Procoagulant Human Monocytes Mediate Tissue Factor/Factor VIIa–Dependent Platelet-Thrombus Formation When Exposed to Flowing Nonanticoagulated Human Blood
Abstract Tissue factor (TF) on monocyte and macrophage surfaces is a nonproteolytic cofactor for factor VIIa (FVIIa)–induced coagulation. Monocyte-derived macrophages in atherosclerotic plaques express TF, which, after plaque disruption or rupture, may complex with FVII/VIIa from the bloodstream, resulting in activation of extrinsic coagulation. We studied the effect of TF expression on human monocytes on arterial thrombus formation in a model system of thrombogenesis. Thawed, cryopreserved human monocytes adherent to plastic coverslips were stimulated with lipopolysaccharide (0.5 μg/mL) to express TF and subsequently exposed to flowing nonanticoagulated human blood in a parallel-plate perfusion chamber. The wall shear rate at the cell surface was 650 seconds−1, corresponding to that of average-sized coronary arteries. The stimulated monocytes elicited pronounced fibrin deposition and platelet-thrombus formation. The platelet-thrombus volume was as large as that triggered by human type III collagen fibrils under similar experimental conditions. In contrast, the monocytes elicited much more fibrin deposition than the collagen surface. However, inclusion of an anti-TF monoclonal antibody that blocks the complexation of FVII/FVIIa with TF virtually abolished the fibrin deposition (P<.03) and reduced platelet-thrombus formation by more than 70% (P<.04). Thus, arterial thrombus formation induced by stimulated monocytes was almost completely blocked by the anti-TF antibody, suggesting that inhibition of TF/FVIIa complex formation on monocytes and macrophages at sites of plaque rupture or after percutaneous transluminal coronary angioplasty procedures may reduce intravascular thrombotic complications.
- Received June 1, 1994.
- Accepted October 20, 1994.
Tissue factor (TF) is a cell-surface transmembrane protein with an extracellular domain that complexes with coagulation factor VIIa (FVIIa) and functions as a cofactor for the catalytic activity of the FVIIa moiety. Such catalytic complexes form at vascular lesions, resulting in proteolytic activation of factors IX and X, subsequent thrombin generation, and eventually fibrin formation and deposition.1 2 3 Platelets are activated concomitantly by thrombin through the platelet-thrombin receptor,4 which enhances platelet recruitment at the lesion.
TF is apparently present in subendothelium,5 6 and disruption of the endothelial layer of blood vessels in experimental animal models activates coagulation and triggers thrombus formation. Inclusion of monoclonal anti-TF antibodies (MAbs) efficiently blocks this thrombus formation,7 8 indicating a prominent role for TF/FVIIa–dependent coagulation in thrombus formation, at least in these models. Furthermore, TF expressed by stimulated endothelium9 and TF present in the endothelial extracellular matrix10 11 activate coagulation and induce thrombus formation at venous9 and arterial10 11 blood flow conditions because anti-TF MAbs efficiently block both coagulation and thrombus formation on these surfaces.9 10 11
Macrophages dominate the lipid core of atherosclerotic plaques,12 and these cells have been shown to express TF.13 14 These observations led us to study the effect of TF expression on immobilized peripheral human blood monocytes on arterial thrombus formation in flowing native human blood. Thawed, cryopreserved human peripheral blood monocytes15 (±lipopolysaccharide stimulation) adherent to plastic coverslips were placed in a parallel-plate perfusion chamber16 17 and exposed to native human blood for 5 minutes at an arterial wall shear rate of 650 seconds−1. The procoagulant monocytes elicited pronounced fibrin deposition and platelet-thrombus formation that was dependent on the expression of TF.
