We have adapted the corn-trypsin inhibitor whole-blood model to include EA.hy926 as an endothelium surrogate to evaluate the vascular modulation of blood coagulation initiated by relipidated recombinant tissue factor (rTf) and a cellular Tf surrogate, lipopolysaccharide (LPS)-stimulated THP1 cells (LPS-THP-1). Compared with bare tubes, EA.hy926 with rTf decreased the rate of thrombin formation, ITS accumulation, and the production of fibrinopeptide A. These phenomena occurred with increased rates of factor Va (fVa) inactivation by cleavages at R506 and R306. Thus, EA.hy926 provides thrombin-dependent protein C activation and APC fVa inactivation. Comparisons of rTf with LPS-THP-1 showed that the latter gave reduced rates for TAT formation but equivalent fibrinopeptide A, and fV activation/inactivation. In the presence of EA.hy926, the reverse was obtained; with the surrogate endothelium and LPS-THP-1 the rates of TAT generation, fibrinopeptide release, and fV activation were almost doubled, whereas cleavage at R306 was equivalent. These observations suggest cooperativity between the 2 cell surrogates. These data suggest that the use of these 2 cell lines provides a reproducible quasi-endothelial quasi-inflammatory cytokine-stimulated monocyte system that provides a method to evaluate the variations in blood phenotype against the background of stable inflammatory cell activator and a stable vascular endothelial surrogate.

Virchow postulated 3 phenomena that lead to venous thrombosis: (1) the components contained within the blood itself lead to systemic hypercoagulability; (2) alteration of the endothelium from injury or dysfunction; and (3) alterations of blood flow.1  These postulates are the foundation for the physiologic understanding of blood coagulation and the maintenance of the hemostatic balance in the venous circulation.

The maintenance of blood fluidity and vascular integrity is actively maintained by collaborations between the vascular wall and the procoagulant and anticoagulant functions in blood. Cell-bound and soluble constituents cooperate to maintain this quasi-equilibrium fluid state. Upon vascular perforation, multiple interactions initiate a local coagulant response, promoting platelet adhesion/aggregation, the generation of membrane-bound catalysts, thrombin generation, and fibrin deposition that overcome local anticoagulants. These processes arrest blood loss with dimensions relevant to the injury. Subsequently, fibrinolysis and cellular proliferation repair2  the injured tissue.

The coagulation processes have been studied in vitro using anticoagulated phlebotomy blood,3  derived plasma,4,5  isolated component systems,6-8  and numerical simulations.9,10  We have used corn-trypsin inhibitor (CTI) stabilized whole blood to study the tissue factor (Tf)–initiated process in the absence of chelating anticoagulants.11,12  Several hundred of these experiments have been performed to characterize the temporal dynamics of various blood clotting reactions in healthy persons, congenital abnormalities, and pharmacologic interventions.11-14  In the “normal” population, extreme phenotypic variations (> 5-fold) in Tf-initiated thrombin generation have been observed.15 

To evaluate the contributions of the vascular wall to the blood coagulation process, studies have used excised blood vessels16-18  and isolated cells.19-21  Badimon22  used human phlebotomy blood in a flow chamber with denuded swine vessels. Although this model is more physiologic in nature, these vessels may vary in protein expression between donor animals and are also complicated by xeno-protein associations.23  We and others have used isolated human umbilical vein endothelial cells (HUVECs),19  human umbilical artery endothelium,20  and human cerebral microvascular endothelium.21  All these cells display altered behavior and phenotype when removed from their natural environment,24  depending upon basement membrane composition,25,26  and show altered protein expression with passage.27  Consequently, thrombomodulin (TM) and von Willebrand factor (VWF) have been shown to vary considerably in isolated HUVECs.28  To address this variability, investigators have used EA.hy926 cells, an immortalized cell line produced by hybridization of HUVECs with lung carcinoma cell A549, to produce cells that display consistent endothelial cell markers over prolonged passages.29,30  EA.hy926 cells are reported to “maintain identical characteristic during at least 100 passages” and express many endothelial proteins.31  EA.hy926 is the most studied and characterized permanent human vascular endothelial cell line.32  Based on these reports, we have evaluated EA.hy926 as a surrogate “endothelium” to explore how the blood coagulation system is modulated by a confluent cellular provider of vascular regulatory elements.

The sources, expression, and regulation of natural Tf by cells and their influence upon blood coagulation processes have been controversial.33-37  Lipopolysaccharide (LPS)-stimulated monocytes that express Tf,38,39  have been used.40,41  However, Tf expression on natural monocytes is highly variable.40-42  LPS-stimulated THP1 cells appear to potentially provide a stable surrogate representing an inflammatory peripheral blood cell source of Tf.

In the present study, we explore the use of EA.hy926 and LPS-stimulated THP1 cells (LPS-THP-1 cells) as surrogates for human endothelial cells and cytokine activated monocytes in a closed system for the simulation of events that might occur during venous stasis. This approach provides a reproducible model that incorporates the components of Virchow's triad necessary for study of cellular regulation of Tf-initiated coagulation in blood. Furthermore, this approach limits the phenotypic variability in blood coagulation observed for the normal population to that of the human donor.15,43 

Materials

Mouse monoclonal anti–human Tf pathway inhibitor (TFPI) antibody (clone no. AHTFPI-5138), sheep polyclonal anti–human TFPI (no. PAHTFPI-S) and goat anti–human VWF (no. PAHVWF-G) antibodies were gifts from Dr R. Jenny (Haematologic Technologies Inc). Monoclonal antibody anti-TM (product no 2375; American Diagnostica). Mouse anti–human VWF-99, anti–Tf-5, anti–Tf-48, anti–TM-41, anti–fVa-17, and α1-Fbg monoclonal and burro polyclonal monospecific anti–prethrombin 1 antibodies were produced and characterized in house.37,44,45  Horseradish peroxidase (HRP)–labeled secondary antibodies were purchased from Southern Biotechnology Associates. Anti-burro (goat anti–equine HRP, no. 6040-05) HRP-labeled secondary antibodies were purchased from Southern Biotechnology Associates. Gel electrophoresis standards were from Novagen. The antibodies used for confocal studies (primary immunoglobulin G antibodies, anti-TM mouse clone [Lab Vision], anti–endothelial cell protein C receptor [EPCR] goat [R&D Systems], and anti-VWF rabbit clone no. A0082 [DAKO] and secondary antibodies, donkey anti–rabbit Alexa Fluor 488 [Invitrogen], donkey anti–goat Alexa Fluor 555 [Invitrogen], and donkey anti–mouse Alexa Fluor 647 [Invitrogen]) were all gifts from Dr Edwin G. Bovill (University of Vermont). Western Lightning Plus ECL (enhanced chemiluminescence substrate; PerkinElmer). Dulbecco Modified Eagle Medium with Glutamax, Alexa Fluor 488, and Alexa Fluor 633 was purchased from Invitrogen. Anti–Tf-5 for confocal microscopy was bound to R-phycoerythrin via a LNK023RPE kit from AbD (Serotec). Millipore polyvinylidene difluoride membrane Immobilon-FL, nitrocellulose, VectorShield with 4,6′-diamidino-2-phenylindole, gentamicin, soluble human recombinant tumor necrosis factor-α (TNF-α), and all other reagents were purchased from Fisher Scientific.

EA.hy926, the endothelial-like hybridoma cells, were a gift from Dr C. J. Edgell (University of North Carolina). HUVECs no. HUVEC-001F-M, fresh HUVEC basal medium no. HUVEC-002, and HUVEC stimulatory supplement no. HUVEC-004 were purchased from AllCells. RPMI 1640 and THP1 monocytic cells were obtained from ATCC. Recombinant Tf1-243 (rTf), a gift from Dr R. Lundblad and Dr S. L. Liu (Hyland Division; Baxter Healthcare), was relipidated in 1-palmitoyl-2-oleoyl-phosphatidyl serine (PS) and 1-palmitoyl-2-oleoyl-phosphatidylcholine (PC) at 25% PS/75% PC (PCPS).46  PC and PS were purchased from Avanti Polar Lipids. TFPI was a gift from Dr S. Hardy (Chiron). Goat anti-mouse HRP-labeled secondary antibody (no. A9917), β-mercaptoethanol, benzamidine-HCl, tris-HCl, HEPES, trifluoroacetic acid, ethylenediaminetetraacetic acid disodium salt dihydrate (EDTA), and LPS from Escherichia coli EH100 were purchased from Sigma-Aldrich. CTI was prepared in house.11  Fetal bovine serum (FBS premium) was purchased from Atlanta Biologicals. Thrombin-antithrombin (TAT) ELISA kits were from Siemens. U-bottom BD Falcon plates were (BD Labware). EPCR immunoassay kits were obtained from R&D Systems. Citrated plasma for clotting assays was produced in house (10 donors).

