If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Department of Anaesthesiology and Intensive Care, Experimental and Clinical Haemostasis, University Hospital Muenster, Muenster, GermanyCenter for Laboratory Medicine, University Hospital Muenster, Muenster, Germany
Corresponding author at: Center for Laboratory Medicine, University Hospital Münster, Albert Schweizer Str 33, D-48129 Münster, Germany. Tel.: +49 251 8356276; fax: +49 251 8356276.
Center for Laboratory Medicine, University Hospital Muenster, Muenster, GermanyDepartment of Medicine, Endocrinology, Metabolism and Geriatrics, University of Modena and Reggio Emilia, Modena, Italy
HIGH-density lipoproteins (HDL) are a negative predictor of platelet-dependent thrombus formation and reduced platelet activation has been observed in vitro in the presence of HDL3, a major HDL fraction. However, mechanisms underlying the anti-thrombotic effects of HDL3 are poorly understood. Scavenger receptors class B represent possible HDL3 binding partners on platelets. We here investigated the role of scavenger receptor class B type I (SR-BI) and CD36 in mediating inhibitory effects of native HDL3 on thrombin-induced platelet activation.
Methods and results
Rhodamine isothiocyanate-labeled HDL3 bound specifically to platelets and HDL3 binding was inhibited by scavenger receptor class B ligands such as phosphatidylserine (PS)- or phosphatidylinositol (PI)-containing liposomes or maleylated albumin (mBSA). By contrast, scavenger receptor class A ligands failed to displace HDL3 from platelets. HDL3, PS- and PI-liposomes, and mBSA inhibited thrombin-induced platelet aggregation, fibrinogen binding, P-selectin expression and mobilization of intracellular Ca2+. In addition, PS- and PI-liposomes emulated HDL3-induced intracellular signaling cascades including diacylglycerol production and protein kinase C activation. The reduction of platelet activation by liposomes was related to their PS or PI content. Moreover, inhibitory effects of native HDL3 were enhanced after enriching lipoproteins with PS, while PS- and PI-poor HDL2 failed to inhibit platelet aggregation and Ca2+ mobilization. Both, HDL3 and PS-containing liposomes failed to inhibit thrombin-induced activation of platelets obtained from SR-BI-deficient mice but not CD36-deficient mice.
Conclusion
We suggest that SR-BI is a functional receptor for native HDL3 on platelets that generates an inhibitory signal for platelet activation. The content of negatively charged phospholipids (PS, PI) in HDL may be an important determinant of their anti-thrombotic potential.
]. Proinflammatory cytokines, chemokines and growth regulatory molecules liberated from platelets promote endothelial dysfunction and alter smooth muscle cell function. In addition, platelets direct leukocyte incorporation into plaques through platelet-mediated leukocyte adhesion. Activated platelets and their aggregates with leukocytes were repeatedly reported in patients with atherothrombotic vascular disease [
High-density lipoproteins (HDL) represent a major anti-atherogenic factor in plasma. Several studies documented inhibitory effects of HDL on the agonist-induced platelet activation both under in vitro and ex vivo conditions [
]. In addition, a strong inverse association exists between plasma HDL levels and recurrent venous thromboembolism, and HDL cholesterol has been identified as an independent predictor of acute platelet thrombus formation [
]. However, mechanisms underlying the inhibitory effects of on platelet activation are not completely understood. We previously demonstrated that HDL3, a major HDL fraction, induce a PKC-dependent inhibition of phosphatidylinositol-specific phospholipase C (PI-PLC) as well as cytoplasm alkalization, ultimately attenuating platelet aggregation, granule secretion and fibrinogen binding in response to thrombin [
HDL3-mediated inhibition of thrombin-induced platelet aggregation and fibrinogen binding occurs via decreased production of phosphoinositide-derived second messengers 1,2-diacylglycerol and inositol 1,4,5-tris-phosphate.
]. The objective of the present study was to specifically investigate the molecular events located upstream to this second messenger cascade in platelets exposed to HDL3. By using functional agonists of the scavenger receptor type B family as well as platelets from animals deficient in CD36 or scavenger receptor B type I (SR-BI) we demonstrate for the first time that native HDL3 modulate platelet activation in a SR-BI-dependent process. Our data further indicate that the content of negatively charged phospholipids such as phosphatidylserine (PS) and phosphatidylinositol (PI) in HDL3 modulates the interaction between HDL3 and SR-BI and thereby determines the anti-thrombotic potential of these lipoproteins.
2. Methods
2.1 Materials
Thrombin, fura 2-AM, rhodamine isothiocyanate (RITC), fluoresceine isothiocyanate (FITC), bovine serum albumin (BSA), phospholipids, and inorganic reagents were from Sigma Chemical Company, Deisenhofen, Germany. [33P]orthophosphoric acid was purchased from NEN-DuPont, Dreieich, Germany. [14C]arachidonic acid (AA) and Sephadex G-25 columns were from GE Healthcare, München, Germany. Silica gel 60 plates and solvents for TLC were obtained from Merck, Darmstadt, Germany. Calphostin C, bis-indolylmaleimide (GF109203X), D609, and GLY-PRO-ARG-PRO (GPRP) were obtained from Merck, Bad Soden, Germany. Highly purified, plasminogen-free fibrinogen was supplied by Enzyme Research Labs. Monoclonal anti-CD62P antibody was from BeckmanCoulter, Krefeld, Germany.
