Introduction

Alzheimer’s disease (AD) is the most common neurodegenerative disorder affecting the elderly and is characterized by progressive decline in cognitive faculties, particularly memory. The two pathognomonic features of this condition are extracellular senile plaques and intracellular neurofibrillary tangles (1), which are composed of fibrillar amyloid β (Aβ) aggregates (2) and hyperphosphorylated Tau proteins (3), respectively. The precise molecular pathogenesis of AD remains to be elucidated (4). However, there is growing evidence for the “amyloid hypothesis” (5), where Aβ is considered pivotal to synaptic loss and neuronal death underlying cognitive decline (6). Several possible mechanisms of Aβ-induced neurotoxicity have been recognized, including mitochondrial dysfunction (7), abnormal calcium homeostasis (8,9), oxidative stress (10), excitotoxicity (11), upregulation of proapoptotic proteins (12,13) and inactivation of growth promoting signaling pathways (13–15).

Aβ peptide ending at residue 40 (Aβ40) also deposits in cerebral vessel walls, leading to cerebral amyloid angiopathy (CAA) (16), an invariable presenting feature of AD (17). Aβ induces apoptosis of vascular endothelial (18) as well as smooth muscle cells (19). Thus, cerebrovascular dysfunction is an early event in pathogenesis of AD (20). Aβ is reported to have multiple binding partners (21), such as receptors for glutamate (22), acetylcholine (23) and nerve growth factor (24), cellular prion protein (25) and Frizzled (15), several of which play critical role in the pathophysiology of AD. However, the influence of these interactions on Aβ toxicity as well as their potential for therapeutic exploitation has not yet been fully explored.

We have recently demonstrated that the active fragment of Aβ elicits strong aggregation of washed human platelets, which was associated with exocytosis of platelet granule contents, activation of surface membrane integrins and spreading of platelets on immobilized matrix, all mediated through RhoA-dependent actomyosin reorganization (26). There have been reports of interaction between amyloid peptide and fibrin clots through binding site located near the C terminus of the fibrinogen β-chain (27), which makes the clot resistant to proteolytic degradation (28–30). These observations present compelling evidence in support of prothrombotic attributes of Aβ. In this study, we demonstrate that fibrinogen is a potent attenuator of Aβ-induced platelet activation in blood and thus serves as physiological deterrent to adverse vascular effects of Aβ.

Materials and Methods

Materials

Aβ25–35 (GSNKGAIIGLM), apyrase, ethylene glycol tetraacetic acid, ethylene-diaminetetraacetic acid (EDTA), aspirin, thrombin, fibrinogen (plasminogen-free), bovine serum albumin (BSA) and minimum essential medium (MEM) nonessential amino acids solution (100×) were purchased from Sigma. Fluorescence-activated cell sorting (FACS) flow sheath fluid and anti-CD62P antibodies were from BD Biosciences. Collagen and Chronolume luciferin-luciferase reagent were procured from Chrono-log. Aβ25–35 Aβ1–42 and Hylite-555 fluor-labeled Aβ42 were purchased from Anaspec. Aβ1–40 solution was a kind gift from Prof. Jay Kant Yadav, Central University of Rajasthan, India. DMEM/F12 (Dulbecco’s modified Eagle medium) was procured from Cell Clone; heat-inactivated fetal bovine serum and antibiotic-antimycotic solutions (100×) were obtained from Gibco. Antibodies against β-actin and phosphorylated myosin light chain phosphatase (MYPT1) were from Sigma and Cell Signaling Technology, respectively. All other reagents were of analytical grade. Type I deionized water (18.2 M Ω ·cm, Millipore) has been used throughout the experiments.

Preparation of Aβ Solution

Working stock (1 mmol/L) of Aβ25–35 was prepared in distilled water and stored at −20°C in aliquots. The solution was incubated at 25°C for 4 h before the experiments. Aβ1-42 (1 mg/mL) was prepared by adding 70–80 µL of 1% NH4OH to 1 mg lyophilized Aβ1-42 powder and immediately diluted in 1× phosphate-buffered saline (PBS) to the desired concentration. Hylite-555-Aβ42 was reconstituted in buffer containing 50 mmol/L Tris (pH 7.4) and 0.1% NH4OH to achieve 0.5 mg/mL and stored at −80°C (26).

