Tissue Factor Preparations Research Articles

Redistribution and Hemostatic Action of Recombinant Activated Factor VII Associated with Platelets

Clinical evidence accumulated from hemophilic patients during prophylaxis with recombinant activated factor VII (rFVIIa) suggests that the duration of the hemostatic action of rFVIIa exceeds its predicted plasma half-life. Mechanisms involved in this outcome have not been elucidated. We have investigated in vitro the redistribution of rFVIIa in platelets from healthy donors, patients with FVII deficiency, and one patient with Bernard-Soulier syndrome. Platelet-rich plasma was exposed to rFVIIa (3 to 60 μg/mL). Flow cytometry, immunocytochemistry, and coagulation tests were applied to detect and quantify rFVIIa. The hemostatic effect of rFVIIa associated to platelets was evaluated using perfusion models. Our studies revealed a dose-dependent association of rFVIIa to the platelet cytoplasm with redistribution into the open canalicular system, and α granules. Mechanisms implicated in the internalization are multiple, involve GPIb and GPIV, and require phospholipids and cytoskeletal assembly. After platelet activation with thrombin, platelets exposed rFVIIa on their membrane. Perfusion studies revealed that the presence of 30% of platelets containing FVIIa improved platelet aggregate formation and enhanced fibrin generation (P < 0.01 versus control). Our results indicate that, at therapeutic concentrations, rFVIIa can be internalized into platelets, where it is protected from physiological clearance mechanisms and can still promote hemostatic activity. Redistribution of rFVIIa into platelets may explain the prolonged prophylactic effectiveness of rFVIIa in hemophilia.

Redistribution of Recombinant Activated Factor VIIA into Platelets and Vascular Bed: Implications on Bioavailability and Hemostatic Mechanisms.

Background: Clinical evidence suggests that hemostatic action of recombinant activated factor VII (rFVIIa) exceeds its predicted plasma half life (around 3 hours). Mechanisms involved in the long lasting effects of rFVIIa for prophylactic treatment of patients with hemophilia and inhibitors have not been fully elucidated. Previous studies from our group have demonstrated that platelets internalize tissue factor preparations containing small amounts of FVII. We have investigated the possible redistribution of rFVIIa in different intracellular compartments focussing more specifically on platelets, endothelium and subendothelial matrix.Methods: We exposed platelet rich plasma, isolated platelets, human umbilical veins and endothelial cells in culture (HUVECs) to rFVIIa. Samples were incubated for up to 2 hours at rFVIIa concentrations from 2-60 mg/ml (normal plasma concentration: 0.5 mg/ml). Flow cytometry techniques were applied to detect surface and intracellular antigens, including FVIIa, in platelets before and after exposure to membrane permeabilizing agents. Immunocytochemical techniques were applied to detect FVIIa in cross sections of umbilical veins and confocal microscopy was used to detect FVIIa in HUVECs. In additional studies, aliquots of washed platelets previously exposed to rFVIIa were incorporated into whole blood anticoagulated with low molecular weight heparin (LMWH) and perfused through a collagen rich damaged vascular surface at a shear rate equivalent to 600/sec.Results: Flow cytometry studies revealed a significantly enhanced presence of intracellular FVIIa in platelets previously exposed to rFVIIa. Fluorescence intensity related to FVIIa was dependent on the concentration of rFVIIa used during the incubation. A more intense labeling for FVIIa was observed in the subendothelial matrix of umbilical veins exposed to rFVIIa. In perfusion studies, presence of platelets previously exposed to rFVIIa improved platelet thrombus formation and enhanced fibrin generation with respect to parallel experiments using platelets not exposed to rFVIIa.Conclusion: Our results indicate that rFVIIa can be internalized into platelets and redistributed into subendothelial compartments. Redistribution of rFVIIa into other cellular compartments may explain the prolonged prophylactic action of rFVIIa in some clinical conditions. The existence of extravascular sources of FVIIa may provide new insights into the physiological and pathological implications of FVIIa on hemostasis mechanisms.

