Circulating platelet-derived microparticles with procoagulant activity may be a potential cause of thrombosis in uremic patients

Kidney International, Vol. 62 (2002), pp. 1757–1763

Circulating platelet-derived microparticles with procoagulant activity may be a potential cause of thrombosis in uremic patients MINORU ANDO, AKIKO IWATA, YASUSHI OZEKI, KEN TSUCHIYA, TAKASHI AKIBA, and HIROSH NIHEI Division of Nephrology, Department of Medicine, Tokyo Metropolitan Geriatric Hospital, and Department of Medicine, Kidney Center, Tokyo Women’s Medical University, Tokyo, and First Institute of New Drug Research, Otsuka Pharmaceutical, Co., Ltd., Tokushima, Japan

Circulating platelet-derived microparticles with procoagulant activity may be a potential cause of thrombosis in uremic patients. Background. Clinical experience indicates that bleeding and thrombotic tendencies co-exist in uremic patients. Numerous studies have shown that platelet functional defects contribute to the bleeding tendency in uremic patients. In contrast, there are no solid studies clarifying the pathogenesis of the prothrombotic state in uremic patients. Platelet-derived microparticles (PMPs), which are small vesicles with procoagulant activity released from activated platelets, are thought to be involved in clinical thrombogenesis. This study addressed the question of why uremic patients are thrombophilic even though they have a bleeding tendency, focusing on the clinical significance of PMPs. Methods. The subjects were pre-dialyzed patients, patients under hemodialysis (HD) or continuous ambulatory peritoneal dialysis (CAPD) therapy, and age-matched healthy controls. Analyses of PMPs were performed using a flow cytometer. Annexin V was used to probe procoagulant activity of PMPs. The impacts of the HD procedure, arteriovenous (AV) fistula, and recombinant human erythropoietin (rHuEPO) treatment on the release of PMPs were additionally assessed. Results. Major results are: (1) PMP counts were significantly greater in each uremic group than in controls. The PMP counts were not different among three types of uremic groups; (2) PMP counts were significantly higher in uremic patients with thrombotic events than in those without thrombotic events; and (3) the HD procedure and existence of AV fistula did not affect PMP counts, but rHuEPO treatment possibly enhanced the PMP release in these patients. Conclusions. Elevated PMP counts may trigger thrombosis in uremic patients. The primary cause of PMP elevation in uremia was not clarified in this study.

Key words: thrombosis, flow cytometric analysis, procoagulant activity, annexin V, bleeding tendency, platelet function, acute arterial occlusion. Received for publication October 10, 2001 and in revised form June 10, 2002 Accepted for publication June 12, 2002

 2002 by the International Society of Nephrology

Uremic patients seem to have two opposite aspects in hemostatic status, which are bleeding and thrombotic tendencies. A number of experimental studies have revealed platelet functional defects in uremia, such as diminished adherence to vascular subendothelium [1, 2], decreased expressions of platelet membrane glycoprotein GIb [3] and dysfunction of glycoprotein GIIb/IIIa [4]. The bleeding tendency in such patients, therefore, may be due to the functional impairments of platelets in normal hemostasis [5, 6]. On the other hand, clinical statistics suggest that uremic patients have a high incidence of thrombotic events [7, 8]. Moreover, a thrombotic tendency, caused by factors such as platelet hyperaggregability and hypercoagulability, has been described in dialysis patients [9–12]. This evidence emphasizes that acute arterial occlusion can be facilitated under a uremic situation. However, to our knowledge there are no studies giving a definite answer to the question of why uremic patients often show a thrombophilic tendency despite their vulnerability in physiological hemostatic functions. Platelets release small particles called platelet-derived microparticles (PMPs) after they are activated by some chemical stimuli or high shear stress [13–19]. The major functional significance of PMPs is associated with its procoagulant activity. PMPs are enriched with membrane receptors for coagulation factor Va, and provide a catalytic surface for the prothrombinase reaction, contributing to the acceleration of thrombin generation [13, 14]. Thus, PMPs may play a pivotal role in acute arterial occlusion in atherosclerotic small arteries or arterioles where pathological levels of fluid shear stress may occur [15–19]. Elevated counts of circulating PMPs have been reported in association with thrombotic disorders, such as cerebrovascular accidents [20], unstable angina [21], and acute myocardial infarction [22]. In addition, PMPs that adhered to vascular endothelium and leukocytes