Peripheral blood mononuclear cells from healthy subjects were isolated by Lymphoprep (Nycomed Pharma), frozen in aliquots of 6×106 cells/mL at −135°C, and thawed according to the method described by Osnes et al.15 This cryopreservation does not affect the potential for TF synthesis. Mononuclear cells were suspended in RPMI-1640 (Whittaker) with 5% fetal calf serum (FCS) (GIBCO, Life Technologies Ltd), and the number of cells was adjusted to 2.4×106/mL, giving approximately 30% monocytes, as determined with nonspecific esterase staining (Bürker-chamber) and flow cytometry with CD14. Two milliliters of the cell suspension (4.8×106 mononuclear cells) were seeded onto 4-cm2 plastic coverslips (Thermanox, Miles Laboratories) placed in multiwell (six-well) plates (NUNC AS). The incubation medium was RPMI-1640 (Whittaker) supplemented with 5% FCS (GIBCO)±0.5 μg/mL lipopolysaccharide from E. Coli OB 55 (Sigma Chemical Co). Incubation was at 37°C and 5% CO2 for 2 or 4 hours. Subsequently, the coverslips were washed three times in phosphate-buffered saline (PBS, 3×2 mL) (Whittaker) to remove nonadherent cells. The adherent cells on the coverslips consisted of more than 90% monocytes, as judged by nonspecific esterase staining.
The coverslips with adherent lipopolysaccharide-stimulated monocytes were incubated for 30 minutes in 2.5 mL fresh RPMI-1640 with or without 25 μg/mL anti-TF MAb (MAb AD 4509, American Diagnostica). This anti-TF MAb blocks the complexation of FVII/FVIIa to TF. The adherent cells were washed three times in RPMI-1640 and immediately used in blood perfusion experiments (see below) or in a one-stage coagulation assay (Thrombotrack 4, Nycomed Pharma) (see below).
The number of adherent monocytes on the coverslips after 2 or 4 hours of incubation and after a 5-minute perfusion was determined. Each coverslip was photographed at a magnification of ×63 with a light microscope (Zeiss Axiophot 405M) in combination with a video camera (DXC-750D, Sony Corp) and color video printer (UP-5000p, Sony) at the upstream edge before and after perfusion experiments with RPMI-1640. The number of adherent monocytes was subsequently counted from the printed images. The monocyte count on the coverslips was similarly determined after 5-minute perfusions with native blood (larger viscous forces) over nonstimulated monocytes and lipopolysaccharide-stimulated and anti-TF MAb–incubated monocytes (2 hours of incubation). The relatively modest thrombus formation under these experimental conditions did not interfere with identification of the Thermanox-adherent monocytes and the cell counting. However, the number of adherent cells after 4 hours of incubation and blood perfusion could not be determined because of thrombus formation that obscured the view of the Thermanox-adherent monocytes.
Cell viability after thawing, seeding on coverslips, lipopolysaccharide stimulation, and incubation with the anti-TF MAb was determined by the trypan blue exclusion test.
Testing for Endotoxin
The growth medium of RPMI-1640 and FCS is tested for endotoxin contamination by the manufacturers and has been routinely tested for endotoxin at the Department of Clinical Chemistry at Ullevål University Hospital, Oslo. The levels of endotoxin have been below the detection limit of 30 pg/mL.18 Thermanox coverslips, provided sterile from the manufacturer, were kept in 96% ethanol to avoid bacterial contamination. Each coverslip was handled under sterile conditions in a laminar air-flow hood. No detectable endotoxin (<30 pg/mL) was measured in growth medium incubated for 4 hours in multiwells in the presence or absence of coverslips.18 However, we cannot exclude the possibility of trace amounts of endotoxin present.
Monocyte Procoagulant Activity
The procoagulant activity of adherent nonstimulated and adherent lipopolysaccharide-stimulated monocytes with and without the inclusion of an anti-TF MAb was determined by a one-stage clotting assay. The cells were scraped off the Thermanox coverslips with a rubber policeman (NUNC AS) and resuspended in 400 μL PBS (Whittaker). Subsequently, 50 μL cell suspension (four parallels) was preincubated with 50 μL of 25 mmol/L CaCl2 for 2 minutes (37°C). Clotting time (in seconds) was measured automatically after addition of 50 μL of a pool of normal human plasma (37°C) by a Thrombotrack 4 coagulometer (Nycomed Pharma). Six dilutions of a standard TF preparation from rabbit brain (Nycomed Pharma) was used to make a standard curve.