Platelet Preparation

Blood from the antecubital vein of healthy volunteers was drawn into citratephosphate-dextrose-adenine anticoagulant (citric acid anhydrous, 15 mmol/L; sodium citrate dihydrate, 86 mmol/L; monobasic sodium phosphate, 16 mmol/L; and dextrose, 130 mmol/L) under informed written consent, strictly as per the recommendations and as approved by the Institutional Ethical Committee of Banaras Hindu University. The study methodologies conformed to the standards set by the Declaration of Helsinki. Washed platelets were prepared from fresh human blood by differential centrifugation as described (31).

Cell Culture and Treatment

Human SH-SY5Y neuroblastoma cell line was cultured either in flasks or in 6- or 96-well plates in DMEM/F12, supplemented with 10% heat-inactivated fetal bovine serum, 1% MEM nonessential amino acids and 1% antibiotic-antimycotic solution at 37°C in a 95% humidified air-5% CO2 incubator. After 24 h, cells (which had reached ~80% confluence) were washed three times with a culture medium and thereafter incubated for 48 h in DMEM/F12 with 1% serum in the presence of Aβ1-42 or Aβ25-35. In the experiments designed to examine the protective effect of fibrinogen, cells were treated with either fibrinogen or BSA (as negative control) before exposure to Aβ1-42 or Aβ25-35. We performed the experiments using cells that had undergone fewer than 15 passages, and all studies were repeated several times with different batches of cells. Cells used for FACS analysis were grown in six-well cluster dishes, whereas those used for cell viability assays were grown in 96-well plates. Serum starvation and vehicle did not affect viability of control samples.

Platelet Aggregation and Dense Granule Secretion

Platelet-rich plasma (PRP) or washed human platelets (with or without fibrinogen) were stirred (1,200 rpm) at 37°C in an optical lumi-aggregometer (Chrono-log model 700-2, Wheecon Instruments) for 1 min, after which collagen (10 µg/mL) or different concentrations of Aβ25–35 were added, and transmittance was recorded. Aggregation was measured as percentage change in light transmission, where 100% refers to transmittance through blank solution for washed platelets or platelet-poor plasma for PRP (31). Secretion of adenine nucleotides from platelet-dense bodies was assessed in parallel with aggregation by measurement of luminescence using Chrono-lume luciferin-luciferase reagent following the manufacturer’s instructions (32).

Measurement of Platelet Intracellular Calcium

PRP was incubated with 2 µmol/L Fura-2/AM for 45 min at 37°C in the dark. Fura-2-loaded platelets were washed and resuspended in buffer B at 108 cells/mL. Fluorescence was recorded in 400-µL aliquots of platelet suspensions at 37°C under nonstirring conditions using a fluorescence spectrophotometer (Hitachi model F-2500). Excitation wavelengths were 340 and 380 nm, and emission wavelength was set at 510 nm. Changes in intracellular free calcium, [Ca2+]i, upon addition of Aβ to either control platelets or platelets preincubated with fibrinogen (2 mg/mL) were monitored from the fluorescence ratio (340/380) using the Intracellular Cation Measurement Program in FL Solutions software (31). Intracellular free calcium was calibrated according to the derivation of Grynkiewicz (33).

Flow Cytometry

Surface expression of P-selectin. Washed human platelets (2 × 108 cells) were incubated at 37 °C for 10 min without stirring in the presence of thrombin (1 U/mL), Aβ25–35 (20 µmol/L) or fibrinogen (2 mg/mL) plus Aβ25-35 (20 µmol/L) and fixed for 30 min with 2% paraformaldehyde. Cells were washed twice in 1 × PBS, following which 5 µL fluorescein isothiocyanate (FITC)-labeled anti-CD62P antibody was added to each sample and incubated for 30 min in the dark at room temperature (RT). Cells were washed with 1 × PBS, resuspended in FACS sheath fluid and analyzed by flow cytometry as described above (31).

ROS measurement. Washed human platelets (2 × 108 cells), pretreated with either fibrinogen or BSA for 5 min, were incubated with either Aβ25–35 or Aβ1–42 for 15 min, followed by incubation with 2,7-dichlorodihydro fluorescent diacetate (H2DCFDA; final concentration 10 µM) for 30 min. Fluorescence was immediately measured using a flow cytometer (Becton Dickinson model FACSCalibur) with excitation wavelength of 488 nm and emission at 530 nm. Gated cells (n = 10,000) were analyzed for each sample.

