Tissue factor (thromboplastin): localization to plasma membranes by peroxidase-conjugated antibodies.

Peroxidase-conjugated antibodies were used to determine the histologic and cytologic localization of bovine and human tissue factor (thromboplastin). Tissue factor antigen was found in highest concentration in the intima of blood vessels, particularly in the plasma membranes of endothelial cells and in human atheromatous plaques. Tissue factor was also found limited to the plasma membranes of many cell types. The presence of tissue factor in the plasma membranes of endothelial cells and atheromata suggests that it may play a significant role in hemostasis and thrombosis.

Characteristics and lipid requirements of coagulant proteins extracted from lung and brain: the specifity of protein component of tissue factor.

Soluble proteins were prepared from bovine lung and brain which, when combined with phospholipids, had the biological characteristics of tissue factor. The proteins were solubilized from delipidated (heptane; butanol-extracted) acetone powders with deoxycholate, and precipitated by (NH(4))(2)SO(4) (30-60% saturation). These soluble proteins, which contained less than 1% phospholipid, were essentially inert in an assay for tissue factor. When they were recombined with phospholipids, however, activity increased by a factor of 500-1000. Phosphatidylethanolamine was the most active specific phospholipid, followed by phosphatidylcholine. Phosphatidylserine was inert. Mixed phospholipids were from two to four times more active than phosphatidylethanolamine. Maximum activity was obtained at phospholipid to protein ratios of 1.5:1 (wt/wt), and when relipidation was performed in deoxycholate followed by dialysis against 0.05 M imidazole-0.375 M NaCl, pH 7.2. The proteins were characterized by gel filtration on Sephadex G-200; the activity eluted as a single peak with an apparent mol wt of 425,000. The brain and lung activities were recovered from chromatography on triethyl-aminoethyl-cellulose as single peaks, although the salt concentration required for elution was different for each protein. This suggests that within each organ there is one tissue factor-protein, but there may be molecular differences between brain and lung tissue factor. Alternatively, the chromatographic differences may simply reflect charge differences imparted to the proteins by residual lipids.

The phospholipid requirement of tissue factor in blood coagulation.

Using a coagulation assay specific for tissue factor, we found that removal of 95% of the tissue factor-phospholipid resulted in a loss of 98% of its biological activity. The activity could be restored, with yields in excess of 100% by combining the extracted tissue factor with either mixed brain phospholipids or highly purified phospholipids. Phosphatidylethanolamine was the most active, followed by phosphatidylcholine. Phosphatidylserine, phosphatidylinositol, and sphingomyelin had little or no activity. In addition, a requirement for unsaturation and the presence of two fatty acids was demonstrated. The activity of phosphatidylcholine was also dependent on the presence of the base. Furthermore, it was shown that activity was not a function of binding of phospholipids to tissue factor, as both active and inactive lipids were equally bound.

Association of tissue factor activity with the surface of cultured cells.

Tissue factor occurs in a dormant state on the surface of cultured normal human fibroblasts and WISH 1 amnion cells. The activity of undisturbed monolayers or cells lifted with brief trypsin treatment (0.125 per cent trypsin for 1 min) increases up to 60-fold upon prolonged digestion with dilute trypsin (0.0025 per cent trypsin for 30 min); activity appears subsequent to cell detachment. Up to 70 per cent of the total cellular tissue factor becomes active under these conditions and is released from the cells. The ruthenium red staining coat of the cells is lost during detachment, but cell viability (more than 90 per cent exclude trypan blue) and cell morphology do not change during the subsequent development of tissue factor activity. Furthermore, less than 10 percent of four intracellular enzymes and less than 20 per cent of two plasma membrane enzymes are released during this period of time. We therefore conclude that cells in culture do have tissue factor activity, that it exists in a latent form, and that total cell disruption is not necessary for this activity to initiate blood coagulation.

Tissue factor is rapidly induced in arterial smooth muscle after balloon injury.