2.2 Animals
Class B, type I scavenger receptor, SR-BI, knockout mice were kindly provided by Dr. M. Krieger (Biology Department, Massachusetts Institute of Technology, Cambridge, MA) [
A targeted mutation in the murine gene encoding the high density lipoprotein (HDL) receptor scavenger receptor class B type I reveals its key role in HDL metabolism.
]. Heterozygous SR-BI knockout mice were crossed to generate homozygous mutant (SR-BI−/−) progeny. The presence of the targeted and/or wild type SR-BI alleles was determined by PCR amplification of DNA extracted from tail biopsies. CD36−/− mice were kindly provided by Dr. M. Febbraio (Department of Medicine, Weill Medical College of Cornell University, New York, USA) [
]. Mice had ad libitum access to water and regular chow diet. All animal protocols used in this study conformed to national law and were approved by the Ethics Committee for Animal Experiments of the Landesamt für Natur, Umwelt und Verbraucherschutz Nordrhein-Westfalen (LANUV).
2.3 Lipoprotein isolation and modification
HDL3 (d = 1.125–1.210 g/mL) and HDL2 (d = 1.063–1.125 g/mL) were isolated from human plasma by discontinuous KBr gradients as described by Havel et al. [
]. HDL fractions were dialyzed against 0.3 mmol/L Tris–HCl, 0.14 mol/L NaCl, 1.0 mmol/L EDTA, pH 7.2, concentrated using the Minicon B15 concentrator (Amicon), and stored in the dark at 4 °C for no longer than 7 days. EDTA was removed by dialysis shortly before experimentation. No effects of storage on the HDL3 electrophoretic mobility and the content of thiobarbituric acid-reactive substances (TBARS) were noted. To enrich HDL3 with anionic phospholipids, PS in amounts corresponding to 0.5% and 1% of estimated total lipoprotein weight was dried under N2 and vigorously vortexed with native HDL3 for 30 min. Lipoproteins were reisolated by ultracentrifugation thereafter. HDL3 was labeled with RITC exactly as described [
]. Briefly, HDL3 (10.0 g/L) were diluted 1:2 (v/v) with 0.2 mol/L Na2HPO4 and 0.02 mg of RITC per 1.0 mg HDL3 was added to the HDL3 solution under continuous stirring. The lipoprotein–RITC mixture was adjusted to pH 9.5 with 0.01 mol/L NaOH. The mixture was incubated in the dark at room temperature for 30 min and RITC–HDL3 were separated from unbound dye on a Sephadex G-25 column and dialyzed against the EDTA-containing buffer, pH 7.2.
Low-density lipoproteins (LDL, d = 1.019–1.063 g/mL) were isolated, dialyzed and stored as described for HDL3. For acetylation, LDL (3.0 mg/mL) were mixed 1:1 (v/v) with saturated sodium acetate on ice under continuous stirring. Thereafter, acetate anhydride was added dropwise within 1 h in a volume (μL) equivalent to protein concentration (μg/mL) × 1.5. The pH was kept constant at 7.4 during the procedure.
HDL-free platelet-rich plasma was obtained by reconstitution of lipoprotein-depleted plasma with homologous very-low-density, intermediary-density, and low-density lipoproteins to original concentrations and with washed platelets (5 × 108/mL).
2.4 Preparation of liposomes and maleylated bovine serum albumin (mBSA)
Chloroform solutions (1 mg/mL) of phosphatidylcholine (PC), PS or PI were mixed in a desired molar ratio, dried under nitrogen and lyophilized. Phospholipids were resuspended in PBS under vigorous vortexing and unilamellar liposomes were produced by extrusion through a polycarbonate membrane (0.1 μm pore diameter). For bovine albumin maleylation, 0.1 mg BSA in 0.2 mol/L Na2B407, pH 8.5 was reacted with an excess maleic anhydride according to a published method [
] and concentrations of PC, PS and PI were determined by high pressure liquid chromatography (HPLC) with light scattering detection exactly as described by Silversand and Haux [
For human platelets, venous blood was drawn from healthy volunteers aged 25–55 years into 5 mL tubes containing 0.5 mL of 0.1 mol/L trisodium citrate solution. For mouse platelets, blood was drawn into silanized capillaries by retro-orbital puncture and anti-coagulated with citrate. Blood was centrifuged at 120 × g for 10 min. Platelet-rich plasma was collected, mixed with acid citrate dextrose (ACD; 6.0 mmol/L citric acid, 12.0 mmol/L trisodium citrate, and 19.4 mmol/L glucose (1:3 v/v)) and centrifuged at 1000 × g for 10 min. Platelets were resuspended in a buffer containing 5.0 mmol/L mmol/L HEPES, 145 mmol/L NaCl, 5.0 mmol/L KCl, 5.0 mmol/L glucose, 1.1 mmol/L NaHCO3, and 0.35% BSA (w/v). Platelets were incubated with indicated concentrations of HDL3, HDL2, liposomes or mBSA for 5 min prior to stimulation with thrombin. Platelet number (5 × 108/mL) was kept constant during experiments.