1757

1758

Ando et al: PMP formation in uremic patients

activate such cells and transport their chemical mediators to those cells, potentially leading to the development of thrombosis and atherosclerosis [23–26]. We hypothesized that an enhancement of PMP generation—if it exists—may participate in thrombogenesis and compensation for the bleeding tendency in uremia. This study used flow cytometry [13, 14, 18] to evaluate the PMPs in healthy subjects and three types of uremic patients: those who were pre-dialyzed, on hemodialysis (HD), and on continuous ambulatory peritoneal dialysis (CAPD) therapy. The procoagulant activity of PMPs were especially evaluated by annexin V, a placental protein with a high affinity and specificity for aminophospholipids, which has been shown to be an appropriate probe for detecting such specific activity of PMPs [27]. In addition, the concomitant influences of the HD procedure, existence of internal arteriovenous (AV) fistula, and chronic recombinant human erythropoietin (rHuEPO) therapy on PMP formation were studied. METHODS Reagents and antibodies Fluorescein isothiocyanate (FITC)-conjugated monoclonal antibody (mAb) to GPIb, NNKY5-5 [28], was a generous gift from Dr. Shosaku Nomura, Kansai Medical School (Osaka, Japan). Phycoerythrin (PE)-conjugated annexin V was purchased from Pharmingen (San Diego, CA, USA). Isotype-matched mAbs of irrelevant specificity (FITC-labeled mouse IgG) were purchased from Becton Dickenson (San Diego, CA, USA). Calibrated latex beads (SPHERO娃 Fluorescent Particles) were purchased from Pharmingen. Subjects Twenty-nine healthy subjects and 46 HD patients, 23 CAPD patients, and 20 pre-dialyzed uremic patients were studied. Diabetic patients were not included to estimate the simple influence of uremia on PMP generation [29]. Patients who had been receiving anti-platelet drugs were excluded. The subjects’ profiles are shown in Table 1. Each patient population was matched for age with the healthy population (mean age, 52.3 ⫾ 5.19 years). A majority of the dialysis patients had been treated with recombinant human erythropoietin (rHuEPO), which was intravenously injected three times a week in the HD patients, and subcutaneously administered once a week in the CAPD and pre-dialyzed patients. The HD patients were dialyzed using columns of different types of membrane materials: polysulfone, cellulose-triacetate, or polymethylmethacrylate. Heparin, as an anticoagulant agent, was continuously injected at a dose of 600 to 800 units per hour during HD treatment.

Blood collection, preparation of washed platelets, and cell labeling Blood samples were obtained before the start of treatment at the first HD of the week in the HD patients, or when patients visited the hospital for their regular medical examination in the CAPD or pre-dialyzed patients. Two milliliters of venous blood were placed into a tube containing 0.2 mL of an EDTA-ACD solution [ethylenediaminetetraacetic acid (EDTA)–2 Na 1.0 g, trisodium citrate 2.2 g, citric acid 807 mg, dextrose 2.2 g in 100 mL of distilled water] and gently inverted twice. Plateletrich plasma (PRP) was prepared by centrifugation at 180 ⫻ g for 20 minutes at room temperature. An aliquot (15 ␮L) of the PRP that had been diluted up to 50 ␮L with 0.1% EDTA-saline was then incubated with 2 ␮L of FITC-GPIb and PE-annexin V for 30 minutes at room temperature in the dark. Nonspecific FITC-labeled mAb was used for the negative controls. Flow cytometry Samples were analyzed using a flow cytometer (FACS Calibur; Becton Dickinson). The light scatter and fluorescence channels were set at a logarithmic gain. The gating methods for PMPs are shown in Figure 1. Only cells and particles positive for GPIb were gated to distinguish platelets and PMPs from electronic noise as follows: The cells labeled with monoclonal antibody (mAb) to FITC-GPIb (R1) were gated on the side scatter (SSC) dot plot versus fluorescence-1 (FL-1) profile, which was determined by comparing with the dot plot profile of the samples treated with non-labeled mAb to GPIb (Fig. 1A). Regions corresponding to PMPs (R3) were then discriminated from the platelet islet contaminated (R2) by reviewing the cells in the gate R1 on SSC dot plot versus the forward scatter (FSC) dot plot profile (Fig. 1B). A known number of fluorescent latex beads (30,000) were added to each sample tube before analysis [30], and the time of analysis was defined by counting a pre-determined bead number (300 counts). The percentage of PMPs positive for PE-annexin V in the cells of the gate R3 was evaluated on histograms (Fig. 1C) constructed by the analysis software, CellQuest (Beckton Dickinson). The bar was determined as the specific positive region of PE-annexin V. The particles in the R3 were designated as crude PMPs. The absolute number of PMPs stained by annexin V was calculated by the following formula: Counts of specific PMPs ⫽ (crude PMP counts in the gate R3) ⫻ (percentage of the stained cells in the gate R3) ⫻ 0.01. These specific PMPs with procoagulant activity were