The protein content of the resuspended cells (200 μL) was determined by a Micro BCA Protein Assay Reagent Kit according to the manufacturer (Pierce Chemical Co). The monocyte procoagulant activity was expressed in arbitrary units of TF activity per microgram of protein because the anti-TF MAb blocked virtually all procoagulant activity. A control MAb (Dako M734, mouse anti-human transferrin receptor MAb, Dacopatts) of the same subclass (IgG1) as the anti-TF MAb was included in the coagulation studies.
The blood donors were healthy nonsmoking individuals, who denied any drug intake for at least 14 days before the blood perfusion experiments. Hematologic parameters (hemoglobin, hematocrit, and leukocyte and platelet counts) were within the normal ranges (Auto Counter AC 920, Swelab Instrument).
Ex vivo perfusion experiments19 were performed in the parallel-plate perfusion model of thrombogenesis, as described in detail previously.16 17 20 A Thermanox plastic coverslip with adherent monocytes was positioned in the perfusion chamber and subsequently exposed to flowing nonanticoagulated human blood for 5 minutes. An antecubital vein was punctured with a 19-gauge butterfly infusion set (Abbot Lab), and the blood was drawn (10 mL/min) directly from the vein through the perfusion chamber by an occlusive roller pump (Gilson Minipuls 3, Villiers Le Bell) placed distally to the perfusion chamber. The wall shear rate at the cell surface was 650 seconds−1. During the perfusion experiments, blood was sampled downstream from the adherent monocytes to assess the plasma levels of fibrinopeptide A (FPA) and β-thromboglobulin (β-TG) after 4 and 4.5 minutes, respectively, of perfusion time (see below). A rubber sleeve wrapped around Silastic tubing (Dow Corning Corp) immediately distal to the perfusion chamber allowed repeated punctures of the tubing without leakage. Each perfusion was terminated with a 20-second perfusion (10 mL/min) with RPMI-1640 at 37°C, followed by a 40-second perfusion (10 mL/min) with freshly prepared fixation solution consisting of 2.5% glutaraldehyde per 0.1 mol/L cacodylate in deionized water (pH 7.4) at 23°C. The specimens were stored in 7% sucrose per 0.1 mol/L cacodylate buffer at 4°C and finally embedded in Epon.21
Semithin Epon sections (1 μm) were prepared perpendicular to the direction of the blood flow 1 mm from the upstream edge of the coverslip.22 The sections were stained with toluidine blue and basic fuchsin. Thrombus formation was quantified as percent surface coverage with fibrin by light microscopy at ×1000 magnification (Standard 25, Zeiss),19 and platelet-thrombus volume (micrometers cubed per micrometers squared) was derived from thrombus area (micrometers squared per micrometer) assessed by computer-assisted morphometry (Kontron Vidas, Eching) at ×2500 magnification, as previously described.23
FPA and β-TG Plasma Levels
The plasma levels of FPA and β-TG were determined in blood samples (0.9 mL) drawn immediately distal to the perfusion chamber. The blood samples were drawn directly into 2-mL syringes (Becton Dickinson) prefilled with 0.1 mL anticoagulant consisting of 1000 IU heparin (LEO, Baldrup) and 1000 KIU Trasylol (Bayer Leverkusen, Germany) per 1 mL saline for FPA and an anticoagulant according to Ludlam and Cash24 for β-TG. Further processing of the blood samples was essentially according to the manufacturers of the respective radioimmunoassay kits: FPA from Imco and β-TG from Amersham, UK.
Statistical analysis was performed with two-tailed Student’s t test and Mann-Whitney U test. A value of P<.05 was considered significant.
Monocyte-Thermanox Adhesion and Cell Viability
Monocytes adhered singly and randomly to the plastic coverslips (Thermanox). Cell aggregate formation on the coverslip was not encountered. The cell density after 2 hours of incubation was on average 500 cells/mm2 (Table 1⇓). When adherent monocytes were perfused with RPMI-1640 for 5 minutes, approximately 50 cells/mm2 were washed off, corresponding to 10% cell loss (n=4, P<.01). Perfusion with native human blood, which has a higher viscous force, increased the cell loss to about 75 cells/mm2, corresponding to 15% cell loss (n=8, P<.05). Neither stimulation with lipopolysaccharide nor inclusion of the anti-TF MAb influenced cell adhesion. However, the number of adherent monocytes increased by 67% to approximately 840 cells/mm2 (n=8, P<.01) when the incubation period was prolonged from 2 to 4 hours. The trypan blue exclusion test revealed less than 1% cell death after the incubation periods and before the perfusion experiments.