2.7 Platelet cholesterol content
To determine total cholesterol content in platelets, cells (5 × 108/mL) were extracted 3 × with ethyl acetate (1:1, v/v), and the solvent was evaporated under nitrogen. The residue was reconstituted in PBS containing 0.05% Tween 20 and the cholesterol content was determined using colorimetric method (cholesterol oxidase/peroxidase) on the clinical laboratory analyzer (Advia 1800, Siemens, Germany)
2.8 Flow cytometric analysis of RITC–HDL3 binding on platelet surface
For analysis of HDL3 binding to the platelet surface, we adopted a procedure previously established in our laboratory for leukocytes [
]. Platelets were pre-incubated at 4 °C with HDL3, liposomes, mBSA or the equal volume of buffer for 30 min, incubated with the indicated amount of RITC–HDL3 for 1 h in a total volume of 0.25 mL, washed and analyzed with a FACScan flow cytometer (Becton Dickinson, Heidelberg, Germany) equipped with a 488 nm argon laser. Each value was determined from the signals collected from 5000 cells, and calculations were based on the mean fluorescence.
2.9 Flow cytometric analysis of fibrinogen binding and CD62 expression on the platelet surface
Platelet surface fibrinogen binding and CD62 expression were determined exactly as described previously [
HDL3-mediated inhibition of thrombin-induced platelet aggregation and fibrinogen binding occurs via decreased production of phosphoinositide-derived second messengers 1,2-diacylglycerol and inositol 1,4,5-tris-phosphate.
Optimally functional fluorescein isothiocyanate (FITC)-labelled fibrinogen for quantitative studies of binding to activated platelets and platelet aggregation.
]. Protein/FITC ratio was 1:5. Platelet suspensions (0.2 mL) pre-incubated for 5 min at room temperature with 0.15 g/L fibrinogen-FITC and with either HDL3, liposomes, mBSA or the equivalent volume of buffer were added to 0.02 mL of a solution containing increasing concentrations of thrombin. The reaction was performed in the presence of 1.25 mmol/L GPRP (an inhibitor of fibrin polymerization) and was stopped after 180 s by fixation in 1% formaldehyde for 30 min. For antigen expression analysis, FITC-coupled monoclonal antibody recognizing CD62 was added after fixation for 45 min.
2.10 Platelet aggregation
Platelet aggregation was monitored in the Elvi611 aggregometer adjusted to its maximal sensitivity as described previously [
HDL3-mediated inhibition of thrombin-induced platelet aggregation and fibrinogen binding occurs via decreased production of phosphoinositide-derived second messengers 1,2-diacylglycerol and inositol 1,4,5-tris-phosphate.
HDL3-mediated inhibition of thrombin-induced platelet aggregation and fibrinogen binding occurs via decreased production of phosphoinositide-derived second messengers 1,2-diacylglycerol and inositol 1,4,5-tris-phosphate.
HDL3-mediated inhibition of thrombin-induced platelet aggregation and fibrinogen binding occurs via decreased production of phosphoinositide-derived second messengers 1,2-diacylglycerol and inositol 1,4,5-tris-phosphate.
]. Briefly, platelets were incubated for 60 min with 0.1 μCi/mL [14C]arachidonic acid at 30 °C. After stimulation with thrombin, aliquots (0.5 mL) were withdrawn and added to 1.5 mL of ice-cold chloroform/methanol (2:1, v/v). Radiolabeled lipids were extracted according to Bligh and Dyer [
]. Plates were analyzed with a BAS 1500 phosphoimager (Fuji Film).
2.13 Determination of protein phosphorylation
Platelets were labeled with 0.1 mCi/mL [33P]orthophosphoric acid for 90 min at 37 °C, washed, and adjusted to 1 × 109/mL. After stimulation with thrombin, the reactions were terminated by adding 10% TCA (1:1 v/v). The precipitate was washed twice, dissolved in 0.1 mL of NaOH (0.2 mol/L), mixed with an equal volume of double-concentrated sample buffer, and subjected to SDS-PAGE. Dried gels were analyzed by phosphoimaging.
2.14 Determination of thrombin activity
Thrombin at indicated concentrations was incubated with liposomes, HDL3 or hirudin (Loxo, Dossenheim, Germany) as positive control for 5 min. Thereafter the thrombin substrate Pefachrome TH (Pentapharm, Basel, Switzerland) was added and the optical density was measured at 405 nm using Tecan Sunrise ELISA reader (Männedorf, Switzerland). Data were calculated as the maximum slope (OD/min).
2.15 Statistical analysis
Data are presented as means ± SD from three separate experiments or as results representative for at least three repetitions, unless indicated otherwise. Comparisons between the groups were performed with two-tailed Student's t-test.
