelet aggregation was dependent on integrin αIIbβ3 and secondary mediators ADP and TxA2, and was antagonized by EM. EB stimulated Syk tyrosine phosphorylation at Y352, which was SFK-dependent and Syk-independent, whereas Y525/526 phosphorylation was SFK-dependent and partially Syk-dependent. Furthermore, phosphorylation of both Syk Y352 and Y525/526 was completely integrin αIIbβ3-independent but, in the case of Y525/526, was partially ADP/TxA2-dependent. Syk activation, observed as Y352/ Y525/Y526 phosphorylation, led to the phosphorylation of direct substrates (LAT Y191, PLCγ2 Y759) and additional targets (Akt S473). PKA/PKG pathways inhibited EB-induced platelet aggregation and Akt phosphorylation but, surprisingly, enhanced Syk and LAT/PLCγ2 tyrosine phosphorylation. A similar PKA/PKG effect was confirmed with convulxin−/GPVI-stimulated platelets. EB-induced InsP1 accumulation/InsP3 production and Ca2+-release were Syk-dependent, but only partially inhibited by PKA/PKG pathways.
EB and EM are specific agonists and antagonists, respectively, of GPIbα-mediated Syk activation leading to platelet aggregation. The cAMP/PKA and cGMP/PKG pathways do not inhibit but enhance GPIbα−/GPVI-initiated, SFK-dependent Syk activation, but strongly inhibit further downstream responses including aggregation. These data establish an important intracellular regulatory network induced by GPIbα.
KeywordscAMP-dependent protein kinase cGMP-dependent protein kinase Glycoprotein receptor GPIb-IX Platelet activation Syk kinase
Protein kinase B
cyclic Adenosine Monophosphate
cyclic Guanosine Monophosphate
G protein-coupled receptors
Inositol monophosphate/ triphosphate
Immunoreceptor tyrosine-based activation motif
Linker of Activated T cells
Protein Kinase A
Protein Kinase G
Src- Family Kinase
Spleen tyrosine kinase
von Willebrand Factor
Plain English summary
As the smallest circulating blood cells, platelets are activated at sites of vascular injury resulting in adhesion, aggregation, and blood coagulation, thereby preventing major blood loss. Pathologically, platelet hypofunction and hyperfunction can result in life-threatening bleeding or thromboinflammatory disorders, respectively. Platelet activation responses with subsequent thrombus formation are tightly controlled by multiple factors, which promote or inhibit platelet activation via membrane receptors and their intracellular effector systems. An essential platelet receptor is the glycoprotein (GP) Ib-IX-V complex, and its subunit GPIbα ligates predominantly von Willebrand factor but also other adhesion proteins and distinct coagulation factors. This enables platelet recruitment to the vessel wall, aggregation and coagulation. Despite this important role, intracellular effects of GPIb-IX-V in platelets and their interaction with other signaling pathways are not well defined. Recently, we characterized the snake venom protein echicetin immobilized on polystyrene beads as a specific GPIb activator. With this tool, we now show that GPIbα causes the activation of the tyrosine kinase Syk in a Src family kinase dependent manner, which results in the generation of further intracellular messengers and ultimately platelet aggregation. Surprisingly, activation of the platelet inhibitory cAMP/PKA and cGMP/PKG pathways enhanced initial Syk phosphorylation/activation, but strongly inhibited GPIbα-induced platelet aggregation, which is distal to Syk activation. These studies establish a new intracellular regulatory network triggered by the activation of GPIbα in human platelets.
Circulating platelets are essential for both physiological and pathological hemostasis and have important roles in inflammatory diseases and cancer [1, 2, 3]. Platelet activating, inhibitory and modulating factors fine tune platelet adhesion to the vessel wall. Physiologically, this fine-tuning prevents excessive bleeding due to vascular injuries, and also prevents or limits pathological thrombus formation / vessel occlusion at sites of injured blood vessels.
Two major groups of platelet activators such as soluble agonists [thrombin, ADP, thromboxane A2 (TxA2)] and adhesion molecules [e.g. von Willebrand factor (vWF), collagen, fibrin, podoplanin] bind to and stimulate specific G-protein-coupled receptors (GPCRs), or cell membrane-spanning adhesion receptors, respectively. These receptors stimulate intracellular signaling pathways and various platelet responses leading to integrin activation (e.g. integrin αIIbβ3), granule secretion, exposure of anionic phospholipids [4, 5, 6] and subsequently firm adhesion, aggregation, thrombin generation and thrombus formation. Conversely, elevation of platelet cAMP or cGMP by endothelial-derived prostacyclin (PGI2) or nitric oxide (NO) inhibits via cAMP-dependent (PKA) and/or cGMP-dependent (PKG) protein kinases, respectively, many of these platelet activation responses at several sites of the activation pathways [7, 8, 9]. Activation of Src-family kinases (SFK) and subsequent protein tyrosine phosphorylation, including membrane proteins containing the “immunoreceptor tyrosine-based activation motif (ITAM)”, initiates receptor-mediated platelet activation via the GPVI/Fcγ chain [10, 11, 12, 13], integrin αIIbβ3/FcγRIIA, CLEC-2, and GPIb-V-IX [5, 14, 15].