Tissue Factor Coagulant Function Is Enhanced by Protein-disulfide Isomerase Independent of Oxidoreductase Activity

Protein-disulfide isomerase (PDI) switches tissue factor (TF) from coagulation to signaling by targeting the allosteric Cys186-Cys209 disulfide. Here, we further characterize the interaction of purified PDI with TF. We find that PDI enhances factor VIIa-dependent substrate factor X activation 5-10-fold in the presence of wild-type, oxidized soluble TF but not TF mutants that contain an unpaired Cys186 or Cys209. PDI-accelerated factor Xa generation was blocked by bacitracin but not influenced by inhibition of vicinal thiols, reduction of PDI, changes in redox gradients, or covalent thiol modification of reduced PDI by N-ethylmaleimide or methyl-methanethiosulfonate, which abolished PDI oxidoreductase but not chaperone activity. PDI had no effect on fully active TF on either negatively charged phospholipids or in activating detergent, indicating that PDI selectively acts upon cryptic TF to facilitate ternary complex formation and macromolecular substrate turnover. PDI activation was reduced upon mutation of TF residues in proximity to the macromolecular substrate binding site, consistent with a primary interaction of PDI with TF. PDI enhanced TF coagulant activity on microvesicles shed from cells, suggesting that PDI plays a role as an activating chaperone for circulating cryptic TF.

Tissue factor-enriched vesicles are taken up by platelets and induce platelet aggregation in the presence of factor VIIa

We investigated the interactions of vesicles containing human tissue factor (TF) with platelets and evaluated responses induced by rFVIIa using standard aggregometry, ultrastructural and flow-cytometry techniques. Washed platelets were exposed to a preparation of placental human TF (pTF) or to a relipidated formulation of recombinant human TF (rTF). Under stirring conditions, pTF induced reversible aggregation with platelets returning to their resting state after 5 minutes. This reversible response to pTF was partially inhibited by antibodies against CD62-P, but not by antithrombin agents, and was not observed with rTF. Sequential ultrastructural studies revealed uptake of both TF preparations by platelets involving traffic of vesicles through channels of the open canalicular system (OCS). Immunocytochemical studies on cryosections identified TF in the OCS, and occasionally in the alpha-granules of the platelets. These processes were faster with pTF than with rTF, but both TF preparations accumulated in platelets at the end of incubation periods. Flow cytometry studies revealed the presence of other cellular antigens (CD62-P, CD14 and CD45) associated to the pTF. Addition of rFVIIa to washed platelets exposed to pTF or rTF, caused a thrombin dependent irreversible platelet aggregation. Our studies demonstrate that platelets possess mechanisms to capture and incorporate TF-rich vesicles. These processes are accelerated by the presence of other cellular antigens in the vesicles. Our findings may explain the hemostatic action of rFVIIa in severely hemodiluted patients, but are also relevant for the understanding of potential implications of TF-associated to platelets in the propagation of thrombus.

Role of Tissue Factor in Platelet Adhesion: Real-Time Perfusion Studies.

Exposure of Tissue Factor (TF) to circulating blood and excessive thrombin generation are major contributors to acute thrombosis. Platelets also play a central role in the pathogenesis of occlusive events. Despite the importance of both components, convincing evidence of direct adherence of platelets onto TF immobilized on a surface is lacking. Here, we studied the interactions of platelets with different TF-rich substrata. Real-time perfusion techniques were applied to visualize thrombus formation and to evaluate the time necessary for platelets to both attach to the surface and to form thrombi. We used three different sources of TF as substrates:Thromborel S®, which is TF purified from human placenta (pTF) (at 0.4 ng/cm2);Innovin®, recombinant TF that has been relipidated (rTF) (at 0.4 ng/cm2); andTF expressed on the membranes of TF-transfected CHO cells (CHO-TF).As negative controls, we used equivalent surfaces sprayed with 0.5% bovine serum albumin or cultured sham-transfected CHO cells (CHO-C-), respectively. Platelet-rich plasma (PRP) from blood anticoagulated with low-molecular-weight heparin was perfused over these substrata at a shear rate of 300s−1 for 5 min. Images were captured with a digital camera at high resolution and streamed to a computer. Continuous real-time photographs were taken during the full 5 min. Single frames were captured at predetermined times and the mean area of the platelet aggregates at these times was measured. Every experiment was replicated at least 4 times. Results are expressed as mean ± S.E.M. Under these conditions, platelets bound and formed thrombi on all substrata containing TF, although the time to aggregate formation and aggregate area varied depending on the substratum. Platelet deposition was minimal on the BSA-coated surfaces or on CHO-C- cells.Initial platelet deposits on pTF were observed after 120 sec of perfusion. After 5 min, thrombi reached an area of 122±41 μm2. Relipidated rTF was less effective in supporting platelet deposition, requiring 135 seconds on average for the initiation of platelet aggregates. After 5 min, the average area of platelet aggregates was 34.3±9.6 μm2. Formation of platelet aggregates on CHO-TF cells was noticeable after 80 seconds. Aggregates formed at 5 min reached an area of 213±58 μm2. In conclusion, our studies demonstrate that platelets adhere to and aggregate on different TF substrata. The presence of other adhesive molecules in cell-derived TF preparations (pTF, CHO-TF) appears to accelerate the interaction of platelets with TF. These observations suggest that in addition to its role in blood coagulation, TF also has role in platelet adhesion by providing an adhesive surface at sites of vessel injury.SAF2003-05780, SGR2005-00952, FISPI040887, FIS CP04-0011, NHLBI 5R01HL064796.

