Intracellular events in platelet activation induced by antiphospholipid antibodies in the presence of low doses of thrombin.код для вставкиСкачать
ARTHRITIS & RHEUMATISM Vol. 50, No. 9, September 2004, pp 2911–2919 DOI 10.1002/art.20434 © 2004, American College of Rheumatology Intracellular Events in Platelet Activation Induced by Antiphospholipid Antibodies in the Presence of Low Doses of Thrombin Mariano Vega-Ostertag, E. Nigel Harris, and Silvia S. Pierangeli Objective. Thrombosis and thrombocytopenia are features of the antiphospholipid syndrome (APS), suggesting that antiphospholipid antibodies (aPL) may bind platelets, causing activation and aggregation of platelets and thrombosis. The intracellular events involved in aPL-mediated platelet activation are not fully understood and are therefore the subject of this study. Methods. IgG fractions and their F(abⴕ)2 fragments were purified from the sera of 7 patients with APS and from the pooled sera of 10 healthy subjects (as controls). Phosphorylation of p38 MAPK, ERK-1/2, and [Ca2ⴙ]-dependent cytosolic phospholipase A2 (cPLA2) was determined in lysates of washed platelets pretreated with low doses of thrombin and aPL or control IgG or their F(abⴕ)2 fragments, by immunoblot. The effects of aPL on platelet aggregation in the presence or absence of a p38 MAPK inhibitor, SB203580, were examined. Thromboxane B2 (TXB2) production was detected by enzyme-linked immunosorbent assay on gel-filtered platelets treated with aPL and thrombin, with or without SB203580. Calcium mobilization studies were done utilizing a fluorometric assay. Results. Treatment of platelets with IgG aPL, or their F(abⴕ)2 fragments, in conjunction with subactivating doses of thrombin resulted in a significant increase in phosphorylation of p38 MAPK. Neither the IgG aPL nor their F(abⴕ)2 fragments increased significantly the phosphorylation of ERK-1/2 MAPKs. Furthermore, pre- treatment of platelets with SB203580 completely abrogated the aPL-mediated enhanced aggregation of the platelets. Platelets treated with F(abⴕ)2 aPL and thrombin produced significantly larger amounts of TXB2 when compared with controls, and this effect was completely abrogated by treatment with SB203580. In addition, cPLA2 was also significantly phosphorylated in platelets treated with thrombin and F(abⴕ)2 aPL. There were no significant changes in intracellular [Ca2ⴙ] when platelets were treated with aPL and low doses of thrombin. Conclusion. The data strongly indicate that aPL in the presence of subactivating doses of thrombin induce the production of TXB2 mainly through the activation of p38 MAPK and subsequent phosphorylation of cPLA2. The ERK-1/2 pathway does not seem to be involved in this process, at least in the early stages of aPL-mediated platelet activation. Antiphospholipid antibodies (aPL) have been associated with recurrent thrombosis (arterial and/or venous) and recurrent pregnancy losses in patients with systemic lupus erythematosus and in those with the antiphospholipid syndrome (APS) (1). Thrombocytopenia is a frequent feature of APS, giving rise to the speculation that aPL may play a pathogenic role in thrombosis by binding platelets and causing platelet activation and aggregation (2). Studies have demonstrated binding of affinitypurified aPL to platelets (2–5). Lellouche and colleagues (6) reported that urinary secretion of the major thromboxane metabolite, 11-dehydrothromboxane B2 (TXB2), was significantly increased in patients with lupus anticoagulant (LAC) as compared with normal controls. Studies from our group have also shown that affinity-purified anticardiolipin antibodies (aCL) from patients with APS, but not from patients with syphilis, enhanced activation of platelets treated with suboptimal doses of ADP, thrombin, or collagen (7). In a recent study by our Supported in part by a Research Center in Minority Institutions grant from the NIH (G12-RR-03034) and a Minority Biomedical Research support grant from the NIH (S02-GM-08248). Mariano Vega-Ostertag, BS, E. Nigel Harris, MD, Silvia S. Pierangeli, PhD: Morehouse School of Medicine, Atlanta, Georgia. Address correspondence and reprint requests to Silvia S. Pierangeli, PhD, Antiphospholipid Standardization Laboratory, Department of Microbiology, Biochemistry and Immunology, Morehouse School of Medicine, Gloster Building 332, 720 Westview Drive SW, Atlanta, GA 30310-1495. Submitted for publication February 24, 2004; accepted in revised form April 13, 2004. 2911 2912 VEGA-OSTERTAG ET AL Figure 1. Possible intracellular signaling platelet pathways activated by antiphospholipid antibodies (aPL). TRAP ⫽ thrombin receptor agonist peptide; PAR ⫽ protein-activated receptor; gp ⫽ glycoprotein; Rc ⫽ receptor; PLC␤ ⫽ phospholipase C␤; PLC␥ ⫽ phospholipase C␥; DG ⫽ diacylglycerol; IP3 ⫽ inositol triphosphate; PKC ⫽ protein kinase C; cPLA2 ⫽ [Ca2⫹]-dependent cytosolic phospholipase A2; AA ⫽ acetylsalicylic acid; TXB2 ⫽ thromboxane B2; TXRc ⫽ thromboxane receptor. group, platelets pretreated with suboptimal doses of thrombin receptor agonist peptide (TRAP) and aPL expressed enhanced levels of activated glycoprotein IIb/ IIIa, indicating platelet activation (8). In another study, rabbit aCL were shown to enhance collagen-induced platelet activation (9). Robbins et al showed that aPL– ␤2-glycoprotein I (␤2GPI) complexes significantly increased production of thromboxane A2 (TXA2), a proaggregatory prostanoid in platelets (10). Platelets contain family members of the MAPKs, including ERK-1 (p44 