ITAM-mediated Syk activation, discovered in immune cells [16, 17], is now established in many mammalian cells including platelets [5, 11, 18]. For human platelets, the presence of two ITAM proteins have been described, Fc receptor γ-chain (FcRγ; gene: FCER1G) and a low affinity IgG receptor FcγRIIa (gene: FCGR2A) [19, 20]. In murine platelets only the FcRγ has been described [21, 22]. Studies with platelets from FcRγ-deficient mice established that this protein is essential for GPVI expression and function . Cytosolic Syk is activated by two distinct overlapping mechanisms designated as ITAM-dependent or Y-phosphorylation-dependent [18, 23, 24, 25]. The Syk Y-phospho-sites Y348/ Y352 and Y525/ Y526 belong to 2 pairs within the interdomain linker and kinase domain, respectively. Syk activation is initiated when these Y-sites are phosphorylated by SFKs or when dually Y-phosphorylated ITAM-containing membrane proteins recruit the two Syk-SH2 domains followed by Syk autophosphorylation [23, 24]. Usually, Syk activation is analyzed by Syk tyrosine phosphorylation (pY352 and pY525/526) and as tyrosine phosphorylation of its substrates (LAT pY191, PLCγ2 pY759). However, there are many additional proteins known to be directly phosphorylated by Syk [26, 27, 28]. Over the last years, hundreds of Syk targets have been identified in proteomic/phosphoproteomic studies and used to generate distinct Syk networks in cancer cells [28, 29], and many of these Syk targets are present in human platelets.
vWF and its major receptor, the glycoprotein Ib-V-IX (GPIb-IX-V) complex, are essential for platelet adhesion and initial thrombus formation at sites of vascular injury under arterial and venous shear conditions [20, 30, 31]. The GPIb-IX-V complex consists of 4 transmembrane proteins (GPIbα, GPIbß, GPIX, GPV; stoichiometry of 2:2:2:1). GPIbα is of special importance since it binds most of the extracellular GPIb-IX-V complex ligands (e.g. vWF, thrombin, FIX, FXII, TSP-1, Mac-1, P-selectin), but also multiple intracellular ligands required for platelet activation . The central role of the GPIb-V-IX complex in hemostasis is demonstrated by the strong bleeding disorder observed with Bernard-Soulier syndrome (BSS) patients who lack the platelet GPIb-V-IX complex, validated in studies with mice deficient in GPIbα [20, 31]. However, there are still many open questions concerning intracellular GPIb signaling and its interaction with other pathways.
When immobilized on the subendothelium of the damaged vessel wall, or presented on activated endothelial cells, the multimeric plasma protein vWF binds via its A1 domain to GPIbα and induces receptor clustering leading to platelet signaling and moderate activation. Without vessel wall components, snake venom toxins such as ristocetin or botrocetin are required for these vWF effects. Multiple intracellular signaling pathways have been proposed to mediate the intracellular effects of GPIbα activation by vWF, the phosphatidylinositol 3-kinase (PI3K)/ protein kinase B (Akt) pathway, cGMP/PKG, the mitogen-activated protein kinase (MAPK) pathways and the FcγR-Syk/PLCγ2 pathway [30, 32, 33, 34, 35, 36]. However, the exact mechanisms have not been fully elucidated and are controversially discussed [20, 33, 37]. GPIb-V-IX is not the only platelet vWF receptor since the integrin αIIbβ3 also ligates vWF via recognition of the RGD-sequence within the C4 domain of vWF, thereby inducing also platelet activation, often together with GPIbα. As vWF is not a selective ligand of GPIbα, studies of GPIbα-selective signaling are rare and have so far been confined to platelets, which adhere to immobilized recombinant dimeric vWF A1 domain or active vWF A1-domains expressed on COS-7 cells [19, 30, 34] that also has limitations. The possible reasons for the often discrepant data and results published for GPIb signaling in human platelets are well reviewed and include the use of various GPIb-V-IX complex ligands, cell types, and biochemical and functional read-out systems [20, 38]. Recently, we developed a GPIbα-specific agonist, the C-type lectin snake venom protein, echicetin, coated on polystyrene beads . Earlier studies showed that echicetin molecules cross-linked by plasma IgMκ caused platelet agglutination and weak aggregation whereas echicetin monomers, when used as specific GPIbα ligands, competed with vWF and thrombin for binding to GPIbα and blocked ristocetin/vWF mediated platelet agglutination . Using a transgenic mouse model, we recently reported that the extracellular GPIbα-domain of murine platelets is essential for echicetin-bead-induced platelet aggregation . In this model the extracellular domain of GPIbα is replaced by the human interleukin 4-receptor (IL4/GPIbα transgenic mice). EB induced aggregation of washed platelets from wildtype mice, but not of washed platelets expressing the IL4/GPIbα mutant protein. These data showed that EB activation of not only human but also mouse platelets requires GPIbα, the predominant receptor for most ligands of the GPIb-IX-V complex.
With these novel tools, echicetin beads (EB) and echicetin monomers (EM), we aimed to clarify whether selective activation of human platelet GPIbα has the capacity to induce activation of the tyrosine kinase Syk and whether these pathways are affected by two major platelet inhibitory pathways, cAMP/PKA and cGMP/PKG. Our results establish a novel, surprising interaction of GPIbα- and PKA/PKG-affected pathways in human platelets.