Subjects

Healthy male donors between the ages of 23 years and 35 years (mean 27 years ± 5 years, n = 5) were recruited and advised according to a protocol approved by the University of Vermont Human Studies Committee. Written consent was obtained in accordance with the Declaration of Helsinki. Subject questionnaires showed no clinical history of coagulopathy nor aspirin or drug use. Repeat blood draws were performed from this donor population.

EA.hy926 adherence and growth in polystyrene reaction tubes

Polystyrene tubes and polypropylene caps11  were treated with a Harrick PDC-3XG plasma cleaner (high setting 20 minutes ambient air). Confluent EA.hy926 cells from two 10-cm diameter culture plates were resuspended in 100 mL growth medium (Dulbecco modified Eagle medium with Glutamax, gentamicin, and 10% FBS premium), and distributed in 2-mL aliquots to the prepared capped polystyrene tubes. Two days growth in a 37°C, 5% CO2 humidified incubator allowed the cells to adequately divide and adhere to achieve confluence in the bottom half of the horizontal tube cylinder (verified under bright-field microscopy). Endothelial growth medium was replaced with growth medium with no FBS at the end of the second day, and the cells were incubated for an additional 15 hours to attain quiescence. The “endothelium” lined tubes were used in whole-blood experiments on the third day after removal of the medium and one wash with fresh medium. The unilaminar “endothelium” from 5 random tubes was removed with versene, centrifuged, resuspended in HBS, stained with trypan blue, and counted (5.2 ± 0.2 × 105 cells/tube, n = 5) using a hemocytometer.

Characterization of EA.hy926 cells

Confluent EA.hy926 cells were labeled with preconjugated anti–VWF-99 Alexa Fluor 488 or primary conjugated anti–TM-41 biotin antibody and secondary streptavidin 633. Confocal comparative analyses between EA.hy926 cells and HUVECs for VWF, TM, and EPCR was accomplished using primary and secondary antibodies provided by Dr Edwin G. Bovill as well as a modified microscopy protocol from Dr Bovill and Dr Taatjes.47,48  Briefly, cells were grown to confluence on 8 chambered slides according to cell-type instructions. Cells were rinsed 3 times with 1X Hanks balanced salt solution (HBS) 20 mM HEPES, 150 mM Nacl and fixed for 5 minutes with 4% paraformaldehyde (10% formalin) in 1X HBS buffer. Cells were rinsed 3 times in 1X HBS before blocking with preimmune donkey serum for 1 hour at room temperature. Primary antibodies TM, EPCR, and VWF were administered for an overnight incubation at 4°C. Secondary antibodies were incubated 1 hour at room temperature. A secondary control incubation including all secondary antibodies in a well containing confluent EA.hy926 and another containing HUVECs showed no immunofluorescence labeling as previously reported for this system when used for blood vessels.47  All images were captured with a Zeiss LSM 510 META confocal laser scanning microscope (Zeiss MicroImaging) with a Plan-Apochromat 63×/1.4 oil-immersion objective. Excitation was performed using a 405-nm laser diode, 488 argon laser, and 543/633 helium/neon laser, and emissions passed through LP650, BP420-480, BP505-530, and BP560-615 filters to acquire images in 12-bit multitrack mode.

For Tf activity assessment,49  cells were grown to confluence, stimulated for 4 hours with LPS (250 ng/mL LPS-THP-1) or treated with versene, counted, centrifuged (200g for 5 minutes) and resuspended in HBS to a concentration of 1 × 106 cells/mL. These cells were added to citrated plasma containing 0.1 mg/mL CTI and 50 μM PCPS. Plasma clotting was initiated with 25 mM CaCl2. Clotting times were determined using the ST8 clotting instrument (Diagnostica Stago).

For analyses of protein expression, cells were grown to confluence, released from the plate with versene, counted, centrifuged (200g for 5 minutes) and resuspended/lysed in 1% Triton X–100/PBS, 1 mM EDTA, 10 mM benzamidine hydrochloride, 200 μM phenylmethanesulphonylfluoride, 40 μM FPRck, 65 kIU aprotinin, 0.5 μg/mL leupeptin, and 0.7 μg/mL pepstatin A. The lysates were mixed, centrifuged at 14 000g, and the supernatants characterized by immunoassays for VWF,45  Tf,37  TM, EPCR, and TFPI. TM was quantitated with anti-TM monoclonal antibody (25 μg/mL; American Diagnostica) coupled beads and probed with biotinylated anti–TM-41 monoclonal antibodies (10 μg/mL) via fluorescence Luminex immunoassay (Luminex Technology).37  Binding was detected with phycoerythrin-streptavidin (5 μg/mL). A nonspecific isotype-matched monoclonal antibody control was included in the reaction mixture. TFPI was quantitated by a sandwich ELISA using anti-TFPI monoclonal antibodies (5 μg/mL) as the capture antibody and a sheep anti-human TFPI antibody (10 μg/mL) as the probe. Binding was detected with HRP-rabbit anti-sheep antibody (1:2000). The EPCR assay was conducted using the R&D Systems protocol.

THP1 cells were treated at a concentration of 2.5 × 106 cells/mL in RPMI 1640 with LPS (250 ng/mL, 4 hours).49,50  Cell-surface and total Tf was quantitated via immunoassay.37  TFPI was quantitated by ELISA.

LPS-THP-1 were titrated into CTI phlebotomy blood with clotting times determined visually (2 observers by subjective evaluation). The clotting activity for 2 × 104 stimulated monocytes/mL was determined to be equivalent to 5 pM rTf/PCPS activity and used for all THP1 cell–based Tf experiments.

Phlebotomy blood experiments

The contributions of EA.hy926 and LPS-THP-1 cells to blood coagulation were evaluated by comparing phlebotomy blood 1) in the presence and absence (bare tubes) of EA.hy926 with 5 pM rTf/PCPS, and 2) in the presence and absence of EA.hy926 with 2 × 104 LPS-THP-1 cells as a source of Tf. Donor blood (1 mL) was delivered into tubes containing CTI (100 μg/mL) and a source of Tf as previously described.11  After LPS prestimulation, THP1 cells were added after 1 rinse in HBS, recentrifugation (200g for 5 minutes) and resuspension in HBS (5 × 106 THP1 cells/mL) in 4-μL aliquots (2 × 104 per aliquot). Tubes (37°C) were rocked (16 cycles per minute) and quenched at various time points over the course of 20 minutes with a cocktail of inhibitors: HBS (1 mL total) containing 50 mM EDTA, 20 mM benzamidine-HCl, and 50 μM FPRck. The zero time point contained the inhibitors before the addition of blood. Clot time was established visually by observation of a thickening of the blood in rocked tubes illustrated by a change in the edge of the fluid cycling back and forth from concave to convex. This methods accuracy is best judged by examination of the interception of the “initiation” and “propagation” phases of the reactions (see Figures 2,Figure 3,Figure 4,Figure 56) that, in general, correspond to the clot times shown for the present experiments and for the previous more than 400 whole-blood experiments done in this laboratory.6,11-14  A control tube with no Tf was used for each experiment. Controls with no EA.hy926 cells clotted over 20 minutes and with EA.hy926 cells clotted over 17 minutes.

The “initiation” and “propagation” phases have been defined and used by this laboratory since approximately 19946  to describe synthetic coagulation proteome and whole-blood experiments in multiple reports11-14  and reviews.51,52  The “initiation” phase corresponds to the initial time course of TAT accumulation (∼ 2 nM) in which the membrane catalyst elements are formed and assembled. The “propagation” phase corresponds to the major thrombin generation/TAT accumulation (∼ 500 nM-800 nM).

Analyses of whole-blood samples

The quenched samples were centrifuged (15 minutes at 1160g) to separate the clot material from the solution phase. These phases were stored separately at −80°C for further analysis. Samples were analyzed for TAT by ELISA.11  Fibrinopeptide A (FPA) using HPLC methodology53  and with Western blotting for prothrombin, fV, fVa, fVa307-506, and fibrinogen-fibrin reaction products under either nonreducing or reducing conditions.11,54  Densitometry was performed using the Fujifilm chemiluminescent imager LAS4000 and Fujifilm MultiGauge imaging software (ver 3.x). Concentrations were determined from serial dilutions of the relevant purified proteins, the absolute concentration for fV was estimated by normalizing the data assuming a mean physiologic value of 20 nM.55  Fibrinopeptide data were normalized from each subject's control data.

Statistical analyses

All data in the manuscript and figures were expressed as mean (± SEM). Prism 4 software (GraphPad Software) was used to perform variable slope curve fitting analyses, linear analyses, and statistical tests: t test, and ANOVA (repeated measures) with Bonferroni posttest. Differences with values for P < .05 were considered statistically significant.