3. Results
3.1 Scavenger receptors class B (SR-B) mediate HDL3 binding to human platelets
To examine the possible role of SR-B as HDL3 receptor on human platelets, we first conducted equilibrium ligand binding analysis using RITC–HDL3 as previously described for leukocytes [
]. As shown in Fig. 1A , increasing amounts of RITC–HDL3 bound to platelets in a saturable fashion. Scatchard analysis of the specific binding curve revealed the dissociation constant (Kd) of 68.42 μg of HDL3 protein/mL close to those reported for the binding of HDL3 to SR-BI, when expressed on mammalian cells [
]. The binding specificity of HDL3 to platelets was further explored by competition experiments with unlabeled SR-B ligands. As expected, native HDL3 effectively suppressed binding of RITC–HDL3 to platelets (Fig. 1B). Similarly, liposomes containing anionic phospholipids PS or PI, which serve as ligands for SR-BI and CD36 [
], inhibited a significant portion of RITC–HDL3 binding to platelets (−50% inhibition at −1.0 g/L native HDL3; Fig. 1C). By contrast, liposomes containing only zwitterionic PC were ineffective at concentrations as high as 5.0 g/L. The propensity of liposomes to displace RITC–HDL3 from platelets was dependent on their anionic phospholipid content with the maximal effect seen at equimolar PC:PS or PC:PI ratios (Fig. 1D). In addition, RITC–HDL3 bound to platelets could be displaced by mBSA, a lipid-free SR-B ligand [
], but not by unmodified BSA (Fig. 1E). As scavenger receptors type A (SR-A) are also expressed on platelets, we tested the effects of SR-A ligands on RITC–HDL3 binding. As shown in Fig. 1F, neither polyinosinic acid and polyadenylic acid, nor fucoidan – specific SR-A ligands – affected RITC–HDL3 binding to human platelets. By contrast, RITC–HDL3 binding was effectively inhibited by acetylated LDL (acLDL), which serves as a binding partner for both SR-A and SR-B receptors (Fig. 1F) [
Fig. 1Scavenger receptor type B but not type A ligands interfere with HDL3 binding to human platelets. (A) Human washed platelets were incubated with increasing concentrations of RITC–HDL3 in the presence (○) or absence (□) of excess unlabeled HDL. (■)–specific HDL3 binding. Data represent means from 3 independent experiments. (B–F) Displacement of RITC–HDL3 from platelets by (B) native HDL3 (C) and (D) PC- (○), PS- (▾), or PI- (▴)-containing liposomes (E) mBSA (♦) or BSA (♢) (F) acetylated LDL (●), poly(I) (○), poly(A) (□), and fucoidan (♢). Data from panels B to F represent mean ± SD from 3 to 5 independent experiments. *p < 0.01; **p < 0.001 PC- vs. PS- or PI-containing liposomes (panel B) or mBSA vs. BSA (panel E).
3.2 SR-B ligands inhibit human platelet activation
To determine whether the engagement of SR-B on the cell surface results in modification of platelet functions, we tested the effects of HDL3, PS- and PI-containing liposomes as well as mBSA on thrombin-induced aggregation and mobilization of intracellular calcium. As shown in Fig. 2A and C , addition of thrombin (0.02 U/mL) to the platelet suspension resulted in parallel aggregation and [Ca2+] elevation and both effects were inhibited after pre-incubation of platelets for 5 min with HDL3 at physiological concentrations (1.0 g/L). Both, PS- and PI-containing liposomes (0.5 g/L, PC:PS and PC:PI ratio 1:1, mol/mol) and mBSA (0.5 g/L) pre-incubated with platelets for 5 min fully emulated the ability of HDL3 to inhibit thrombin-induced aggregation and [Ca2+] elevation, whereas PC-liposomes and unmodified BSA (not shown) exerted no effect. The capacity of liposomes to inhibit platelet aggregation and calcium mobilization was dependent on anionic phospholipids and reached the maximum at equimolar PC:PS or PC:PI ratios (Fig. 2B and D).
Fig. 2Scavenger receptor type B ligands inhibit thrombin-induced activation of human platelets. Human washed platelets (A–F) or HDL-deprived platelet-rich plasma (G and H) were stimulated with 0.02 U/mL thrombin or with increasing concentrations of thrombin (Thr, ○) in the absence or presence of HDL3 (1.0 g/L, □), (PC-liposomes (0.5 g/L)), PS-liposomes (0.5 g/L, ▾), PI-liposomes (0.5 g/L ▴), or mBSA (0.5 g/L, ♦). (A and B) Aggregation, (C and D) Ca2+ mobilization, (E and G) fibrinogen binding, (F and H) CD62 expression. Panels A and C show original tracings superimposed for comparison. Maximal stimulation with thrombin was set as 100%. Data from panels B, D, E, F, G and H represent mean ± SD from 3 to 6 independent experiments. *p < 0.01; **p < 0.001 thrombin vs. thrombin + HDL3, +liposomes or +mBSA. Panel E: significant differences between thrombin and thrombin + HDL3, +liposomes or +mBSA were observed at 0.01 U/mL (p < 0.01), 0.03 U/mL (p < 0.01), and 0.05 U/mL (p < 0.01). Panel F: significant differences between thrombin and thrombin + HDL3, +liposomes or +mBSA were observed at 0.01 U/mL (p < 0.05), 0.03 U/mL (p < 0.01), and 0.05 U/mL (HDL3 and PI-liposomes – p < 0.05; mBSA – p < 0.01). Panel G: significant differences between thrombin and thrombin + HDL3 or +liposomes were observed at 0.05 U/mL (p < 0.01) and 0.1 (p < 0.01; HDL3 and PI-liposomes – p < 0.05). Panel H: significant differences between thrombin and thrombin + HDL3 or +liposomes were observed at 0.05 U/mL (p < 0.05) and 0.1 (p < 0.01).