Tissue factor (TF) is a major activator of the coagulation cascade and may play a role in initiating thrombosis after intravascular injury. To investigate whether medial vascular smooth muscle provides a source of TF following arterial injury, the induction of TF mRNA and protein was studied in balloon-injured rat aorta. After full length aortic injury, aortas were harvested at various times and the media and adventitia separated using collagenase digestion and microscopic dissection. In uninjured aortic media, TF mRNA was undetectable by RNA blot hybridization. 2 h after balloon injury TF mRNA levels increased markedly. Return to near baseline levels occurred at 24 h. In situ hybridization with a 35S-labeled antisense rat TF cRNA probe detected TF mRNA in the adventitia but not in the media or endothelium of uninjured aorta. 2 h after balloon dilatation, a marked induction of TF mRNA was observed in the adventitia and media. Using a functional clotting assay, TF procoagulant activity was detected at low levels in uninjured rat aortic media and rose by approximately 10-fold 2 h after balloon dilatation. Return to baseline occurred within 4 d. These data demonstrate that vascular injury rapidly induces active TF in arterial smooth muscle, providing a procoagulant that may result in thrombus initiation or propagation.

Agonist-mediated tissue factor expression in cultured vascular smooth muscle cells. Role of Ca2+ mobilization and protein kinase C activation.

Tissue factor (TF) is a low molecular weight glycoprotein that initiates the clotting cascade and is considered to be a major regulator of coagulation, hemostasis, and thrombosis. TF is not expressed in the intima or media of normal adult blood vessels. Accordingly, it has been hypothesized that the initiation of intravascular coagulation may require the "induced" expression of TF in the vessel wall. We report that TF mRNA and protein are rapidly and markedly induced in early and late passaged vascular smooth muscle cells (VSMC) by growth factors (serum, platelet-derived growth factor, epidermal growth factor), vasoactive agonists (angiotensin II), and a clotting factor (alpha-thrombin). The induction of TF mRNA by these agents is dependent upon mobilization of intracellular Ca2+ and is blocked by Ca2+ chelation. In contrast to other growth factor-responsive genes, such as KC and c-fos, downregulation of protein kinase C activity by prolonged treatment with phorbol esters fails to block agonist-mediated TF induction. This raises the possibility that protein kinase C activation may not be necessary for TF mRNA induction in VSMC. VSMC may play a role in the generation or propagation of thrombus through the induction of TF, particularly in settings, such as those associated with acute vessel injury, where the endothelium is denuded and the VSMC are exposed to circulating blood.

Tissue factor expression in human arterial smooth muscle cells. TF is present in three cellular pools after growth factor stimulation.

Tissue factor (TF) is a transmembrane glycoprotein that initiates the coagulation cascade. Because of the potential role of TF in mediating arterial thrombosis, we have examined its expression in human aortic and coronary artery smooth muscle cells (SMC). TF mRNA and protein were induced in SMC by a variety of growth agonists. Exposure to PDGF AA or BB for 30 min provided all of the necessary signals for induction of TF mRNA and protein. This result was consistent with nuclear runoff analyses, demonstrating that PDGF-induced TF transcription occurred within 30 min. A newly developed assay involving binding of digoxigenin-labeled FVIIa (DigVIIa) and digoxigenin-labeled Factor X (DigX) was used to localize cellular TF. By light and confocal microscopy, prominent TF staining was seen in the perinuclear cytoplasm beginning 2 h after agonist treatment and persisting for 10-12 h. Surface TF activity, measured on SMC monolayers under flow conditions, increased transiently, peaking 4-6 h after agonist stimulation and returning to baseline within 16 h. Peak surface TF activity was only approximately 20% of total TF activity measured in cell lysates. Surface TF-blocking experiments demonstrated that the remaining TF was found as encrypted surface TF, and also in an intracellular pool. The relatively short-lived surface expression of TF may be critical for limiting the thrombotic potential of intact SMC exposed to growth factor stimulation. In contrast, the encrypted surface and intracellular pools may provide a rich source of TF under conditions associated with SMC damage, such as during atherosclerotic plaque rupture or balloon arterial injury.

