Author + information
- Received November 18, 1997
- Revision received August 10, 1998
- Accepted September 10, 1998
- Published online January 1, 1999.
- David J Schneider, MD, FACCa,* (, )
- Douglas J Taatjes, PhDb,
- Diantha B Howard, MSa and
- Burton E Sobel, MD, FACCa
- ↵*Address for correspondence: Dr. David J. Schneider, Cardiovascular Division, Colchester Research Facility, University of Vermont, 55A South Park Drive, Colchester, Vermont 05446
Objectives. To determine whether augmented activation (degranulation) of platelets might contribute to the association between higher concentrations of fibrinogen and risk of myocardial infarction, we characterized adenosine diphosphate (ADP)-induced expression of P-selectin by platelets in whole blood as a function of this exposure to selected concentrations of fibrinogen.
Background. An increased risk of myocardial infarction has been associated with increased concentrations of fibrinogen.
Methods. Fibrinogen was added to blood anticoagulated with corn trypsin inhibitor (a specific inhibitor of Factor XIIa without effect on other coagulation factors). Degranulation of platelets was identified by flow cytometry.
Results. Addition of fibrinogen to blood did not activate platelets under basal conditions (without ADP). By contrast, a concentration-dependent increase in ADP and thrombin receptor agonist peptide (TRAP)-induced activation occurred with increasing concentrations of fibrinogen. Increased ADP-induced degranulation was apparent with the addition of 100 mg/dl of fibrinogen (p ≤ 0.001 for 1.5 μmol/liter ADP, n = 10 subjects). Inhibition by abciximab of binding of fibrinogen to the surface glycoprotein IIb-IIIa did not attenuate the observed augmentation of reactivity induced by fibrinogen. Augmented degranulation was associated with uptake of fibrinogen into α-granules without surface binding despite pretreatment with abciximab as shown by laser scanning confocal microscopy.
Conclusions. Fibrinogen in blood augments degranulation of platelets in response to ADP and is accompanied by uptake of fibrinogen into α-granules. Thus, elevated concentrations of fibrinogen secondary to inflammation implicated in cardiovascular risk may operate, in part, by increasing reactivity of platelets.
Increased concentrations of fibrinogen in blood have been associated with an increased risk of cardiovascular events independent of other known risk factors (1–5). Its positive predictive value for cardiovascular events exceeded that for cholesterol in the Northwick Park Heart Study (3). One proposed mechanism is a predisposition to exuberant thrombosis complicating atherosclerotic plaque rupture because increased concentrations of fibrinogen have been associated with an increased deposition of fibrin accompanying activation of coagulation factors (6,7). In addition, increased fibrinogen may predispose to aggregation of platelets secondary to surface glycoprotein binding and cross-linking of activated platelets (8).
Activation of platelets is complex and entails 1) a conformational change in the surface glycoprotein IIb-IIIa that exposes a binding site for fibrinogen on the activated conformer that facilitates aggregation (9); 2) degranulation, involving release of the contents of α-granules and of dense granules (not necessarily concordantly); 3) a procoagulant response involving exposure of negatively charged phospholipids on the surface of the platelet that facilitates assembly of coagulation factor complexes tenase and prothrombinase (10); and 4) change in the shape of the platelet with pseudopod extension. The α-granule degranulation occurs by fusion of the vesicles with the plasma membrane and leads to exposure on the platelet surface of membrane-bound granular proteins including P-selectin (11).
We have recently characterized a method to define activation of platelets in whole blood with the use of flow cytometry (12). We have found that the conditions used facilitate identification of agents that alter the reactivity of platelets. Because of the possibility that fibrinogen would alter platelet reactivity independent of its capacity to aggregate already activated platelets through binding to glycoprotein IIb-IIIa, the present studies were performed to characterize effects of fibrinogen on the threshold for activation of platelets. Activation of platelets was characterized by α-granule degranulation reflected by surface expression of P-selectin. Our studies were designed to determine 1) whether elevated concentrations of fibrinogen in blood alter reactivity of platelets in addition to the augmentation of aggregation of previously activated platelets, and 2) whether the threshold for degranulation is lowered despite inhibition of binding of fibrinogen to the surface glycoprotein IIb-IIIa by abciximab (ReoPro).