Ando et al: PMP formation in uremic patients

1759

Fig. 2. Comparison of PMP counts between three types of uremic patients (HD, CAPD, and pre-dialyzed uremic patients) and controls. Each patient population was matched for age with the controls. The significant differences were obtained between each uremic group and the controls. The PMP counts were not different among three types of uremic patients. Statistically significant difference is marked. **P ⬍ 0.01 vs. control.

simply designated as “PMPs” in the following description unless otherwise stated. The study was approved by the institutional review board of the hospital and was carried out according to the principles of the Declaration of Helsinki. Informed consent was obtained from all of the subjects included in this study. Statistics Data were expressed as mean ⫾ SD unless otherwise stated. The differences between mean values were evaluated by the two-tailed Student t test for comparing two groups and by the analysis of variance (ANOVA) followed by Bonferroni-Dunn for comparing more than three groups. P ⬍ 0.05 was considered statistically significant. RESULTS PMP counts in uremic patients The PMP counts were studied in 89 non-DM uremic patients and 29 age-matched controls. The PMP counts were significantly higher in each uremic group (pre-dialyzed, 142.3 ⫾ 49.5; HD, 195.3 ⫾ 103.1; CAPD, 161.0 ⫾ 47.9) than in the controls (104.9 ⫾ 46.3). The PMP counts were not different among three types of uremic groups (Fig. 2). The PMP counts did not correlate with platelet

Fig. 1. (A) Cells stained with monoclonal antibody (mAb) to FITCGPIb were gated as R1 on the side scatter (SSC) dot plot versus the fluorescence-1 (FL-1) profile. R1 was determined by referring to the SSC versus FL-1 profile of the samples treated with non-labeled mAb to GPIb. (B) The cells in R1 were reviewed on SSC dot plot versus the forward scatter (FSC) dot plot profile. Regions corresponding to platelet-derived microparticles (PMPs; R3) were discriminated by size

from the platelet islet shown as R2. (C ) The percentage of PMPs positive for PE-annexin V in the cells of R3 was evaluated with a histogram analysis. The bar was determined as the specific positive region of PE-annexin V by comparison with the histograms of the samples treated with non-labeled Ab to annexin V.

1760

Ando et al: PMP formation in uremic patients

Fig. 3. Comparison of PMP counts between uremic patients with thrombotic events and those without such events. Both patients were matched for age. PMP counts were significantly higher in the uremic group with thrombotic events than in those who had not experienced thrombotic events. Statistically significant difference is marked. **P ⬍ 0.01 vs. uremic group without thrombotic events.

counts and serum levels of albumin, urea nitrogen, and creatinine in any uremic group. PMP counts in uremic patients with and without thrombotic events A clinical link between PMP counts and thrombotic events were examined in all uremic subjects. The thrombotic events were retrospectively investigated. Overall, the number of patients who had suffered thrombotic events for the last five years was 15 out of 89 patients (17.4%; 9 of 46 HD patients, 3 of 23 CAPD patients, and 3 of 20 pre-dialyzed patients). The PMP counts in the patients with thrombotic events (248.5 ⫾ 91.1) were significantly greater than in the patients without such events (159.5 ⫾ 74.5; Fig. 3). Effects of rHuEPO on the PMP count in dialysis patients The PMP counts were compared between the HD patients receiving chronic rHuEPO treatment (recipients; 220.0 ⫾ 106.5, N ⫽ 30) and those receiving no such treatment (non-recipients; 148.9 ⫾ 80.5, N ⫽ 16). The PMP counts of the EPO recipients were significantly higher than those of the EPO non-recipients (Fig. 4). Both groups were matched for age (59.2 ⫾ 12.8 vs. 55.4 ⫾ 11.4 years old). The mean dose of rHuEPO used was 2380 ⫾ 921 units a week. Moreover, the significant difference in the PMP counts between the EPO recipients (178.7 ⫾ 50.3, N ⫽ 13) and non-recipients (138.0 ⫾ 34.3, N ⫽ 10) was found in the subgroup of CAPD (Fig. 4). This comparative study was not done in the subgroup of pre-dialysis patients because of the small number of