EA.hy926 cells

Confocal microscopy showed the presence of TM, EPCR, and VWF on EA.hy926 cells and HUVECs (Figure 1). Tf was not detectable after LPS stimulation (250 ng/mL, 4 hours) either by immunoassay at a concentration of 2.5 × 105 cells/mL or by activity at 1 × 106 cells/mL (< 10 fM).35,49  Following TNF-α stimulation (10 ng/mL, 4 hours), Tf expression above background was observed via immunoassay but at concentrations below the quantitation limit (0.49 pM per 5 × 105 EA.hy926 cells/mL). The molar concentrations of EA.hy926 expressed TM, VWF, EPCR, and TFPI were determined by immunoassay. These results (Table 1) are similar to those previously published for HUVECs.56-58 

Figure 1

Fluorescent confocal micrographs of HUVECs and EA.hy926 hybridoma cells. All images were obtained at a magnification ×400 using a 40X oil immersion objective under identical settings for focus and image capture. EPCR (green), VWF (white), TM (red), and 4,6′-diamidino-2-phenylindole (DAPI; blue) are shown in split xy along with a merged image (composite) as viewed by Zeiss LSM 5 image viewer software.

Figure 1

Fluorescent confocal micrographs of HUVECs and EA.hy926 hybridoma cells. All images were obtained at a magnification ×400 using a 40X oil immersion objective under identical settings for focus and image capture. EPCR (green), VWF (white), TM (red), and 4,6′-diamidino-2-phenylindole (DAPI; blue) are shown in split xy along with a merged image (composite) as viewed by Zeiss LSM 5 image viewer software.

Table 1

Characterization of EA.hy926 cell and HUVEC types

ProteinEA.hy926 cells
HUVECs
Molecules/cellMolar concentration for 5 × 105 cells/mL solution, MMolecules/cellMolar concentration for 5 × 105 cells/mL solution, MReference*
VWF 2.5 × 105 2.0 × 10−10 2.6 × 105 2.2 × 10−10 58  
TM 8.7 × 104 7.3 × 10−11 5 × 104, 1 × 105 4.2 × 10−11, 8.3 × 10−11 56, 85  
EPCR 3.9 × 105 3.2 × 10−10 7 × 105 5.8 × 10−10 85  
TFPI 2.6 × 105 2.2 × 10−10 7.8 × 104 6.5 × 10−11 57  
LPS-induced Tf ND ND 5.0 × 103 4 × 10−12 68  
TNFα-induced Tf <6.0 × 102 <4.9 × 10−13 2.5 × 104 2.0 × 10−11 69  
ProteinEA.hy926 cells
HUVECs
Molecules/cellMolar concentration for 5 × 105 cells/mL solution, MMolecules/cellMolar concentration for 5 × 105 cells/mL solution, MReference*
VWF 2.5 × 105 2.0 × 10−10 2.6 × 105 2.2 × 10−10 58  
TM 8.7 × 104 7.3 × 10−11 5 × 104, 1 × 105 4.2 × 10−11, 8.3 × 10−11 56, 85  
EPCR 3.9 × 105 3.2 × 10−10 7 × 105 5.8 × 10−10 85  
TFPI 2.6 × 105 2.2 × 10−10 7.8 × 104 6.5 × 10−11 57  
LPS-induced Tf ND ND 5.0 × 103 4 × 10−12 68  
TNFα-induced Tf <6.0 × 102 <4.9 × 10−13 2.5 × 104 2.0 × 10−11 69  

Data are expressed in concentrations based on 5 ×105 EA.hy926 cells/mL present in the whole-blood model. EA.hy926 cells were stimulated with 250 ng/mL LPS for 4 hours or 10 ng/mL TNF-α for 4 hours in attempts to elicit expression of Tf. All other values reflect nonstimulated conditions and total expression.

*

Reference data for HUVECs only.

ND indicates not detected.

The Tf issue on endothelial cells is controversial. Several recent reviews59-61  document the current understanding of Tf including cell types and anatomical locations. Tf has been reported by several studies to be present on endothelium in vitro,62-65  yet conclusive proof of endothelium Tf synthesis is lacking.66-68  A study of TNF-α–activated HUVECs69  suggests 1 fM/cm2 or a 20 pM concentration (Table 1).

Recombinant Tf with EA.hy926 cells

When blood (5 donors) was treated with 5 pM rTf/PCPS in the presence and absence of EA.hy926, the clot time did not vary significantly (2.9 ± 0.14 minutes vs 3.0 ± 0.17 minutes, P = .71). Thrombin generation (TAT formation) in response to 5 pM rTf/PCPS stimulus with or without EA.hy926 is shown in Figure 2A as the mean of 5 experiments (± SEM). Consistent with the visually observed clot time (Figure 2A arrows), the initiation phase of thrombin generation (∼ 2 nM thrombin) is not changed by EA.hy926. Thus the contribution of EA.hy926 TFPI (Table 1) was negligible, as would be anticipated from previous studies.54  In addition, computer modeling9,10,70  predicted only a 16-second increase in the initiation phase duration (data not shown). In contrast, EA.hy926 shortens the duration of the propagation phase (8.9 minutes to 5.2 minutes), decreases the maximum rate of thrombin generation (60 ± 5.5 nM/min to 34 ± 6.6 nM/min, P = .019), and produces less total thrombin (512 ± 28 nM to 203 ± 12 nM, P < .001). The decrease in the rate of thrombin generation during the propagation phase in the presence of EA.hy926 is similar to that previously observed for the purified proteome with soluble TM7  and for HUVEC19  with products corresponding to R507-709 and R307-506,71  which most likely are attributed to the generation of activated protein C (APC) by thrombin with EA.hy926 TM and EPCR (Figure 1).72 

Figure 2

rTf-activated CTI-inhibited blood with or without EA.hy926 cells. Bare tubes (■) and tubes containing a unilaminar layer of EA.hy926 (○) were used to measure (A) TAT formation over time after CTI-inhibited fresh blood activation with 5 pM rTf, (B) FPA production, (C) normalized fibrinogen consumption with a representative Western blot of a single subject and (D) fVa generation (left) and inactivation (right) as quantified by densitometry of Western blot analysis using α-fV no. 17 HC antibody showing fVa HC and fVa307-506 HC cleavage product (bottom). FVa1-506 HC cleavage is also shown but was not quantified. Clot times in all experiments are indicated (■ = 3.0 ± 0.17 minutes and ○ = 2.9 ± 0.14 minutes). Dashed line in a and b represents 95% confidence limits.

Figure 2

rTf-activated CTI-inhibited blood with or without EA.hy926 cells. Bare tubes (■) and tubes containing a unilaminar layer of EA.hy926 (○) were used to measure (A) TAT formation over time after CTI-inhibited fresh blood activation with 5 pM rTf, (B) FPA production, (C) normalized fibrinogen consumption with a representative Western blot of a single subject and (D) fVa generation (left) and inactivation (right) as quantified by densitometry of Western blot analysis using α-fV no. 17 HC antibody showing fVa HC and fVa307-506 HC cleavage product (bottom). FVa1-506 HC cleavage is also shown but was not quantified. Clot times in all experiments are indicated (■ = 3.0 ± 0.17 minutes and ○ = 2.9 ± 0.14 minutes). Dashed line in a and b represents 95% confidence limits.

Figure 2B illustrates the normalized FPA generation with rTf in bare tubes compared with EA.hy926. The initial rates of FPA generation in both circumstances are similar (5.1 ± 1.3 μM/min vs 4.2 ± 0.9 μM/min); however, the ultimate relative FPA release is somewhat reduced in the presence of the endothelial cell surrogate, consistent with the earlier termination of thrombin generation (Figure 2A). As observed in Figure 2C and in earlier studies,12  at clot time, virtually all fibrinogen products are removed from the fluid phase of the blood indicating that the initial clot is a heterogeneous mixture of fibrin 1, fibrin 2, and fibrinogen because at clot time only 30% to 40% of total FPA is released (Figure 2B).53 

Western blot analyses for prothrombin consumption with rTf with and without EA.hy926 after 5 pM rTf/PCPS stimulation showed that in the absence of EA.hy926 the rate of prothrombin consumption is greater (−0.07 ± 0.01 μM/min) compared with the reaction with EA.hy926 (0.00 ± 0.02 μM/min, P = .003; linear analysis) from time points 6 to 20 minutes. A 20-minute time point comparison shows a greater remaining prothrombin concentration of 1.12 ± 0.17μM in the presence of EA.hy926 versus 0.25 ± 0.07μM (P = .002) without EA.hy926 (data not shown).

Western blot analyses of the fV activation/inactivation process in 5 pM rTf/PCPS initiated reactions in the presence and absence of EA.hy926 are shown in Figure 2D. In both cases, fVa heavy chain (HC) is observed by 1 minute. Densitometric analyses (n = 5) shows that although the initial rate of fV activation is not different (with or without EA.hy926 cells, 4.7 ± 0.7 nM/min vs 4.5 ± 1.4 nM/min; Figure 2d). With EA.hy926, a 2-fold decrease in the maximum level of fVa HC is observed (20 vs 10 nM, P < .001). As expected,71  inactivation of fVa via the R506 cleavage occurs more rapidly in the presence of EA.hy926 (5.2 ± 0.8 minutes vs 7.6 ± 0.9 minutes) and the initial rate of the fVa R307-506 cleavage product formation is accelerated (2.9 ± 0.44 nM/min vs 0.57 ± 0.13 nM/min; P < .001). With EA.hy926, the 30-kDa fVai fragment, fVa307-506 is seen earlier (4.0 ± 0.7 minutes vs 9.2 ± 0.8 minutes) consistent with the functional contribution of the EA.hy926 provided TM and EPCR (Table 1)71-73  to the APC R306, R506 cleavages.