For measurement of intracellular ROS in SH-SY5Y and N2a neuroblastoma cells, 105 cells were seeded per well in six-well plates in DMEM/F12. After 24 h of treatment with either fibrinogen or BSA (negative control), cells were detached with 1 mmol/L EDTA and washed with PBS. H2DCFDA was added to cells at a final concentration of 20 µmol/L and incubated for 30 min. Fluorescence was measured with flow cytometry as described above.

Mitochondrial Viability Measurement by MitoTracker Red Dye

For determination of mitochondrial viability, 105 cells were seeded per well in six-well plates in DMEM/F12. After 24 h of treatment with either fibrinogen or BSA, cells were detached with 1 mmol/L EDTA and washed twice with PBS, followed by incubation with MitoTracker dye (final concentration 500 nmol/L) for 45 min. Fluorescence was measured immediately using the flow cytometer in FL2 channel.

Immunoblotting

Platelet proteins were separated on 10% SDS-PAGE gels and electrophoretically transferred to polyvinylidene fluoride membrane using the TE 77 PWR semidry system (GE Healthcare). Membranes were blocked with 5% nonfat dry milk in Tris-buffered saline (10 mmol/L Tris-HCl and 150 mmol/L NaCl, pH 8.0) containing 0.05% Tween-20 (TBST) for 1 h at RT. Blots were incubated overnight with respective primary antibodies (anti-MYPT1, 1:1000; anti-β-actin, 1:5000), followed by three washings with TBST for 5 min each. Membranes were then placed in horseradish peroxidase-labeled anti-IgG secondary antibodies appropriately diluted in blocking buffer or TBST for 1 h. Blots were similarly washed, and antibody binding was detected using the ECL detection kit. Images were acquired on a multispectral imaging system (BioSpectrum 800 Imaging System, UVP) and quantified using VisionWorks LS software (UVP).

Cell Viability Assay

Cells were plated at a density of 1 × 104 cells in 96-well plates and cultured for 24 h before treatment. Cell viability was measured using 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay as previously described (34). Absorbance at 540 nm was measured using a microplate reader (BioTek Model Synergy H1). Viability of vehicle-treated control groups (not exposed to Aβ25–35) was defined as 100%.

For determination of cell viability using calcein/AM, 105 cells were seeded per well in a six-well plate in DMEM/F12. After 24 h of treatment, cells were detached using 1 mmol/L EDTA and washed with PBS. Calcein/AM was added to cells at a final concentration of 1 µg/mL, followed by incubation for 30 min. Fluorescence was analyzed immediately using the flow cytometer with excitation wavelength 488 nm and emission at 530 nm. Gated cells (n = 10,000) were analyzed for each sample.

Study of Aβ Interaction with Platelets and Neuronal Cells

Washed platelets were incubated with 5 µL of FITC-labeled anti-CD41a antibody and 5 µmol/L Hylite-555-Aβ42 for 15 min, with or without fibrinogen or BSA, washed with PBS and resus-pended in FACS sheath fluid. Fluorescence data were collected in FL1 and FL2 channel in the flow cytometer using four-quadrant logarithmic amplification of 10,000 events as described above. Untreated platelets served as control.

For confocal microscopy, platelets loaded with 1 µmol/L calcein/AM were treated with Hylite-555-Aβ42 in the presence of either fibrinogen or BSA or saline and were fixed overnight with 2% paraformaldehyde. Cells were examined under a Zeiss LSM 700 laser scanning confocal microscope with 10 × objective and 1 AU pinhole size. Images were acquired and analyzed using ZEN imaging software (Carl Zeiss) (26).

For the study of Aβ-neuronal cell interaction, SH-SY5Y cells were grown on glass chamber slides (Lab-Tek). After 24 h, cells were washed twice with culture medium, and fresh medium was added. Cells were incubated with fibrinogen, BSA or saline, followed by addition of Hylite-555-Aβ42. After another 24 h, cells were fixed with 2% paraformaldehyde, washed with PBS and visualized under a laser scanning confocal microscope (LSM 700, Zeiss) (26).