Fig. 4. Comparisons of platelet-derived microparticle (PMP) counts between recombinant human erythropoietin (EPO) recipients and nonrecipients in dialysis patients. The significant differences between the recipients and non-recipients were obtained in each subgroup of dialysis patients. Statistically significant difference is marked. *P ⬍ 0.05 vs. nonrecipients.

rHuEPO recipients, as shown in Table 1. No significant correlation was obtained between the PMP counts and rHuEPO dose administered in each dialysis subgroup. Regardless of rHuEPO use, the PMP counts were significantly higher in the EPO non-recipients of HD (N ⫽ 30; 148.9 ⫾ 80.5), CAPD (N ⫽ 10, 138.0 ⫾ 34.3), and predialyzed (N ⫽ 18; 133.1 ⫾ 42.3) patients than in the controls. In addition, PMP counts were significantly higher (P ⫽ 0.0260) even in the uremic patients without thrombotic events receiving no rHuEPO treatment (N ⫽ 38, 131.6 ⫾ 48.7) than in the controls (N ⫽ 29, 104.9 ⫾ 46.3). Influence of HD therapy on PMP generation Changes in PMP counts during HD treatment. The time courses in PMP counts were examined in 14 HD patients before and 10 minutes, 30 minutes, and 2, 3, and 4 hours after starting the HD treatment. Blood samples were collected from the outlet port of the HD circuit tube before the dialysis column. The counts of PMP did not significantly change during HD treatment. The data are shown in Figure 5A. Changes in PMP count before and after dialysis columns. The changes of PMP counts before and after dialysis columns were specially examined to better understand the influence of blood contact with membrane materials on PMP counts at two hours after the start of HD. No significant changes in PMP counts were observed before and after the columns (before, 188.7 ⫾ 22.3; after, 193.1 ⫾ 41.5; Fig. 5B). The HD membrane materials used were polysulfone (N ⫽ 10) and cellulose triacetate (N ⫽ 4). There were no significant differences

1761

Ando et al: PMP formation in uremic patients Table 1. Subject profiles

Number Age years Dialysis duration years rHuEPO recipients/non-recipients Hematocrit % Platelet counts ⫻ 109/L Serum albumin g/dL Serum urea nitrogen mg/dL Serum creatinine mg/dL

Pre-dialyzed patients

HD patients

CAPD patients

Controls

20 54.8 ⫾ 10.1 — 2/18 31.9 ⫾ 3.8 225.3 ⫾ 60.1 3.5 ⫾ 0.73 53.4 ⫾ 8.89a 6.2 ⫾ 4.8a

46 57.9 ⫾ 12.3 12.3 ⫾ 6.7 30/16 32.1 ⫾ 7.2 196.2 ⫾ 75.4 3.4 ⫾ 1.8 80.2 ⫾ 11.1 10.3 ⫾ 2.8

23 50.8 ⫾ 8.31a 5.8 ⫾ 3.2a 13/10 31.8 ⫾ 6.6 214.8 ⫾ 61.0 3.1 ⫾ 1.7 56.6 ⫾ 10.2a 11.1 ⫾ 3.8

29 52.3 ⫾ 5.19 — — 44.1 ⫾ 5.3 232.6 ⫾ 51.3 4.3 ⫾ 0.8 17.3 ⫾ 1.81 1.1 ⫾ 0.1

Data are expressed as mean ⫾ SD. Abbreviations are: HD, hemodialysis; CAPD, continuous ambulatory peritoneal dialysis; rhuEPO, recombinant human erythropoietin. a P ⬍ 0.05 versus HD patients

Table 2. Effects of internal arteriovenous (AV) fistula on platelet-derived microparticles (PMP) counts Parameters PMP counts (N ⫽ 10)