Recombinant Tf versus LPS-THP-1 in bare tubes

Figure 3 shows clot times for 2 titrations (from 644 to 3.3 × 105 cells/mL and from 250 to 2.0 × 104 cells/mL) of LPS-THP-1 (LPS stimulated at 250 ng/mL, 4 hours) into CTI-inhibited blood. These experiments demonstrate the reproducibility of LPS-induced Tf activity by THP1 cells. LPS-THP-1 (2 × 104 cells/mL) yield equivalent clotting times (3.1 ± 0.1 minutes) to that of 5 pM rTf/PCPS (Figure 3 dashed line). The concentration of surface Tf antigen per 2 × 104 LPS-THP-1 cells/mL was found to be 250 fM; a value similar to that previously reported.37 

Figure 3

Clotting of CTI-inhibited fresh blood triggered with LPS-THP-1. CTI-inhibited whole blood was triggered to clot by varying concentrations of LPS-THP-1 (▴, n = 5 blood donors). A total of 2 × 104 LPS-THP-1 cells/mL are shown to have equivalent clotting activity to the 5pM rTf). Clot time represented by the horizontal dashed line (n = 5). An experiment using another population of titrated LPS-THP-1 cells illustrates a very similar titration curve indicating reproducibility of LPS-THP-1 activity (open gray box, n = 1 donor). A total of 5 × 104 nonstimulated THP1 cells/mL (■, n = 5 donor) is equivalent in clotting activity to 2.5 × 103 stimulated THP1 cells/mL. Nonstimulated THP1 cells show 5% of stimulated THP1 activity.

Figure 3

Clotting of CTI-inhibited fresh blood triggered with LPS-THP-1. CTI-inhibited whole blood was triggered to clot by varying concentrations of LPS-THP-1 (▴, n = 5 blood donors). A total of 2 × 104 LPS-THP-1 cells/mL are shown to have equivalent clotting activity to the 5pM rTf). Clot time represented by the horizontal dashed line (n = 5). An experiment using another population of titrated LPS-THP-1 cells illustrates a very similar titration curve indicating reproducibility of LPS-THP-1 activity (open gray box, n = 1 donor). A total of 5 × 104 nonstimulated THP1 cells/mL (■, n = 5 donor) is equivalent in clotting activity to 2.5 × 103 stimulated THP1 cells/mL. Nonstimulated THP1 cells show 5% of stimulated THP1 activity.

LPS-THP-1 (2 × 104 cells/mL) and 5 pM rTf/PCPS with CTI blood in the absence of EA.hy926 showed no difference in clot time (3.0 ± 0.34 minutes vs 3.0 ± 0.17 minutes, P = 1.0). Although both Tf sources yield similar visual clotting times (Figure 4A arrow) and initiation phase durations (Figure 4A), the TAT accumulation profiles show some variation. The maximum rate of thrombin generation is decreased by 2.3-fold when LPS-THP-1 cells are substituted for rTf (60 ± 5.5 nM/min to 26 ± 2.8 nM/min, P < .001). Furthermore, although the maximum concentration of TAT at 20 minutes is not significantly different in LPS-THP-1–activated blood, TAT generation had not achieved a plateau value at the termination of the experiment. The rTf stimulated prothrombin consumption (time points 6-20 minutes; Figure 4B) occurs at a faster rate (−0.07 ± 0.01 μM/min) than with LPS-THP-1 (−0.03 ± 0.01 μM/min, P = .016). Normalized FPA generation in blood activated with either Tf source is similar (Figure 4C).

Figure 4

CTI-inhibited fresh blood activated with rTf or LPS-THP-1 in the absence of EA.hy926. Bare tubes containing rTf (■) or LPS-THP-1 (♦) activated whole blood was used to measure (A) TAT formation over time, (B) prothrombin consumption, (C) FPA production, and (D) fVa generation (left), inactivation (right), and representative Western blot. FVa1-506 HC cleavage is also shown but was not quantified. Clot times in all experiments are indicated (■ = 3.0 ± 0.17 minutes and ♦ = 3.0 ± 0.34 minutes). Dashed line in panels A and C represents 95% confidence limits.

Figure 4

CTI-inhibited fresh blood activated with rTf or LPS-THP-1 in the absence of EA.hy926. Bare tubes containing rTf (■) or LPS-THP-1 (♦) activated whole blood was used to measure (A) TAT formation over time, (B) prothrombin consumption, (C) FPA production, and (D) fVa generation (left), inactivation (right), and representative Western blot. FVa1-506 HC cleavage is also shown but was not quantified. Clot times in all experiments are indicated (■ = 3.0 ± 0.17 minutes and ♦ = 3.0 ± 0.34 minutes). Dashed line in panels A and C represents 95% confidence limits.

The rate of fV activation was equivalent with either 5 pM rTf/PCPS and LPS-THP-1 (4.7 ± 0.7 nM/min, 4.3 ± 0.6 nM/min, P = .70; Figure 4D). However, with rTf/PCPS, the formation of fVa307-506 was slightly accelerated relative to LPS-THP-1 (0.57 ± 0.13 nM/min vs 0.39 ± 0.06 nM/min, P < .001) possibly because of the influence of the added PCPS in the former because cleavage at R306 is PCPS dependent.74 

EA.hy926 with LPS-THP-1 or rTf

In the presence of EA.hy926, LPS-THP-1 blood clotted slightly more rapidly than rTf activated blood (2.3 ± 0.10 minutes vs 2.9 ± 0.14 minutes, P = .01).

In the presence of EA.hy926, blood activated with LPS-THP-1 (2 × 104 cells/mL) shows a trend toward an accelerated rate of TAT accumulation (63 ± 15 nM/min) compared with 5 pM rTf/PCPS (34 ± 6.6 nM/min, P = .047; Figure 5A). The maximum thrombin accumulated is also slightly greater for the LPS-THP-1 reactions compared with rTf/PCPS reactions (281 ± 52 nM vs 203 ± 12 nM, respectively). In contrast, compared with bare tubes (Figure 4A), the presence of EA.hy926 (Figure 5A) with LPS-THP-1 increased the maximum rate of TAT generation from 26 ± 2.8 nM/min to 63 ± 15 nM/min, P = .008) and shortens the duration of the propagation phase (> 17 to ∼ 7 minutes). With EA.hy926, TAT formation is unchanged from time points 7 to 20 minutes (Figure 5A), whereas a slower continued TAT generation occurred without EA.hy926 (rate, 26 ± 2.8 nM/min, P < .001; Figure 4A). In the absence of EA.hy926, slower thrombin generation continued beyond 15 minutes such that at the 20-minute point, TAT concentrations are not statistically different (without EA.hy926, 439 ± 87 nM;with EA.hy926, 281 ± 52 nM, P = .26; Figure 4A).

Figure 5

CTI-inhibited fresh blood activated with rTf or LPS-THP-1 in the presence EA.hy926. Tubes containing a unilaminar layer of EA.hy926 with rTf (○) or LPS-THP-1 (▵) activated whole blood were used to measure (A) TAT formation, (B) prothrombin consumption, (C) FPA production, and (D) fVa generation (left), inactivation (right), and representative Western blot. FVa1-506 HC cleavage is shown but not quantified. Clot times in all experiments are indicated (▵ = 2.3 ± 0.10 minutes and ○ = 2.9 ± 0.14 minutes). Dashed line in a and c represents 95% confidence limits.

Figure 5

CTI-inhibited fresh blood activated with rTf or LPS-THP-1 in the presence EA.hy926. Tubes containing a unilaminar layer of EA.hy926 with rTf (○) or LPS-THP-1 (▵) activated whole blood were used to measure (A) TAT formation, (B) prothrombin consumption, (C) FPA production, and (D) fVa generation (left), inactivation (right), and representative Western blot. FVa1-506 HC cleavage is shown but not quantified. Clot times in all experiments are indicated (▵ = 2.3 ± 0.10 minutes and ○ = 2.9 ± 0.14 minutes). Dashed line in a and c represents 95% confidence limits.

In the presence of EA.hy926, with rTf or LPS-THP-1 activation, the rates of prothrombin consumption are not significantly different but the trend to faster consumption is in favor of LPS-THP-1 (Figure 5B).

