Expression and modulation of tissue factor and tissue factor pathway inhibitor in an endothelial cell based model

Abstract

Haemostasis is a complex physiological process involving cellular and plasma protein components that interact to keep the blood fluid under normal conditions and prevent blood loss after vessel injury by promoting clot formation. Primary haemostasis encompasses the activation and aggregation of platelets and is supported by secondary haemostasis, in which the coagulation factors of the plasma interact in a complex series of reactions. Secondary haemostasis is initiated by the exposure of tissue factor (TF) to the blood after vessel injury. TF forms a complex with activated factor VII (FVIIa), which in turn activates factor X (FXa) and ultimately results in fibrin formation. The TF-FVIIa complex and FXa are tightly regulated by tissue factor pathway inhibitor (TFPI), a trivalent Kunitz-type protease inhibitor. The endothelium, consisting of endothelial cells (ECs), constitutes the inner lining of all blood vessels. As such, it is in constant contact with the blood and plays a major role in haemostasis by synthesising and storing both pro- and anti- coagulant substances, including TF and TFPI. Release of TFPI from ECs is increased after exposure to both unfractionated and low molecular weight heparins, though the mechanisms are not clearly defined. TFPI circulates in plasma, predominantly bound to lipoproteins, though the effect of the three major lipoproteins [low density (LDL), very low density (VLDL) and high density (HDL)] on the release of TFPI from ECs is not well established. Furthermore, previous studies have not systematically investigated the effect of these lipoproteins on both TF and TFPI. The initial aim of this project was to establish assays for the measurement of TF activity and TFPI antigen to supplement the TFPI activity assay that is well established in our laboratory.These assays were then used to determine the effects of heparin and the major lipoproteins on the expression of TF and the release of TFPI on/from ECs. Human umbilical vein endothelial cells (HUVECs) were used as the EC model because their collection and isolation is well established and they have biochemical and physiological properties representative of in vivo conditions. A TF activity assay, based on a previously published method, was successfully modified and validated for the measurement of cell surface TF (standard curve R2 = 0.997). Despite exhaustive attempts, adaptation of this assay for plasma TF was unsuccessful, raising doubts regarding the plasma fractionation procedure of the originally published assay [Fukuda, C., Iijima, K. and Nakamura, K. (1989). "Measuring tissue factor (factor III) activity in plasma." Clinical Chemistry 35(9): 1897†1900]. A novel insect cell expression system was used to produce well defined recombinant TFPI standards for use in TFPI activity and antigen assays. For the first time, truncated TFPI variants, containing the first Kunitz domain only, the first and second Kunitz domains only, and the first through third Kunitz domains minus the carboxyl terminus, were successfully produced in insect cells, though the full length molecule was not. Possible reasons for this include codon bias, protein instability and/or the signal peptide used. An ELISA to measure TFPI antigen was designed using a monoclonal anti†TFPI antibody directed against the N†terminus for protein capture and a polyclonal anti†TFPI antibody for detection. The assay was successfully optimised (standard curve R2 = 0.978, intra†assay CV = 4.8%), however it produced inaccurate results (normal range = 498.7 ± 156.3 ng/mL), probably due to the antibody combination used.TF and TFPI activity assays were used to determine the effect of both unfractionated and low molecular weight heparins (UFH and LMWH, respectively) on the release of TFPI and the expression of TF from/on ECs. A significant increase in the secretion of functional TFPI from ECs due to heparin (0 U/ml vs 1 and 10 U/mL) was demonstrated only in the presence of serum (UFH: 9.0 mU/mL vs 18.3 and 18.4 mU/mL, p < 0.0001; LMWH: 8.8 mU/mL vs 13.3 and 21.4 mU/mL, p < 0.05), suggesting, for the first time, that a component of serum is required for the heparin†dependent release of TFPI. The effect of LDL, VLDL and HDL on the release of TFPI and the expression of TF from/on ECs was also investigated. All three lipoprotein fractions increased the secretion of functional TFPI after one hour incubation (LDL: 12.5 μg/mL, p < 0.01; 25 μg/mL, p < 0.05; VLDL: 50 μg/mL, p < 0.01; HDL: 50 μg/mL, p < 0.05). This is the first data to demonstrate a HDL†dependent increase in released TFPI. After 24 hours, both LDL and VLDL decreased levels of secreted functional TFPI (LDL: 25 μg/mL, p < 0.01; 50 μg/mL, p < 0.01; VLDL: 12.5 μg/mL, p < 0.01), probably due to the oxidation and subsequent association of both lipoprotein species with TFPI. Surprisingly, both LDL and VLDL decreased cell surface TF, though this effect was not dose dependent. These results suggest that the major lipoproteins have a short term anticoagulant effect which is reversed in the longer term due to lipid oxidation. In summary, this thesis describes the successful adaptation of a chromogenic assay for the measurement of cell surface TF activity and the production of truncated TFPI variants.Both will be used for the measurement of TF and TFPI, their association with thrombus formation and propagation, and investigations into potential therapeutic applications of TFPI. The results presented in this thesis extend the current knowledge on the expression and release of TF and TFPI on/from ECs by heparin, highlighting the importance of serum in the heparin dependent release of TFPI in vitro. Furthermore, it describes for the first time the effects of the major lipoprotein fractions on TFPI release and TF expression. The data support novel mechanisms by which LDL and VLDL are procoagulant, and HDL anticoagulant. This study provides a foundation for future research of the TF pathway in cellular models, which is critical in increasing the understanding of the pathogenesis and treatment of thrombotic disease. vitro. Furthermore, it describes for the first time the effects of the major lipoprotein fractions on TFPI release and TF expression. The data support novel mechanisms by which LDL and VLDL are procoagulant, and HDL anticoagulant. This study provides a foundation for future research of the TF pathway in cellular models, which is critical in increasing the understanding of the pathogenesis and treatment of thrombotic disease.