Materials and methods
Lyophilized snake venom of E. carinatus sochureki was from Latoxan, France. Lyophilized convulxin (isolated from Crotalus durissus terrificus) was from Enzo life sciences, Lausen, Switzerland. Human vWF isolated from plasma (Wilate®) was from Octapharma GmbH, Langenfeld, Germany. Ristocetin was from Loxo GmbH, Dossenheim, Germany. Affinity chromatography column, protein A sepharose-4B column coated with rabbit polyclonal antibodies against echicetin was from Dr. Alexei Navdaev, Würzburg, Germany. Diethylaminoethyl (DEAE) resin matrix, Toyopearls 650 S was from Tosoh Bioscience GmbH, Germany. The resin was packed in Tricorn 5/50 column from GE healthcare life sciences, Germany. Syk inhibitors, OXSI-2, 2,3-Dihydro-3-[(1-methyl-1H-indol-3-yl)methylene]-2-oxo-1H-indole-5-sulfonamide were purchased from Merck, Germany and PRT060318 (PRT-318), (2-(1R,2S)-2-aminocyclohexylamino)-4-(m-tolylamino)pyrimidine-5-carboxamide), was from Selleckchem, Germany. PP2 and its inactive analogue PP3 were from Abcam, England. Tirofiban (Aggrastat®) was from Iroko Cardio LLC, USA. MRS2179 was from Viozol, Eching, Germany, AR-C69931 from The Medicines Company, Parsippany, NJ, USA and SQ-29548, was from Cayman chemical, MI, USA. Wortmannin was purchased from Biozol, Eching, Germany. Iloprost (Ilomedine®) and riociguat were from Bayer, Germany. Bovine serum albumin (BSA) was from Capricorn Scientific GmbH, Germany. Clarity™ Western ECL Substrate was from BioRad Laboratories, USA. Rabbit monoclonal antibodies against phosphorylated Syk Y525/526 and polyclonal antibodies against phosphorylated Syk Y352, LAT Y191, PLCγ2 Y759, Akt S473, VASP S239, α-actinin were from Cell Signaling Technologies, USA. Mouse monoclonal antibodies against Syk, PLCɣ2 and Akt were from Santa Cruz Biotechnology, USA. Antibodies against β-actin were from Abcam, Engalnd. Anti- phosphotyrosine mouse antibodies were from Merck, Schwalbach, Germany. Secondary antibodies HRP-conjugated goat anti-rabbit and anti-mouse IgG were from BioRad Laboratories Hercules, USA. IP-One ELISA kit (96 wells) was from Cisbio, Codolet, France. Fluo-3, AM, intracellular cytoplasmic Ca2+ indicator was from Life Technologies, USA.
Echicetin purification and echicetin-beads preparation
Echicetin was purified from E. carinatus sochureki lyophilized venom by affinity chromatography followed by DEAE anion exchange chromatography and validated by mass spectrometry analysis. For affinity chromatography protein A sepharose-4B column coated with rabbit polyclonal antibodies directed against echicetin (generated by A. Navdaev) was used. Echicetin was eluted using 0.2 M acetate buffer pH 2.7. The eluent buffer was exchanged into 10 mM Tris buffer pH 8.0 (buffer A) and then applied to DEAE anion exchange column. Elution of echicetin was performed by a 0 to 1 M gradient of NaCl in buffer A, under a flow rate of 1 ml/min. Fraction eluted at 120 mM NaCl consisted equally of α and β subunit and was used in all the experiments. Silver staining and mass spectrometry analysis were performed in order to confirm the purity of echicetin. Echicetin beads (EB) were prepared as reported  and coated for all experiments used with 0.3 mg/ml echicetin.
Samples from peak 1 and peak 2 were prepared under reducing conditions (by adding Laemmli buffer) then boiled at 95 °C for 10 min. Proteins of both peaks were separated by electrophoresis using 15% SDS-PAGE gels. Gels were stained using InstantBlue™. Bands were cut and digested using trypsin. Protein sequences were analyzed by mass spectrometry in the mass spectrometry core facility at the University Medical Center of the Johannes Gutenberg University, Mainz.
Preparation of washed human platelets
Venous blood was collected as citrated whole blood after informed consent from healthy volunteers who did not take any medication for at least 10 days before blood collection. Studies using human platelets from healthy volunteers and from a patient with Glanzmann thrombosthenia caused by a homozygous point mutation in ITGA2B c.621C > T; p.T176I [42, 43] were approved by the local ethics committees (Study No. 837.302.12; 25.07.12; FF109/2015). EGTA (2 mM final concentration) was added to the whole blood before centrifuging at 200 x g for 10 min at room temperature (RT) to get platelet-rich plasma (PRP). PRP was diluted 1:1 with CGS buffer (120 mM NaCl, 12.9 mM Tri-Na-citrate, 30 mM glucose, pH 6.5) then centrifuged at 400 x g for 10 min at RT. The platelet pellet was resuspended with Hepes buffer (150 mM NaCl, 5 mM KCl, 1 mM MgCl2, 10 mM Glucose, 10 mM Hepes) pH 7.4. Agglutination followed by fibrinogen-dependent aggregation of washed human platelets was induced by adding human vWF (2.5 μg/ml final concentration) plus ristocetin (0.5 mg/ml final concentration) as previously described . Residual platelet aggregation was calculated by subtraction of overall maximum light transmission in the absence of tirofiban minus maximum light transmission in the presence of tirofiban. Washed platelets were placed for 20 min at 37 °C for resting. Human washed platelets (3 × 108 platelets/ml) were pre-incubated at 37 °C with different inhibitors: Syk inhibitors, OXSI-2 or PRT-318, or Src family kinase inhibitor PP2 or its inactive analogue PP3 or wortmannin. In addition, ADP and TxA2 receptors blockers were used (AR-C69931, MRS2179, SQ-29548), cAMP-elevating agents iloprost and cGMP-elevating soluble guanylate cyclase (sGC) stimulator riociguat, respectively. Platelet aggregation was triggered with 0.5% (v/v) echicetin beads for 1, 2 and 5 min under continuous stirring (1000 rpm) at 37 °C in an Apact4S Plus aggregometer (DiaSys Greiner, Flacht, Germany). Control samples mentioned as basal (in graphs) or time zero (in blots and graphs) refer to washed platelets placed under continuous stirring for 1 min in absence of any agonist.