MAPK), ERK-2 (p42 MAPK), and p38 MAPK (Figure 1). The MAPK p38 is a member of a family of proline-directed serine/threonine kinases that is dual-phosphorylated on a threonine and tyrosine residue, separated by 1 single amino acid (11,12). In platelets, p38 MAPK is activated by stress, such as heat and osmotic shock, arsenite, H2O2, ␣-thrombin, collagen, and thromboxane analog (11,12), and is involved in the phosphorylation of [Ca2⫹]-dependent cytosolic phospholipase A2 (cPLA2), with subsequent production of TXB2 (Figure 1). Thrombin has also been shown to induce phosphorylation of ERK-1/2, involving protein kinase C (PKC), phospholipase C␤ (PLC␤), and the intracellular mobilization of [Ca2⫹] (Figure 1) (13–15). Although several studies have shown that aPL enhance platelet activation in vitro in the presence of low doses of agonists (ADP, thrombin, collagen, or TRAP) (7–10,16), the intracellular events involved in this process are not understood. To address this question, we examined the effects of aPL on phosphorylation of p38 MAPK, ERK-1/2 MAPKs, and cPLA2 on intracellular [Ca2⫹] mobilization and on TXB2 production in the presence of subactivating doses of thrombin. The effects of the specific inhibitor for p38 MAPK, SB203580 (4-[4-fluorophenyl]-2[4-methylsulfinylphenyl]-5-[4-pyridyl] 1-imidazole), on aPL-mediated enhancement of platelet aggregation and on TXB2 production in the presence of thrombin were also determined. PATIENTS AND METHODS Purification of IgG and preparation of F(abⴕ)2 fragments. IgG from 7 patients with APS (in accordance with the Sapporo criteria ) and from the pooled sera of 10 healthy INTRACELLULAR EVENTS IN aPL-MEDIATED THROMBOSIS subjects (as controls) were affinity purified by protein G (Gamma BindG Type 3; Amersham-Pharmacia, Piscataway, NJ) by elution with 0.5M acetic acid (pH 3) and neutralization of the fractions in 1M Tris buffer (pH 9). IgG were subsequently dialyzed in a 0.1M acetate buffer, pH 4.5, for 2 hours at 4°C, and digestion with pepsin to obtain F(ab⬘)2 fragments was performed by incubation with pepsin-agarose beads (Sigma-Aldrich, St. Louis, MO) for 18 hours at 37°C. The beads were separated by centrifugation for 10 minutes at 4,000 revolutions per minute, and the supernatants were dialyzed against phosphate buffer, pH 7.4, at 4°C. Intact Fc fragments and nondigested antibody were removed by protein G–sepharose chromatography. F(ab⬘)2 fragments showed a single band at ⬃110 kd in nonreduced sodium dodecyl sulfate– polyacrylamide gel electrophoresis (SDS-PAGE) (utilizing silver staining). IgG fractions and F(ab⬘)2 were passed through a 0.22-m filter and were checked for the absence or presence of endotoxin (lipopolysaccharide [LPS]) by the Limulus amebocyte lysate assay (Amebolysate; ICN Biomedical, Costa Mesa, CA). Determination of antibodies. The aCL and antiphosphatidylserine (aPS) antibodies in the IgG fractions were determined by enzyme-linked immunosorbent assay (ELISA) as previously described (18), using an anti-human IgG antiserum labeled with alkaline phosphatase (␥-chain specific). Similarly, binding of the F(ab⬘)2 fragments to cardiolipin was determined by ELISA, using an anti-human and secondary antibody cocktail (1:1,000 dilution) labeled with alkaline phosphatase (Sigma-Aldrich). Titers of aCL were reported in IgG phospholipid (GPL) units, and levels of aPS antibodies were reported in net optical density (OD) units. Anti-␤2GPI antibodies were detected by ELISA as previously described, with some modifications, using irradiated Costar microtiter plates (Corning, Corning, NY) (19). Plates were coated with 20 g/ml purified ␤2GPI, and blocked with 5% ovalbumin–phosphate buffered saline (PBS) solution. Anti-human IgG (␥-chain specific) and anti- and anti- secondary antibodies labeled with alkaline phosphatase were used. Anti-␤2GPI activities were determined at 405 nm, with results expressed in net OD units. LAC activities of the IgG and of the F(ab⬘)2 fragments were determined by a modified activated partial thromboplastin time (APTT) as previously described (20). Isolation of platelets. Platelets were obtained from the blood of adult healthy volunteers who had not received medication for at least 10 days. Blood was collected in acid– citrate–dextrose anticoagulant (9/1 volume/volume). Plasmarich platelets (PRPs) were obtained by centrifugation for 20 minutes at 120g at room temperature (21). Preparation of washed platelets. PRPs were centrifuged at 800g for 10 minutes at room temperature and resuspended in HEPES buffer (10 mM HEPES, 140 mM NaCl, 3 mM KCl, 0.5 mM MgCl2, 5 mM NaHCO3, 10 mM dextrose, pH 7.4) supplemented with 0.4 units/ml apyrase and 1 mM aspirin. The platelet count was adjusted to 1 ⫻ 109 platelets/ml and these preparations were used in the immunoblot assays and in the calcium mobilization experiments. Preparation of gel-filtered platelets (GFPs). For the aggregation studies, GFPs were used. Platelets were filtered through a Sepharose 2B column equilibrated with HEPES buffer as previously described (21) without inhibitors, and the 2913 platelet count was adjusted to 250 ⫾ 50 ⫻ 106/ml for the aggregation studies and 300 ⫻ 106 platelets/ml for the TXB2 assay. Platelets were maintained at 37°C in a closed (capped) tube and analyzed within 2 hours after drawing. Immunoblot analysis for phosphorylation of p38 and p44/42 MAPKs. One hundred–microliter aliquots of washed platelets were incubated in microtubes for 5 minutes at 37°C, under static conditions, with 50, 100, or 200 g/ml IgG aPL from