Fluorogenic Assays for Functional Factor VIIa and Tissue Factor.

Blood coagulation is initiated when cryptic tissue factor (TF) becomes exposed on the surface of vascular cells where it can bind circulating factor VIIa (fVIIa). The resulting fVIIa/TF enzyme complex catalyzes the activation of certain blood zymogens that propagate the coagulation event. As a key enzyme for the initiation of coagulation, circulating fVIIa levels have been viewed as an indicator of hemostatic potential and may also be an indicator for the risk of developing cardiovascular disease. The formation of the fVIIa/TF complex is also the basis of specific coagulation assays. Natural or synthetic TF is employed in the prothrombin time (PT assay) where it is used to initiate coagulation in vitro. Therefore, fVIIa and TF play important roles both in vivo and in vitro. Despite such important roles, there have been no simple methods described for measuring the active forms of either protein. This report describes the initial development of an assay for fVIIa and TF that employs the use of an Aminonaphthalenesulfonamide-based (ANSN) fluorogenic substrate. The ANSN substrate is hydrolyzed by fVIIa in a TF-dependent manner, and by titrating fVIIa into a system containing excess TF, the assay becomes sensitive to the fVIIa concentration. Inversely, by titrating TF into a system containing excess fVIIa, the assay becomes sensitive to the TF concentration. This approach has led to the development of both a rapid kinetic and an end-point assay method for each analyte providing detection ranges down to 156 and 1.56 pM respectively. The utility of these assays is displayed by the following applications. Thromboplastin reagents are generally qualified against a W.H.O. standard and are assigned a numerical value known as the International Sensitivity Index (ISI). Theoretically the ISI value should be affected by the quantity and quality of TF in each preparation. An inverse relationship between ISI value and the amount of functional TF would support this theory. To prove this relationship the TF assay was employed to examine two thromboplastin reagents with ISI values of 1.84 and 1.01. The concentration of functional TF measured by this assay was 10.8 nM and 14.6 nM respectively. Because most commercial thromboplastin reagents are prepared by adding TF on a mass basis (as opposed to a functional basis), target ISI values are often missed, leading to an excessive number of failed lots. The TF functional assay provides a method for assigning a specific activity to each TF preparation. By adding TF on a functional basis, lot-to-lot variability would be minimized and target ISI values would be easily achieved. To substantiate this claim, we have determined the functional activity of seven different lots of purified recombinant human TF. With each lot analyzed at a fixed concentration, we observed a broad range of specific activity (range: 8,250 U/sec/mg -14,500 U/sec/mg). These data support the need for a functional TF assay to be incorporated in the manufacturing process of thromboplastin reagents. Lastly, the utility of the fVIIa assay was demonstrated in an experiment using factor VII-deficient plasma samples that were spiked with known quantities of fVIIa. Plasma samples spiked with fVIIa in the range of 0.5-2nM returned mean assay values within 71–100% of the expected values. These data demonstrate the feasibility of using the fVIIa assay to monitor fVIIa levels in hemophilia patients receiving recombinant fVIIa therapy. Further studies will address the circulating concentrations of fVIIa and TF in “normal” and disease state plasma samples.