Western blot analysis
Western blot samples were prepared by immediately adding Laemmli buffer to washed platelets inside the cuvettes and boiled at 95 °C for 10 min. Platelet proteins were separated by electrophoresis using 8% SDS-PAGE gels then transferred to polyvinylidene difluoride (PVDF) membranes before blocking the membranes for one hour with 5% BSA in TBS-T (20 mM Tris, 140 mM NaCl, 0.1% Tween, pH 7.6) at room temperature. Membranes were incubated overnight under gentle shaking with 1:1000 diluted primary antibodies. Membranes were washed three times with TBS-T and incubated for two hours at RT with the appropriate secondary antibodies diluted 1:5000 in 5% BSA. Membranes were washed again for 3 times with TBS-T before developing the blots by ECL detection. Control blots with total proteins (Syk, PLCγ2 and Akt) derive from same samples from the corresponding phosphorylated form. α-actinin was also used as loading control for distinct p-Akt blots. Blots with p-LAT, p-VASP were stripped and reprobed with anti-β-actin antibody. Blots were cropped following the corresponding molecular weight of the protein of interest.
Inositol monophosphate (InsP1) measurement
The production of inositol triphosphate (InsP3) was measured by the level of the InsP1 accumulated in washed platelets using the IP-One ELISA kit according to the manufacturer’s instructions. Human washed platelets adjusted to 3 × 108/ml in presence of LiCl (1 mM), inhibiting the degradation of InsP1 into myo-inositol. Washed platelets were activated by EB in the absence or presence of effectors, under stirring conditions. Platelets were lysed after 5 min of stimulation then centrifuged at 16,000 x g for 10 min at 4 °C. Platelets lysates were incubated with InsP1- HRP conjugate and anti- InsP1 monoclonal antibody for 3 h.
Intracellular Ca2+-release measurement
Human platelets were prepared as already mentioned above and adjusted to 3 × 108/ml in Hepes buffer. Washed platelets were then pre-incubated with a Ca2+ indicator dye, fluo-3 acetoxymethyl (AM) esters (5 μM) for 30 min at 37 °C. Intracellular Ca2+-release was monitored for 2 min by flow cytometry after stimulation with EB without additional supplementation of extracellular Ca2+.
Experiments were performed at least three times with at least three different healthy donors. The data were presented as mean ± standard deviation (S.D.). Statistical analysis was performed using GraphPad Prism 7 for Windows (GraphPad Software, San Diego, CA). The 2-tailed Student’s t-test was used for comparison of two groups, one-way and two-way ANOVA and Tukey’s multiple comparison tests were used, when appropriate, for comparison of more than two groups. P < 0.05 was considered as significant.
Echicetin coated polystyrene beads activate Syk tyrosine kinase and stimulate αIIbβ3 integrin-dependent platelet aggregation
Initial experiments with conventionally purified echicetin from the snake venom Echis carinatus sochureki produced variable effects on platelets and prompted us to modify the purification procedure (Additional file 1: Figure S1). The final DEAE column produced 2 major protein peaks, which were characterized by mass spectrometry. Only peak 1 contained pure echicetin heterodimer (with α + β subunits), whereas peak 2 contained primarily the ß subunit and some unidentified bands (Additional file 1: Figure S1A, B). Echicetin eluted in peak 1 was used for all experiments reported here with consistent results, both with echicetin-coated beads (EB) or as echicetin monomer (α/ß subunits) (EM).
In agreement with our earlier work , the αIIbβ3 integrin antagonist tirofiban completely inhibited EB-induced platelet aggregation of washed platelets. EM, which alone selectively bind to GPIbα without inducing platelet activation, completely inhibited EB-induced platelet aggregation in a competitive manner. As negative-control, BSA coated beads did not show any effect on platelet aggregation (Additional file 1: Figure S1C, D). We also tested whether EM affect platelet aggregation induced by other main platelet receptors. Echicetin did not affect the aggregation induced by collagen (Additional file 2: Figure S2A, B), ADP (Additional file 2: Figure S2C, D), TxA2 (Additional file 2: Figure S2E, F), TRAP6 (Additional file 2: Figure S2G, H), thrombin at high concentration (Additional file 2: Figure S2I, J). A small inhibitory effect was observed with aggregation induced by a low concentration of thrombin (Additional file 2: Fig. S2K, L), which possibly reflects the thrombin binding site of GPIbα [20, 44].