APS patients or control IgG, or with 200 g/ml of F(ab⬘)2 aPL from APS patients or control F(ab⬘)2 IgG. Platelets were then treated with thrombin (0.005 units/ml). Platelets to be used as controls were treated with PBS and 1 unit/ml thrombin. Platelets were subsequently lysed with Laemmli buffer (Bio-Rad, Richmond, CA) and 2-mercaptoethanol, and lysates were heated at 95°C for 10 minutes and centrifuged for 10 minutes at 11,000g. Samples were subjected to SDS-PAGE in a 12% gel. After the electrophoresis, proteins were transferred to nitrocellulose membranes at 10V for 30 minutes in a semidry transfer cell. The membrane was then blocked and subsequently incubated with one of the following antibodies: p38 MAPK (Thr180/Tyr182) 28b10 mouse anti-human monoclonal antibody (Cell Signals Technology, Beverly, MA) or p38 MAPK rabbit polyclonal antibody (Santa Cruz Biotechnology, Santa Cruz, CA), or with phosphorylated ERK-1/2 (Thr202/ Tyr204) E10 mouse anti-human monoclonal antibody (Cell Signals Technology) or ERK-1/2 rabbit polyclonal antibody (Santa Cruz Biotechnology), followed by incubation with secondary antibody/goat anti-mouse or goat anti-rabbit IgG labeled with peroxidase (21) and detection by chemiluminescence (Amplight/BioRad, Hercules, CA). The intensity of the bands was quantitated by densitometric analysis using Gel Pro Analyzer software, version 4.5 (Fotodyne, Hartland, WI). Phosphorylation of cPLA2. One hundred microliters of purified platelets in a concentration of 1 ⫻ 109 platelets/ml (prepared as described above) was incubated with 50 l (200 g/ml final concentration) of F(ab⬘)2 aPL from APS patients (n ⫽ 5) or 50 l control F(ab⬘)2 IgG or PBS for 5 minutes, and then stimulated with 0.005 units/ml thrombin, or treated with PBS and 1 unit/ml thrombin as a positive control. The reaction was stopped by the addition of an equal volume of 2⫻ immunoprecipitation buffer (100 mM Tris HCl, pH 7.4, 300 mM NaCl, 2 mM EGTA, 5 g/ml leupeptin, 5 g/ml aprotinin, 2 mM phenylmethylsulfonyl fluoride, 2 mM Na3VO4, 2 mM NaF, 2% Nonidet P40, 0.5% sodium deoxycholate). The lysates were cleared by treating them with protein G– Sepharose for 45 minutes. The cleared supernatants were incubated with 1 g of cPLA2 antibody (Santa Cruz Biotechnology) for 2 hours at 4°C. The immune complexes were precipitated by addition of 90 l of protein G–Sepharose for 45 minutes. After brief centrifugation, immunoprecipitates were washed 3 times with 1 ml of 1⫻ immunoprecipitation buffer and then resuspended with 25 l of SDS–sample buffer, and subjected to SDS-PAGE in a 10% gel for 210 minutes. The proteins in the gel were transferred to a nitrocellulose paper and were immunoblotted as described previously, using the same cPLA2 antibody (rabbit polyclonal cPLA2; Santa Cruz Biotechnology). The phosphorylated form of cPLA2 is detected as a slower-moving second band. Thrombin-induced anti-cPLA2 was detectable after 2 minutes of stimulation (approximately half of the protein was in the phosphorylated form) (21). 2914 VEGA-OSTERTAG ET AL Aggregation studies. Aggregation of platelets was performed as follows: a 420-l aliquot of GFPs and 1 l of CaCl2 (1 mM final concentration) were placed in the aggregometer cuvettes at 37°C with 30 l of IgG aPL antibodies from APS patients (final concentration 200 g/ml) or 30 l of control IgG (final concentration 200 g/ml) for 5 minutes. In some experiments, platelets were pretreated for 20 minutes with SB203580 (0.1 and 1 M) or with 1% dimethylsulfoxide (22). Aggregation was then started in a dual channel aggregometer (Minigator II; Payton Scientific, Buffalo, NY) and registered in a linear recorder. Platelet aggregation was evaluated as the amplitude of the aggregation curves (expressed in millimeters) after 5 minutes. TXB2 assay. GFPs were adjusted to a concentration of 3 ⫻ 108 platelets/ml in HEPES buffer and incubated for 5 minutes with (Fab⬘)2 aPL fragments from APS patients or control F(ab⬘)2 IgG fragments or PBS in the presence or the absence of 1 M/liter SB203580, and then stimulated with 0.005 units/ml thrombin for 3 minutes. The production of TXB2 was stopped by the addition of indomethacin (0.1 mM) in HEPES buffer, and the preparations were centrifuged at 10,000 rpm for 30 seconds at 4°C. The supernatants were rapidly separated and stored at ⫺20°C until analyzed. TBX2 was analyzed in the supernatants utilizing a commercial enzyme immunoassay system (Amersham-Pharmacia) as indicated by the manufacturer. Intracellular [Ca 2ⴙ ] mobilization experiments. Washed platelets were incubated for 20 minutes at 37°C with 5 M visible light–excitable [Ca2⫹] indicator Fura Red– acetoxymethyl ester (AM) and 5 M calcium green–AM (Molecular Probes, Eugene, OR). Subsequently, platelets were centrifuged at 800g for 10 minutes at room temperature and resuspended in HEPES buffer together with 0.4 units/ml apyrase VII, 0.1 M prostaglandin E1, and 1 mM acetylsalicylic acid, and standardized at a platelet count of 1 ⫻ 109/ml. One hundred–microliter aliquots of platelets were then incubated with 200 g/ml F(ab⬘)2 aPL from APS patients or 200 g/ml control F(ab⬘)2 IgG or PBS for 5 minutes, and then stimulated with 0.01 units/ml thrombin. Each experiment included a calibration curve utilizing 0.01, 0.1, 0.5, and 1 units/ml thrombin. The emission peaks of calcium green and Fura Red signal intensities were determined at 530/30 nm and 640/40 nm, respectively (21). Elevations in [Ca2⫹] result in increased fluorescence intensity of calcium green and decreased fluorescence