Clot Retraction Studies

Washed platelets (2 × 108 cells in 500 µL) were treated with either ADP (20 µmol/L) or Aβ25–35 (20 µmol/L) in the presence of calcium chloride (1 mmol/L) for 5 min at RT. Fibrinogen was added at a final concentration of 2 mg/mL either before or after addition of Aβ25–35. Atroxin (0.1 µg/mL) was added finally to initiate fibrinogen cleavage and clot formation. Contents were incubated at 37 °C. Clot retraction kinetics was observed, and images were captured at different time intervals and analyzed as described (35).

In Vitro Hemolysis Assay

In vitro hemolysis was carried out as described (31). To study the protective role of fibrinogen against Aβ-induced hemolysis, red blood cells (RBCs) were preincubated with fibrinogen (2 mg/mL) followed by exposure to varying concentrations of Aβ25–35 (5–20 µmol/L). RBCs suspended in deionized water or PBS served as positive and negative controls, respectively. Samples were incubated at 37°C for 1 h, followed by centrifugation at 10,000g for 10 min. Hemoglobin absorbance in the supernatant was measured at 540 nm in a microplate reader at 37°C. Percent hemolysis was calculated using the following equation:

$$\rm{Hemolysis}\;(\% )\;{\rm{ = }}\>{{{\rm{Sample\; absorbance\;at\;540\; nm}} - \;{\rm{Absorbance\; of}}\;{\rm{negative\; control}}\;{\rm{at}}\;{\rm{540}}\;{\rm{nm}})} \over {{\rm{Absorbance}}\;{\rm{of}}\;{\rm{positive}}\;{\rm{control}}\;{\rm{at}} - \;{\rm{Absorbance}}\;{\rm{of}}\;{\rm{negative}}\;{\rm{control}}\;{\rm{at}})}} \times \>100.$$

Statistical Analysis

All data are representative of at least three independent experiments. Two-tailed Student t test was used for evaluation of significance, and values of p < 0.05 were considered significant.

All supplementary materials are available online at www.molmed.org .

Results

Aβ-Induced Platelet Activation and Clot Retraction Are Attenuated by Plasma Fibrinogen

We recently reported that active fragment of Aβ containing amino acid residues 25–35 (Aβ25–35; 20 µmol/L) induces aggregation of washed platelets (suspended in buffer) through RhoA-dependent modulation of myosin light chain phosphorylation and actomyosin organization (26). Intriguingly, we also found that Aβ25–35, even at concentrations as high as 50 µmol/L, failed completely to aggregate platelets suspended in plasma (Figure 1A). In contrast, physiological agonists such as collagen and ADP induced robust platelet aggregation under similar experimental conditions as expected. As Aβ has lately been shown to interact with fibrinogen (27), we subsequently studied the effect of fibrinogen added at physiological concentrations (36) to washed platelets in the presence of Aβ. Fibrinogen (2 mg/mL) significantly impaired platelet aggregation (p < 0.0001; tracing 3, Figure 1B), ATP release (p = 0.0021; tracing 3′, Figure 1B) and P-selectin exposure (p = 0.047; Figure 1E) induced by Aβ25–35 (20 µmol/L), while addition of BSA at similar concentrations to platelet suspension had no effect (tracings 4 and 4′, Figure 1B). Further, inhibition of Aβ25–35-induced platelet aggregation by fibrinogen was dose dependent (Figure 1C), with nearly total abrogation of aggregation at fibrinogen level 3 mg/mL. Fibrinogenmediated inhibition of platelet aggregation could be overcome by raising the concentration of Aβ25–35 (Figure 1D), which was suggestive of specific interaction between Aβ peptide and fibrinogen. Comparable results were obtained with full-length Aβ peptide. Aβ40-induced platelet aggregation and secretion was significantly inhibited by fibrinogen (Supplementary Figure S1).