Identification and characterization of murine alternatively spliced tissue factor.

Tissue factor (TF) is a transmembrane glycoprotein that initiates coagulation and plays a critical role in regulating hemostasis and thrombosis. We have recently reported a naturally occurring, soluble form of human tissue factor (asTF) generated by alternative splicing. This splice variant has a novel C-terminus with no homology to that of the full-length TF (flTF), lacks a transmembrane domain, and is active in the presence of phospholipids. Mouse models offer unique opportunities to examine the relative importance of flTF and asTF in mediating thrombosis, the response to arterial injury, and ischemic damage. To that end, we have identified and characterized murine asTF (masTF). Like the human splice variant, masTF lacks a transmembrane domain and has a unique C-terminus. We have generated antibodies specific to masTF and murine flTF (mflTF) to examine the expression of both forms of TF. masTF antigen is widely and abundantly expressed, with a pattern similar to that of mflTF, in adult tissues, in experimentally induced thrombi, and during development. These studies demonstrate that masTF contributes to the pool of total TF and may thus play an important role in mediating TF-dependent processes.

Initiation of blood coagulation: the tissue factor/factor VIIa complex.

Tissue factor (TF), a transmembrane glycoprotein, functions as an essential activator of the serine protease factor VIIa. This enzymatic complex is considered to be the principal initiator of in vivo coagulation. Recent studies emphasize the role of the TF/VIIa complex in a number of pathophysiological processes, such as Gram-negative sepsis, coronary artery disease and neointimal hyperplasia after angioplasty. Monocytes/macrophages are important contributors to some of these diseases and there have been new insights into the biology of TF regulation in monocytes. In the light of its structural similarity to cytokine receptors, there has been frequent speculation that TF has a role in intracellular signaling, a suggestion that is supported by some recent studies that propose a true receptor function for TF.

Tissue factor, the blood, and the arterial wall.

Thrombogenic tissue factor (TF) on cell-derived microparticles is present in the circulating blood of patients with acute coronary syndromes. Recently, we reported that leukocytes transfer TF-positive particles to platelet thrombi, making them capable of triggering and propagating thrombus growth. This observation changes the original dogma that vessel-wall injury and exposure of tissue factor within the vasculature to blood is sufficient for the occurrence of arterial thrombosis. The transfer of TF-positive leukocyte particles is dependent on the interaction of CD15 and TF with platelet thrombi. The inhibition of TF transfer and TF activity suggests a novel therapeutic approach to the prevention of thrombosis that may prove to be effective in disorders associated with increased blood TF.

Biological mechanisms underlying RU 486 clinical effects: inhibition of endometrial stromal cell tissue factor content.

Despite the pronounced hemorrhagic effects of RU 486 administration on luteal phase and early gestational endometrium, no information is available on the effect of RU 486 on endometrial hemostatic potential. The expression of endometrial stromal cell tissue factor (TF), the primary initiator of hemostasis, has been shown to be progestationally regulated in vivo and in vitro. To evaluate the effects of RU 486 on progestin-enhanced TF expression, confluent stromal cell cultures derived from proliferative phase endometria were exposed to vehicle control, 10(-8) mol/L estradiol (E2), 10(-6) mol/L dexamethasone, 10(-7) mol/L medroxyprogesterone acetate (MPA), E2 plus MPA, E2 plus 10(-6) mol/L progesterone (P), or 10(-6) mol/L RU 486 alone or with E2 plus MPA or E2 plus P for 3-4 days. Compared to the vehicle control, E2, dexamethasone, and RU 486 alone had no effect on the content of immunoreactive and functionally active TF protein, whereas MPA increased and the combination of E2 and MPA further increased TF protein content. Similarly, E2 and P enhanced the stromal cell TF content. These progestin effects were blocked by RU 486. Similar results were obtained for steady state TF messenger ribonucleic acid (mRNA) levels. Possible RU 486-mediated reversal of progestin-enhanced stromal cell TF expression was assessed by incubating confluent cultures in E2 plus MPA for 3-10 days to enhance TF content, then washing the cultures and reexposing them to either E2 plus MPA or to RU 486 alone or with E2 plus MPA for 3, 4, or 7 days. Exposure to RU 486 alone or with E2 plus MPA greatly reduced levels of stromal cell TF protein and mRNA expression compared to those in cultures maintained in E2 plus MPA. These findings demonstrate that RU 486 not only blocks but also reverses in vitro progestin-enhanced stromal cell TF protein and mRNA expression, suggesting an additional mechanism for RU 486-induced menses and early abortion.