AV fistula side

Non-AV fistula side

213.8 ⫾ 80.6

205.5 ⫾ 88.56

Data are expressed as mean ⫾ SD.

ence in the PMP counts between the blood samples from the arms with and without AV fistula (Table 2). DISCUSSION

Fig. 5. (A) Time course changes of PMP counts in circulation during four-hour HD treatments. PMPs were measured before HD, and 10 minutes, 30 minutes, 2 hours, 3 hours, and 4 hours after the HD treatment. Blood samples were collected from the artery-side port of circuit tube before HD column. The PMP counts did not significantly change during HD treatment. (B) Changes in the PMP counts before and after HD column of different membrane materials. PMP counts were specially measured before and after the HD column at 2 hours after HD treatment started. No significance between them was observed. These studies were performed in 14 HD patients who used polysulfone (N ⫽ 10) and cellulose triacetate (N ⫽ 4) membrane materials. Data are expressed as mean ⫾ SE.

between the data of two types of membrane materials (data not shown). Influence of internal AV fistula on PMP counts This examination was performed in 10 HD patients with an internal AV fistula. Blood samples were obtained from upstream veins close to AV fistula and almost simultaneously obtained from cubital veins of the opposite arm without AV fistula. There was no significant differ-

The major findings of this study are (1) the PMP counts were significantly elevated in a resting state in uremic patients such as pre-dialyzed, HD, and CAPD patients, compared with the age-matched healthy subjects, the PMP counts were not significantly different among three types of uremic groups; (2) the HD procedure and existence of internal AV fistula did not affect the PMP counts, but rHuEPO treatment was possibly involved in the further increase of PMP release; and (3) the PMP counts in the uremic patients who had experienced thrombotic events were much higher than those in the patients without such events. This result indicates that an extreme increase of PMP counts may be responsible for a high incidence of thrombosis in uremic patients. However, the primary cause of PMP elevation in uremia was not clarified in this study. Our data showed that the counts of PMP enriched with procoagulant activity are elevated in non-diabetic uremic patients, as compared with the controls, regardless of uremic stages or dialysis modality. Nomura et al previously showed that PMP counts were increased in uremic patients [31, 32], which was likely to support our current results. However, their data were deficient in several points. The profiles of their uremic subjects did not include the stage of uremia, modality of dialysis, ages, and complications (for example, diabetes mellitus). Moreover, their data were based upon so-called crude PMPs that might include a lot of debris and non-func-

1762

Ando et al: PMP formation in uremic patients

tioning PMPs, as we described in our Methods section. In contrast, we clarified the patients’ characteristics and identified vesicles with procoagulant activity as PMPs. We believe that our current report is comprehensive and elaborates on the clinical significance of PMPs in uremic patients. Our study shows the biological relevance of elevated PMP counts in uremia. PMPs provide competent surfaces for the assembly of coagulation factors and may interact with components of the vessel walls, contributing to thrombus and atheroma formation [13, 14, 23–26]. Although the clinical significance of circulating PMPs in the pathogenesis of thrombotic diseases is not sufficiently acknowledged yet, recent clinical evidence indicates that an increase of circulating PMP counts is involved in the incidence of ischemic heart diseases and cerebral strokes [20–22]. Our data show a positive link between PMP counts and thrombotic events in uremic patients, suggesting that an increase of PMP release in the circulatory system underlies the high incidence of thrombotic diseases in uremic patients [7, 8]. Alternatively, PMP could be a new marker or predictor of clinical thrombosis in uremia. We expected that changes of PMP release might contribute to the pathogenesis of intra-dialyzer coagulation and vascular access thrombosis in HD patients. Several studies suggested that mechanical stress, particularly high shear stress by a cone-plate viscounter, stimulates the shedding of PMPs from platelets [15, 18]. Thus, we speculated that similar phenomena would occur in the HD procedure that may impose repetitive mechanical stress on platelets during extracorporeal circulation. However, the data proved opposite to this speculation. AV fistulas also were expected to stimulate platelets and to facilitate PMP production by physical mechanism induced while arterial blood ran through narrow anastomosis [33, 34], but the data did not support this speculation either. These studies suggest that PMP’s actions may be irrelevant to common thrombotic episodes associated with HD treatment. The PMP counts of EPO non-recipients in each uremic group were greater than those of the controls, showing that the basal release of PMP with the subject in a resting state was elevated in uremia. However, the PMP counts of EPO recipients were much higher than those of EPO non-recipients in both HD and CAPD subgroups (Fig. 4). This fact allowed us to speculate that rHuEPO may serve as an enhancer of PMP release. While rHuEPO is known to increase platelet numbers and platelet reactivity to several stimuli [35], there are no reports showing the clinical relevance of rHuEPO to PMP production. Therefore, further examinations are needed to consolidate our speculation, such as, longitudinal observation of PMP counts after commencing rHuEPO therapy, and studies are now underway in our laboratory.