The rate of FPA generation was accelerated in the presence of EA.hy926 when blood was activated with LPS-THP-1 compared with rTf (Figure 5C). An increased rate of FPA release (9.3 ± 0.9 μM/min vs 4.2 ± 0.9 μM/min, P = .001) with an earlier point for FPA detection (2 vs 3 minutes) was evident consistent with the accelerated clot time observed for the EA.hy926 system after LPS-THP-1 activation. In bare tubes, both Tf sources produced FPA to the same ultimate level at similar rates (Figure 4C).

Densitometric analyses of fibrinogen-fibrin blots, in the presence or absence of EA.hy926, showed no significant difference in the rate of fibrinogen removal as expected from the similar clot times with THP1 activation in the presence of EA.hy926 (data not shown).

In the presence of EA.hy926, activation with either Tf source did not significantly alter the maximum rates of fVa generation (rTf/PCPS, 4.5 ± 1.4 nM/min; LPS-THP-1, 6.4 ± 0.6 nM/min, P = .18); however, a lag was seen with rTf (Figure 5D). Under the experimental conditions used, the initial inactivation of fVa (R506) associated with rTf/PCPS stimulation with EA.hy926 allowed only 3 time points, limiting the power of the analysis. In the presence of EA.hy926, 5 pM rTf/PCPS caused slightly accelerated formation of fVai307-506 over the LPS-THP-1–activated process (rTf/PCPS, 2.9 ± 0.44 nM/min; LPS-THP-1, 2.1 ± 0.32 nM/min, P = .019) suggesting more efficient APC function, probably as a consequence of the PCPS effect on APC cleavage at R306.74 

When compared with bare tubes (Figure 4D), the contribution of LPS-THP-1 on fV activation and inactivation in the presence of EA.hy926 (Figure 5D) showed both an accelerated rate of activation (6.4 ± 0.6 nM/min vs 4.3 ± 0.6 nM/min, P = .035) and initial inactivation (2.1 ± 0.32 nM/min vs 0.39 ± 0.06 nM/min). The maximum concentrations of fVa formed differed slightly (with EA.hy926, 18.3 nM vs without EA.hy926, 20 nM), whereas the peak concentration occurred earlier with EA.hy926 (3 minutes vs 8 minutes, respectively). The faster rate of fVai307-506 formation with LPS-THP-1 activation with EA.hy926 compared with the same Tf source in bare tubes (2.1 ± 0.32 nM/min vs 0.39 ± 0.06 nM/min, P < .001) is illustrated by comparison of Figures 5D and 4D.

Table 2 summarizes the results of experiments with the 2 Tf sources with and without EA.hy926. The following differences stand out with respect to the combination of LPS-THP-1 in the presence of EA.hy926: 1) clot time is shorter, 2) the delayed rate of TAT generation observed with LPS-THP-1 in bare tubes is recovered to a rate equivalent to rTf in bare tubes, and 3) rTf gave a slower TAT generation rate with the endothelial cell surrogate. The ultimate accumulated level of thrombin induced by both Tf sources is reduced with the endothelial cell surrogate. The combination of monocyte and endothelial surrogates increases the rate of fV activation. The rate of fVa inactivation is accelerated with EA.hy926 with both Tf sources compared with bare tubes. A dramatic increase in the release of FPA is observed with a combination of LPS-THP-1 and EA.hy926 compared with all systems.

Table 2

Summary of experimental results for 2 Tf sources with and without EA.hy926 cells

Tf source without EA.hy926Clot time, minTAT Rate, nM/minTAT maximum, nMFVa rate, nM/minFVa307-506 rate, nM/minFPA rate, μM/min
    rTf 3.0 ± 0.17 60 ± 5.5 512 ± 28 4.7 ± 0.7 0.57 ± 0.13 5.1 ± 1.3 
    LPS-THP-1 3.0 ± 0.34 26 ± 2.8 439 ± 87 4.3 ± 0.6 0.39 ± 0.06 4.2 ± 0.25 
Tf source without EA.hy926Clot time, minTAT Rate, nM/minTAT maximum, nMFVa rate, nM/minFVa307-506 rate, nM/minFPA rate, μM/min
    rTf 3.0 ± 0.17 60 ± 5.5 512 ± 28 4.7 ± 0.7 0.57 ± 0.13 5.1 ± 1.3 
    LPS-THP-1 3.0 ± 0.34 26 ± 2.8 439 ± 87 4.3 ± 0.6 0.39 ± 0.06 4.2 ± 0.25 
Tf Source with EA.hy926
    rTf 2.9 ± 0.14 34 ± 6.6 203 ± 12 4.5 ± 1.4 2.9 ± 0.44 4.2 ± 0.9 
    LPS-THP-1 2.3 ± 0.1 63 ± 15 281 ± 52 6.4 ± 0.6 2.1 ± 0.32 9.3 ± 0.9 
Tf Source with EA.hy926
    rTf 2.9 ± 0.14 34 ± 6.6 203 ± 12 4.5 ± 1.4 2.9 ± 0.44 4.2 ± 0.9 
    LPS-THP-1 2.3 ± 0.1 63 ± 15 281 ± 52 6.4 ± 0.6 2.1 ± 0.32 9.3 ± 0.9 

Rates are shown (± SEM).

Reproducibility of the surrogate cell system

A primary objective of the present study was to develop reproducible endothelial and inflammatory monocyte surrogates to evaluate the individual phenotypic variations in blood composition and response to a Tf stimulus mimicking venous stasis with an activated inflammatory cell. To ensure that these data in the cohort study represented by the mean product formation over time are reproducible and would not mask the observations for a specific subject within the cohort, we compared data for 2 studies conducted 8 days apart for a single individual's blood evaluated using the monocyte/endothelial surrogate environment. Previous studies15,43  have shown that most individuals' blood is highly reproducible with respect to Tf stimulation in bare tubes. These data presented in Figure 6 illustrate that this is also true for a single subject when blood in contact with a surrogate endothelial cell layer is confronted with the surrogate monocyte Tf presentation. Presented are data for the 2 experiments conducted 8 days apart for generation of TAT, FPA, fVa HC, and the fVai APC degradation product 307-506. These data obtained for the single subject are reproducible over time and the surrogate inflammatory cell activator and vascular bed are representative of the population study presented in Figure 5.

Figure 6

Surrogate cell reproducibility. Data represent replicate of experiments for a single subject with EA.hy926 and LPS-THP-1 conducted on 2 occasions (▴, ▵) with an 8-day interval between experiments. These data illustrate the reproducibility of (A) TAT formation, (B) FPA formation, (C) fV activation, and (D) fVa inactivation in these duplicate experiments. These data show the reproducibility of the method with respect to the use of the 2 surrogates and also show these observations for the single subject are representative of those for the cohort.

Figure 6

Surrogate cell reproducibility. Data represent replicate of experiments for a single subject with EA.hy926 and LPS-THP-1 conducted on 2 occasions (▴, ▵) with an 8-day interval between experiments. These data illustrate the reproducibility of (A) TAT formation, (B) FPA formation, (C) fV activation, and (D) fVa inactivation in these duplicate experiments. These data show the reproducibility of the method with respect to the use of the 2 surrogates and also show these observations for the single subject are representative of those for the cohort.

Our data demonstrate that (1) the EA.hy926 cell line provides a stable endothelial cell surrogate presenting a complement of known coagulation-related endothelial cell surface proteins, which produces consistent anticoagulant responses suggestive of a functioning endothelium with Tf-activated phlebotomy blood; (2) LPS-THP-1 can function as a reproducible surrogate for a stimulated circulating cell source of Tf; (3) EA.hy926 and LPS-THP-1 provide a reasonable surrogate for venous stasis coupled with a blood derived inflammatory stimulus; and (4) there appears to be a cooperative cell-cell effect between EA.hy926 and an LPS-THP-1, not seen when rTf/PCPS is used as the activator.

The activation of blood by rTf/PCPS versus LPS-THP-1 produces very different coagulation profiles with EA.hy926 including: faster and greater thrombin formation, increased fVa generation, and increased rate of FPA release. Complete fVa inactivation (fVai307-506 generation) is faster in the presence of EA.hy926 regardless of the Tf source illustrating the presence of a functional TM-EPCR protein C activation system.19,72  In contrast, in the absence of EA.hy926, blood clotting times with both Tf sources are similar as are FPA release and fV activation and inactivation, although thrombin is generated at a slower rate during the propagation phase with LPS-THP-1.

EA.hy926 and LPS-THP-1 correspond to reproducible surrogates for the known procoagulant and anticoagulant functions displayed by endothelial cells presenting similar concentrations of VWF, TM, EPCR, and TFPI. The principle reason for using these cells as surrogates for the respective natural cells is because of the variability of the cells from natural sources,75-81  which in combination with the high variability of human donor blood would lead to uninterpretable data. The reproducibility problem for endothelial cells from various natural sources and from passage to passage is well documented. EA.hy926 appears to provide endothelial cell–like function with respect to APC anticoagulation.