intensity of Fura Red, at 485/20 nm excitation wavelength. Changes in [Ca2⫹] were evaluated on the basis of a ratio of signal intensity at the G-channel (calcium green, sensitive peak at 530 nm) relative to the R-channel (red, sensitivity peak at 640 nm) (21). Statistical analysis. Data are presented as the mean ⫾ SD. Student’s t-test was used to compare mean values between treated platelets and control platelets. P values less than or equal to 0.05 were considered significant. RESULTS Characterization of IgG aPL and F(abⴕ)2 aPL. All of the IgG aPL samples from the 7 APS patients Figure 2. Effects of different concentrations of thrombin on phosphorylation of p38 MAPK and ERK-1/2. Washed platelets were treated with various concentrations of thrombin as indicated (0–5 units/ml thrombin). Phosphorylation of p38 MAPK (P38-P) (A) or phosphorylation of ERK-1/2 (ERK1/ERK2-P) (B) was detected by immunoblot in the platelet lysates using specific monoclonal antibodies to the phosphorylated form of the enzymes, as described in Patients and Methods. were positive for aCL (⬎100 GPL units) and all of the IgG samples from the healthy controls were negative for aCL (⬍10 GPL units), when preparations were tested at a 200 g/ml protein concentration. The anti-␤2GPI antibody activities of the IgG aPL and control IgG were a mean ⫾ SD 0.870 ⫾ 0.218 net OD units and 0.140 ⫾ 0.019 net OD units, respectively, when samples were tested at a 100 g/ml protein concentration. Six of the 7 IgG aPL samples were positive for LAC (APTT ratio of patient plasma to normal plasma 1.21:1.50). All F(ab⬘)2 aPL samples were positive for binding to cardiolipin (25–186 GPL units/100 g protein), and the anti-␤2GPI activities of the F(ab⬘)2 aPL and control F(ab⬘)2 IgG were a mean ⫾ SD 0.505 ⫾ 0.111 net OD units and 0.068 ⫾ 0.024 net OD units, respectively (significantly different; P ⫽ 0.0023). Binding to phosphatidylserine was tested by ELISA and was a mean ⫾ SD 0.506 ⫾ 0.112 net OD units for the F(ab⬘)2 aPL preparations and 0.018 ⫾ 0.005 net OD units for control F(ab⬘)2 IgG (significantly different; P ⫽ 0.0002). All IgG preparations and F(ab⬘)2 fragments tested negative for LPS in the Limulus amebocyte lysate assay. Effects of aPL on phosphorylation of p38 MAPK in platelets. The treatment of washed platelets with varying concentrations of thrombin (0.005–5 units/ml) produced a dose-dependent phosphorylation of p38 MAPK (Figure 2A). Treatment of platelets with control IgG and 0.005 units/ml thrombin produced phosphorylation that was not different from that of platelets treated with 0.005 units/ml thrombin alone. In contrast, platelets treated with IgG aPL (n ⫽ 7) at 200 g/ml protein concentration and 0.005 units/ml thrombin produced a significant increase in phosphorylation of p38 MAPK (among the 7 IgG aPL samples, fold increases of 4.4, 7.8, 7.4, 5.8, 7.5, 6.3, and 5.0, when compared with INTRACELLULAR EVENTS IN aPL-MEDIATED THROMBOSIS Figure 3. Effects of IgG antiphospholipid antibodies (IgGaPL) on phosphorylation of p38 MAPK (P38P). Washed platelets were treated with 200 g/ml of IgGaPL (n ⫽ 7) or IgG from normal healthy controls (IgGNHS) and stimulated with 0.005 units/ml thrombin, or treated with phosphate buffered saline (PBS) alone or with 0.2 units/ml thrombin alone (A). For comparison, platelets were treated with IgGaPL (n ⫽ 7) or IgGNHS or PBS alone in the absence of thrombin (B). A positive control comprising platelets treated with 0.2 units/ml thrombin alone was also included in the experiments. Phosphorylation of p38 MAPK was detected by immunoblot using specific monoclonal antibodies to the phosphorylated form of the enzyme, as described in Patients and Methods. The results shown are representative of 5 experiments, using platelets from different donors. platelets treated with control IgG and thrombin) (Figure 3A). A 7.6-fold increase was observed in platelets treated with 0.2 units/ml of thrombin alone (positive control). The effect was dependent on the amount of aPL used. For example, for sample IgGaPL1, a 4.4-, 3.0-, and 1.9-fold increase in p38 MAPK phosphorylation was observed with 200, 100, and 50 g/ml protein concentrations, respectively. Treatment of platelets with IgG aPL alone (n ⫽ 7) (in the absence of thrombin) did not produce any effect on p38 MAPK phosphorylation (Figure 3B). To determine whether the effects produced by IgG aPL were due to interactions of the antibodies with platelets either through the Fab fragment or through the Fc portion of the immunoglobulins, the effects of F(ab⬘)2 fragments on phosphorylation of p38 MAPK on platelets were determined by immunoblot. As shown in Figures 4A and B, 4 of the 5 F(ab⬘)2 aPL preparations produced a significant increase of phosphorylation (12.6-, 4.6-, 3.7-, and 12.6-fold increases) when compared with platelets treated with control F(ab⬘)2 IgG and 0.005 units/ml thrombin. No increase in phosphorylation of p38 MAPK was observed in platelets treated with control F(ab⬘)2 IgG and thrombin when compared with platelets treated with thrombin alone. 