Clotting Assays: TF/FVIIa–Induced Coagulation
Nonstimulated monocytes adherent to plastic coverslips for 2 hours showed very little procoagulant activity, whereas a threefold increase in activity was observed after 4 hours of adhesion (P<.0001) (Fig 1⇓). The anti-TF MAb reduces this activity by more than 90% (n=4, P<.005). In contrast, the lipopolysaccharide-stimulated adherent monocytes elicited considerable procoagulant activity after incubation for 2 hours and almost twice as much after 4 hours (P<.0001). When adherent cells stimulated with lipopolysaccharide for 2 or 4 hours were incubated with the anti-TF MAb, the coagulant activity was almost abolished (P<.0001). Inclusion of the anti–transferrin receptor MAb did not significantly affect the cellular procoagulant activity.
Perfusion studies with Thermanox-adherent monocytes using nonanticoagulated human blood were performed for 5 minutes at a wall shear rate of 650 seconds−1. Thrombus formation was morphometrically assessed as percent surface coverage with fibrin (fibrin deposition) and as platelet-thrombus volume.
Fig 2A⇓ gives data on monocyte, collagen, and Thermanox-induced fibrin deposition. Lipopolysaccharide-stimulated cells triggered substantial fibrin deposition, which was most pronounced on and around cells stimulated for 4 hours (P<.03). Fibrin deposition was almost completely abolished when the cells were incubated with the anti-TF MAb (P<.03). Nonstimulated cells elicited virtually no detectable fibrin deposition. When blood was perfused over uncoated Thermanox coverslips or over coverslips coated with type III collagen fibrils, virtually no detectable fibrin deposition was observed.
Fig 2B⇑ summarizes the data on platelet-thrombus volume elicited by monocytes, collagen fibrils, and Thermanox. Blood perfusions of 2-hour-adherent and nonstimulated monocytes resulted in very little platelet-thrombus formation, whereas 4-hour-adherent and nonstimulated monocytes triggered pronounced platelet-thrombus formation. Monocytes stimulated with lipopolysaccharide for 2 hours triggered pronounced formation of platelet thrombi compared with the nonstimulated cells (P<.03). There was no difference in platelet-thrombus formation on cells stimulated with lipopolysaccharide for 2 versus 4 hours or between monocytes that were not stimulated and those stimulated for 4 hours.
However, inclusion of the anti-TF MAb inhibited platelet-thrombus formation on cells stimulated 2 and 4 hours with lipopolysaccharide by 86% (P<.05) and 71% (P<.01), respectively.
Type III collagen fibrils triggered platelet-thrombus formation as efficiently as the monocytes. However, no platelet thrombi formed on uncoated Thermanox coverslips.
Plasma levels of FPA were measured distal to the perfusion chamber (Table 2⇓). There was no significant difference in the average plasma levels of FPA after perfusions of cells stimulated for 2 hours with lipopolysaccharide or nonstimulated cells. Nor did blocking of the TF/FVIIa complex formation with the anti-TF MAb produce any significant change. The average plasma levels of FPA increased with time of monocyte adherence to the Thermanox and with time of lipopolysaccharide stimulation, although not significantly. However, prolonged time of adherence with and without stimulation resulted in about a twofold increase in adherent cells, which presumably affected the FPA levels.
Inclusion of the anti-TF MAb reduced the average FPA levels but not significantly.
Plasma levels of β-TG were measured distal to the perfusion chamber (Table 2⇑). The levels were increased (P<.01) after 4 hours of polysaccharide stimulation of monocytes compared with the nonstimulated cells. The increased β-TG levels paralleled the increase in platelet-thrombus volume (Fig 2B⇑). In contrast, no increase was observed after perfusion of cells stimulated for 2 hours, despite the fact that the measured thrombus volume was increased significantly over that of nonstimulated cells (Fig 2B⇑).
Inclusion of the anti-TF MAb significantly decreased the plasma β-TG levels for cells stimulated for 4 hours with lipopolysaccharide (P<.04) but not for the cells stimulated for 2 hours.