Effects of (Pre-)analytical Variables on Activated Protein C Resistance Determined Via a Thrombin Generation-based Assay

The normalized activated protein C sensitivity ratio (nAPC-sr) determined with an assay that quantifies the effect of APC on thrombin formation initiated via the extrinsic coagulation pathway identifies hereditary and acquired defects of the protein C system. We investigated the influence of assay conditions (analytical variables) and plasma handling (pre-analytical variables) on nAPC-sr obtained with this APC resistance test. The effect of the analytical variables (CaCl2, phospholipid and APC concentrations and the concentration and source of tissue factor) was determined in pooled normal plasma. Inhibition of thrombin formation by APC was dependent on the APC concentration and was also affected by the tissue factor, Ca2+ and phospholipid concentrations. Thus, strict standardization of reactant concentrations is required to obtain reproducible nAPC-sr. Three different tissue factor preparations were compared by determining nAPCsr in plasma samples obtained from 90 healthy individuals. nAPC-sr were similar for all three tissue factor preparations although, compared with the noncommercially available tissue factor used in earlier studies, values determined with commercial tissue factor preparations showed larger variation. Pre-analytical variables, investigated in plasma of nine volunteers (3 normal individuals and 6 individuals with an APC-resistant phenotype) were: concentration of anticoagulant (3.2% vs. 3.8% trisodiumcitrate), time before processing of blood (0, 4 and 24 h), centrifugation speed, storage temperature of plasma (-20 degrees C vs. -80 degrees C) and sample thawing. Multiple linear regression analysis showed that only the citrate concentration affected the nAPC-sr, which was higher in samples collected in 3.2% trisodiumcitrate than in samples collected in 3.8% trisodiumcitrate.

Magnesium and manganese ions accelerate tissue factor-induced coagulation independently of factor IX.

The purpose of the present study was to assess the effect of magnesium and manganese ions on tissue factor (TF)-induced coagulation and the possible role of factor IX therein. When magnesium chloride or manganese chloride were added in low concentrations to normal human plasma, the human (recombinant) TF-induced coagulation time was shortened. At higher concentrations, magnesium and manganese prolonged the TF-induced coagulation time. Maximum shortening of the coagulation time was obtained at a concentration of 0.5 mmol/l Mn or 2 mmol/l Mg in plasma. Shortening of the TF-induced coagulation time by magnesium and manganese was also observed in factor IX-deficient plasma. A comparison was made between TF preparations from human, rabbit, and bovine brain. The accelerating effect of magnesium was greater with human than with rabbit brain TF. Using bovine brain TF, the clotting time was not shortened by magnesium. Activated factor X-induced coagulation of normal plasma was not accelerated by magnesium. From these experiments, it is inferred that activation of factor X by factor VII-TF can be accelerated by magnesium and manganese ions independently of factor IX.

Influence of Three Types of Automated Coagulometers on the International Sensitivity Index (ISI) of Rabbit, Human, and Recombinant Human Tissue Factor Preparations

Five tissue factor reagents and three types of automated instruments for prothrombin time (PT) determination were studied in an international multicenter collaborative exercise. The purpose of this work was to determine the international sensitivity index (ISI) for each combination of reagent and instrument against the international reference preparation RBT/90. Each type of instrument was used by 3 or 4 centers to assess the interlaboratory variation of the ISI. The interlaboratory variation of the ISI for each combination of reagent and instrument ranged between 0.4% and 7.8% coefficient of variation. For three reagents, the mean ISI values for ACL (nephelometric) and STA (mechanical) were practically identical, but the mean ISI values for MLA (photo-optical) were at least 10% higher. For two other reagents prepared from rabbit tissue, the mean ISI values increased in the order ACL, STA, MLA. The widest range of mean ISI values was noted with one rabbit tissue factor reagent: 1.68 for ACL and 2.00 for MLA. Exclusion of patient specimens with INR <1.5 and INR >4.5 determined by the international reference preparation resulted in a slight decrease of the mean ISI. The interlaboratory variation of the International Normalized Ratio (INR) was assessed from the results obtained with common lyophilized and deep-frozen plasmas. The use of instrument-specific ISI values resulted in reduced interlaboratory variation of the INR. It is recommended that thromboplastin manufacturers provide instrument-specific ISI values.