Blood was obtained from 19 human subjects that had not taken aspirin or any nonsteroidal anti-inflammatory drugs for at least 10 days. All subjects gave written informed consent before phlebotomy. The study protocol was approved by the Institutional Review Board of the University of Vermont.
Phlebotomy was performed with the use of a two syringe technique whereby the first 3 ml of blood was discarded. Activation of platelets was delineated in blood drawn directly into a syringe containing 32 μg/ml of corn trypsin inhibitor (CTI, 9:1 v/v), an agent that inhibits Factor XIIa without affecting other coagulation factors (12,13). Unlike CTI, other anticoagulants, particularly chelators of calcium such as citrate, alter activation of platelets (12). Additional aliquots of blood were drawn into syringes containing CTI (32 μg/ml) plus 4 μg/ml, 10 μg/ml and 100 μg/ml of ReoPro (Centocor BV, Leiden, Netherlands, 9:1 v/v). The concentration of fibrinogen was assayed by the method of Clauss (14)in plasma anticoagulated with sodium citrate (0.129 mol/liter, pH 6.0) that was stored at −80°C until assay.
Activation of platelets
Fibrinogen (free of plasminogen and von Willebrand Factor, homogeneous by SDS–PAGE, and greater than 95% clottable; Enzyme Research, South Bend, Indiana) was dialyzed exhaustively against phosphate buffered saline (PBS) to remove citrate. Whole blood was treated with PBS (control) and with PBS plus fibrinogen to increase the concentration of fibrinogen in selected aliquots by 50 mg/dl, 100 mg/dl, 200 mg/dl, 300 mg/dl or 500 mg/dl. Albumin and ferritin (Sigma, St. Louis, Missouri) were resuspended in PBS, and molar concentrations equivalent to 100 mg/dl, 200 mg/dl and 500 mg/dl of fibrinogen were added to blood to determine the specificity of effects of fibrinogen.
Activation of platelets was induced as described previously (12). Briefly, 5 μl of blood was added to microcentrifuge tubes containing HEPES-Tyrode’s buffer (5 mmol/liter HEPES, 137 mmol/liter NaCl, 2.7 mmol/liter NaHCO3, 0.36 mmol/liter NaH2PO4, 2 mmol/liter CaCl2, 4 mmol/liter MgCl2and 5 mmol/liter dextrose, pH 7.4), a fluorescein isothiocyanate (FITC)-conjugated IgG directed against glycoprotein IIb-IIIa (HP1-1D, 4 μg/ml), a phycoerythrin (PE)-conjugated IgG directed against P-selectin (Becton Dickinson, San Jose, California, 0.6 μg/ml), and selected concentrations of adenosine diphosphate (ADP) (0, 0.2, 0.8 and 1.5 μmol/liter) and thrombin receptor agonist peptide (20 μmol/liter thrombin receptor agonist peptide [TRAP], Bachem). The HP1-1D binds to an activation-independent epitope on glycoprotein IIb-IIIa and was used to identify platelets. The P-selectin expression was used identify α-granule degranulation (11). Blood was added to the reaction tubes within 15 min after phlebotomy.
After addition of blood, the reaction solution was mixed gently and incubated at room temperature for 15 min. Subsequently, the platelets were fixed, and the red blood cells were lysed by addition of Optilyse-C (Immunotech, Westbrook, Maine). Thus, platelets were exposed to the added fibrinogen for approximately 30 min before fixation. Control tubes (containing FITC-HP1-1D and nonfractionated PE-IgG) were used for detection of nonspecific antibody association with platelets. The association of antibodies with platelets was detected with the use of a fluorescence-activated cell sorter (Becton Dickinson) as previously described (12).
Binding of fibrinogen to platelets was detected with the use of flow cytometry. Fibrinogen was labeled with FITC as has been described previously (15). The less stringent labeling conditions used with celite FITC (Calbiochem, La Jolla, California) preserve functional activity of fibrinogen and do not alter binding of fibrinogen to activated platelets (15). Binding of fibrinogen was detected by the addition of 5 μl of blood to microcentrifuge tubes containing HEPES-Tyrode’s buffer, FITC-fibrinogen (0.1 mg/ml), a PerCP conjugated IgG (2.3 μg/ml) directed against glycoprotein IIIa (anti-CD61, Becton Dickinson), and selected concentrations of ADP (0, 0.2, 0.8 and 1.5 μmol/liter). Binding of anti-CD61 to glycoprotein IIIa is independent of activation and does not interfere with binding of fibrinogen to platelets (information from supplier). The population of platelets identified by HP1-1D is identical to that identified by anti-CD61. After addition of blood to the reagent tubes, samples were mixed gently and incubated for 15 min at room temperature. Platelets were fixed, and red blood cells were lysed with Optilyse C. Control tubes with albumin-FITC and anti-CD61 PerCP were used to detect nonspecific association of proteins with platelets.