Monocytes adhered singly and randomly to the Thermanox coverslips. However, confluence at 2 or 4 hours of incubation was not achieved with the number of cells seeded, thus leaving some Thermanox material exposed to blood flow. The platelet thrombi always formed in association with adherent monocytes and deposited fibrin, giving the en face appearance of more clustering of the thrombotic deposits than on collagen fibrils (Fig 3⇓). Inspection of the sections revealed fibrin strands sprouting out from monocytes and frequently covering parts of the uncoated Thermanox coverslip. Platelet thrombi formed on fibrin strands in close proximity to adherent monocytes (Fig 4A⇓) or apparently directly on the monocyte surface (Fig 4B⇓). Occasionally, fibrin strands formed on lipopolysaccharide-stimulated monocytes incubated with the anti-TF MAb, but in general fibrin deposition and platelet-thrombus formation were suppressed (Fig 4C⇓).
The normal vessel wall possesses TF that complexes with FVII/FVIIa from plasma and acts as a cofactor for activation of the extrinsic coagulation.5 6 TF also is found in atherosclerotic plaques25 and generally in association with monocyte-derived macrophages.13 14 The location of TF in healthy and sick vascular compartments separated from the blood flow by the endothelial layer only and the apparent role of FVII/FVIIa in thrombogenesis26 have spurred much interest in TF/FVIIa–induced coagulation and thrombosis.
The role of TF associated with lipopolysaccharide- or cytokine-stimulated endothelium or corresponding extracellular endothelial matrices,9 10 normal arterial intima, or deeper compartments of the vessel wall7 8 has been investigated in various studies of thrombogenesis. Blocking TF/FVIIa activity by anti-TF MAbs interrupts this thrombus formation efficiently.7 8 9 10 11 However, peripheral blood monocytes also express TF when stimulated, and together with TF of monocyte-derived macrophages in arterial plaques, these cells may provide the primary source of TF in several pathological conditions,27 28 such as disseminated intravascular coagulation, plaque rupture, and plaque disruption after percutaneous transluminal coronary angioplasty (PTCA) procedures. Therefore, we found it of interest to investigate the role of TF expression of monocytes15 in arterial thrombus formation. For that purpose, human monocytes were immobilized on plastic coverslips and exposed to flowing nonanticoagulated human blood in a parallel-plate perfusion chamber device16 17 20 at an arterial wall shear rate of 650 seconds−1.
We observed that lipopolysaccharide-stimulated monocytes elicited pronounced fibrin deposition and platelet-thrombus formation. However, inclusion of an anti-TF MAb, which blocks the complexation of TF with FVII/FVIIa, almost abolished fibrin deposition and efficiently reduced platelet-thrombus formation. The procoagulant state of these cells was also confirmed by the one-stage clotting assay and by the inclusion of the same anti-TF MAb, which efficiently blocked the procoagulant activity. An indifferent antibody of the same subclass had no significant effect, indicating that the anti-TF MAb specifically affected the monocyte TF procoagulant activity only. Thus, the monocyte procoagulant activity and its significant impact on arterial thrombus formation were predominantly TF/FVIIa dependent.
However, nonstimulated monocytes adherent for 4 hours to plastic coverslips expressed low levels of TF activity. Induction of TF by adherence has been confirmed on the cellular mRNA level (unpublished observation). Because endotoxin was not detected in either media or on coverslips, these findings indicate that the adherence process itself may initiate TF expression. Also, this observation is in accordance with findings reported by others,29 30 showing that adherence of monocytes and macrophages to artificial surfaces triggers cellular procoagulant activity in the absence of exogenous agonists. It is interesting to note, however, that these non–lipopolysaccharide-stimulated adherent monocytes, which expressed low levels of TF and elicited platelet-thrombus formation under arterial blood flow conditions, may trigger similar events when adhering to graft material in vivo and thus play a role in artificial graft failure.