EB-induced Syk tyrosine phosphorylation requires src family kinases (SFKs)
Syk inhibitors demonstrate an essential role of Syk in EB-induced platelet aggregation and a differential regulation of Syk Y525/526 and Y352 phosphorylation
EB-induced platelet aggregation requires the secondary mediators ADP and TxA2
EB-mediated Syk activation does not require integrin αIIbβ3 signaling
cAMP/ cGMP elevation inhibits EB-induced platelet aggregation but not Syk activation, an effect also observed with convulxin-treated platelets
Then we also used lower concentrations of iloprost (1 nM) and riociguat (10 μM), which partially inhibit EB-induced platelet aggregation, to test if the observed protein phosphorylation effects are still present compared to 2 nM iloprost and 20 μM riociguat, respectively. For 1 nM iloprost (Additional file 6: Figure S6A-E) as well as for 10 μM riociguat (Additional file 7: Figure S7A-E) we observed also increased and prolonged phosphorylation of Syk Y525/526, Syk Y325 and PLCγ2 Y759 as well as diminished Akt S473 phosphorylation (Additional file 6: Figure S6D, F and Additional file 7: Figure S7D, F), indicating similar effects of iloprost and riociguat on EB-induced platelet signaling.
The surprising differential effects of the inhibitory pathways (PKA/PKG) on EB-simulated platelet aggregation and Syk activation prompted us to study also a classical mechanism of Syk activation in human platelets, the convulxin/GPVI pathway [10, 11, 12]. For the analysis of PKA/PKG effects on GPVI-mediated signaling pathways, we used 50 ng/ml convulxin to stimulate washed human platelets. Convulxin-stimulated platelet aggregation was completely inhibited by iloprost and riociguat (Additional file 8: Figure S8A, B). Then, convulxin-induced Syk phosphorylation in the presence of a vehicle control, iloprost or riociguat was analyzed. The Syk activation marker Y525/526 and the regulatory site Y352 were well but only transiently phosphorylated in response to convulxin (Additional file 8: Figure S8C). Strikingly, convulxin-stimulated Syk Y525/526 phosphorylation was significantly increased/prolonged (hyperphosphorylation) by iloprost and riociguat compared to the vehicle control (Additional file 8: Figure S8C, D), which was supported by LC-MS/MS (data not shown). With Syk Y352 phosphorylation, there was little effect after 1 min of cvx stimulation, but at 2 and 5 min there was also significantly increased and prolonged phosphorylation in the presence of both inhibitory agents (Additional file 8: Figure S8C, E).
EB-stimulated intracellular InsP1 accumulation and Ca2+-release are only partially inhibited by iloprost and riociguat
In this study we demonstrated that cAMP/PKA and cGMP/PKG pathways cause dichotomous regulation of GPIbα-mediated Syk stimulation and activation of human platelets. To ensure selective GPIbα-activation the specific GPIbα-ligand echicetin was used as multimer complex, coated on polystyrene beads (EB), leading to integrin αIIbβ3-dependent aggregation of human platelets [39, 40, 41]. Echicetin monomer (EM) antagonized all observed EB effects but not GPVI/FcRγ or GPCR-mediated platelet aggregation.
EB induced platelet Syk phosphorylation and aggregation, which requires SFKs as these effects are prevented by the SFK inhibitors PP2 and dasatinib. Syk activation by EB in human platelets as assessed by Y352 and Y525/526 phosphorylation, was rapid but transient, indicating substantial dephosphorylation. Interestingly, dephosphorylation of murine Syk at Y346 (~ Y352 in the human kinase) by the TULA-2 protein-tyrosine phosphatase suppressed its activation in murine platelets .
As reported by us recently, EB strongly activated murine platelets, expressing only FcRγ-chain but not FcRγIIA, which required GPIbα . Our previous quantitative proteomic studies with human platelets demonstrated expression of FCER1G (8170 copies), FCGR2A (990 copies) and Syk (4900 copies) . Within a major, still ongoing phosphoproteomic study (to be published) we recently detected multiple tyrosine-phosphorylated proteins in EB-stimulated human platelets including dually Y-phosphorylated FCER1G (Y56 and Y76, both 1.8 fold increased phosphorylation). Tyrosine phosphorylation of FCGR2A was not detected.
Altogether, the data indicate that EB activation of the human platelet GPIb-complex stimulates Syk phosphorylation via an SFK-dependent mechanism, which is antagonized by EM. SFK-dependent Syk activation involves direct Syk Y352 phosphorylation and may also involve the platelet ITAM protein FcRγ-chain.
To validate a functional role of Syk for GPIbα-signaling in human platelets, we tested two different Syk inhibitors, OXSI-2 and PRT-318. These have been described as Syk and platelet function inhibitors in studies of convulxin/GPVI signaling [48, 49, 50]. PRT-318 also prevented heparin-induced thrombocytopenia (HIT) and thrombosis in a mouse model . Another novel orally bioavailable Syk inhibitor (BI1002494) prevented arterial thrombosis and thromboinflammatory brain infarction in a mouse model to a similar extent as did, in another mouse model, platelet-specific Syk deficiency . In humans, the Syk inhibitor fostamatinib was recently FDA-approved for the treatment of thrombocytopenia in adult patients with chronic immune thrombocytopenia (ITP) .