2915 Effects of aPL on phosphorylation of ERK-1/2. Treatment of platelets with different concentrations of thrombin (0.005–5 units/ml thrombin) produced a dosedependent phosphorylation of ERK-1/2 (Figure 2B). Phosphorylation of ERK-1/2 was observed starting at 0.2 units/ml thrombin. Neither the platelets treated with F(ab⬘)2 aPL (n ⫽ 5) at 200 g/ml protein concentration and 0.005 units/ml thrombin nor the platelets treated with control F(ab⬘)2 IgG and 0.005 units/ml thrombin induced phosphorylation of ERK-1/2 after 2 minutes of treatment (Figure 4C). Effects of a specific inhibitor of p38 MAPK on aPL-mediated platelet aggregation. To confirm whether the phosphorylation of p38 MAPK is involved in aPLmediated platelet activation, aggregation studies were performed in the presence and in the absence of the specific inhibitor of p38 MAPK, SB203580. As shown in Figure 5, platelets treated with IgG aPL and 0.005 units/ml thrombin produced an aggregation of 55%, whereas there was no aggregation in platelets treated with control IgG and 0.005 units/ml thrombin. Pretreatment of the platelets with 0.1 M or 1 M SB203580 abrogated the aggregatory effects of aPL by 89% and 100%, respectively. Effects of aPL on phosphorylation of cPLA2. Treatment of washed platelets with 1 unit/ml thrombin induced strong phosphorylation of cPLA2 after 2 min- Figure 4. Effects of F(ab⬘)2 aPL on phosphorylation of p38 MAPK (p38-P) and ERK-1/2 (ERK1/ERK2-P). Washed human platelets were incubated with 200 g/ml of F(ab⬘)2 IgGNHS or with F(ab⬘)2 aPL (n ⫽ 5) and stimulated with 0.005 units/ml thrombin or with PBS and 0.005 units/ml or 1 unit/ml thrombin. Lysates of the platelets were immunoblotted using specific antibodies for the phosphorylated forms of p38 MAPK (A) and of ERK-1/2 (C) and the unphosphorylated form of p38 MAPK (B). Each blot is representative of the results of 5 experiments, using platelets from different donors. See Figure 3 for other definitions. 2916 VEGA-OSTERTAG ET AL Table 1. Effects of F(ab⬘)2 antiphospholipid antibodies (aPL) on thromboxane B2 (TXB2) production in platelets* Treatment 0.005 units/ml thrombin F(ab⬘)2 IgG-NHS ⫹ 0.005 units/ml thrombin F(ab⬘)2 aPL ⫹ 0.005 units/ml thrombin F(ab⬘)2 aPL ⫹ 0.005 units/ml thrombin ⫹ 1 M SB203580 Figure 5. Effects of a specific inhibitor of p38 MAPK on aPLmediated platelet aggregation. Gel-filtered platelets were treated with IgG aPL (200 g/ml) and thrombin (0.005 units/ml) in the presence or absence of 1 M or 0.1 M SB203580 or with IgG NHS (200 g/ml) and thrombin (0.005 units/ml). Platelet aggregation was determined in an aggregometer over a period of 5 minutes, as described in Patients and Methods. Scale is indicated at the bottom left, with axes indicating the aggregation percentage per minute. See Figure 3 for definitions. utes (Figure 6). Four of the 5 F(ab⬘)2 aPL fragments induced significant phosphorylation of cPLA2 in platelets pretreated with 0.005 units/ml thrombin. Neither treatment with 0.005 units/ml thrombin nor treatment with control F(ab⬘)2 IgG and 0.005 units/ml thrombin induced phosphorylation of cPLA2 (Figure 6). Figure 6. Effects of IgG antiphospholipid antibodies (aPL) on phosphorylation of [Ca2⫹]-dependent cytosolic phospholipase A2 (cPLA2). Washed platelets were treated with 200 g/ml of F(ab⬘)2 aPL (n ⫽ 5) or F(ab⬘)2 IgG from normal healthy subjects (IgGNHS) or phosphate buffered saline (PBS), and then stimulated with 0.005 units/ml of thrombin for 2 minutes. Platelets were also treated with PBS and 1 unit/ml thrombin as a positive control. The phosphorylated (cPLA2-P) and the nonphosphorylated form of the enzyme cPLA2 were detected in the platelet lysates by immunoprecipitation and immunoblot, as described in Patients and Methods. TXB2 production, mean ⫾ SD pg/ml 230.0 ⫾ 12.5 251.0 ⫾ 12.7 535.0 ⫾ 50.9† 280.3 ⫾ 57.0‡ * Platelets were treated with 0.005 units/ml thrombin and with F(ab⬘)2 aPL (n ⫽ 5) or F(ab⬘)2 IgG from normal healthy subjects (IgG-NHS) (n ⫽ 5), and TXB2 was measured by enzyme-linked immunosorbent assay, as described in Patients and Methods. In some experiments, platelets were pretreated with 1 M SB203580. Experiments were repeated 4 times. † P ⬍ 0.0027 versus platelets treated with F(ab⬘)2 IgG-NHS plus thrombin. ‡ P not significant versus platelets treated with F(ab⬘)2 IgG-NHS plus thrombin. Effects of aPL on TXB2 production. The production of TXB2 was significantly increased in platelets treated with F(ab⬘)2 aPL and thrombin, when compared with platelets treated with control F(ab⬘)2 IgG and thrombin (mean ⫾ SD levels of TXB2 535.0 ⫾ 50.9 pg/ml versus 251.0 ⫾ 12.7 pg/ml; P ⫽ 0.0027) or with thrombin alone (230.0 ⫾ 12.5 pg/ml). The mean values of TXB2 produced by platelets treated with control F(ab⬘)2 IgG and those treated with 0.005 units/ml thrombin alone were not statistically significantly different (Table 1). The effect of F(ab⬘)2 aPL on TXB2 production was completely abrogated by pretreatment of the platelets with SB203580 (mean ⫾ SD levels of TXB2 280.3 ⫾ 57.0 pg/ml), indicating that the activation of p38 MAPK is involved in aPL-mediated production of TXB2 by platelets (Table 1). Effects of aPL on intracellular [Ca2ⴙ] mobilization. The treatment of platelets with different concentrations of thrombin (0.01, 0.1, 1, and 5 units thrombin) produced a significant and dose-dependent increase in intracellular calcium, with a peak between 2 and 3 minutes of treatment (G-channel to R-channel mean ratios 2.1, 2.8, 3.2, and 5.7, respectively) (Figure 7). There was no significant increase in intracellular calcium concentrations for a period of 5 minutes when platelets were treated with IgG aPL and 0.01 units/ml thrombin, as compared with platelets treated with 0.01 units/ml thrombin alone or platelets treated with control IgG and thrombin (mean G-channel to R-channel ratios at 3 minutes 2.22, 2.1, and 2.2, respectively) (Figure 7). The INTRACELLULAR EVENTS IN aPL-MEDIATED THROMBOSIS Figure 7. Effects of aPL on intracellular calcium ([Ca2⫹]) mobilization. Washed