We propose that a uremic milieu itself may cause an increase in circulating PMPs. High shear stress by a coneplate viscounter, which is thought to be most involved in the in vitro PMP formation, is analogous with clinical levels of fluid shear stress that may occur in atherosclerotic arteries [15–19]. It would be natural to think that such physiological stress is easy to induce in uremic patients who commonly have hypertension and systemic arterio-atherosclerosis [8]. Alternatively, unknown uremic substances that are accumulated despite or by repetitive dialysis treatment may additionally impact the process of PMP formation from activated platelets. In conclusion, the counts of circulating PMPs were elevated in three types of uremic patients compared with the healthy controls. The primary cause of PMP elevation was not clarified, but rHuEPO treatment could be involved as an enhancer of PMP release. In a uremic setting overlapped by unfavorable factors such as uremic toxins, pro-inflammatory cytokines, hypertension, and systemic atherosclerosis, an increase of PMPs may be a potential trigger of acute thrombotic accidents, even under circumstances where the hemostatic functions of platelets are impaired. ACKNOWLEDGMENTS We thank the research staff at Otsuka Pharmaceutical Co., Ltd. (Tokushima, Japan) for their excellent technical support. Reprint requests to Minoru Ando, M.D., Division of Nephrology, Department of Medicine, Tokyo Metropolitan Geriatric Hospital, 35-2, Sakae-Cho, Itabashi-Ku, Tokyo, 173-0015, Japan. E-mail: [email protected]

REFERENCES 1. Castillo R, Lozano T, Escolar G, et al: Defective platelet adhesion on vessel subendothelium in uremic blood. Blood 68:337– 342, 1986 2. Escolar G, Cases A, Bastida E, et al: Uremic platelets have a functional defect affecting the interaction of von Willebrand factor with glycoprotein IIb-IIIa. Blood 76:1336–1340, 1990 3. Sloand EM, Sloand JA, Prodouz K, et al: Reduction of platelet glycoprotein Ib in uraemia. Br J Haematol 77:375–381, 1991 4. Gawaz MP, Dobos G, Spath M, et al: Impaired function of platelet membrane glycoprotein IIb-IIIa in end-stage renal disease. J Am Soc Nephrol 5:36–46, 1994 5. Remuzzi G: Bleeding in renal failure. Lancet 1:1205–1208, 1988 6. Eknoyan G, Wacksman SJ, Glueck HI, et al: Platelet function in renal failure. N Engl J Med 280:677–681, 1969 7. Bloembergen WE, Port FK, Mauger EA, et al: A comparison of mortality between patients treated with hemodialysis and peritoneal dialysis. J Am Soc Nephrol 6:177–183, 1995 8. London GM, Druke TB: Atherosclerosis and arteriosclerosis in chronic renal failure. Kidney Int 51:1678–1695, 1997 9. Viener AM, Aviram M, Better OS, et al: Enhanced in vitro platelet aggregation in hemodialysis patients. Nephron 43:139–143, 1986 10. Nakamura Y, Chida Y, Tomura S: Enhanced coagulation-fibrinolysis in patients on regular hemodialysis treatment. Nephron 58: 201–204, 1991 11. Barton CH, Vaziri ND: Central retinal vein occlusion associated with hemodialysis. Am J Med Sci 277:39–46, 1979 12. Turney JH, Woods HF, Fewell MR, et al: Factor VIII complex in uraemia and effects of haemodialysis. BMJ 282:1653–1656, 1981 13. Sims PJ, Faioni EM, Wiedmer T, et al: Complement proteins C5b-9

Ando et al: PMP formation in uremic patients

14. 15. 16. 17. 18. 19.

20.

21.

22.

23.