OXSI-2 (2 μM) and PRT-318 (1 μM) abolished EB-induced aggregation and strongly inhibited EB–induced Syk Y525/526, but not Syk Y352, phosphorylation. These inhibitors also prevented EB-induced phosphorylation of the direct Syk substrate (PLCγ2 at Y759) and a further downstream effector, Akt S473. Syk Y525/526 phosphorylation, primarily due to autophosphorylation [23, 24], is expected to be blocked by Syk inhibitors. In contrast, Syk Y348/Y352 phosphorylation catalyzed by SFKs in intact cells initiates Syk activation , and is not blocked by Syk inhibitors but by the SFK inhibitor PP2, as observed here in our studies. Absent inhibition of platelet agonist–induced Syk Y348/Y352 phosphorylation, catalyzed by SFKs, has been used as one specificity criteria for Syk inhibitors [48, 49]. For the overall regulation/activation of Syk, the sites Y348/Y352 are considered more important than the activation loop sites Y525/Y526 since mutations of latter sites did not reduce Syk kinase activity [23, 24]. This indicates that Syk tyrosine phosphorylation alone cannot be equated with Syk kinase activity. The strong inhibitory effects of Syk inhibitors on EB-induced Syk Y525/526, PLCγ2 Y759 phosphorylation, and EB-induced aggregation, indicate that EB activate human platelets by a Syk-dependent process.
Based on the inhibitory effects of the Syk inhibitors, EB-induced Akt phosphorylation requires Syk and therefore appears to be downstream of Syk activity. However, Akt is certainly not an immediate Syk target/direct substrate but is most likely phosphorylated in response to additional adapter proteins affecting PI3K , which is confirmed by our results that the PI3K inhibitor wortmannin abolished EB-mediated Akt S473 phosphorylation without inhibition of Syk tyrosine phosphorylation.
A recent paper reported that Syk activity is dispensable for platelet GPIb-IX-V signaling induced by ristocetin/vWF  by showing unaffected phosphorylation of Syk at Y352 and of Akt at S473 by the Syk inhibitor PRT-318. Using the same Syk inhibitor we confirmed that ristocetin/vWF-induced Syk Y352 phosphorylation is not affected. However, we could show that PRT-318 inhibited phosphorylation of Syk at Y525/526, of the Syk substrate LAT at Y191 and of Akt at S473 in washed human platelets, indicating that Syk plays an important role in EB- as well in ristocetin/vWF-mediated GPIbα-signaling of human platelets.
The receptors GPVI and CLEC-2 activate platelets by a Syk-dependent mechanism and require the release of secondary mediators (ADP,TxA2) for a full response, with certain differences . Therefore, the role of the secondary mediators for the EB responses studied here was investigated. EB–induced platelet aggregation was abolished when the ADP receptors (P2Y12 and P2Y1) and the TxA2 receptor were blocked. In contrast, Syk phosphorylation was not at all (Y352) or only partially (Y525/Y526) reduced, whereas PLCγ2 Y759 and Akt S473 phosphorylation was strongly reduced. We have no clear explanation for the partial inhibition of Syk 526/526 phosphorylation by the secondary mediators except to speculate that this site may also be affected by pathways other than SFKs. This has also been proposed for PLCγ2 and its phosphorylation . In our platelet phosphoproteomic studies, ADP only stimulated Syk serine phosphorylation with no detectable effect on Syk tyrosine phosphorylation , which was confirmed in our ongoing experiments (Makhoul S et al., unpublished data). Overall, our data show that ADP and TxA2 have no major effect on EB-induced Syk phosphorylation but, in contrast, are required for further downstream effects including PLCγ2 Y759/Akt S473 phosphorylation and aggregation.
In addition to ADP and TxA2, other pathways such as integrin αIIbβ3 activation influence various steps of platelet activation including ITAM (FcγRIIA)-dependent Syk stimulation [18, 60]. To evaluate the possible contribution of αIIbβ3 for EB-induced signaling, we compared platelets from normal controls and αIIbβ3-deficient platelets from a patient with Glanzmann thrombasthenia. There was no difference between normal and αIIbβ3-deficient platelets with respect to EB-induced Syk activation as indicated by Y352 and Y525/Y526 phosphorylation. Similarly, the αIIbβ3 inhibitor tirofiban did not affect EB-induced Syk tyrosine phosphorylation and phosphorylation of the Syk substrate PLCγ2 at Y759. These data show that EB/GPIb-mediated phosphorylation and activation of Syk is integrin αIIbβ3-independent.
Platelet functions are tightly regulated by a network of intracellular pathways consisting of protein kinases/protein phosphatases and their substrates [6, 8, 61], but the regulation of specific GPIb-mediated signaling in human platelets by the PKA/PKG inhibitory pathways is unclear. PKA-stimulated GPIbß phosphorylation (S166, now S191) has been observed in multiple studies with human platelets [52, 62, 63]. When tested in the Chinese hamster ovary cell (CHO), PKA phosphorylation of GPIbß S166 correlated with reduced binding of vWF to the GPIb-IX complex  whereas other studies did not observe a functional effect of this phosphosite . Clearly, PKA phosphorylation of GPIbß S166 (S191) and of other sites within the GPIb-IX complex need to be re-investigated in future studies. In our present studies, we investigated primarily the regulation of GPIbα-mediated signaling and showed that both iloprost (cAMP pathway) and riociguat (cGMP pathway) dose-dependently inhibited EB-induced aggregation of washed human platelets, with maximal inhibition at 5 nM iloprost and 20 μM riociguat, respectively. These clinically-used drugs and our conditions have been extensively used, and characterized to achieve strong, but selective activation of PKA (iloprost) and PKG (riociguat) in human platelets, also monitored by established substrates [8, 52, 53, 54]. Using different concentrations of iloprost (1 nM, 2 nM) and riociguat (10 μM, 20 μM), which inhibited significantly EB-induced aggregation, they did not inhibit, but instead enhanced/prolonged EB-induced Syk activation observed as Syk Y352 and Y525/526 phosphorylation (hyperphosphorylation). Enhanced phosphorylation of the direct Syk substrates PLCγ2 (Y759) and LAT (Y191) was also detected. In contrast, EB-induced Akt S473 phosphorylation was abolished. These results show that the PKA/PKG pathways do not prevent EB-induced Syk activation (Y352/ Y525/526) and Syk activity (PLCγ2/LAT), whereas further downstream effects (Akt phosphorylation, platelet aggregation) are strongly inhibited. The observation that cAMP and cGMP abolish GPIbα-mediated aggregation but not Syk activation under these conditions suggests that the inhibitory PKA/PKG effects occur downstream of Syk activation, not at the level of Syk activation/Syk activity.
