Cytosolic Calcium Changes in a Process of Platelet Adhesion and Cohesion on a von Willebrand Factor-Coated Surface Under Flow Conditions: Presented in part at the American Society of Hematology Meeting in Miami Beach, FL, December 4-8, 1998 (abstr no. 1423).

PLATELET THROMBUS formation represents an essential physiologic defense mechanism in primary hemostasis to arrest bleeding at sites of vessel disruption and can contribute to generation of pathologic intravascular thrombosis.1,2 von Willebrand factor (vWF) plays a central role in both physiologic and pathologic thrombus formation.3 Indeed, the critical contribution of vWF to processes of platelet plug generation, especially under high shear rate conditions, has become clear through platelet functional studies that take physiologic blood flow situations into consideration.4-6 Recent analysis of platelet function on a vWF-coated glass surface in a real-time flow system under high shear rates showed that the interaction between surface-immobilized vWF and platelet glycoprotein (GP) Ib leads to the rapid but transient capture of flowing platelets onto the surface.7 This initial tethering of platelets results in platelet translocation along the surface. During translocation, the GP IIb-IIIa complex is assumed to become gradually activated, presumably through an intraplatelet signal generated from the vWF-GP Ib interaction. This activation, in turn, leads to firm platelet adhesion through binding of activated GP IIb-IIIa to immobilized vWF. Cohesion of platelets firmly adhering to the surface occurs thereafter through the binding of soluble adhesive proteins, such as vWF or fibrinogen, to activated GP IIb-IIIa, resulting in platelet aggregate accumulation on a vWF-coated surface.7 8 

Most studies to clarify the molecular mechanisms by which platelets become activated during platelet aggregation processes have been conducted in static or closed stirring experimental conditions.9 10 However, little is known about the precise mechanisms involved in the real-time process of platelet adhesion and cohesion on a thrombogenic surface under physiologic blood flow conditions, perhaps because of the technical difficulties in evaluating the state of platelet activation under flow conditions.

To address this issue, we analyzed changes in platelet cytosolic Ca²⁺ concentrations ([Ca²⁺]i), a well-known indicator of platelet activation, in a colorimetric system that allows visualization of [Ca²⁺]i changes during the platelet adhesive process on a vWF-coated surface under flow conditions. We show that the initial phase of GP IIb-IIIa activation, which may be generated through the interaction between surface-immobilized vWF and GP Ib during translocation and is necessary for firm platelet adhesion to the vWF-coated surface, is independent of [Ca²⁺]i elevation. However, platelet [Ca²⁺]i elevation, presumably triggered by binding of activated GP IIb-IIIa to immobilized vWF (firm adhesion), is a prerequisite for the second phase of GP IIb-IIIa activation necessary for subsequent platelet cohesion.

Visible light-excitable Ca²⁺ indicator Fura Red acetoxymethyl ester (AM), Calcium Green-1 AM, and intracellular Ca²⁺ chelator 5,5′dimethyl 1,2-bis (2-aminophenoxy) ethane-N,N,N′,N′-tetraacetic acid (dimethyl BAPTA) AM were purchased from Molecular Probes, Inc (Eugene, OR). The Fab fragment of human/mouse chimeric anti–GP IIb-IIIa monoclonal antibody c7E3, which totally inhibits the ligand-binding functions of GP IIb-IIIa at concentrations up to 0.3 μmol/L,11,12 was purchased from Eli Lilly and Co (Indianapolis, IN). AP1, an anti-GP Ib monoclonal antibody that completely blocks the vWF-GP Ib interaction at concentrations up to 0.1 μmol/L, was a kind gift from Dr Thomas Kunicki (The Scripps Research Institute, La Jolla, CA).6,13The F(ab′)₂ fragment of AP1 was prepared by pepsin digestion of IgG at low pH and the collection of flow-through fractions in protein A-Sepharose (Pharmacia-LKB Japan, Tokyo, Japan) column chromatography as described.6,14 The antithrombin agent argatroban was supplied by Mitsubishi Chemical Corp (Tokyo, Japan). Apyrase (grade VIII) was from Sigma-Aldrich Japan Co (Tokyo, Japan), and bovine serum albumin (fraction V) was from Calbiochem (La Jolla, CA). Human native vWF containing the highest molecular weight multimers, as judged by sodium dodecyl sulfate (SDS)-1.5% agarose gel electrophoresis,15 was purified from cryoprecipitates, as described.16-18 