platelets were incubated with 200 g/ml F(ab⬘)2 from healthy controls or F(ab⬘)2 aPL (n ⫽ 5) for 5 minutes and stimulated with 0.01 units/ml of thrombin. Changes in [Ca2⫹] were evaluated based on a mean ratio of the signal intensities of G-channel to R-channel (G/R) over 1-minute (Min.) intervals of time. A ⫽ platelets treated with 1 unit/ml thrombin; B ⫽ platelets treated with 0.5 units/ml thrombin; C ⫽ platelets treated with 0.1 units/ml thrombin; D ⫽ platelets treated with F(ab⬘)2 IgG from healthy controls plus 0.01 units/ml thrombin; E ⫽ platelets treated with F(ab⬘)2 aPL and 0.01 units/ml thrombin; F ⫽ platelets treated with 0.01 units/ml thrombin alone. See Figure 6 for other definitions. experiments were run 7 times and the results shown in Figure 7 correspond to 1 representative experiment. DISCUSSION Studies have shown conclusively that aPL are thrombogenic in in vivo animal models (23–25). The prothrombotic properties of aPL may be explained in part by their ability to enhance the activation of platelets. Moreover, aPL have been shown to increase production of TXB2 in the presence of low doses of ADP, collagen, or thrombin (7,10,26,27). A recent study by our group showed a significant increase in the expression of activated glycoprotein IIb/IIIa on platelets treated with aPL and TRAP (8). Furthermore, aPL-enhanced thrombosis in vivo can be abrogated by infusions of a glycoprotein IIb/IIIa antagonist (1B5) monoclonal antibody and in ␤3-null (glycoprotein IIb/IIIa–deficient) mice (Vega-Ostertag M, et al: unpublished observations). In this study, we confirmed that aPL have a direct effect on platelet activation in the presence of thrombin, and we examined the intracellular pathways involved in this process. The activation of platelets by thrombin has been shown to involve more than one pathway, comprising the p38 MAPK pathway, including the downstream calciumdependent phosphorylation of cPLA2, or the PLC␤ 2917 pathway, involving activation of PKC, phosphorylation of ERK1/ERK2 (Figure 1), and the intracellular mobilization of [Ca2⫹]. In both cases, TXB2 is produced and platelets are activated (11–15). In this study, we found that phosphorylation of p38 MAPK mediates aPL-induced activation of platelets in vitro. The degree of phosphorylation of this enzyme varied among the 7 different preparations of aPL (4.4– 7.9-fold increase) and the F(ab⬘)2 aPL fragments (3.7– 12.6-fold increase over the control) used in this study. This variability is not surprising. Antiphospholipid antibodies are known to be heterogeneous in specificity and function, and have been shown to bind negatively charged phospholipids, ␤2GPI, prothrombin, annexin V, and other proteins of the coagulation cascade. A wide variety of functions have been attributed to aPL, from activation of endothelial cells, up-regulation of tissue factor in monocytes, and platelet activation. The present data conclusively show that aPL induce phosphorylation of p38 MAPK after pretreatment with low (subactivating) doses of thrombin. Furthermore, these effects are dependent on the dose of antibody utilized and are abrogated by pretreatment of the platelets with the specific inhibitor of the enzyme, SB203580, as shown in the aggregation studies and in the TXB2 production experiments. We also show that aPL up-regulate production of TXB2 in platelets. This is consistent with the study by Opara et al that recently demonstrated an increase in platelet TXB2 production and in aggregation by aCL– ␤2GPI complexes (9,28). Furthermore, in our study, we show that pretreatment of platelets with SB203580 completely abrogates the production of TXB2 induced by aPL and low doses of thrombin, and that cPLA2 is significantly phosphorylated in platelets treated with low doses of thrombin and F(ab⬘)2 aPL, an event downstream of p38 MAPK activation (Figure 1). Therefore, the data conclusively show involvement of that enzyme in aPL-mediated platelet activation. These data are also consistent with the suggestions by Gonzalez-Buritica et al, who, in 1988, hypothesized that phospholipase A2 plays a role in platelet activation in patients with aPL (29). The ERK-1/2 phosphorylation pathway may also be initiated in platelets by thrombin (Figure 1) (14,15). Interestingly, the data from our study show that treatment with aPL and low doses of thrombin does not induce phosphorylation of ERK-1/2. Initial in vitro studies performed in transfected cells and in HeLa cells suggested that p42 MAPK phosphorylates cPLA2 at 2918 Ser505, which lies within a consensus sequence for MAPK (Pro-Xaa-Ser/Thr-Pro). Subsequently, other investigators reported the concomitant activation of MAPK and phosphorylation of cPLA2 in stimulated cells (30–34). Several recent reports, however, dissociated cPLA2 phosphorylation from MAPK activation in platelets. One study showed that phosphorylation of ERK-1/2 is not required for phosphorylation of cPLA2 in thrombin-stimulated platelets (35). Those authors concluded that cPLA2 is the physiologic target of p38 MAPK, and that ERK-1/2 phosphorylation of cPLA2 is not required for its receptor-mediated activation in platelets (35). Borsch-Haubold et al have shown no effect on phosphorylation of cPLA2 or release of thromboxane when the specific inhibitor of PKC, Ro31-8220, was used in platelets stimulated with thrombin or collagen (12,35–38). Similarly, the same group of authors showed that inhibition of MAPK kinase using PD98059 did not affect platelet responses to the physiologic stimuli, thrombin and