LPS stimulation of monocytes has been reported to elicit the secretion of TNF-α and interleukin-1, which may modify the function of endothelium.82  Furthermore, simultaneous thrombin and LPS stimulation of monocytes increases the release of TNF-α and interleukin-1β.82  In the present study, we observe that although EA.hy926 Tf expression is not detectable via LPS administration, the significant differences observed for LPS-THP-1 with or without EA.hy926 suggest a cooperative process between the surrogate inflammatory cell and the surrogate endothelium. However, TNF-α–induced expression of Tf is an unlikely mechanism because Tf surface expression occurring 3 minutes after stimulation could not occur via de novo synthesis and Tf is not stored within the cell. The increase in EA.hy926–LPS-THP-1 procoagulant activity potentially may provide in vitro experimental access to the relationship between sepsis, inflammation, and coagulation. These data suggest that stimulated monocytes may directly activate or intensify the endothelial inflammatory phenotype and contribute to vascular thrombosis.

A finding replicated in the present study related to whole-blood activation by Tf is the absence of quantitative conversion of prothrombin to thrombin. In addition, a significant fraction of thrombin B chain escapes covalent association with antithrombin. The laboratory of Dr Hemker83  identified that a significant fraction of thrombin is bound to α2-macroglobulin, and our studies suggest accumulation of the thrombin α1-antitrypsin complex.84  Furthermore, significant destructive prothrombin cleavage by thrombin leads to production of prothrombin fragment 1 and prethrombin 1.

The use of these 2 transformed cell lines provides a reproducible method to evaluate the variations in blood phenotype against the background of stable blood cell activator and vascular endothelial surrogates. The EA.hy926–LPS-THP-1 system thus provides a useful method for characterizing the biochemical dynamics of how an individual's blood would respond to an inflammatory cell presented Tf under conditions of venous stasis.

The publication costs of this article were defrayed in part by page charge payment. Therefore, and solely to indicate this fact, this article is hereby marked “advertisement” in accordance with 18 USC section 1734.

We thank Matthew Whelihan and Daniel LaJoie for technical assistance and Dr Thomas Orfeo for intellectual input. We thank Jolanta Amblo for technical assistance and expertise in immunochemical assays and Dr Jeffrey Spees and Marilyn Wadsworth for assistance with confocal microscopy. Special thanks to Winifred Trotman for expertise in fluorescent confocal microscopy image capture for comparative analysis of EA.hy926 cells and HUVECs and to Dr Beth Bouchard for use of laboratory equipment and intellectual input.

This study was supported by grants P01 HL46703 and T32 HL07594 by the National Institutes of Health.

National Institutes of Health

Contribution: J.E.C. designed and performed research, analyzed data, and contributed to the writing of the paper; K.E.B.-Z. designed research, analyzed data, and contributed to the writing of the paper; S.B. performed research and contributed to the writing of the paper; and K.G.M. contributed to experimental design, analytical tools, data analyses, and the writing of the paper.

Conflict-of-interest disclosure: K.G.M. is Chairman of the Board of Haematologic Technologies. The remaining authors declare no competing financial interests.

Correspondence: Kenneth G. Mann, PhD, Department of Biochemistry, Colchester Research Facility Rm 235, University of Vermont, Colchester, VT 05446; e-mail: kenneth.mann@uvm.edu.

1
Virchow
 
R
Matzdorff
 
AC
Bell
 
WR
Thrombose und Embolie. Gefassentzundung und septische Infektion.
Gesammelte Abhandlungen zur wissenschlaflichen Medicin, Frankfurk am Main: Von Meidinger & Sohn
1998
, vol. 
1856
 
Canton, Massachusetts
Science History Publications
(pg. 
219
-
732
)
2
Naldini
 
A
Carney
 
DH
Pucci
 
A
Pasquali
 
A
Carraro
 
F
Thrombin regulates the expression of proangiogenic cytokines via proteolytic activation of protease-activated receptor-1.
Gen Pharmacol
2000
, vol. 
35
 
5
(pg. 
255
-
259
)
3
Chavez
 
JJ
Foley
 
DE
Snider
 
CC
, et al. 
A novel thrombelastograph tissue factor/kaolin assay of activated clotting times for monitoring heparin anticoagulation during cardiopulmonary bypass.
Anesth Analg
2004
, vol. 
99
 
5
(pg. 
1290
-
1294
table of contents
4
Beguin
 
S
Lindhout
 
T
Hemker
 
HC
The effect of trace amounts of tissue factor on thrombin generation in platelet rich plasma, its inhibition by heparin.
Thromb Haemost
1989
, vol. 
61
 
1
(pg. 
25
-
29
)
5
Hron
 
G
Kollars
 
M
Binder
 
BR
Eichinger
 
S
Kyrle
 
PA
Identification of patients at low risk for recurrent venous thromboembolism by measuring thrombin generation.
JAMA
2006
, vol. 
296
 
4
(pg. 
397
-
402
)
6
Lawson
 
JH
Kalafatis
 
M
Stram
 
S
Mann
 
KG
A model for the tissue factor pathway to thrombin. I. An empirical study.
J Biol Chem
1994
, vol. 
269
 
37
(pg. 
23357
-
23366
)
7
van't Veer
 
C
Golden
 
NJ
Kalafatis
 
M
Mann
 
KG
Inhibitory mechanism of the protein C pathway on tissue factor-induced thrombin generation. Synergistic effect in combination with tissue factor pathway inhibitor.
J Biol Chem
1997
, vol. 
272
 
12
(pg. 
7983
-
7994
)
8
Butenas
 
S
van't Veer
 
C
Mann
 
KG
Evaluation of the initiation phase of blood coagulation using ultrasensitive assays for serine proteases.
J Biol Chem
1997
, vol. 
272
 
34
(pg. 
21527
-
21533
)
9
Hockin
 
MF
Jones
 
KC
Everse
 
SJ
Mann
 
KG
A model for the stoichiometric regulation of blood coagulation.
J Biol Chem
2002
, vol. 
277
 
21
(pg. 
18322
-
18333
)
10
Jones
 
KC
Mann
 
KG
A model for the tissue factor pathway to thrombin. II. A mathematical simulation.
J Biol Chem
1994
, vol. 
269
 
37
(pg. 
23367
-
23373
)
11
Rand
 
MD
Lock
 
JB
van't Veer
 
C
Gaffney
 
DP
Mann
 
KG
Blood clotting in minimally altered whole blood.
Blood
1996
, vol. 
88
 
9
(pg. 
3432
-
3445
)
12
Brummel
 
KE
Paradis
 
SG
Butenas
 
S
Mann
 
KG
Thrombin functions during tissue factor-induced blood coagulation.
Blood
2002
, vol. 
100
 
1
(pg. 
148
-
152
)
13
Cawthern
 
KM
van't Veer
 
C
Lock
 
JB
DiLorenzo
 
ME
Branda
 
RF
Mann
 
KG
Blood coagulation in hemophilia A and hemophilia C.
Blood
1998
, vol. 
91
 
12
(pg. 
4581
-
4592
)
14
Orfeo
 
T
Brummel-Ziedins
 
KE
Gissel
 
M
Butenas
 
S
Mann
 
KG
The nature of the stable blood clot procoagulant activities.
J Biol Chem
2008
, vol. 
283
 
15
(pg. 
9776
-
9786
)
15
Brummel-Ziedins
 
KE
Pouliot
 
RL
Mann
 
KG
Thrombin generation: phenotypic quantitation.
J Thromb Haemost
2004
, vol. 
2
 
2
(pg. 
281
-
288
)
16
Badimon
 
JJ
Weng
 
D
Chesebro
 
JH
Fuster
 
V
Badimon
 
L
Platelet deposition induced by severely damaged vessel wall is inhibited by a boroarginine synthetic peptide with antithrombin activity.
Thromb Haemost
1994
, vol. 
71
 
4
(pg. 
511
-
516
)
17
Badimon
 
L
Badimon
 
JJ
Lassila
 
R
Heras
 
M
Chesebro
 
JH
Fuster
 
V
Thrombin regulation of platelet interaction with damaged vessel wall and isolated collagen type I at arterial flow conditions in a porcine model: effects of hirudins, heparin, and calcium chelation.
Blood
1991
, vol. 
78
 
2
(pg. 
423
-
434
)
18
Badimon
 
L
Badimon
 
JJ
Turitto
 
VT
Vallabhajosula
 
S
Fuster
 
V
Platelet thrombus formation on collagen type I. A model of deep vessel injury. Influence of blood rheology, von Willebrand factor, and blood coagulation.
Circulation
1988
, vol. 
78
 
6
(pg. 
1431
-
1442
)
19
Hockin
 
MF
Kalafatis
 
M
Shatos
 
M
Mann
 
KG
Protein C activation and factor Va inactivation on human umbilical vein endothelial cells.
Arterioscler Thromb Vasc Biol
1997
, vol. 
17
 
11
(pg. 
2765
-
2775
)
20
Turner
 
N
Nolasco
 
L
Tao
 
Z
Dong
 
JF
Moake
 
J
Human endothelial cells synthesize and release ADAMTS-13.
J Thromb Haemost
2006
, vol. 
4
 