Confocal laser scanning microscopy
The association with and localization in platelets of exogenous fibrinogen were determined in blood that was anticoagulated with CTI or with CTI spiked with ReoPro (10 μg/ml) and exposed to 300 mg/dl of FITC-fibrinogen for 15 min. Platelets were fixed and red blood cells were lysed with Optilyse C. The plasma membranes of the platelets were rendered permeable by addition of 0.1% Triton X-100 (Sigma) and anti-CD62 (P-selectin) was added. The platelet suspension was incubated for 15 min and then applied to a glass microscopic slide for 15 min. The slide was then washed and an Alexa 568 conjugated goat anti-mouse IgG (Molecular Probes, Eugene, Oregon) was applied for 30 min. After additional washes the samples were air-dried and a cover-slip was applied with 1% n-propyl gallate (Sigma) in 50% glycerol: 50% PBS. The slides were evaluated with a BioRad (Hercules, California) MRC 1000 confocal scanning laser system equipped with a krypton/argon laser mounted on an Olympus BX50 microscope.
Platelets were imaged with the use of a 100X phase-contrast oil immersion lens (numerical aperture = 1.3) and an electronic zoom factor of 2.4. Phase-contrast (nonconfocal) images were acquired with a transmitted light detector attachment. The FITC-fibrinogen was visualized with the use of 488-nm laser excitation, and the Alexa 568 conjugated secondary antibody localizing P-selectin was visualized with 568-nm laser excitation. Co-localization analysis was performed with the multiply function in Macro Programming Language Software (BioRad). This function multiplies each pixel in the active display box by the corresponding pixel of a second image. After multiplication, the resulting image displays color only in areas where fluorescence was present in both images to co-localize two fluorochromes.
Analysis of data
Values are means ± SEM. Repeated measures analysis of variance (ANOVA) using a mixed model with block compound symmetry that addressed the heterogeneity of variances across concentrations of ADP was performed to test the effect on P-selectin expression of ADP in combination with fibrinogen, albumin, ferritin and ReoPro concentrations. After definition of significance with ANOVA, Dunnett’s test was used to compare each concentration of fibrinogen, albumin and ferritin with control. Results with TRAP were analyzed with the use of paired Student ttests. Significance was defined as p ≤ 0.05.
Results are from assays from each individual donor with respect to each concentration of ADP performed in triplicate in each case. The effect of fibrinogen on the activation of platelets induced by ADP was determined in blood taken from 19 subjects, 11 male and 8 female normal volunteers ranging in age from 20 to 55 years and who had no significant medical history and were not taking medications. The hematocrit (range 39% to 44%), hemoglobin (range 12 to 15 mg/dl), white blood count (range 3,700 to 9,000/mm3) and platelet count (range 140,000 to 290,000/mm3) were normal. The average concentration of fibrinogen was 257 ± 66 (SD).
Compared with control conditions, fibrinogen significantly increased the activation of platelets in response to ADP (Figs. 1 and 2). ⇓⇓A significant interaction (p < 0.001) was observed between ADP and fibrinogen. No activation (<1%) of platelets was observed after addition of PBS (control) or after addition of fibrinogen in the absence of ADP. The P-selectin expression in response to 0.2 μmol/liter ADP was increased significantly when 300 and 500 mg/dl of fibrinogen were added to blood; P-selectin expression in response to 0.8 μmol/liter ADP was increased significantly when 200, 300 and 500 mg/dl of fibrinogen were added to blood. Also, P-selectin expression in response to 1.5 μmol/liter ADP was increased significantly when 100, 200, 300 and 500 mg/dl of fibrinogen were added to blood.