collagen, indicating a role for p38 MAPK in primary activation of human platelets, independent of ERK-1/2 (35–39). Altogether, these results are in agreement with our findings and provide evidence against a role for ERK-1/2 in primary aPL-mediated platelet aggregation. However, we do not exclude the possibility that these MAPKs may play a role in postaggregation events in platelets mediated by aPL. We did not find significant changes in intracellular [Ca2⫹] when platelets were treated with F(ab⬘)2 aPL and low doses of thrombin, supporting the hypothesis that the PLC␤ pathway and the downstream phosphorylation of ERK-1/2 activation are not involved in aPLmediated platelet activation (see Figure 7). The p38 MAPK–dependent activation of cPLA2 in platelets appears to be dependent on intracellular calcium concentrations. In our studies, although no significant changes in calcium concentrations were observed when aPL were added to the system, treatment of platelets with 0.01 units/ml thrombin induced a modest increase in calcium (as shown in Figure 7). We speculate that the observed change may have been sufficient to initiate cPLA2 activation, but these observations would need to be further evaluated to confirm this hypothesis. The effects of aPL on phosphorylation of p38 MAPK and of cPLA2 and production of TXB2 by platelets reported in this study are due to the influence of the F(ab⬘)2 fragment and not to the Fc portion of the antibody. This finding is consistent with the findings of the study by Robbins et al, which showed that F(ab⬘)2 VEGA-OSTERTAG ET AL aCL fragments significantly stimulated TXB2 production in platelets (10). The present study did not focus on establishing the nature of the receptors to which aPL bind on platelets. Studies have shown that aPL bind only to platelets previously exposed to low doses of agonists or to platelets that have been frozen and thawed repeatedly, exposing negatively charged phospholipids (i.e., phosphatidylserine) (2). However, the precise nature of the receptors for aPL in platelets is not known. Opara et al previously hypothesized that ␤2GPI might mediate aCL binding to the activated platelet cell surface by binding with phosphatidylserine, thereby promoting increased platelet activation by the aCL–␤2GPI complexes (28). A recent study demonstrated that dimeric ␤2GPI binds to members of the low-density lipoprotein receptor family in platelets and induces increased platelet adhesion to collagen (40). This effect was increased by addition of anti-␤2GPI monoclonal antibodies to the system and was abrogated by inhibition of thromboxane synthesis. In summary, our study is the first to show that aPL-mediated platelet activation occurs selectively through the p38 MAPK pathway. Upon priming of the platelets with aPL and low doses of thrombin, cPLA2 is phosphorylated and TXB2 is produced. PKC and ERK1/2 activation do not seem to be involved in this response. These findings may be important in designing new approaches to targeted treatment of thrombosis in APS patients. ACKNOWLEDGMENT We are grateful to Dr. Jacob Rand of Albert Einstein School of Medicine (New York, NY) for his helpful comments on this work. REFERENCES 1. Harris EN, Asherson RA, Gharavi AE, Morgan SH, Derue G, Hughes GR. Thrombocytopenia in SLE and related autoimmune disorders: association with anti-cardiolipin antibody. Br J Haematol 1985;59:227–30. 2. Khamashta MA, Harris EN, Gharavi AE, Derue G, Gil A, Vazquez JJ, et al. Immune mediated mechanism for thrombosis: antiphospholipid antibody binding to platelet membranes. Ann Rheum Dis 1988;47:849–54. 3. Out HJ, deGroot P, van Vliet M, de Gast G, Nieuwenhuis K, Derksen RH. Antibodies to platelets in patients with antiphospholipid antibodies. Blood 1991:77:2655–9. 4. Shatter Y, Tall Y, Greenberg A, Brenner B. Platelet activation in patients with antiphospholipid syndrome. Blood Coag Fibrinol 1998;9:653–7. 5. Hasselaar P, Derksen RH, Blokzijl L, deGroot PG. Crossreactivity INTRACELLULAR EVENTS IN aPL-MEDIATED THROMBOSIS 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. of antibodies directed against cardiolipin, DNA, endothelial cells and blood platelets. J Thromb Haemost 1990;63:169–73. Lellouche F, Martinuzzo M, Said P, Maclouf J, Carreras LO. Imbalance of thromboxane/prostacyclin biosynthesis in patients with lupus anticoagulant. Blood 1991;78:2894–9. Campbell A, Pierangeli SS, Wellhausen S, Harris EN. Comparison of the effects of anticardiolipin antibodies from patients with the antiphospholipid syndrome and with syphilis on platelet activation and aggregation. J Thromb Haemost 1995;73:529–34. Espinola RG, Pierangeli SS, Gharavi AE, Harris EN. Hydroxychloroquine reverses platelet activation induced by human IgG antiphospholipid antibodies. J Thromb Haemost 2002;87:518–22. Ling YL, Wang CT. Activation of human platelets by the rabbit anticardiolipin antibodies. Blood 1992;80:3135–43. Robbins DL, Leung S, Miller-Blair DJ, Ziboh V. Effect of anticardiolipin/␤2 glycoprotein I complexes on production of thromboxane A2 by platelets from patients with the antiphospholipid syndrome. J Rheumatol 1998;25:51–6. Buschbeck M, Ghomashchi F, Gelb MH, Watson SP, BorschHaubold AG. Stress stimuli increase calcium-induced arachidonic acid release through phosphorylation of cytosolic phospholipase A2. Biochemistry 1999;344:359–66. Borsch-Haubold AG, Ghomashchi F, Pasquet S, Goedert M, Cohen P, Gelb MH, et al. Phosphorylation of cytosolic phospholipase A2 in platelets is mediated by multiple stress-activated protein kinase