Platelet activation is associated with the conformational change of the integrin αIIbβ3 enabling fibrinogen binding and the secretion of granular content accompanied by exposure of activation-dependent glycoproteins such as P-selectin (CD62) on the platelet surface. Therefore, we next compared the effects of HDL3 and SR-B ligands on thrombin-induced fibrinogen binding and CD62 surface expression. As shown in Fig. 2E and F, exposure of platelets to increasing concentrations of thrombin increased the surface binding of FITC-conjugated fibrinogen and anti-CD62 with the binding maxima observed at agonist concentrations between 0.01 U/mL and 0.05 U/mL. Pre-incubation of platelets with HDL3 (1.0 g/L), PS- or PI-containing liposomes (0.5 g/L, PC:PS and PC:PI ratio 1:1, mol/mol), or mBSA (0.5 g/L) for 5 min prior to stimulation with the agonist significantly reduced both fibrinogen binding and CD62 surface exposure at lower thrombin concentrations (up to 0.05 U/mL). At higher thrombin concentrations, however, both HDL3 and SR-B agonists exerted no appreciable inhibitory effect.
We next examined whether the inhibitory effects of SR-B agonists on platelet activation are also exerted in a more physiological environment. To this aim, we determined the effect of HDL3 as well as PS- and PI-containing liposomes on fibrinogen binding and CD62 surface exposure in platelet-rich plasma deprived of HDL3. As shown in Fig. 2G and H, reconstitution of HDL3-free plasma with native lipoproteins reduced both thrombin-induced fibrinogen binding and CD62 surface exposure at lower agonist concentrations. Similar inhibitory effects were observed, when PS- or PI-containing liposomes were used instead of HDL3.
To address the possibility that SR-B agonists inhibit platelet activation by directly binding thrombin and thereby diminishing enzyme activity, the effect of HDL3, PS- or PI-containing liposomes on thrombin-mediated degradation of the specific enzyme substrate was examined. As shown in Table 1, neither native lipoproteins nor liposomes reduced the enzymatic activity of thrombin.
Table 1Effect of SR-B agonists on the thrombin activity.
Thrombin activity (optical density at 450 nm/min)
Thrombin concentrations (U/mL)
0.05
0.1
0.2
Control
0.112 ± 0.001
0.153 ± 0.002
0.236 ± 0.005
HDL3 (1.0 mg/mL)
0.152 ± 0.003
0.235 ± 0.004
0.359 ± 0.008
PS-liposomes (0.5 mg/mL)
0.150 ± 0.002
0.188 ± 0.001
0.345 ± 0.006
PI-liposomes (0.5 mg/mL)
0.131 ± 0.003
0.198 ± 0.003
0.318 ± 0.006
Hirudin (10 U/mL)
0.060 ± 0.001
0.061 ± 0.001
0.060 ± 0.001
Values represent means ± SD from three determinations.
Finally, we sought to examine whether the negative effect of SR-B agonists on platelets could be mediated by their cholesterol efflux capacity. However, neither native HDL3 nor PS- or PI-containing liposomes decreased the platelet total cholesterol content during the 5 min incubation period, which was sufficient to produce inhibitory effects on platelet activation (not shown).
3.3 The inhibitory effect of HDL3 on platelet activation is enhanced by PS enrichment
As pointed above, anionic phospholipids serve as SR-B ligands and determine the inhibitory effects of liposomes on platelet activation. To specifically address the significance of anionic phospholipids for HDL3–platelet interactions, we enriched HDL3 with various amounts of PS. The anionic phospholipid (PS + PI) content in HDL3 particles increased from 23.3 ± 2.7 μg/mg protein in native lipoproteins to 43.9 ± 4.5 μg/mg protein and 72.8 ± 5.1 μg/mg protein in HDL3 enriched with 0.5% (w/w) and 1.0% (w/w) PS, respectively. By contrast, the content of phosphatidylcholine was not significantly changed. When incubated for 5 min with the platelet suspension prior to stimulation with thrombin (0.02 U/mL), PS-enriched HDL3 (0.5 g/L) more potently inhibited platelet aggregation and intracellular calcium mobilization than native HDL3 (Fig. 3A and B) . Moreover, the capacity of PS-enriched HDL3 to attenuate thrombin-induced [Ca2+]i elevation was related to the PS content in lipoprotein particles (Fig. 3C and D).