Progestational regulation of human endometrial stromal cell tissue factor expression during decidualization.

Decidualized endometrial stromal cells from mid- to late secretory phase and decidual cells from gestational human endometrium displayed prominent immunohistochemical staining for tissue factor, a primary initiator of hemostasis. Consistent with the regulation by progesterone of the decidualization process in vivo, medroxy-progesterone acetate elevated the tissue factor content of primary stromal cell cultures 8-fold over basal values. This increase paralleled the release of immunoreactive PRL, a marker of decidualization. The induced, as well as basal, tissue factor displayed full functional activity and the expected electrophoretic mobility (46 kilodaltons). Moreover, Northern analysis of RNA from cultured stromal cells indicated that medroxyprogesterone acetate increased tissue factor mRNA levels approximately 10-fold relative to control levels. In contrast, cultured stromal cell tissue factor protein content and mRNA levels were unaffected by exogenous estradiol. These findings indicate that enhancement of endometrial stromal cell tissue factor content is associated with progesterone induction of the decidualization process. In humans, trophoblastic invasion of the endometrial vasculature during blastocyst implantation risks hemorrhage. Therefore, increases in perivascular decidual cell tissue factor expression could serve to promote periimplantational endometrial hemostasis.

Steroid-modulated stromal cell tissue factor expression: a model for the regulation of endometrial hemostasis and menstruation.

This study examined steroidal regulation of tissue factor expression by cultured endometrial stromal cells. Confluent stromal cell cultures derived from cycling human endometria were exposed to vehicle control, 10(-8) mol/L estradiol (E2), 10(-8)-10(-6) mol/L medroxyprogesterone acetate (MPA), or both E2 and MPA for 2-24 days in serum-containing medium. The progestin enhanced immunoreactive and functionally active stromal cell tissue factor content, achieving peak effects by 8-12 days of culture. Although E2 alone was ineffective, it augmented MPA-enhanced tissue factor content by 8 days of culture, with continued increases beyond 20 days. Dose-dependent effects on tissue factor protein content were observed between 10(-8)-10(-6) mol/L MPA added alone or together with E2. The content of tissue factor mRNA was also increased by MPA and synergistically increased by E2 plus MPA. Similar steroidal effects on stromal cell tissue factor protein and mRNA content were observed using a defined medium. After optimal induction of tissue factor expression by E2 plus MPA, removal of these steroids reduced levels of stromal cell tissue factor mRNA and protein, with virtually complete reversal by day 7 of withdrawal. These time-course and dose-response relationships establish in vitro conditions with which to dissect factors controlling endometrial hemostasis, whereas the observed effects of steroid withdrawal establish a novel model for the study of mechanisms regulating normal and abnormal uterine bleeding.

Inhibition of tissue factor limits the growth of venous thrombus in the rabbit.