pathways. Eur J Biochem 1999;265:195–203. Nadal-Wollbold F, Pawlowski M, Levy-Toledano S, Berrou E, Rosa JP, Brycjaert M. Platelet ERK2 activation by thrombin is dependent on calcium and conventional protein kinases C but not Raf-1 or B-Raf. FEBS Lett 2002;531:475–82. Levy-Toledano S, Gallet C, Nadal F, Bryckaert M, Maclouf J, Rosa J-P. Phosphorylation and dephosphorylation mechanisms in platelet function: a tightly regulated balance. J Thromb Haemost 1997;78:226–33. Levy-Toledano S. Platelet signal transduction pathways: could we organize them into a “hierarchy”? Haemostasis 1999:29:4–15. Wang L, Su CY, Chou KY, Wang CT. Enhancement of human platelet activation by the combination of low concentrations of collagen and rabbit anticardiolipin antibodies. Br J Haematol 2002;118:1152–62. Wilson WA, Gharavi AE, Koike T, Lockshin MD, Branch DW, Piette JC, et al. International consensus statement on preliminary classification criteria for definite antiphospholipid syndrome: report of an international workshop. Arthritis Rheum 1999;42; 1309–11. Harris EN. Annotation: antiphospholipid antibodies. Br J Haematol 1990;74:1–9. Palomo I, Pereira J, Alarcon M, Larrain AM, Pinochet C, Vasqez M, et al. Antiphospholipid antibodies in Chilean patients with systemic lupus erythematosus. J Lab Clin Med 2002;14:336–41. Exner T, Ricard KA, Kronenberg H. A sensitive test demonstrating lupus anticoagulant and its behavioral patterns. Br J Haematol 1985;59:227–30. Hackeng CM, Relou IA, Pladet MA, Gorter G, van Rijn HJ, Akkerman JW. Early platelet activation by low density lipoprotein via p38MAP kinase. J Thromb Haemost 1999;82:1749–56. Davies SP, Reddy H, Caivano M, Cohen M. Specificity and mechanism of action of some commonly used protein kinase inhibitors. Biochemistry 2000:351:95–105. Pierangeli SS, Liu X, Barker JH, Anderson G, Harris EN. Induction of thrombosis in a mouse model by IgG, IgM and IgA immunoglobulins from patients with the antiphospholipid syndrome. J Thromb Haemost 1995;74:1361–7. 2919 24. Espinola RG, Liu X, Colden-Stanfield M, Hall J, Harris EN, Pierangeli SS. E-selectin mediates pathogenic effects of antiphospholipid antibodies. J Thromb Haemost 2003;1:843–8. 25. Jankowski M, Vreys L, Wittevrongel C, Boon D, Vermylen J, Hoylaerts MD, et al. Thrombogenicity of ␤2glycoprotein I dependent antiphospholipid antibodies in a photochemically induced thrombosis model in the hamster. Blood 2002;101:157–62. 26. Martinuzzo ME, Maclouf J, Carreras LO, Levy-Toledano S. Antiphospholipid antibodies enhance thrombin-induced platelet activation and thromboxane formation. J Thromb Haemost 1993: 70:667–71. 27. Arvieux J, Rousell B, Pouzol P, Colomb MG. Platelet activating properties of murine monoclonal antibodies to ␤2glycoprotein I. J Thromb Haemost 1993;70:336–41. 28. Opara R, Robbins DL, Ziboh VA. Cyclic-AMP agonists inhibit antiphospholipid/␤2 gycoprotein I induced synthesis of human platelet thromboxane A2 in vitro. J Rheumatol 2003;30:55–9. 29. Gonzalez-Buritica H, Khamashta MA, Hughes GR. Is there a role for phospholipases in the anti-cardiolipin syndrome? Clin Exp Rheumatol 1988;6:341–2. 30. Kramer R, Roberts M, Manetta EF, Hyslop JV, Jakubowski JA. Thrombin-induced phosphorylation and activation of Ca2⫹. J Biol Chem 1993;268: 26796–804. 31. Hefner Y, Borsch-Haubold AG, Murakami M, Wilde JI, Pasquet S, Schielts D, et al. Serine 727 phosphorylation and activation of cytosolic phospholipase A2 by MNK1-related protein kinases. J Biol Chem 2000;275:37542–51. 32. Lin LL, Wartmann M, Lin AY, Knopf JL, Seth A, Davis RJ. cPLA2 is phophorylated and activated by MAP kinase. Cell 1993:72:269–78. 33. Nemenoff RA, Winitz S, Qian NX, Van Putten V, Johnson GL, Heasley LE. Phosphorylation and activation of a high molecular weight form of phospholipase A2 by p42 microtubule-associated protein 2 kinase and protein kinase. J Biol Chem 1993;268:1960–4. 34. Borsch-Haubold AG, Bartoli F, Asselin J, Dudler T, Kramer RM, Apitz-Castro R, et al. Identification of the phosphorylation sites of cytosolic phospholipasse A2 in agonist-stimulated human platelets and HeLA cells. J Biol Chem 1998;273:4449–58. 35. Borsch-Haubold AG, Kramer RM, Watson SP. Cytosolic phospholipase A2 is phosphorylated in collagen- and thrombin-stimulated human platelets independent of protein kinase C and mitogen-activated protein kinase. J Biol Chem 1995;270:25885–98. 36. Saklatvala J, Rawlinson L, Waller RJ, Sarsfield S, Lee JC, Morton LF, et al. Role for p38 mitogen-activated protein kinase in platelet aggregation caused by collagen or a thromboxane analog. J Biol Chem 1996;271:6586–9. 37. Borsch-Haubold AG, Kramer RM, Watson SP. Inhibition of mitogen-activated protein kinase kinase does not impair primary activation of human platelets. Biochemistry 1996;318:207–12. 38. Kramer RM, Roberts EF, Um SL, Borsch-Haubold AG, Watson SP, Fisher MJ, et al. p38 mitogen-activated protein kinase phosphorylates cytosolic phospholipase A2 (cPLA2) in thrombinstimulated platelets: evidence that proline-directed phosphorylation is not required for mobilization of arachidonic acid by cPLA2. J Biol Chem 1996;271:27723–9. 39. Borsch-Haubold AG, Pasquet S, Watson SP. Direct inhibition of cyclooxygenase-1 and -2 by the kinase inhibitors SB203580 and PD98059. J Biol Chem 1998;273:28766–72. 40. Lutters BC, Derksen RH, Tekelnenburg WL, Lenting PJ, Arnout J, De Groot PG. Dimers of ␤2-glycoprotein I increase platelet deposition to collagen via interaction with phospholipids and the apolipoprotein E receptor 2. J Biol Chem 2003;278:33831–8.