Figure 1
figure 1

Effect of fibrinogen on Aβ25–35-induced platelet activation. (A) Tracings 1, 2, 3 and 4 represent platelet aggregation in PRP induced by Aβ25–35 (20 and 50 µmol/L), ADP (20 µmol/L) and collagen (10 µg/mL), respectively. The corresponding bar diagram is shown in the lower panel. (B) Tracings 1–4 represent aggregation, and tracings 1′-4′ represent corresponding dense granule secretion in washed platelets stimulated with either thrombin (1 U/mL; tracings 2 and 2′) or Aβ25_35 (20 µmol/L; tracings 1, 1′, 3, 3′, 4 and 4′). Tracings 3 and 3′ represent platelets pretreated with fibrinogen (2 mg/mL), whereas tracings 4 and 4′ represent platelets pretreated with BSA (2 mg/mL). Corresponding bar diagram is shown in the lower panel. (C) Tracings 1, 2, 3 and 4 represent aggregation of washed platelets induced by Aβ25_35 (20 µmol/L) in the presence of 3, 2, 1.5 and 0.75 mg/mL fibrinogen, respectively. Corresponding bar diagram is shown in the lower panel. (D) Tracings 1, 2, 3 and 4 represent aggregation of washed platelets induced by Aβ25–35 (20,30,40 and 60 µmol/L, respectively) in the presence of fibrinogen (2 mg/mL). Corresponding bar diagram is shown in the lower panel. (E) Flow cytometric analysis of expression of P-selectin on surface of platelets treated with different reagents and agonists as indicated. (F) Retraction of fibrin clot by platelets stimulated with either Aβ25–35 (20 µmol/L) or ADP (20 µmol/L) in presence of fibrinogen (2 mg/mL) added either before or after exposure to agonists as indicated.

Full size image

We next studied whether fibrinogen can prevent retraction of the fibrin clot induced by Aβ-treated platelets. Platelets were stimulated with Aβ25-35, followed by addition of fibrinogen and atroxin. The latter is snake venom, which is known to elicit clot formation by cleavage of fibrinogen. Cytoskeletal contractility in platelets activated by Aβ25–35 evoked progressive compaction of the fibrin clot for up to 60 min. Strikingly, retraction of the clot was precluded when platelets were incubated with fibrinogen before addition of Aβ25–35 (Figure 1F). Pretreatment with fibrinogen, however, had no inhibitory effect on retraction of the clot when Aβ was substituted with ADP.

Agonist stimulation of platelets is known to be associated with increases in cytosolic Ca+2 and reactive oxygen species (ROS). In agreement with earlier reports (37,38), Aβ25–35 brought about a sharp rise in both intracellular ROS (Figure 2A) and Ca+2 (Figure 2B). The rise was almost completely abrogated by pretreatment of cells with fibrinogen (2 mg/mL; p = 0.0024). We have recently demonstrated that the stimulatory effect of Aβ on platelets is mediated through the RhoA-MYPT1-myosin light chain (MLC) axis (26). As expected, Aβ40-aggregated platelets were associated with enhanced phosphorylation of MYPT1. Fibrinogen (3 mg/mL), too, abolished Aβ-induced RhoA signaling in these platelets, as evidenced by the lack of rise in the level of phospho-MYPT1 upon Aβ40 treatment (Figure 2C, D).

Figure 2
figure 2

(A) Aβ25–35 induced rise in intracellular ROS in the absence and presence of fibrinogen (2 mg/mL) as indicated. (B) Aβ25_35 induced rise in intracellular calcium in the absence and presence of fibrinogen (2 mg/mL) as indicated. (C) Immunoblots representing MYPT1 phosphorylation in washed platelets stimulated with Aβ40 (20 µmol/L for 5 min) in the presence or absence of fibrinogen (3 mg/mL). (D) Corresponding bar diagram representing the densitometric analyses of immunoblots for p-MYPT1. *p < 0.05 with respect to control platelets; #p < 0.05 with respect to Aβ40-treated platelets.

Full size image

We next asked whether the effect of Aβ25–35 on RBCs, the other abundant cell type in blood, too, was precluded by fibrinogen. As expected, Aβ25–35 induced lysis of red blood cells suspended in isotonic buffer (39), which was absolutely mitigated in the presence of 2 mg/mL fibrinogen (p < 0.0001; Supplementary Figure S2).

Fibrinogen Prevents Interaction of Aβ with Platelets and Neuronal Cells

We employed fluorescently labeled Aβ peptide (Hylite-555-Aβ42) to explore Aβ-platelet interaction in the presence or absence of fibrinogen. As we have demonstrated earlier (26), platelets incubated with Hylite-555-Aβ42 exhibited gains in FL2 fluorescence in flow cytometric analysis, indicative of association of Aβ with platelets. The percentage of FL2-positve events within platelet gate was significantly reduced when cells were pretreated with either fibrinogen (2 mg/mL; p = 0.02; Figure 3D) or plasma (p = 0.016; Figure 3F) but not with BSA (2 mg/mL; control; Figure 3E).