Antibody mediated inhibition of tissue factor (TF) function reduces thrombus size in ex vivo perfusion of human blood over a TF-free surface at venous shear rates suggesting that TF might be involved in the mechanism of deep vein thrombosis. Moreover, TF-bearing monocytes and polymorphonuclear (PMN) leukocytes were identified in human ex vivo formed thrombi and in circulating blood. To understand the role of TF in thrombus growth, we applied a rabbit venous thrombosis model in which a collagen-coated thread was installed within the jugular vein or within a silicon vein shunt. The effect of an inhibitory monoclonal antirabbit TF antibody (AP-1) or Napsagatran, a specific inhibitor of thrombin, was quantified by continuously monitoring 125I-fibrinogen incorporation into the growing thrombi. The antithrombotic effect obtained with the anti-TF antibody was comparable to the effect observed with the thrombin inhibitor napsagatran suggesting that in this animal model the thrombus propagation is highly TF dependent. Immunostaining revealed that TF was mostly associated with leukocytes within the thrombi formed in the jugular vein or in the silicon vein shunt. Ex vivo perfusion experiments over collagen-coated coverslips demonstrated the presence of TF-bearing PMN leukocytes in circulating blood. The results suggest that in rabbits venous thrombus growth is mediated by clot-bound TF and that blocking the TF activity can inhibit thrombus propagation.

Biological control of factor VII.

The tissue factor pathway is initiated by factor VII in the presence of tissue factor. The first proteolytic reaction involves cleavage of factor X by factor VII. Activated factor X, the product of this reaction, can activate factor VII by cleavage of a specific bond. The apparent coagulant activity of factor VII then rises about 60-fold. Activated factor X can also inactivate factor VII by catalyzing the cleavage of a second bond which results in a three chain molecule. Fragments of Hageman factor and perhaps kallikrein can also activate factor VII. Hageman factor, however, does not catalyze the inactivating cleavage of factor VII at a significant rate. Recent data showing that the tissue factor pathway can activate the intrinsic system are discussed. We have shown that activated factor X, which can be generated by the tissue factor pathway, can feed back and activate factor IX in a calcium and phospholipid requiring reaction.

Tissue factor: then and now.

Much has been learned about tissue factor during the past 25 years and much is yet to be learned. On the biochemical level, the association of TF:VIIa with factor X is yet to be resolved. The mechanism by which TF is transported from the nucleus to the plasma membrane has yet to even be approached. Thus, those of us studying this interesting and fundamentally important molecule have much work in the future.

The effect of flow on hemostasis and thrombosis.

While dilution of procoagulants has generally been proposed as the mechanism by which flow reduces coagulation at surfaces, such a mechanism has never been verified experimentally and, in fact, there are theoretical grounds for suspecting the validity of such a hypothesis (29). It is quite plausible that flow may have direct effects on certain enzyme or polymerization kinetics involved in thrombosis, in addition to the well-defined effect that flow has an enhancing transport of reactants and products to and from the vessel wall. Such effects of flow on immobilized enzymes have occasionally been observed, but never studied with respect to coagulative processes (30). The study of the effects of flow on hemostasis and thrombosis, while numerous, are still in their infancy. As noted above, increasing shear increases the rate of formation of factor Xa in a tubular reactor. In the presence of factors VIII and IX, there is also a shear-induced enhancement of Xa production (31). These studies indicate that at least some coagulation reactions are accelerated in the presence of high shear. However, it has been observed that fibrin formation is diminished at increasing shear rates (20). This implies that at least one step of the coagulation cascade is being inhibited by high shear. One possibility is that fibrin monomer is being removed by the high local flow conditions, although the concomitant reduction in fibrinopeptide A argues against this interpretation. Another possibility, not yet tested, is that thrombin itself is removed by flow.(ABSTRACT TRUNCATED AT 250 WORDS)

Factors affecting the interaction of tissue factor/factor VII with factor X in a heterogeneous tubular reactor.