6
(pg. 
1396
-
1404
)
21
Shatos
 
MA
Orfeo
 
T
Doherty
 
JM
Penar
 
PL
Collen
 
D
Mann
 
KG
Alpha-thrombin stimulates urokinase production and DNA synthesis in cultured human cerebral microvascular endothelial cells.
Arterioscler Thromb Vasc Biol
1995
, vol. 
15
 
7
(pg. 
903
-
911
)
22
Badimon
 
L
Turitto
 
V
Rosemark
 
JA
Badimon
 
JJ
Fuster
 
V
Characterization of a tubular flow chamber for studying platelet interaction with biologic and prosthetic materials: deposition of indium 111-labeled platelets on collagen, subendothelium, and expanded polytetrafluoroethylene.
J Lab Clin Med
1987
, vol. 
110
 
6
(pg. 
706
-
718
)
23
Roussel
 
JC
Moran
 
CJ
Salvaris
 
EJ
Nandurkar
 
HH
d'Apice
 
AJ
Cowan
 
PJ
Pig thrombomodulin binds human thrombin but is a poor cofactor for activation of human protein C and TAFI.
Am J Transplant
2008
, vol. 
8
 
6
(pg. 
1101
-
1112
)
24
Grau
 
GE
Mili
 
N
Lou
 
JN
, et al. 
Phenotypic and functional analysis of pulmonary microvascular endothelial cells from patients with acute respiratory distress syndrome.
Lab Invest
1996
, vol. 
74
 
4
(pg. 
761
-
770
)
25
Zhu
 
DZ
Cheng
 
CF
Pauli
 
BU
Mediation of lung metastasis of murine melanomas by a lung-specific endothelial cell adhesion molecule.
Proc Natl Acad Sci U S A
1991
, vol. 
88
 
21
(pg. 
9568
-
9572
)
26
Milici
 
AJ
Furie
 
MB
Carley
 
WW
The formation of fenestrations and channels by capillary endothelium in vitro.
Proc Natl Acad Sci U S A
1985
, vol. 
82
 
18
(pg. 
6181
-
6185
)
27
Hampel
 
B
Fortschegger
 
K
Ressler
 
S
, et al. 
Increased expression of extracellular proteins as a hallmark of human endothelial cell in vitro senescence.
Exp Gerontol
2006
, vol. 
41
 
5
(pg. 
474
-
481
)
28
Moore
 
KL
Andreoli
 
SP
Esmon
 
NL
Esmon
 
CT
Bang
 
NU
Endotoxin enhances tissue factor and suppresses thrombomodulin expression of human vascular endothelium in vitro.
J Clin Invest
1987
, vol. 
79
 
1
(pg. 
124
-
130
)
29
Edgell
 
CJ
McDonald
 
CC
Graham
 
JB
Permanent cell line expressing human factor VIII-related antigen established by hybridization.
Proc Natl Acad Sci U S A
1983
, vol. 
80
 
12
(pg. 
3734
-
3737
)
30
Edgell
 
CJ
Haizlip
 
JE
Bagnell
 
CR
, et al. 
Endothelium specific Weibel-Palade bodies in a continuous human cell line, EA.hy926.
In Vitro Cell Dev Biol
1990
, vol. 
26
 
12
(pg. 
1167
-
1172
)
31
Schaefer
 
HI
van't Hooft
 
FM
van der Laarse
 
A
Growth characteristics of a permanent human endothelial cell line.
In Vitro Cell Dev Biol
1992
, vol. 
28A
 
7-8
(pg. 
465
-
467
)
32
Bouis
 
D
Hospers
 
GA
Meijer
 
C
Molema
 
G
Mulder
 
NH
Endothelium in vitro: a review of human vascular endothelial cell lines for blood vessel-related research.
Angiogenesis
2001
, vol. 
4
 
2
(pg. 
91
-
102
)
33
Balasubramanian
 
V
Grabowski
 
E
Bini
 
A
Nemerson
 
Y
Platelets, circulating tissue factor, and fibrin colocalize in ex vivo thrombi: real-time fluorescence images of thrombus formation and propagation under defined flow conditions.
Blood
2002
, vol. 
100
 
8
(pg. 
2787
-
2792
)
34
Mezzano
 
D
Matus
 
V
Saez
 
CG
Pereira
 
J
Panes
 
O
Tissue factor storage, synthesis and function in normal and activated human platelets.
Thromb Res
2008
, vol. 
122
 
suppl 1
(pg. 
S31
-
S36
)
35
Butenas
 
S
Bouchard
 
BA
Brummel-Ziedins
 
KE
Parhami-Seren
 
B
Mann
 
KG
Tissue factor activity in whole blood.
Blood
2005
, vol. 
105
 
7
(pg. 
2764
-
2770
)
36
Bouchard
 
BA
Mann
 
KG
Butenas
 
S
No evidence for tissue factor on platelets.
Blood
2010
, vol. 
116
 
5
(pg. 
854
-
855
)
37
Parhami-Seren
 
B
Butenas
 
S
Krudysz-Amblo
 
J
Mann
 
KG
Immunologic quantitation of tissue factors.
J Thromb Haemost
2006
, vol. 
4
 
8
(pg. 
1747
-
1755
)
38
Edwards
 
RL
Rickles
 
FR
Bobrove
 
AM
Mononuclear cell tissue factor: cell of origin and requirements for activation.
Blood
1979
, vol. 
54
 
2
(pg. 
359
-
370
)
39
Monroe
 
DM
Roberts
 
HR
Hoffman
 
M
Platelet procoagulant complex assembly in a tissue factor-initiated system.
Br J Haematol
1994
, vol. 
88
 
2
(pg. 
364
-
371
)
40
Egorina
 
EM
Sovershaev
 
MA
Bjorkoy
 
G
, et al. 
Intracellular and surface distribution of monocyte tissue factor: application to intersubject variability.
Arterioscler Thromb Vasc Biol
2005
, vol. 
25
 
7
(pg. 
1493
-
1498
)
41
Tracy
 
PB
Robinson
 
RA
Worfolk
 
LA
Allen
 
DH
Procoagulant activities expressed by peripheral blood mononuclear cells.
Methods Enzymol
1993
, vol. 
222
 (pg. 
281
-
299
)
42
Sovershaev
 
MA
Egorina
 
EM
Gruber
 
FX
Olsen
 
JO
Osterud
 
B
High tissue factor-expressing human monocytes carry low surface CD36: application to intersubject variability.
J Thromb Haemost
2007
, vol. 
5
 
12
(pg. 
2453
-
2460
)
43
Brummel-Ziedins
 
K
Vossen
 
CY
Rosendaal
 
FR
Umezaki
 
K
Mann
 
KG
The plasma hemostatic proteome: thrombin generation in healthy individuals.
J Thromb Haemost
2005
, vol. 
3
 
7
(pg. 
1472
-
1481
)
44
Foster
 
WB
Tucker
 
MM
Katzmann
 
JA
Miller
 
RS
Nesheim
 
ME
Mann
 
KG
Monoclonal antibodies to human coagulation factor V and factor Va.
Blood
1983
, vol. 
61
 
6
(pg. 
1060
-
1067
)
45
Butenas
 
S
Parhami-Seren
 
B
Mann
 
KG
The influence of von Willebrand factor on factor VIII activity measurements.
J Thromb Haemost
2009
, vol. 
7
 
1
(pg. 
132
-
137
)
46
Higgins
 
DL
Mann
 
KG
The interaction of bovine factor V and factor V-derived peptides with phospholipid vesicles.
J Biol Chem
1983
, vol. 
258
 
10
(pg. 
6503
-
6508
)
47
Brooks
 
EG
Trotman
 
W
Wadsworth
 
MP
, et al. 
Valves of the deep venous system: an overlooked risk factor.
Blood
2009
, vol. 
114
 
6
(pg. 
1276
-
1279
)
48
Taatjes
 
DJ
Wadsworth
 
MP
Zaman
 
AK
Schneider
 
DJ
Sobel
 
BE
A novel dual staining method for identification of apoptotic cells reveals a modest apoptotic response in infarcted mouse myocardium.
Histochem Cell Biol
2007
, vol. 
128
 
3
(pg. 
275
-
283
)
49
Butenas
 
S
Parhami-Seren
 
B
Gissel
 
MT
Amblo
 
J
Brummel-Ziedins
 
K
Mann
 
KG
The nature of LPS-stimulated monocyte tissue factor activity.
Blood
2008
, vol. 
112
 
11
(pg. 
375
-
376
Abstract 1022
50
Henriksson
 
CE
Hellum
 
M
Landsverk
 
KS
Klingenberg
 
O
Joo
 
GB
Kierulf
 
P
Flow cytometry-sorted non-viable endotoxin-treated human monocytes are strongly procoagulant.
Thromb Haemost
2006
, vol. 
96
 