Blood was collected as described12,19 from nonsmoking healthy donors who had not taken any medication for at least 2 weeks using argatroban at a final concentration of 125 μg/mL as anticoagulant. Platelet-rich plasma (PRP) was obtained by centrifugation at 180g for 10 minutes. Apyrase was added (final concentration, 3 U/mL) to PRP and the mixture was centrifuged on a 40% albumin cushion at 700g for 2 minutes.6 20 The platelet fraction was resuspended in HEPES buffer (137 mmol/L NaCl, 2.7 mmol/L KCl, 1 mmol/L MgCl₂, 3 mmol/L NaH₂PO₄, 5.5 mmol/L glucose, 0.35% albumin, 3.5 mmol/L N-2-hydroxyethylpiperazine N′-2-ethanesulfonic acid, pH 7.2) containing 3 U/mL apyrase, 30 μmol/L Fura Red AM, and 20 μmol/L Calcium Green-1 AM at 37°C for 30 minutes. Ca²⁺ indicator-loaded platelets were centrifuged again and resuspended in HEPES buffer (pH 7.4). Erythrocytes were washed 3 times in HEPES buffer. Ca²⁺ indicator-loaded platelets and washed erythrocytes were mixed together and adjusted to 2 to 3 × 10⁷ platelets/mL and a 35% hematocrit.

The perfusion chamber used varies the shear rate in a linear manner as originally described by Usami et al.21 In brief, this chamber was designed to reproduce shear rates starting from a predetermined maximum value at the entrance and decreasing to 0 at the exit.21 Naflon tape (Iuchi Inc, Osaka, Japan) was used to adjust flow path height to 0.1 mm. Glass coverslips (24 × 50 mm; Matsunami Glass, Osaka, Japan) were coated with 200 μL of suspension of purified vWF (100 μg/mL), placed in a humid environment at room temperature for 60 minutes, and rinsed with 10 mL of HEPES buffer before perfusion. The perfusion chamber was assembled and mounted on BX60 Olympus epifluorescence microscope (Tokyo, Japan). The inner surface of the vWF-coated area was focused with a 100× objective lens (UPlanApo 100× oil Iris; Olympus), the focus depth of which was 0.49 μm. Thus, the [Ca²⁺]i changes as well as adhesive processes of platelets were observed preferentially in platelets adjacent to the vWF surface. Microscopy was equipped with an epifluorescent illumination apparatus (BX-FLA; Olympus; the excitation filter/dichroic mirror/barrier filter combination was MBP490/DM505/BA515IF, respectively) attached to a color charge-coupled device (CCD) camera (U-VPT-N; Olympus).

Before perfusion, CaCl₂ (final concentration, 1 mmol/L) was added to cell suspensions containing erythrocytes and Ca²⁺indicator-loaded platelets in HEPES buffer. Reconstituted blood containing platelets were aspirated through the chamber by a syringe pump (Model 935; Harvard Apparatus, South Natick, MA) at a constant flow rate of 0.285 mL/min and 37°C situations in a thermostatic air bath (Model UI-50; Iuchi Inc). In the present study, platelets were perfused through the chamber a single time and not recirculated. [Ca²⁺]i changes and the entire adhesive and cohesive process were recorded with a Hi-8 video cassette recorder (VL-HL1; Sharp Inc, Osaka, Japan), with a video rate of 30 frames per second, ie, time resolution of 0.033 seconds for each frame. Only platelets interacting with the vWF-coated surface were analyzed in this study. Freely flowing platelets in perfused blood can be discriminated from platelets interacting with the vWF surface by (1) the effective focus depth directed to the vWF surface and (2) the limit of time resolution of videotapes. Therefore, the background signals (flowing platelets) that were usually less than 20 pixel values can be separated from platelets interacting with the vWF surface (>140 pixel values) and subtracted as background by a computer-assisted image analysis using Win ROOF software (Mitani Corp, Fukui, Japan), as described.12 19 

The emission peaks of Calcium Green-1 and Fura Red were 531 nm and 637 to 657 nm, respectively.22 Elevated [Ca²⁺]i results in increased fluorescence intensity of Calcium Green-1 and decreased fluorescence intensity of Fura Red at an excitation wavelength of 490 nm. Thus, the combined use of these 2 Ca²⁺ indicators allows colorimetric analysis of cytosolic calcium changes, with ordinal fluorescence in which platelets glow from red (→orange→yellow-green→) to green with increasing platelet [Ca²⁺]i.22 Analogue records of video tape images were digitized by a frame grabber (DIG98; Detect, Tokyo, Japan) and analyzed by an image processing application (Win ROOF), as described.12,19 Changes in platelet [Ca²⁺]i were evaluated based on a ratio of signal intensity at G-channel (sensitive peak at 534 nm, green) relative to R-channel (sensitive peak at 594 nm, red) from digitized images. In this regard, we used in this study considerably higher concentrations of Ca²⁺ indicators (30 μmol/L Fura red AM and 20 μmol/L Calcium Green-1 AM) than usual to obtain sufficient signal intensity, which is crucial for precise colorimetric evaluation of [Ca²⁺]i changes during observation periods. These concentrations of Ca²⁺ indicators were selected based on our preliminary experiments, in which various concentrations of Ca²⁺ indicators on [Ca²⁺]i changes under flow on a collagen- or vWF-coated glass surface were tested. In addition, to address the possibility that the combination loading of such high concentrations of Ca²⁺ indicators seriously alters platelet functions, we compared the entire adhesive and cohesive process of Ca²⁺ indicator-loaded platelets with that of platelets labeled with mepacrine at a concentration (10 μmol/L) assumed to preserve normal platelet functions.5 7 No significant differences between these 2 approaches were observed in the extent of platelet translocation or final platelet surface coverage (see below). Next, to minimize the effect of photo-activation of platelets or photo-bleaching of fluorescence, our observation of platelet adhesive processes was performed by limiting the continuous photo-exposure time. In fact, preliminary experiments confirmed that 3-second consecutive illumination did not induce significant photo-activation of platelets and that any photo-bleaching had no effect on evaluation of platelet [Ca²⁺]i.