A novel reactor recently described for studying phospholipid-dependent blood coagulation reactions under flow conditions similar to those occurring in the vasculature has been further characterized. The reactor is a capillary whose inner wall is coated with a stable phospholipid bilayer (or two bilayers) containing tissue factor, a transmembrane protein that is required for the enzymatic activation of factor X by factor VIIa. Perfusion of the capillary at wall shear rates ranging from 25 s-1 to 1,200 s-1 with purified bovine factors X and VIIa led to steady state factor Xa levels at the outlet. Assay were performed using a chromogenic substrate, Spectrozyme TMFXa, or by using a radiometric technique. In the absence of Ca2+ or factor VIIa there was no product formation. No difference was noted in the levels of factor Xa achieved when non-activated factor VII was perfused. Once steady state was achieved further factor Xa production continued in the absence of factor VIIa implying a very strong association of factor VIIa with the tissue factor in the phospholipid membrane. In agreement with static vesicle-type studies the reactor was sensitive to wall tissue factor concentration, temperature and the presence of phosphatidylserine in the bilayer.

Regulation of the procoagulant response to arterial injury.

The last few years have provided increasing evidence to support a major role for TF in the initiation and propagation of thrombosis after acute arterial injury. Although thrombotic occlusion occurs in a small minority of patients undergoing acute coronary interventions or bypass surgery, mural thrombi are likely to be present in almost all cases. These thrombi may stimulate SMC and promote the development of intimal hyperplasia and luminal narrowing. The use of inhibitors of TF and factor VIIa, therefore, may not only be valuable for inhibiting thrombus formation associated with acute arterial interventions, but may also have benefit in attenuating intimal hyperplasia. Although this paper focuses on the role of TF in establishing a procoagulant state after arterial injury, the fibrinolytic system undoubtedly plays a role in balancing the effects of increased TF production in the arterial wall. This is underscored by the success of activators of fibrinolysis (tissue plasminogen activator, streptokinase, urokinase) in revascularization in the setting of acute myocardial infarction and is reviewed elsewhere. Likewise, local regulation of TFPI in the atherosclerotic plaque and injured vessel wall may be important in attenuating the effects of increased TF synthesis and accumulation. It has been assumed that the primary source of active TF after arterial injury is either SMC or invading macrophages and that active TF is anchored to the surface of these cells. Recent data have suggested that the majority of cell-associated TF is either encrypted on the cell surface or present in an intracellular pool. Arterial injury may, therefore, involve the de-encryption of surface TF or the release of intracellular TF. In addition, active vascular TF may be present in microparticles that are not anchored to the arterial wall and may be washed into the circulation. The procoagulant state may be further accentuated by the accumulation of bloodborne TF at sites of arterial injury and in developing thrombi. This TF is likely to arise from circulating leukocytes, including neutrophils and monocytes. These studies suggest that the cellular processing of TF may be an important target for inhibiting thrombotic complications associated with arterial injury and acute coronary events.

Activation of factor X by factor VIIa complexed with human-mouse tissue factor chimeras requires human exon 3.

In an attempt to define sequence elements in human and mouse tissue factor (TF) that are responsible for the species specificity observed in their interaction with human factor VIIa (HVIIa), we constructed human-mouse chimeric TF cDNAs, inserted them into plasmid vectors, and induced their expression in E. coli. Assays for procoagulant activity were carried out with the resulting E. coli lysates using (HVIIa) human and mouse (MVIIa). The ratio of the procoagulant activities, HVIIa/MVIIa, revealed that human TF exon 3 was essential for activity when the TF:VIIa complex was formed with HVIIa. By ligating the maltose binding protein (MBP) gene to TF cDNAs it was possible to construct, express and purify MBP-TF chimeras as well as to estimate their specific activities. With selected MBP-TF chimeras and HVIIa we determined kinetic parameters for the activation of human factor X. Replacement of exon 3 in human TF cDNA with the corresponding exon from mouse TF cDNA resulted in both lower affinity for HVIIa and failure to convert bound HVIIa into a potent protease.