1
(pg. 
29
-
37
)
51
Mann
 
KG
Thrombin formation.
Chest
2003
, vol. 
124
 
3 suppl
(pg. 
4S
-
10S
)
52
Mann
 
KG
Orfeo
 
T
Butenas
 
S
Undas
 
A
Brummel-Ziedins
 
K
Blood coagulation dynamics in haemostasis.
Hamostaseologie
2009
, vol. 
29
 
1
(pg. 
7
-
16
)
53
Brummel
 
KE
Butenas
 
S
Mann
 
KG
An integrated study of fibrinogen during blood coagulation.
J Biol Chem
1999
, vol. 
274
 
32
(pg. 
22862
-
22870
)
54
van't Veer
 
C
Mann
 
KG
Regulation of tissue factor initiated thrombin generation by the stoichiometric inhibitors tissue factor pathway inhibitor, antithrombin-III, and heparin cofactor-II.
J Biol Chem
1997
, vol. 
272
 
7
(pg. 
4367
-
4377
)
55
Tracy
 
PB
Eide
 
LL
Bowie
 
EJ
Mann
 
KG
Radioimmunoassay of factor V in human plasma and platelets.
Blood
1982
, vol. 
60
 
1
(pg. 
59
-
63
)
56
Maruyama
 
I
Majerus
 
PW
The turnover of thrombin-thrombomodulin complex in cultured human umbilical vein endothelial cells and A549 lung cancer cells. Endocytosis and degradation of thrombin.
J Biol Chem
1985
, vol. 
260
 
29
(pg. 
15432
-
15438
)
57
Sato
 
N
Kokame
 
K
Miyata
 
T
Kato
 
H
Lysophosphatidylcholine decreases the synthesis of tissue factor pathway inhibitor in human umbilical vein endothelial cells.
Thromb Haemost
1998
, vol. 
79
 
1
(pg. 
217
-
221
)
58
Wilbourn
 
B
Harrison
 
P
Lawrie
 
A
, et al. 
Unique expression of von Willebrand factor by type IIA von Willebrand's disease endothelial cells.
Br J Haematol
1992
, vol. 
81
 
3
(pg. 
401
-
406
)
59
Butenas
 
S
Orfeo
 
T
Mann
 
KG
Tissue factor in coagulation: Which? Where? When?
Arterioscler Thromb Vasc Biol
2009
, vol. 
29
 
12
(pg. 
1989
-
1996
)
60
Osterud
 
B
Bjorklid
 
E
Sources of tissue factor.
Semin Thromb Hemost
2006
, vol. 
32
 
1
(pg. 
11
-
23
)
61
Butenas
 
S
Orfeo
 
T
Mann
 
KG
Tissue factor activity and function in blood coagulation.
Thromb Res
2008
, vol. 
122
 
suppl 1
(pg. 
S42
-
S46
)
62
Breider
 
MA
Yang
 
Z
Tissue factor expression in bovine endothelial cells induced by Pasteurella haemolytica lipopolysaccharide and interleukin-1.
Vet Pathol
1994
, vol. 
31
 
1
(pg. 
55
-
60
)
63
Yang
 
HL
Lu
 
FJ
Wung
 
SL
Chiu
 
HC
Humic acid induces expression of tissue factor by cultured endothelial cells: regulation by cytosolic calcium and protein kinase C.
Thromb Haemost
1994
, vol. 
71
 
3
(pg. 
325
-
330
)
64
Faucette
 
KJ
Fitzgerald
 
LA
Liu
 
L
Parker
 
CJ
Rodgers
 
GM
Characterization of the interactions between procoagulant albumin and human endothelial cells.
Blood
1993
, vol. 
82
 
9
(pg. 
2684
-
2692
)
65
van der Logt
 
CP
Dirven
 
RJ
Reitsma
 
PH
Bertina
 
RM
Expression of tissue factor and tissue factor pathway inhibitor in monocytes in response to bacterial lipopolysaccharide and phorbolester.
Blood Coagul Fibrinolysis
1994
, vol. 
5
 
2
(pg. 
211
-
220
)
66
Osterud
 
B
Tissue factor expression in blood cells.
Thromb Res
2010
, vol. 
125
 
suppl 1
(pg. 
S31
-
S34
)
67
Rodgers
 
GM
Broze
 
GJ
Shuman
 
MA
The number of receptors for factor VII correlates with the ability of cultured cells to initiate coagulation.
Blood
1984
, vol. 
63
 
2
(pg. 
434
-
438
)
68
Reuning
 
U
Preissner
 
KT
Muller-Berghaus
 
G
Two independent binding sites on monolayers of human endothelial cells are responsible for interaction with coagulation factor VII and factor VIIa.
Thromb Haemost
1993
, vol. 
69
 
2
(pg. 
197
-
204
)
69
Camera
 
M
Giesen
 
PL
Fallon
 
J
, et al. 
Cooperation between VEGF and TNF-alpha is necessary for exposure of active tissue factor on the surface of human endothelial cells.
Arterioscler Thromb Vasc Biol
1999
, vol. 
19
 
3
(pg. 
531
-
537
)
70
Hockin
 
MF
Cawthern
 
KM
Kalafatis
 
M
Mann
 
KG
A model describing the inactivation of factor Va by APC: bond cleavage, fragment dissociation, and product inhibition.
Biochemistry
1999
, vol. 
38
 
21
(pg. 
6918
-
6934
)
71
Kalafatis
 
M
Rand
 
MD
Mann
 
KG
The mechanism of inactivation of human factor V and human factor Va by activated protein C.
J Biol Chem
1994
, vol. 
269
 
50
(pg. 
31869
-
31880
)
72
Esmon
 
CT
The protein C pathway.
Chest
2003
, vol. 
124
 
3 suppl
(pg. 
26S
-
32S
)
73
Mann
 
KG
Hockin
 
MF
Begin
 
KJ
Kalafatis
 
M
Activated protein C cleavage of factor Va leads to dissociation of the A2 domain.
J Biol Chem
1997
, vol. 
272
 
33
(pg. 
20678
-
20683
)
74
Kalafatis
 
M
Mann
 
KG
Role of the membrane in the inactivation of factor Va by activated protein C.
J Biol Chem
1993
, vol. 
268
 
36
(pg. 
27246
-
27257
)
75
Jaffe
 
EA
Nachman
 
RL
Becker
 
CG
Minick
 
CR
Culture of human endothelial cells derived from umbilical veins. Identification by morphologic and immunologic criteria.
J Clin Invest
1973
, vol. 
52
 
11
(pg. 
2745
-
2756
)
76
Castellot
 
JJ
Addonizio
 
ML
Rosenberg
 
R
Karnovsky
 
MJ
Cultured endothelial cells produce a heparinlike inhibitor of smooth muscle cell growth.
J Cell Biol
1981
, vol. 
90
 
2
(pg. 
372
-
379
)
77
Goldsmith
 
JC
McCormick
 
JJ
Yen
 
A
Endothelial cell cycle kinetics. Changes in culture and correlation with endothelial properties.
Lab Invest
1984
, vol. 
51
 
6
(pg. 
643
-
647
)
78
Adragna
 
NC
Changes in cation composition of aortic endothelial cells in culture.
Tissue Cell
1988
, vol. 
20
 
3
(pg. 
313
-
329
)
79
Fry
 
GL
Modification of phenotypic-expression of human vascular endothelium in culture.
Arteriosclerosis
1988
, vol. 
8
 
5
(pg. 
A666
-
A666
)
80
Rodriguez-Morata
 
A
Garzon
 
I
Alaminos
 
M
, et al. 
Cell viability and prostacyclin release in cultured human umbilical vein endothelial cells.
Ann Vasc Surg
2008
, vol. 
22
 
3
(pg. 
440
-
448
)
81
Brown
 
ML
Deykin
 
D
Passage state affects arachidonic acid content and eicosanoid release in porcine aortic endothelial cells.
Arterioscler Thromb
1991
, vol. 
11
 
1
(pg. 
167
-
173
)
82
Hoffman
 
M
Cooper
 
ST
Thrombin enhances monocyte secretion of tumor necrosis factor and interleukin-1 beta by two distinct mechanisms.
Blood Cells Mol Dis
1995
, vol. 
21
 
2
(pg. 
156
-
167
)
83
Kessels
 
H
Willems
 
G
Hemker
 
HC
Analysis of thrombin generation in plasma.
Comput Biol Med
1994
, vol. 
24
 
4
(pg. 
277
-
288
)
84
Downing
 
MR
Bloom
 
JW
Mann
 
KG
Comparison of the inhibition of thrombin by three plasma protease inhibitors.
Biochemistry
1978
, vol. 
17
 
13
(pg. 
2649
-
2653
)
85
Xu
 
J
Esmon
 
NL
Esmon
 
CT
Reconstitution of the human endothelial cell protein C receptor with thrombomodulin in phosphatidylcholine vesicles enhances protein C activation.
J Biol Chem
1999
, vol. 
274
 
10
(pg. 
6704
-
6710
)