24.

cause release of membrane vesicles from the platelet surface that are enriched in the membrane receptor for coagulation factor Va and express prothrombinase activity. J Biol Chem 263:18205– 18212, 1988 Sims PJ, Wiedmer T, Esmon CT, et al: Assembly of the platelet prothrombinase complex is linked to vesiculation of the platelet plasma membrane. J Biol Chem 264:17049–17057, 1989 Ikeda Y, Handa M, Kawano K, et al: The role of von Willebrand factor and fibrinogen in platelet aggregation under varying shear stress. J Clin Invest 87:1234–1240, 1991 O’Brien JR: Shear-induced platelet aggregation. Lancet 335:711– 712, 1990 Butler J: Shear stress platelet activation [Letter]. Lancet 346: 841, 1995 Miyazaki Y, Nomura S, Miyake T, et al: High shear stress can initiate both platelet aggregation and shedding of procoagulant containing microparticles. Blood 88:3456–3464, 1996 Holme PA, Orvim U, Hamers MJAG, et al: Shear-induced platelet activation and platelet microparticle formation at blood flow conditions as in arteries with a severe stenosis. Arterioscler Thromb 17: 646–653, 1996 Geiser T, Sturzenegger M, Genewein U, et al: Mechanisms of cerebrovascular events as assessed by procoagulant activity, cerebral microemboli, and platelet microparticles in patients with prosthetic heart valves. Stroke 29:1770–1777, 1998 Katopodis JN, Kolodny L, Jy W, et al: Platelet microparticles and calcium hemostasis in acute coronary ischemias. Am J Hematol 54:95–101, 1997 Gawaz M, Neumann JF, Ott I, et al: Platelet function in acute myocardial infarction treated with direct angioplasty. Circulation 93:229–237, 1996 Barry OP, Pratico D, Savani RC, et al: Modulation of monocyteendothelial cell interactions by platelet microparticles. J Clin Invest 102:136–144, 1998 Mallat Z, Hugel B, Ohan J, et al: Shed membrane microparticles

25. 26. 27.

28. 29. 30. 31. 32. 33. 34. 35.

1763

with procoagulant potential in human atherosclerotic plaques. A role for apoptosis in plaque thrombogenicity. Circulation 99:348– 353, 1999 Barry OP, Pratico D, Lawson JA, et al: Transcellular activation of platelets and endothelial cells by bioactive lipids in platelet microparticles. J Clin Invest 99:2118–2127, 1997 Iwamoto S, Kawasaki T, Kambayashi J, et al: The release mechanism of platelet-activating factor during shear-stress induced platelet aggregation. Biochem Biophys Res Commun 239:101–105, 1997 Dachary-Prigent J, Freyssinet J-M, Pasquet J-M, et al: Annexin V as a probe of aminophospholipid exposure and platelet membrane vesitation: A flow cytometry study showing a role for free sulfhydryl groups. Blood 81:2554–2565, 1993 Yanabu M, Ozaki Y, Nomura S, et al: Tyrosine phosphorylation and p72syk activation by an anti-glycoprotein Ib monoclonal antibody. Blood 89:1590–1598, 1997 Suzuki M, Kawakatsu T, Miyake T, et al: Platelet-derived microparticles in alloxan-induced diabetes in rabbits. Haemostasis 26: 228–232, 1996 Cobes V, Dignat-George F, Mutin M, et al: A new flow cytometry method of platelet-derived microvesicle quantitation in plasma. (Letter) Thromb Haemost 77:220, 1997 Nomura S, Shouzu A, Nishikawa M, et al: Significance of plateletderived microparticles in uremia. (Letter) Nephron 63:485, 1993 Nomura S, Komiyama Y, Miyake T, et al: Amyloid ␤-protein precursor-rich platelet microparticles in thrombotic disease. Thromb Haemost 72:519–522, 1994 Girerd X, London G, Boutouyrie P, et al: Remodeling of the radial artery in response to a chronic increase in shear stress. Hypertension 27(part 2):799–803,1996 Driss AB, Benessiano J, Poitevin P, et al: Arterial expansive remodeling induced by high flow rates. Am J Physiol 272:H851– H858, 1997 Stohlawetz PJ, Dzirlo L, Hergovich N, et al: Effects of erythropoietin on platelet reactivity and thrombopoiesis in humans. Blood 95:2983–2989, 2000