During the platelet adhesive process on a vWF-coated surface under flow conditions, platelets that initially translocate along the surface gradually become firmly adhered to the surface. The extent of firm platelet adhesion was determined by counting translocating platelets, defined as those moving along the surface at a distance greater than their diameters in 1 second as evaluated based on the logical OR algorithm of superimposed images (10 frames in 1 second) and Win ROOF computer software, within a defined area. Platelet adherence to the surface was evaluated using Win ROOF software to determine the percentage of the area covered by adhering platelets in a defined area after background subtraction and binarization of each image.

Platelet [Ca²⁺]i changes during the adhesive process on a vWF-coated surface were evaluated at shear rates of 50 s⁻¹, 750 s⁻¹, or 1,500 s⁻¹, representing low, intermediate, or high shear rate, respectively (Fig 1). After 8 minutes of blood perfusion at 50 s⁻¹, platelet adhesion was sparse and platelets appeared mostly reddish, indicating only slight [Ca²⁺]i elevation. Even at a shear rate of 750 s⁻¹, most platelets remained red, although a larger number of platelets, some of which certainly cohered, adhered to the surface. At 1,500 s⁻¹, the surface area covered by platelets was greatly increased, and green-colored platelets (high [Ca²⁺]i), most of which cohered, were more abundant. These observations indicate the shear-dependency of [Ca²⁺]i elevation as well as platelet adhesion and cohesion on a vWF-coated surface. Thus, all further experiments to examine real-time changes in [Ca²⁺]i during the platelet adhesive and cohesive process were performed at a wall shear rate of 1,500 s⁻¹.

The platelet adhesive and cohesive process on a vWF-coated surface involves (1) an initial platelet attachment to the vWF-coated surface; (2) translocation of platelets along the surface mediated by interaction between immobilized vWF and GP Ib; (3) firm adhesion to the surface mediated by the interaction between immobilized vWF and activated GP IIb-IIIa; and (4) platelet cohesion mediated by binding of soluble adhesive proteins to activated GP IIb-IIIa, as determined under experimental conditions involving direct fluorescence-labeling of platelets in whole blood.7,8 All of these events were recapitulated in the present system using reconstituted blood containing washed platelets labeled with fluorescent Ca²⁺indicators (Fig 2). Our approach also enabled evaluation of the platelet [Ca²⁺]i changes in the real-time platelet adhesive and cohesive process. Under flow conditions at a shear rate of 1,500 s⁻¹, no elevation in [Ca²⁺]i was observed during translocation of platelets, but [Ca²⁺]i increased apparently after platelets firmly adhered to the surface (Fig 2). However, the time-lag from firm adhesion to the [Ca²⁺]i elevation was highly variable among adhering platelets, with some firmly adhering platelets turning to green in a few seconds, whereas others remained red throughout the observation period (results not shown). A similar heterogeneity was also reported in the case of platelet adhesion on a fibrinogen-coated surface under low shear rate conditions.23 Platelet cohesion then occurred, using green platelets with fully elevated [Ca²⁺]i as a base (Fig 2). In this regard, adhesive proteins, vWF, or fibrinogen released from platelets were thought to play a role, because reconstituted blood free of plasma components was used in this study.

To confirm the time point and significance of the platelet [Ca²⁺]i elevation in the platelet adhesive process on a vWF-coated surface, we evaluated the effect of the intracellular Ca²⁺ chelator dimethyl BAPTA, an inhibitory anti-GP Ib, or anti–GP IIb-IIIa antibody on [Ca²⁺]i in our system. As judged by the visual observation at 8 minutes after perfusion, [Ca²⁺]i chelation did not block firm platelet adhesion after platelet translocation, but it did abrogate [Ca²⁺]i elevation and subsequent cohesion (Fig 3). This phenomenon was also confirmed by the time point observation at 30 seconds, 1 minute, 2 minutes, 4 minutes, and 6 minutes after the beginning of perfusion (results not shown). Neither [Ca²⁺]i elevation nor firm platelet adhesion was observed by the blockage of functional sites on GP IIb-IIIa (Fig 3). Indeed, statistical image analysis (Fig 4) supported the visual findings.