Figure 3
figure 3

Interaction of fluorescently labeled Aβ with platelets. (A) Washed platelets labeled with CD41a-FITC (WPCD41a); (B) washed platelets treated with Hylite-555-labeled Aβ42 (Aβ-FL); (C) CD41a-FITC-labeled washed platelets incubated with fluorescently labeled Aβ42 (Aβ-FL); (D) CD41a-FITC-labeled washed platelets preincubated with fibrinogen (Fg; 2 mg/mL) followed by treatment with fluorescently labeled Aβ42 (Aβ-FL); (E) CD41a-FITC-labeled washed platelets preincubated with BSA (2 mg/mL) followed by treatment with fluorescently labeled Aβ42 (Aβ-FL); (F) CD41a-FITC-labeled platelets suspended in plasma (PRPCD41a) treated with fluorescently labeled Aβ42 (Aβ-FL).

Full size image

In concurrence, confocal microscopy of platelets treated with Hylite-555-Aβ42 revealed intracellular accumulation of Aβ, which was significantly reduced in the presence of fibrinogen, and Hylite-555-Aβ42 fluorescence was found to be scattered/dispersed extracellularly (Figure 4). Pretreatment with saline or BSA, in place of fibrinogen, did not prevent Aβ internalization.

Figure 4
figure 4

Intracellular localization of Aβ42 in human platelets. (A) Human platelets, fluorescently labeled with both Hylite-555-Aβ42 (red) and calcein/AM (green), were pretreated with saline (upper panel), fibrinogen (middle panel) or BSA (lower panel). (B) The bars represent the number of platelets positive and negative for Aβ42 for every 200 platelets counted. #p < 0.05 with respect to Aβ42-treated cells in the absence of fibrinogen.

Full size image

Next, we explored the influence of fibrinogen on the association between Hylite-555-labeled Aβ42 and SH-SY5Y neuroblastoma cells by confocal microscopy (40). Consistent with the above observations, pretreatment with fibrinogen but not BSA brought about significant reductions in Aβ-neuronal cell interaction (p < 0.0001; Figure 5).

Figure 5
figure 5

Intracellular localization of Aβ42 in neuronal cells. (A) SH-SY5Y cells labeled with Hylite-555-Aβ42 pretreated with saline (upper panel), fibrinogen (middle panel) or BSA (lower panel). Arrows indicate localization of Aβ42 (red) inside cells. (B) The bars represent the number of SH-SY5Y cells positive and negative for Aβ42. Data are representative of about 250 cells counted in 10 low-power (10×) fields. #p < 0.05 with respect to Aβ42-treated cells in the absence of fibrinogen.

Full size image

As fibrinogen attenuated Aβ-induced platelet activation and hemolysis, we hypothesized that it could similarly mitigate neurotoxicity of Aβ. Aβ25-35 brought about a dose-dependent decline in viability of SH-SY5Y neuroblastoma cells as estimated by MTT assay (41,42). Fibrinogen (2 mg/mL) remarkably ameliorated the decrease in cell viability at all concentrations of Aβ25–35 studied (Supplementary Figure S3). The neurotoxicity of Aβ has been attributed to mitochondrial dysfunction (7) and oxidative stress (10). Aβ25-35 (20 µmol/L) triggered a drop in neuronal cell mitochondrial transmembrane potential and a rise in intracellular ROS levels, which were abrogated by 23% and 50%, respectively, in the presence of fibrinogen (2 mg/mL; Supplementary Figure S3).

Discussion

This study demonstrates that fibrinogen attenuates platelet activation and cytotoxicity induced by active fragment of Aβ (Aβ25–35) by preventing Aβ (Aβ42) interaction with cells. We had recently reported that Aβ provokes strong stimulation of human platelets through RhoA-dependent modulation of MLC phosphorylation and actomyosin organization (26). This observation was supported by the presence of “preactivated” platelets in mouse models of AD (43). Platelets preincubated with Aβ, too, exhibited enhanced adhesion to injured carotid artery in mice (37). Taken together, the above reports are suggestive of Aβ-platelet interaction resulting in a prothrombotic phenotype that contributes to CAA associated with AD.