Syk-mediated PLCγ2 activation and InsP3 and 1,2-diacylglycerol (DAG) generation are essential components of the GPVI-pathway in human and murine platelet, and most GPVI/ITAM-effects are thought be mediated by Syk [10, 13, 18], perhaps with some special exceptions such as generation of reactive oxygen species . In contrast, specific GPIbα-stimulation of the Syk/PLCγ2 system with InsP3/DAG generation/intracellular Ca2+-release and its functional roles remained unclear until now [20, 30, 35, 65]. Our data conclusively show that EB, via GPIbα−/Syk activation, induced a marked increase of InsP1 as marker of InsP3 generation and Ca2+-release in human platelets, which was completely GPIbα-dependent (response abolished by EM), Syk-dependent (abolished by PRT-318) and integrin αIIbβ3-independent (unaffected by tirofiban). However, iloprost (cAMP) and riociguat (cGMP) pre-treatment only partially inhibited EB-stimulated InsP1 increase and Ca2+-release. These only partial inhibitory effects differ from the known strong PKA/PKG-mediated inhibition of the InsP3/Ca2+ response by platelet GPCR (ADP, thrombin, TxA2) activation, which occurs at several sites including GPCR signaling, PLCβ, and the InsP3 receptor or its associated protein IRAG [8, 9, 61, 66, 67]. ADP−/ thrombin- / TxA2-stimulated Ca2+-release from intracellular stores in human platelets was essentially abolished by the PKA and PKG pathway [66, 67]. Our present data therefore suggest that cAMP/cGMP inhibit EB/GPIbα-platelet activation partially at the level of the InsP3/Ca2+ response but perhaps stronger at sites downstream of InsP3 production and Ca2+ elevation. Interestingly, earlier studies using PLCγ2-deficient murine platelets and vWF/botrocetin suggested that PLCγ2 is not required  or only moderately involved in GPIb-V-IX signaling  indicating that additional and/or compensating mechanisms exist, perhaps activation of other phospholipases. In our present studies, EB-induced aggregation, PLCγ2 Y759 phosphorylation, Akt S473 phosphorylation as well as InsP1 accumulation/InsP3 generation and Ca2+-release were found to be dependent on the secondary mediators ADP and TxA2, different from the initial Syk phosphorylation/activation. It is very likely that important functional responses, such as Ca2+-release from intracellular stores, are heterogeneous and controlled by several pathways and proteins. A possible hotspot for the integration of various in-going Syk signals and out-going responses is the membrane protein LAT, an important Syk substrate and component of ITAM/Syk signaling in immune cells and also in platelets [10, 68].
Interestingly, we observed similar effects of cAMP/PKA and cGMP/PKG pathway stimulation on GPVI-mediated Syk phosphorylation. Whereas convulxin-stimulated platelet aggregation was completely inhibited by iloprost (cAMP) and riociguat (cGMP), convulxin-induced Syk phosphorylation (Y352 and Y525/526) was not inhibited, but instead clearly enhanced/prolonged (hyperphosphorylation) by iloprost or riociguat pretreatment. Convulxin-stimulated PLCγ2 Y759 phosphorylation was also enhanced by the cAMP/cGMP pathways, whereas Akt S473 phosphorylation was inhibited. In line with this observation, stimulation of CLEC-2 (a hemi-ITAM containing special adhesion receptor) on activated mouse platelets was shown to be only weakly inhibited by the cAMP, and not by the cGMP pathway . Our cvx/GPVI signaling data agree with the GPVI-signaling data of a recent published study, which reported that the cAMP/PKA pathway inhibited distal but not proximal collagen/GPVI-mediated signaling events in human platelets .
Presently, the mechanism(s) for the PKA/PKG-mediated increase of Syk Y352 and Y525/526 phosphorylation (hyperphosphorylation) and kinase activity is a topic of further investigation in our laboratory. Interestingly, a possibly related “Syk Y-hyperphosphorylation” has been observed in response to PKC inhibitors . Recently, a large spectrum of additional Syk serine and tyrosine phosphorylation sites, interacting proteins and multisite ubiquitylation has been described [46, 72]. Clearly, additional regulatory mechanisms exist, which may control phosphorylation/dephosphorylation and membrane interactions of Syk, and which need to be addressed in future studies.
Our data establish that selective activation of GPIbα in human platelets by using echicetin-beads results in SFK-dependent Syk activation, subsequent Syk-mediated signaling, and ultimately integrin-dependent aggregation. In contrast to others  we obtained evidence that Syk plays an important role in EB- as well in ristocetin/vWF-mediated GPIbα-signaling of human platelets.