Fig. 3Enrichment in phosphatidylserine (PS) potentiates inhibitory effect of HDL3 on the thrombin-induced activation of human platelets. (A–D) Human washed platelets were stimulated with thrombin (0.02 U/mL) in the presence or absence of native HDL3 (0.5 g/L) or HDL3 enriched with 0.5% (w/w) PS (HDL3-PS/0.5) or 1% (w/w) PS (HDL3-PS/1.0). (E–F) Platelets were pre-incubated with native HDL3 or HDL2 (1.0 g/l) prior to thrombin stimulation. Panels A and C show original tracings superimposed for comparison. Maximal stimulation with thrombin was set as 100%. Data from panels B and D–F represent means ± SD from 3 independent experiments. *p < 0.05; **p < 0.01 ***p < 0.001 thrombin vs. thrombin + HDL3 +HDL2 or +PS-enriched HDL3.
As HDL subfractions contain variable amounts of phospholipids, we additionally determined the PS content in HDL2. This HDL subfraction contains substantially less anionic phospholipids (PS + PI) than HDL3 (5.3 ± 0.9 μg/mg protein), but a comparable amount of PC. Moreover, thrombin-induced platelet aggregation and [Ca2+]i elevation were not significantly changed in platelets exposed to HDL2 (Fig. 3E and F).
3.4 SR-B ligands promote DAG release and activate PKC in human platelets
We previously demonstrated that the phosphatidylcholine-specific phospholipase C (PC-PLC)-dependent release of DAG and the ensuing PKC activation are prerequisites for attenuated functional responses of platelets exposed to HDL3 [
HDL3-mediated inhibition of thrombin-induced platelet aggregation and fibrinogen binding occurs via decreased production of phosphoinositide-derived second messengers 1,2-diacylglycerol and inositol 1,4,5-tris-phosphate.
]. Therefore, we next compared the effects of HDL3 and SR-B agonists on DAG production and PKC activity. In agreement with previous observations, physiological concentrations of HDL3 (1.0 g/L) triggered the formation of [14C]DAG in [14C]AA-labeled platelets (Fig. 4A) . Both PS- and PI-containing liposomes (0.5 g/L, PC:PS and PI:PC ratio 1:1, mol/mol) fully emulated the stimulating effect of HDL3 on [14C]DAG production, while liposomes containing only PC were without effect. Moreover, incubation of platelets for 1 min with HDL3, PS- or PI-containing liposomes resulted in phosphorylation of the 40–47 kD major protein substrate, which is considered to be a marker of PKC activation in platelets (Fig. 4B). To assess the relevance of the DAG-dependent PKC activation for the inhibitory effects exerted by PS- or PI-containing liposomes on platelet activation, platelet suspensions were pre-incubated with D609 (20.0 μmol/L), a PC-PLC inhibitor, or with staurosporin (5.0 μmol/L) or bis-indoylmaleylimide (5.0 μmol/L), two structurally distinct PKC inhibitors, prior to exposure to HDL3 or liposomes and stimulation with thrombin (0.02 U/mL). As shown in Fig. 4C, the inhibitory effects of both HDL3 and PS- or PI-containing liposomes on thrombin-induced Ca2+ mobilization were abolished in platelets pre-treated with PC-PLC or PKC inhibitors.
Fig. 4Scavenger receptor type B ligands induce intracellular signaling cascades in human platelets. (A and B) Human washed platelets labeled with [14C]arachidonic acid or [33P]orthophosphoric acid were incubated in the presence of HDL3 (1.0 g/L, □), PC-liposomes (0.5 g/L, ○), PS-liposomes (0.5 g/L,▾) or PI-liposomes (0.5 g/L, ▴) and analyzed for [14C]DAG (A) and protein phophorylation (1 min incubation; B) as described under Section 2. The arrow indicates p43 PKC phosphorylation substrate. Data represent mean ± SD or are representative for 3 independent experiments. (C) Human platelets pre-incubated for 10 min with D609 (20 μmol/L), staurosporine (Str, 10 μmol/L), or bisindoylmaleimide (Bis, 1.0 μmol/L) were stimulated with thrombin (0.02 U/mL) in the absence or presence of HDL3, PS-liposomes or PI-liposomes and Ca2+ mobilization was determined. as described under Section 2. Data represent mean ± SD from 3–5 independent experiments. *p < 0.05; **p < 0.01 thrombin + HDL3 vs. thrombin + HDL3 + inhibitor.
3.5 SR-BI mediates the inhibition of platelet activation by HDL3 and SR-B agonists in mouse platelets
HDL3 and anionic liposomes serve as ligands for SR-BI and CD36, which are both expressed on the platelet surface. To specifically address the identity of class B scavenger receptors mediating the HDL3- or anionic liposomes-dependent inhibition of platelet activation, we initially examined the effects of PS-containing liposomes on FITC-conjugated fibrinogen and anti-CD62 binding in mouse platelet-rich plasma (PRP). As shown in Fig. 5A–F , addition of thrombin to PRP promoted fibrinogen binding and CD62 surface expression in a concentration-dependent fashion in platelets from wild type (WT), SR-BI−/− and CD36−/− mice. Pre-incubation of PRP for 5 min with PS-containing liposomes (0.5 g/L, PC:PS ratio 1:1, mol/mol) significantly attenuated platelet activation in PRP from WT and CD36−/− mice but not SR-BI−/− mice. Next, we examined the influence of native HDL3 and PS-containing liposomes on the thrombin-induced fibrinogen and anti-CD62 binding in washed murine platelets. Fig. 5G–J demonstrates that both responses were stimulated by thrombin in a concentration-dependent manner and attenuated in WT platelets pre-incubated for 5 min with either HDL3 (1.0 g/L) or PS-containing liposomes (0.5 g/L, PC:PS ratio 1:1, mol/mol). By contrast, both HDL3 and PS-containing liposomes failed to affect thrombin-induced fibrinogen and anti-CD62 binding in washed platelets obtained from SR-BI−/− mice.