Fibrinogen increased the extent of degranulation of platelets induced by TRAP. The P-selectin expression in response to 20 μmol/liter TRAP was nearly doubled after addition of 500 mg/dl of fibrinogen to blood (percentage of platelets expressing P-selectin, control = 24.5 ± 7.2, addition of 500 mg/dl of fibrinogen = 42 ± 6.6, p = 0.01, n = 7 subjects).
To determine whether the effects of fibrinogen were specific, albumin and ferritin were added to blood in molar concentrations equivalent to 100 mg/dl, 200 mg/dl and 500 mg/dl of fibrinogen before determination of degranulation in response to ADP (Fig. 3). Although the addition of 6 μmol/liter albumin (equivalent to 200 mg/dl of fibrinogen) led to significant (p = 0.04) increase in ADP-induced P-selectin expression, no change in ADP-induced P-selectin expression was seen with 3 and 15 μmol/liter albumin. The addition of ferritin did not alter ADP-induced P-selectin expression.
To clarify mechanisms by which fibrinogen led to increased ADP-induced degranulation of platelets, the binding of fibrinogen to the surface glycoprotein IIb-IIIa was inhibited pharmacologically. We found that the addition of greater than 4 μg/ml of ReoPro to blood inhibited binding of FITC-fibrinogen to platelets (Fig. 4). The addition of 4 μg/ml, 10 μg/ml and 100 μg/ml of ReoPro to blood did not alter the augmentation of expression of P-selectin observed when 500 mg/dl of fibrinogen was added to blood followed by the exposure of the platelets to ADP (Fig. 5).
As can be seen in Figure 4, despite inhibition of binding of fibrinogen to the surface glycoprotein IIb-IIIa, the exposure of platelets to 10 mg/dl of FITC-fibrinogen resulted in approximately 5% to 15% of platelets exhibiting associated FITC-fibrinogen. Results with confocal scanning laser microscopy demonstrated that the FITC-fibrinogen was localized in α-granules (n = 5 subjects, Fig. 6). No FITC-fibrinogen was observed on the surface of quiescent platelets that had not been exposed to ADP. Thus, the exposure of platelets to fibrinogen led to its uptake into α-granules. When blood was pretreated with ReoPro (10 μg/ml) before exposure of platelets to FITC-labeled fibrinogen, the FITC-fibrinogen was again localized in the α-granules (n = 5 subjects, Fig. 6). Accordingly, inhibition of binding of fibrinogen to the surface glycoprotein IIb-IIIa by ReoPro does not inhibit uptake of fibrinogen into α-granules of nonactivated platelets.
In the present study we found that the addition of fibrinogen to whole blood increased degranulation of platelets induced by ADP and TRAP. Fibrinogen, per se, did not activate platelets under basal conditions (without the addition of an agonist). Accordingly, fibrinogen is not a direct agonist but rather a factor that augments α-granule degranulation induced by agonists such as ADP and TRAP. The increased predisposition to agonist-induced degranulation of platelets induced by fibrinogen may contribute to the increased risk of cardiovascular events in subjects with increased concentrations of fibrinogen (16).
Fibrinogen and glycoprotein IIb-IIIa
Binding of fibrinogen to the activated surface glycoprotein IIb-IIIa is responsible for cross-linking and hence aggregation of platelets (9,17). To determine whether binding of fibrinogen could potentiate ADP-induced degranulation of platelets, surface binding was inhibited by the addition of ReoPro to blood. Addition of fibrinogen augmented the ADP-induced degranulation of platelets despite verified inhibition of surface binding of fibrinogen to the activated IIb-IIIa conformer. With the use of confocal scanning laser microscopy we found that the fibrinogen did not bind to the surface of quiescent platelets. By contrast, the exposure of platelets to added fibrinogen resulted in uptake of the fibrinogen into α-granules. This uptake was not abolished by pretreatment with ReoPro. Thus, uptake into α-granules of fibrinogen may potentiate degranulation of platelets.
Fibrinogen and α-granules
One potential mechanism by which the uptake of fibrinogen into α-granules could potentiate degranulation of platelets is through a mass effect. That is, an increase in the mass of proteins in the α-granules could potentiate degranulation. Additional mechanisms could include 1) changes in intra-platelet concentration or localization of calcium necessary for degranulation, 2) changes in the intraplatelet concentrations of proteins or nucleotides such as cAMP or cGMP that alter the reactivity of platelets, and 3) changes in the expression of surface receptors for agonists.