Syk pathway components display a variable dependency on secondary mediators (ADP, TxA2) and variable regulation by cAMP/cGMP inhibitory pathways, revealing two distinct directions of GPIbα−/Syk signaling. The initial GPIbα-caused Syk phosphorylation, activation and direct substrate phosphorylation is independent of the integrin αIIbβ3 and secondary mediators. Then, two distinct platelet inhibitory mediators, cAMP/PKA and cGMP/PKG, completely block one part of GPIbα−/GPVI-stimulated Syk signaling responses (aggregation, PI3K pathway), but increase the initial receptor-induced Syk activation and phosphorylation of two important direct Syk substrates, PLCγ2 and LAT. The observation that cAMP and cGMP abolish GPIbα-mediated aggregation but not Syk activation under our conditions suggests that the inhibitory PKA/PKG effects occur downstream of Syk activation, not at the level of Syk activation/Syk activity. Our results indicate the presence of distinct Syk signaling/effector systems, which differ in their interaction with other signaling pathways. The physiological and pathophysiological significance of these different Syk effector systems needs further investigation. This is clearly warranted in the light of the functional importance of Syk in platelets and immune cells, and in the light of Syk and related tyrosine protein kinases as important pharmacological target.
SG was supported by a grant from Ministry of Science and Higher Education of the Russian Federation (№АААА-А18-118012290371-3). The authors thank Dr. A. Navdaev (CSL Behring AG, Zurich) for his help and advice in preparing echicetin proteins and Dr. Suzanne M. Lohmann for English editing.
SM performed most of the experiments, analyzed and interpreted data and wrote the manuscript (part of her doctoral thesis). KT performed, analyzed and interpreted initial experiments (part of her doctoral thesis). SG designed experiments, interpreted data and revised the manuscript. ST was responsible for mass spectrometry experiments and data. DP recruited patients. UW designed the study, interpreted data and wrote the manuscript. KJ designed the study and experiments, interpreted data, wrote and revised the manuscript. All authors read and approved the final manuscript.
The study was supported by the German Federal Ministry of Education and Research (BMBF 01EO1003 and 01EO1503 (KJ, UW)).
Ethics approval and consent to participate
All studies using human platelets from healthy volunteers and from a patient with Glanzmann thrombasthenia caused by a homozygous point mutation in ITGA2B c.621C > T; p.T176I [42, 43] were approved by the local ethics committees (Study No. 837.302.12; 25.07.12; FF109/2015).
Consent for publication
The authors declare that they have no competing interests.
- 6.Brass LF, Ma P, Tomaiuolo M, Diamond SL, Stalker TJ. A Systems Approach to the Platelet Signaling Network and the Hemostatic Response to Injury. Platelets in Thrombotic and Non-Thrombotic Disorders: Springer. 2017:367–78.Google Scholar
- 8.Makhoul S, Walter E, Pagel O, Walter U, Sickmann A, Gambaryan S, et al. Effects of the NO/soluble guanylate cyclase/cGMP system on the functions of human platelets. Nitric oxide : biology and chemistry. 2018;76:71–80.Google Scholar
- 14.Gardiner EE, Andrews RK. Platelet adhesion. In: Gresele P, Kleiman NS, Lopez JA, Page CP, editors. Platelets in thrombotic and non-thrombotic disorders: pathophysiology, pharmacology and therapeutics: an update. Cham: Springer International Publishing; 2017. p. 309–19.Google Scholar
- 20.López JA. The platelet glycoprotein Ib-IX-V complex. In: Gresele P, Kleiman NS, Lopez JA, Page CP, editors. Platelets in thrombotic and non-thrombotic disorders: pathophysiology, pharmacology and therapeutics: an update. Cham: Springer International Publishing; 2017. p. 85–97.Google Scholar
- 21.Stegner D, Haining EJ, Nieswandt B. Targeting glycoprotein VI and the Immunoreceptor tyrosine-based activation motif signaling pathway. Arteriosclerosis Thrombosis and Vascular Biology. 2014;34(8):1615–20.Google Scholar
- 42.Geffen JPV, Brouns SLN, Batista J, McKinney H, Kempster C. Nagy M, et al. Haematologica: High-throughput elucidation of thrombus formation reveals sources of platelet function variability; 2018.Google Scholar
- 56.Reppschlager K, Gosselin J, Dangelmaier CA, Thomas DH, Carpino N, McKenzie SE, et al. TULA-2 protein phosphatase suppresses activation of Syk through the GPVI platelet receptor for collagen by dephosphorylating Tyr(P)(346), a regulatory site of Syk. J Biol Chem. 2016;291(43):22427–41.PubMedPubMedCentralGoogle Scholar
- 58.van Eeuwijk JMM, Stegner D, Lamb DJ, Kraft P, Beck S, Thielmann I, et al. The Novel Oral Syk Inhibitor, Bl1002494, Protects Mice From Arterial Thrombosis and Thromboinflammatory Brain Infarction. Arteriosclerosis Thrombosis and Vascular Biology. 2016;36(6):1247−+.Google Scholar
- 65.Mangin P, Yuan Y, Goncalves I, Eckly A, Freund M, Cazenave JP, et al. Signaling role for phospholipase C gamma 2 in platelet glycoprotein Ib alpha calcium flux and cytoskeletal reorganization. Involvement of a pathway distinct from FcR gamma chain and fc gamma RIIA. J Biol Chem. 2003;278(35):32880–91.PubMedGoogle Scholar
- 67.Geiger J, Nolte C, Walter U. Regulation Of Calcium Mobilization And Entry In Human Platelets By Endothelium-Derived Factors. Am J Physiol. 1994;267(1):C236-CC44.Google Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.