Fig. 5Inhibitory effect of PS-liposomes on the thrombin-induced activation of murine platelets is abolished in SR-BI deficiency. Platelet-rich plasmas (PRP) from wild type (A and B), SR-BI−/− (C and D) or CD36−/− (E and F) mice or washed platelets from wild type (G and H) or SR-BI−/− (I and J) mice were incubated with increasing concentrations of thrombin in the absence (○) or presence of HDL3 (1.0 g/L, □) or PS-liposomes (0.5 g/L, ▾). A, C, E, G, I fibrinogen binding, B, D, F, H, J CD62 expression. Data represent mean ± SD from 4–6 independent experiments. #p < 0.05 HDL3 vs. control, *p < 0.05 PS-liposomes vs. control.
Most work to date strongly supports a direct inhibitory effect of HDL or their major fraction, HDL3, on platelet activation and the subsequent formation of venous and arterial thrombi [
]. However, the mechanism underlying the anti-platelet effect of HDL remains enigmatic. Our combined data provide several pieces of evidence consistent with the direct involvement of scavenger receptor type BI in the interaction between HDL and platelets. First, structurally distinct synthetic SR-BI ligands interfered with HDL3 binding to platelets, while ligands of scavenger receptor type A were ineffective in this respect. Second, SR-BI ligands fully mimicked the biological effects of HDL3 with regard to inhibition of thrombin-induced washed platelet aggregation, mobilization of intracellular calcium, granule secretion, and binding of fibrinogen to the platelet surface. In addition, synthetic SR-BI ligands substituted native HDL with respect to platelet activation in platelet-rich plasma. Third, the intracellular signaling cascade comprising the DAG liberation from phosphatidylcholine and PKC activation was triggered equally effectively in platelets exposed to HDL3 or SR-BI ligands. Fourth, the inhibitory effects of HDL3 and synthetic SR-BI ligands on thrombin-induced fibrinogen binding and granule secretion were completely abolished in SR-BI-deficient murine platelets. By contrast, these effects were preserved in platelets deficient in CD36. Taken together, these findings corroborate the hypothesis that SR-BI is a functional HDL3 receptor on platelets mediating the inhibitory effects of these lipoproteins on agonist-induced platelet activation.
Previous studies regarding the identity of the HDL receptor on the platelet surface led to conflicting results. Early ligand-blotting experiments suggested that the integrin αIIbβ3 may serve as an HDL binding partner [
]. However, observations demonstrating unaltered binding of HDL particles to platelets in the presence of anti-αIIbβ3 antibodies or αIIbβ3 ligands such as vitronectin and von Willebrand factor argued against a role of this integrin as a platelet receptor for HDL [
]. A splice variant of ApoER2/LRP8, a member of the LDL receptor family, has been proposed as a platelet receptor for the apolipoprotein E (apoE)-rich HDL subfraction [
]. However, HDL obtained from apoE-deficient individuals retained an inhibitory effect on platelet activation thereby excluding apoER2/LRP8 as the sole HDL receptor on platelets [
]. SR-BI expression on the platelet surface has been identified in two independent studies but no formal evidence documenting its role in mediating HDL–platelet interactions has been presented to date [
]. The HDL3 binding characteristics and the results of competition experiments shown here for the first time point to scavenger receptor class B as a binding partner for native HDL3 on human platelets. Moreover, the results of functional studies documenting abolished inhibitory effects of HDL3 on agonist-stimulated fibrinogen binding and granule secretion in murine platelets lacking SR-BI provide unequivocal evidence that this receptor is actually required for native HDL3 to inhibit platelet activation. Our findings are in concordance with previous results of Imachi et al., who noted a strong inverse relationship between the SR-BI abundance in human platelets and the extent of ADP-induced platelet aggregation [
]. Together with the results of the present study, these findings suggest that HDL3 exert a continuous suppressing effect on platelet reactivity via SR-BI and may provide a mechanistic explanation for the increased incidence of thrombosis in subjects with hypo-α-lipoproteinemia.
In addition to SR-BI, HDL has been demonstrated to serve as a ligand for CD36 another member of the class B scavenger receptor superfamily present on the platelet surface. However, the absence of CD36 in platelets failed to influence the inhibitory effect of HDL3 on thrombin-induced platelet fibrinogen binding and granule secretion. CD36 on macrophages facilitates cellular uptake of oxidized lipoproteins and therefore is believed to play a pro-atherogenic role, while SR-BI is atheroprotective by virtue of participating in the reverse cholesterol transport. CD36 on platelets has been previously reported to bind both oxidized LDL and/or phospholipids thereby resulting in the enhanced platelet response to agonists [
]. By demonstrating blunted platelet reactivity as a consequence of SR-BI engagement, the present study suggests that both class B scavenger receptors may play an opposite role not only in context of atherosclerosis but also thrombosis: while CD36 promotes a prothrombotic phenotype, SR-BI might counteract thrombus formation by suppressing platelet activation.