Human fibroblast tissue factor is inhibited by lipoprotein-associated coagulation inhibitor and placental anticoagulant protein but not by apolipoprotein A-II

Studies of proteins that inhibit tissue factor activity have generally been conducted using either an extracted tissue homogenate (“thromboplastin”) or tissue factor protein reconstituted into phospholipid vesicles rather than with tissue factor expressed in cell membranes (its physiological environment). In the present study, a human fibroblast cell strain was used to evaluate the effects of lipoprotein associated coagulation inhibitor (LACI), placental anticoagulant protein (PAP), and apolipoprotein A-II (apo A-II) on human tissue factor in cell membranes. LACI was tested from 7.8 to 500 pmol/L on fibroblasts cultured at cell densities ranging from 3,500 to 9,925 cells/well, and caused a progressive inhibition of tissue factor activity. PAP was tested from 3.9 nmol/L to 1 mumol/L at cell densities ranging from 4,500 to 15,400 cells/well and caused up to 83% inhibition of tissue factor activity. Inhibition by these proteins appeared to be influenced by cell density as well as whether the cells were intact or disrupted. Apo A-II, up to 1 mumol/L, did not inhibit the tissue factor activity of intact or disrupted fibroblasts at any cell density examined even though it did inhibit the activity of tissue factor in phospholipid vesicles. Of these inhibitors of tissue factor-dependent activation of factor X, LACI was the most effective in suppressing the generation of factor Xa activity. The effects obtained with apo A-II are clearly dependent on the nature of the tissue factor preparation with which it is tested. The disparity between the inhibitory effect of apo A-II on the activity of tissue factor reconstituted into lipid vesicles and the absence of effect on the activity of tissue factor remaining in cell membranes serves to reemphasize the necessity of reexamining results obtained with model systems using as nearly physiological reagents as possible.

Human Fibroblast Tissue Factor Is Inhibited by Lipoprotein-Associated Coagulation Inhibitor and Placental Anticoagulant Protein but not by Apolipoprotein A-II

Studies of proteins that inhibit tissue factor activity have generally been conducted using either an extracted tissue homogenate ("thromboplastin") or tissue factor protein reconstituted into phospholipid vesicles rather than with tissue factor expressed in cell membranes (its physiological environment). In the present study, a human fibroblast cell strain was used to evaluate the effects of lipoprotein associated coagulation inhibitor (LACI), placental anticoagulant protein (PAP), and apolipoprotein A-II (apo A-II) on human tissue factor in cell membranes. LACI was tested from 7.8 to 500 pmol/L on fibroblasts cultured at cell densities ranging from 3,500 to 9,925 cells/well, and caused a progressive inhibition of tissue factor activity. PAP was tested from 3.9 nmol/L to 1 mumol/L at cell densities ranging from 4,500 to 15,400 cells/well and caused up to 83% inhibition of tissue factor activity. Inhibition by these proteins appeared to be influenced by cell density as well as whether the cells were intact or disrupted. Apo A-II, up to 1 mumol/L, did not inhibit the tissue factor activity of intact or disrupted fibroblasts at any cell density examined even though it did inhibit the activity of tissue factor in phospholipid vesicles. Of these inhibitors of tissue factor-dependent activation of factor X, LACI was the most effective in suppressing the generation of factor Xa activity. The effects obtained with apo A-II are clearly dependent on the nature of the tissue factor preparation with which it is tested. The disparity between the inhibitory effect of apo A-II on the activity of tissue factor reconstituted into lipid vesicles and the absence of effect on the activity of tissue factor remaining in cell membranes serves to reemphasize the necessity of reexamining results obtained with model systems using as nearly physiological reagents as possible.

Human tissue factor contains thioester-linked palmitate and stearate on the cytoplasmic half-cystine.