Despite the pronounced effect of the anti-TF MAb on monocyte-mediated coagulation and thrombus formation, platelets apparently adhered directly to the monocytes, which subsequently resulted in platelet-thrombus formation. These platelet thrombi formed on the cell surface in areas that on morphological inspection revealed no signs of fibrin deposition. However, we cannot exclude the possibility that the cells bound and activated factor X,28 31 resulting in minute undetectable fibrin deposition that served as a nidus for platelet-thrombus formation, or that the 1-μm-thick sections revealed platelet thrombi growing outside the fibrin mesh and thus gave the impression of direct platelet-monocyte adhesion. Nevertheless, from our experiments, it appeared that TF is the key initiator of monocyte-mediated coagulation and arterial thrombus formation, at least during the early stage of thrombus formation.
The material of the Thermanox coverslips is virtually nonthrombogenic.17 Exposure of this material to flowing nonanticoagulated human blood results in minimal activation of platelets and coagulation. Fibrin deposits are not detected, and platelets are generally making contact with the Thermanox surface by a small portion of their plasma membrane. Platelet aggregates are never observed.17 Thus, it is obvious that fibrin deposition and platelet-thrombus formation elicited by the Thermanox-adherent monocytes were due entirely to the cells. The larger thrombus formation triggered by 4-hour-adherent and/or stimulated cells relative to the 2-hour-adherent cells was apparently related to higher amounts of expressed TF and to the higher density of adherent cells. The size of these platelet thrombi was in the range of that previously observed on human type III collagen fibrils.17 32 Thus, procoagulant monocytes trigger platelet-thrombus formation as efficiently as fibrillar collagen.
Plasma activation markers of coagulation (FPA) and platelets (β-TG) collected distally to the perfusion chamber after 4 hours of lipopolysaccharide stimulation showed a nearly sixfold and fourfold increase in the average FPA and β-TG levels, respectively. Inclusion of the anti-TF MAb reduced these levels to those measured with nonstimulated cells. Thus, the levels of the activation markers measured complemented the morphological data obtained with adherent cells stimulated with lipopolysaccharide for 4 hours. However, the situation was different for the 2-hour-adherent cells. Lipopolysaccharide stimulation increased procoagulant activity and fibrin deposition without affecting the plasma FPA levels. Furthermore, the platelet-thrombus formation was enhanced more than 10-fold, but this was not reflected in the modest nonsignificant rise of the plasma β-TG. The reason remains unknown. However, it should be emphasized that the morphometric approach is an end point measurement, while measurements of the plasma activation markers represent observations taken at one defined point of time during thrombus formation. Thus, dissimilarities between such measured parameters may occur.
Our data suggest that thrombotic complications following situations such as plaque disruption after PTCA, rupture of arterial plaques, or artificial graft failure can be reduced by blocking monocyte procoagulant activity. However, it might prove necessary to include a platelet antagonist as well because other thrombogenic components such as collagens or even the monocyte itself33 may trigger platelet-dependent thrombus formation.
We acknowledge Merete Thune Wiiger, Biotechnology Centre of Oslo, University of Oslo, for sharing her experience with measurements of cell procoagulant activity.
Nemerson Y. Tissue factor and hemostasis. Blood. 1988:71:1-8.
Østerud B, Rapaport SI. Activation of factor IX by the reaction product of tissue factor and factor VII: additional pathway for initiating blood coagulation. Proc Natl Acad Sci U S A. 1977;74:5260-5267.
Bauer KA, Bryon LK, ten Cate H, Hawiger JJ, Rosenberg RD. Factor IX is activated in vivo by the tissue factor mechanism. Blood. 1990;76:731-736.
Hung DT, Vu TK, Wheaton VI, Ishii K, Coughlin SR. Cloned platelet thrombin receptor is necessary for thrombin-induced platelet activation. J Clin Invest. 1992;89:1350-1353.
Weiss HJ, Turitto VT, Baumgartner HR, Nemerson Y, Hoffmann T. Evidence for the presence of tissue factor activity on subendothelium. Blood. 1989;73:968-975.
Weiss HJ, Hoffmann T, Turitto VT, Nemerson Y. Further studies on the presence of functional tissue factor activity on subendothelium of normal human and rabbit arteries. Thromb Res. 1994;5:313-326.