Recently, Valijaveettil et al. reported that oxidatively modified HDL (oxHDL) act as a potent inhibitor of platelet activation and aggregation by physiologic agonists and, based on the results of experiments with SR-BI-deficient platelets or anti-SR-BI blocking antibody, attributed these effects to SR-BI [
]. These findings are in major contradiction to previously published data, which demonstrated that Cu2+- or hypochlorite-oxidized HDL autonomously induce platelet aggregation or at least enhance aggregation brought about by other agonists [
]. In addition, increased granule secretion and elevated intracellular Ca2+ levels were observed in platelets exposed to oxHDL. The results of the present study demonstrating that the negative charge of the ligand, both native (HDL3) and synthetic (liposomes), is critical to SR-BI-mediated suppression of platelet reactivity may help to at least partially resolve this discrepancy. Conceivably, exposure of HDL to more mild oxidative conditions, such as short-term incubation with Cu2+, preferentially leads to generation of oxidized phospholipids, which would be expected to increase the negative charge of HDL particles and thereby their affinity to SR-BI and the propensity to suppress platelet reactivity. On the other hand, strong HDL oxidation by hypochlorite might alter the apolipoprotein moiety of HDL particles generating an effective ligand for the prothrombotic CD36.
Recently, Calkin et al. found that reconstituted HDL particles (rHDL) containing apoA-I and phospholipids dramatically reduced platelet free cholesterol content and inhibited the ADP-induced platelet aggregation and granule secretion in a receptor-independent fashion [
]. However, native HDL particles were much less effective in removing cholesterol from platelets and in much contrast to the present study failed to affect platelet aggregation in response to ADP. While reasons for these discrepancies are currently unclear, it has to be stressed that native HDL in concentrations considerably below physiological levels were used by Calkin et al. In addition, these authors used rHDL pre-incubation times between 30 min and 2 h, which were apparently necessary to effectively deplete platelet cholesterol. It cannot be excluded, therefore, that anti-platelet effects of native and reconstituted HDL particles are mediated by entirely distinct mechanisms: while native HDL influences platelet function via interaction with specific receptors and induction of intracellular signaling cascades, rHDL decreases platelet reactivity by promoting cholesterol efflux and altering physicochemical properties of the plasma membrane.
The recognition that the negative charge is a critical determinant of the HDL-mediated suppression of platelet reactivity may have far-reaching consequences for understanding the anti-thrombotic effects attributed to these lipoproteins. The results of the present study predict that HDL particles/subclasses abundant in phosphatidylserine and/or phosphatidylinositol are more efficacious in preventing platelet activation and thereby might better countervail the formation of intravascular thrombi. Actually, only minor inhibitory effects on thrombin-induced platelet activation were seen in the presence of HDL2, which is a lipoprotein subfraction characterized by a low amount of negatively charged phospholipids. Differences in phosphatidylserine and/or phosphatidylinositol content in HDL particles might also account for the individual susceptibility to thrombosis in subjects with low HDL concentrations. Unfortunately, concentrations of anionic phospholipids in patients with atherothrombotic vascular disease have not been examined to date. Likewise, little is known about molecular mechanisms controlling phosphatidylserine and/or phosphatidylinositol transfer to and their exchange among HDL particles. Delineation of these mechanisms might improve the assessment of the risk of thrombosis and provide a platform for development of new drugs for its treatment.
In conclusion, the results of the present study suggest that SR-BI is a functional receptor for native HDL3 on platelets that generates inhibitory signals for platelet activation. The content of negatively charged phospholipids in HDL may be an important determinant of their anti-thrombotic potential.
Acknowledgements
This work was supported by Innovative Medizinische Forschung (IMF) Foundation (grant no 119827 to JRN) and The Netherlands Heart Foundation (grant 2006B107 to SJAK). M.V.E. is an Established Investigator of The Netherlands Heart Foundation (grant 2007T056). U.J.F.T. is supported by the Netherlands Organization for Scientific Research (VIDI grant 917-56-358).
References
Weber C.
Platelets and chemokines in atherosclerosis: partners in crime.
HDL3-mediated inhibition of thrombin-induced platelet aggregation and fibrinogen binding occurs via decreased production of phosphoinositide-derived second messengers 1,2-diacylglycerol and inositol 1,4,5-tris-phosphate.
A targeted mutation in the murine gene encoding the high density lipoprotein (HDL) receptor scavenger receptor class B type I reveals its key role in HDL metabolism.
Optimally functional fluorescein isothiocyanate (FITC)-labelled fibrinogen for quantitative studies of binding to activated platelets and platelet aggregation.
00043-8&id=gr1.jpg){kind=link}
00043-8&id=gr2.gif){kind=link}
00043-8&id=gr3.gif){kind=link}
00043-8&id=gr4.jpg){kind=link}
00043-8&id=gr5.jpg){kind=link}