Intriguingly, we found that Aβ25–35, unlike physiological agonists such as collagen and ADP, failed to induce aggregation of platelets suspended in plasma. As plasma fibrinogen has been reported to be one of the binding partners for Aβ (27), we evaluated the influence of fibrinogen on Aβ-induced platelet aggregation. Fibrinogen brought about dose-dependent declines in aggregatory response elicited by Aβ25–35 (20 µmol/L), with nearly total abrogation of aggregation at fibrinogen concentrations ≥2 mg/mL (normal plasma fibrinogen level being 2–4 mg/mL), which was reversed by Aβ25–35 added in increasing doses. Fibrinogen attenuated Aβ-induced platelet responses such as exocytosis of granule contents and rise in cytosolic Ca+2 and ROS. Aβ-induced RhoA signaling was also abolished by fibrinogen. Consistent with these findings, intracellular accumulation of exogenous Aβ, as was reported earlier (26), was found by confocal microscopy to be dramatically impeded in the presence of fibrinogen, although the association of Aβ with platelets was only moderately diminished as evidenced by flow cytometry. The majority of platelet-associated Aβ remained extracellular in the presence of fibrinogen. Thus, inhibition of Aβ-induced platelet responses by fibrinogen could possibly be attributable to lack of Aβ internalization.

Platelets contribute to >90% amyloid precursor protein in circulation (44), which, upon proteolytic cleavage by β- and γ-secretases, yields Aβ40 (45). Despite low plasma levels, high local concentrations of Aβ are achieved (28) at the site of thrombus formation owing to release from stimulated platelets (46,47) and at the sites of CAA (48) or atherosclerotic plaques (49,50), which can initiate vicious cycles of platelet stimulation and Aβ release and potentially lead to massive thrombosis. Based on observations presented in this study, it is tempting to speculate that fibrinogen acts as a physiological “shield” to preclude Aβ-induced activation of platelets, thus keeping the prothrombotic attributes of amyloid peptide strictly under check and providing steady protection against thrombotic vascular occlusion. It has, however, been reported that fibrin clots are rendered resistant to lysis following interaction with amyloid peptide (28–30). Thus, even as the probability of Aβ-induced platelet activation is minimized in the presence of fibrinogen, Aβ would prolong the half-life of an eventual clot. This hypothesis is consistent with our previous findings, that although Aβ was not thrombogenic on its own, it exacerbated pulmonary thromboembolism induced by collagen-epinephrine in a mouse model (26).

Next, we asked whether fibrinogen could, too, protect neuronal cells from Aβ-mediated toxicity similar to platelets. In fact, preincubation with fibrinogen did significantly impair the association of Aβ peptide with neurons and reversed the drop in neuronal cell viability and mitochondrial transmembrane potential induced in the presence of Aβ. Although fibrinogen cannot normally gain access to the central nervous system, altered blood-brain barrier permeability associated with AD can allow extravasation of plasma proteins, including fibrinogen, into brain parenchyma (51). Further, there is increasing evidence for a crucial role of dynamic exchange of Aβ between brain and plasma during AD pathogenesis (52). Here we show that fibrinogen can possibly serve as a peripheral sink sequestering Aβ in plasma and minimizing Aβ-mediated toxicity.

It has recently been demonstrated that fibrin(ogen) deposits in cerebral vasculature contribute to the pathogenesis of AD (29,30,53). This was attributed to resistance of fibrin clots to degradation upon interaction with Aβ, which in turn would lead to reduced cerebral blood flow. There are also reports to suggest that fibrin deposited in AD brain could promote vascular pathology, which is dependent on fibrin(ogen)’s ability to activate microglial CD11b receptors (51). Thus, our findings, when considered in the context of the above reports, suggest that fibrinogen has dualistic role in the pathogenesis of prothrombotic phenotype as well as AD.

Conclusion

We conclude that fibrinogen circulating peripherally in plasma sequesters Aβ, serving as a physiological deterrent to Aβ-induced thrombogenecity and platelet activation. Peptides or small molecules that could similarly sequester Aβ and prevent its interaction with cells can be potentially therapeutic in AD and CAA.

Disclosure

The authors declare that they have no competing interests as defined by Molecular Medicine, or other interests that might be perceived to influence the results and discussion reported in this paper.