Jang IK, Gold HK, Leinbach RC, Fallon JT, Collen D, Wilcox JN. Antithrombotic effect of a monoclonal antibody against tissue factor in a rabbit model of platelet-mediated arterial thrombosis. Arterioscler Thromb. 1992;12:948-954.
Pawashe A, Golino P, Guiseppe A, Migliaccio F, Ragni M, Pascucci I, Chiariello M, Garen A, Konigsberg W, Ezekowitz MD. Inhibition of intravascular thrombosis in the rabbit model by a monoclonal antibody against tissue factor. Thromb Haemost. 1993;69:abstract 2264.
Kirchhofer D, Sakariassen KS, Clozel M, Tschopp TB, Hadvary P, Nemerson Y, Baumgartner HR. Tissue factor mediated fibrin deposition on stimulated endothelial cells is not a requirement for leucocyte adherence under venous blood flow conditions. Blood. 1993;81:2050-2058.
Zwaginga JJ, Sixma JJ, de Groot PhG. Activation of endothelial cells induces platelet thrombus formation on their matrix: studies of a new in vitro thrombosis model with low molecular weight heparin as anticoagulant. Arteriosclerosis. 1990;10:49-61.
Jonasson L, Holm J, Bondjers G, Hansson GK. Regional accumulation of T cells, macrophages, and smooth muscle cells in the human atherosclerotic plaque. Arteriosclerosis. 1986;6:131-138.
Wilcox JN, Smith KM, Schwartz SM, Gordon D. Localization of tissue factor in normal vessel wall and in the atherosclerotic plaque. Proc Natl Acad Sci U S A. 1989;86:2839-2843.
Drake TA, Ruf W, Morrissey JH, Edginton TS. Functional tissue factor is entirely surface expressed on lipopolysaccharide-stimulated human blood monocytes and a constitutively tissue factor producing neoplastic cell line. J Cell Biol. 1989;109:389-394.
Osnes LTN, Westvik Å-B, Kierulf P. Procoagulant and profibrinolytic activities of cryopreserved human monocytes. Cryobiology. In press.
Sakariassen KS, Joss R, Muggli R, Kuhn H, Tshopp TB, Sage H, Baumgartner HR. Collagen type III induced ex vivo thrombogenesis in humans: role of platelets and leukocytes in deposition of fibrin. Arteriosclerosis. 1990;10:276-284.
Brandtzæg P, Kierulf P, Gaustad P, Skulberg A, Bruun JN, Halvorsen S, Sørensen E. Plasma endotoxin as a predictor of multiple organ failure and death in systemic meningococcal disease. J Infect Dis. 1989;159:195-207.
Sakariassen KS, Roald HE, Salatti JA. Ex vivo models for studying thrombosis: special emphasis on shear rate dependent blood-collagen interaction. In: Hwang NHC, Turitto VT, Yen MRT, eds. Advances in Cardiovascular Engineering. New York, NY: Plenum Press; 1992:151-174.
Sakariassen KS, Baumgartner HR. Axial dependence of platelet-collagen interactions in flowing blood: upstream thrombus growth impairs downstream platelet adhesion. Arteriosclerosis. 1989;9:33-42.
Sakariassen KS, Muggli R, Baumgartner HR. Growth and stability of thrombi in flowing citrated blood: assessment of platelet surface interactions with computer-assisted morphometry. Thromb Haemost. 1988;60:390-398.
Muhlfelder T, Teodrescu V, Essesse I, Niemetz J. Procoagulant activity of atheromatous plaques from human arteries. Thromb Haemost. 1993;69:abstract 2272.
Østerud B. Initiation mechanisms: activation induced by thromboplastin. In: Zwaal RFA, Hemker HC, eds. Blood Coagulation. Amsterdam, Netherlands: Elsevier Science Publishers BV; 1986:129-139.
Altieri DC. Coagulation assembly on leukocytes in transmembrane signalling and cell adhesion. Blood. 1993;81:569-579.
Worfolk L, Robinson RA, Tracy PB. Factor Xa interacts with two sites on monocytes with different functional activities. Blood. 1992;80:1989-1997.
Rinder HM, Bonan JL, Rinder CS, Ault KA, Smith BR. Activated and unactivated platelet adhesion to monocytes and neutrophiles. Blood. 1991;78:1760-1769.