The protein content of α-granules of platelets is derived from endogenously synthesized megakaryocyte proteins and from endocytosis of circulating plasma proteins. Megakaryocytes do not synthesize fibrinogen (18,19). Fibrinogen has been identified in α-granules of platelets after intravenous injection (20,21). Our results suggest that the prevailing concentration of fibrinogen in blood can influence not only uptake of fibrinogen into α-granules but also agonist-induced degranulation of platelets.
Mechanisms of uptake of fibrinogen
Mechanisms by which fibrinogen is taken up into α-granules are incompletely understood. Treatment of guinea pigs with Kistrin, an antagonist of both the integrins glycoprotein IIb-IIIa and αvβ3, inhibits the uptake of biotinylated fibrinogen (22). Our data demonstrate that augmented activation of platelets occurs after exposure to fibrinogen despite pretreatment with ReoPro at a concentration that inhibits binding of fibrinogen to platelets.
ReoPro inhibits intracellular trafficking of fibrinogen and binding of fibrinogen to αvβ3(23,24). Thus, three possibilities may account for the increased ADP-induced degranulation of platelets produced by fibrinogen: 1) uptake mediated by binding of fibrinogen to other integrins not inhibited by ReoPro or mediated by the pool of integrins (IIb-IIIa and αvβ3) that are not associated with the added ReoPro; 2) uptake of fibrinogen into α-granules independent of binding to surface integrins; or 3) mechanisms independent of uptake of fibrinogen into α-granules.
The increased ADP-induced degranulation of platelets induced by fibrinogen may contribute, in part, to the impact of inflammation on risk of myocardial infarction. Recently, protective effects of aspirin in the primary prevention Physicians’ Health Study were found to be most prominent in those subjects who had increased concentrations in blood of another acute-phase reactant, C-reactive protein (25). Subjects with evidence of previous infection with chlamydia, cytolomegalovirus, and heliobacter are at increased risk of cardiac events (26–29). Any process that causes an increase in the concentrations in blood of acute-phase reactants is likely to increase concentrations of fibrinogen and, as judged from our results, increase the propensity for degranulation of platelets. Thus, for example, pregnancy is associated with both increased concentrations of fibrinogen and increased reactivity of platelets (30). Elucidation of mechanisms responsible for increased propensity for degranulation of platelets induced by fibrinogen may provide insights useful for development of novel pharmacologic agents designed to attenuate their sensitivity to increased concentrations of fibrinogen in blood.
The authors thank Paula B. Tracy, PhD, for helpful discussion and review of the manuscript, and Julie L. Ludko, Colette Charland, Matthew Hofmann, and Marilyn Wadsworth for expert technical assistance.
☆ These studies were supported in part by a grant from the National Science Foundation (NSF EPS-9703938).
- adenosine diphosphate
- fluorescein isothiocyanate
- phosphate buffered saline
- thrombin receptor agonist peptide
- Received November 18, 1997.
- Revision received August 10, 1998.
- Accepted September 10, 1998.
- American College of Cardiology
- Baker I.A,
- Eastham R,
- Elwood P.C,
- Etherington M,
- O’Brian J.R,
- Sweetham P.M
- Yarnell J.W,
- Baker I.A,
- Sweetnam P.M,
- Bainton D,
- O’Brien J.R,
- Whitehead P.J,
- et al.
- Naski M.C,
- Shafer J.A
- Sims P.J,
- Ginsberg M.H,
- Plow E.F,
- Shattil S.J
- McEver R.P,
- Martin M.N
- Schneider D.J,
- Tracy P.B,
- Mann K.G,
- Sobel B.E
- Rand M.D,
- Lock J.B,
- Veer C.V,
- Gaffney D.P,
- Mann K.G
- Marguerie G.A,
- Edgington T.S,
- Plow E.F
- Louarche F,
- Debili N,
- Cramer E,
- Breton-Gorius J,
- Vainchenker W
- Nakada M.T,
- Jordan R.E,
- Knight D.M
- Gupta S,
- Leatham E.W,
- Carrington D,
- Mendall M.A,
- Kaski J.C,
- Camm A.J
- Mendall M.A,
- Goggin P.M,
- Molineaux N,
- et al.