The state of the five half-cystine residues in human tissue factor (TF) has been characterized. The results indicate that the four half-cystines in the extracellular domain of TF form two disulfide bonds and the half-cystine in the cytoplasmic region is acylated by palmitic acid and stearic acid. The extracellular disulfide cross-links, Cys49-Cys57 and Cys186-Cys209, were deduced from the analysis of tryptic peptides. Acylation of the cytoplasmic half-cystine was demonstrated by purifying and characterizing fibroblast TF from cells labeled with [3H]palmitic acid. Radiolabeled fibroblast TF was observed by autoradiography following sodium dodecyl sulfate-polyacrylamide gel electrophoresis. The tritiated material covalently bound to the protein was identified as [3H]palmitate and [3H]stearate by reverse-phase high-pressure liquid chromatography. Deacylation of TF with hydroxylamine resulted in the spontaneous generation of disulfide-linked TF dimers. This result suggests that the disulfide-linked TF dimer, a minor component of most TF preparations, and the recently described heterodimeric form of TF are artifacts produced by deacylation of Cys245 and subsequent interchain disulfide bond formation.

Activation of intrinsic or extrinsic blood coagulation in experimental venous thrombosis and disseminated intravascular coagulation: Pathogenetic differences

Clinical observations and the findings in certain experimental animal stuides indicate that disseminated intravascular coagulation (DIC) and venous thromboembolism are not usually associated suggesting that different pathogenetic mechanisms may be responsible for their production. This thesis was evaluated in rabbits into which various procoagulant mixtures were infused. Ellagic acid, homologous serum and a factor IXa concentrate each induced a shortening of the partial thromboplastin time (PTT) and venous stasis thrombi but not defibrination and little systemic fibrin monomer (FM) elaboration. An entirely different pattern of response consisting of prolongation of the PTT, FM elaboration and defribrination but no thrombosis was found after the infusion of three tissue factor preparations, Xa or thrombin. It is proposed that activation of the intrinsic pathway of blood coagulation results in a “hypercoagulable state” which results in the formation of venous thrombi but does not cause DIC. The latter is triggered by activation of the tissue or final common pathway and is associated with elaboration of an anticoagulant activity which prolongs the PTT and which may prevent the formation of venous thrombi.

Procoagulant activity of thrombosthenin preparations.

<h2>Abstract</h2> 3 x precipitated bovine thrombosthenin preparation exerted a powerful procoagulant effect similar to platelet factor 3 activity. In the presence of thrombosthenin the recalcification time, the kaolin clotting time and the Stypven time considerably shortened, the partial thromboplastin time only slightly altered and the activated partial thromboplastin time did not change at all. Thrombosthenin highly accelerated the generation of intrinsic thromboplastin but it was not active in a factor IXa generation test. It neither possessed tissue factor like property nor influenced the activity of tissue factor preparation or the action of thrombin on purified fibrinogen. Thrombosthenin contained 109 μg phospholipid/mg protein the composition of which was quantitatively determined by thin layer chromatography. The procoagulant activity as well as the phospholipid content were retained after agarose gel filtration. It is suggested that thrombosthenin having an associated phospholipid component significantly contributes to the clot promoting effect of activated platelets.

The hydrolysis of biologically active peptides by bovine lung tissue factor (thromboplastin).

Tissue factor apoprotein and relipidated tissue factor preparations extensively hydrolyze bradykinin, Lys-bradykinin, Met-Lys-bradykinin, substance P, [Asp1, Ile5]-angiotensin II, [Asp1, Ile5]-angiotensin I, and human fibrinopeptide A while acting more slowly on [Sar1, Ile5]-angiotensin II, [Me2Gly1, Ile5]-angiotensin II, bradykinin potentiating pentapeptide from B. jararaca, luteinizing hormone-releasing hormone, melanocyte stimulating hormone-release-inhibiting factor (Pro-Leu-Gly-NH2), and oxytocin. No hydrolysis of thyrotropin-releasing factor or bradykinin potentiating nonapeptide from B. jararaca is observed. Relipidated and apoprotein tissue factor act at identical rates under the conditions of the assay. Dansylation and chromatography of tissue factor-peptide incubation mixtures further indicate that relipidated and apoprotein tissue factor also hydrolyze peptides by identical mechanisms. No fewer than six bonds are hydrolyzed in bradykinin while the angiotensins and substance P are degraded to constituent amino acids. Only the N-terminal alanine is released from fibrinopeptide A. 2-Mercaptoethanol greatly inhibits the hydrolysis of bradykinin by relipidated tissue factor.