Both firm adhesion of individual platelets and subsequent platelet cohesion on a vWF-coated surface have been thought to require activation of GP IIb-IIIa. Our results demonstrate that platelet [Ca²⁺]i elevation is a downstream phenomenon and not a prerequisite for firm platelet adhesion to a vWF-coated surface. Although intraplatelet signaling plays a role in firm platelet adhesion, as evidenced by previous observations that prostaglandin E₁ (PGE₁)-induced elevation of intraplatelet cAMP levels blocks the firm adhesion,7 8 our results show that the first step in GP IIb-IIIa activation, which occurs during platelet translocation by the inside-out signaling from interaction between immobilized vWF and GP Ib, is independent of platelet [Ca²⁺]i elevation. In contrast, platelet [Ca²⁺]i elevation appears to be necessary for subsequent platelet cohesion on the surface, in which platelet-platelet engagement is achieved by binding of soluble adhesive ligands to activated GP IIb-IIIa.

Current studies on the activation of integrins have raised the possibility that there are at least 2 distinct phases in modulation of GP IIb-IIIa (also known as integrin αIIbβ3), namely an increase in affinity or an increase in avidity.24 An increase in affinity is caused by conformational changes in the heterodimer itself, resulting in a greater strength of ligand binding, whereas an increase in avidity is induced by clustering or multimerization of GP IIb-IIIa on the plasma membrane.24 Although alternative interpretations may be possible, it is interesting to assume that firm adhesion of individual platelets is a consequence of the affinity modulation of GP IIb-IIIa, independent of [Ca²⁺]i elevation, and that platelet cohesion requires the avidity modulation of GP IIb-IIIa, a more drastic GP IIb-IIIa activation. Platelet [Ca²⁺]i elevation may serve to trigger this avidity modulation of GP IIb-IIIa. Indeed, recent studies using cell types other than platelets suggested that [Ca²⁺]i elevation triggers clustering of several membrane receptors, including other members of integrin family.25-27 When [Ca²⁺]i is elevated, the Ca²⁺-binding protein complex activates the myosin light chain kinase, which is known to activate actin-activated myosin ATPase.28-30 The contraction of actin fibers evoked by the myosin ATPase may play a role in clustering of membrane receptors on the cell surface.

With regard to mechanisms by which [Ca²⁺]i is elevated in platelets during platelet aggregation processes in a closed suspension system, Chow et al,31 using a cone-and-plate type viscometer, demonstrated that cooperative functions of GP Ib and GP IIb-IIIa are essential for platelet [Ca²⁺]i elevation in platelet aggregation mediated by high shear stress. On the other hand, Ikeda et al,32 in a similar experimental system, reported that platelet [Ca²⁺]i elevation is absolutely dependent on the vWF-GP Ib interaction without GP IIb-IIIa functions. The basis for these discrepant findings remains uncertain, but may rest in the different blocking agents used in the respective inhibition studies. Although our experimental system involving observation of platelet adhesive and cohesive process on a surface is quite different from systems that analyze platelet aggregation in the soluble phase, our findings that platelet [Ca²⁺]i elevation occurs only after firm adhesion indicates that the binding of GP IIb-IIIa to immobilized vWF is indispensable in this regard. An outside-in signaling from the GP IIb-IIIa complex, generated when occupied by ligands, is likely the trigger for platelet [Ca²⁺]i elevation. Indeed, earlier studies, albeit in static experimental conditions, indicated the critical involvement of GP IIb-IIIa functions in activation of platelet membrane calcium channels.33 34 

Although the vWF-GP Ib interaction is also involved in [Ca²⁺]i elevation by capturing rapidly flowing platelets onto a surface under high shear rate conditions, our results confirmed that the contribution of this interaction is only indirect, indicating the lack of any substantial role for the inside-out signaling generated from GP Ib in [Ca²⁺]i elevation during platelet translocation.

In conclusion, our approach involving observations of changes in platelet [Ca²⁺]i during a real-time platelet adhesive and cohesive process under flow conditions showed distinct phase-specific mechanisms of platelet activation (adhesion and cohesion) that have not been demonstrable in previous static or closed stirring experiments. Studies under physiologic flow conditions to determine in detail the process of platelet activation will further contribute to an understanding of the complex mechanisms involved in mural platelet thrombogenesis.

The authors thank Marina Hoffman for editorial assistance.