Effect of Cholesterol Lowering on Intravascular Pools of TFPI and Its Anticoagulant Potential in Type II Hyperlipoproteinemia
Abstract Tissue factor pathway inhibitor (TFPI) inhibits the extrinsic coagulation system. A major pool of TFPI is associated with the vascular endothelium and can be mobilized into the circulation by heparin. In circulating blood, TFPI is mainly associated with LDL (80%), whereas 10% to 20% is carrier free. In this study, heparin administration caused a 2.2-fold and a 7.5-fold increase in TFPI activity and TFPI antigen, respectively, in 25 patients with phenotypes IIa and IIb hyperbetalipoproteinemia. Because the antigen determination of TFPI almost exclusively measures carrier-free TFPI, more than 90% of the heparin-induced increase in TFPI activity was caused by mobilization of carrier-free TFPI from the vascular endothelium. Therapeutic lowering of total cholesterol (a decrease of 31.1±11.6%, P<.001) by 40 mg/d lovastatin in 17 patients with hyperbetalipoproteinemia was accompanied by a parallel decrease in TFPI activity (of 27.7±24.2%, P<.001) because of a reduction in LDL-TFPI complexes. However, drug intervention did not affect carrier-free TFPI or the magnitude of the vascular pool of TFPI that could be mobilized into the circulation by heparin. Moreover, this reduction of LDL-TFPI complexes did not reduce the anticoagulant potency of TFPI in plasma or of the vascular endothelial pool. The results of this study may imply that the anticoagulant potency of TFPI is associated with its carrier-free form in plasma or on the endothelium and that downregulation of LDL affects neither the size nor the anticoagulant potency of the endothelial pool of TFPI.
- Received November 14, 1994.
- Accepted April 12, 1995.
Exposure of tissue factor (TF) to circulating blood is believed to trigger blood coagulation during normal hemostasis1 and to play a major role in thrombogenesis associated with atherosclerosis.2 TF pathway inhibitor (TFPI) is a protease inhibitor that may function as a natural anticoagulant regulating TF-induced coagulation.3 Requiring factor Xa, TFPI neutralizes the catalytic activity of factor Xa and inactivates the factor VIIa–TF complex.4 5 TFPI is the major inhibitor of the factor VIIa–TF activity formed either in vitro with TF in suspension6 or with TF expressed on the surface membranes of perturbed endothelial cells.7 Studies in rabbits indicate that TFPI functions as a natural anticoagulant by inhibiting the factor VIIa–TF catalytic activity formed in circulating blood exposed to a low concentration of TF.8
The endothelium is the primary site of TFPI synthesis under normal conditions,9 and TFPI exists in four intravascular pools in humans.3 In plasma, more than 80% is bound to lipoproteins (pool I), primarily LDL, whereas 10% to 20% is carrier free (pool II). A major pool is probably associated with the vascular endothelium (pool III), and is released into the circulation within minutes of intravenous injection of heparin.10 Heparin-releasable TFPI functions as an important physiological anticoagulant and is probably involved in the therapeutic action of heparins.11 The magnitude of TFPI release varies twofold to 10-fold because of differences in methodology for the determination of TFPI.10 12 The endothelial pool may have a crucial regulatory function because the endothelial surface serves as a binding site for factor VII/VIIa, mediating the activation of the extrinsic coagulation system.1 Lastly, platelets contain a small amount of TFPI (pool IV), about 8% of the plasma pool.
A positive correlation between the total plasma TFPI activity and LDL cholesterol/apolipoprotein B has been reported in both normal subjects13 and hypercholesterolemic patients.14 15 No consistent correlations have been found to HDL cholesterol, triglycerides, or apo A.14 15 Subjects with familial hypercholesterolemia have higher TFPI levels than age- and sex-matched normolipemic subjects.15 Their TFPI activity normalizes during cholesterol-lowering therapy14 15 because of a specific drop in LDL-TFPI complexes in plasma.15 In contrast, an increase in TFPI due to an enhancement in plasma LDL-TFPI complexes was recently demonstrated in monkeys fed a cholesterol-rich diet.16 Moor et al17 reported increased levels of plasma TFPI activity in young male postinfarction patients, compared with population-based controls, and the TFPI activity correlated with dense LDL and HDL particles. The biological link between lipoproteins and the extrinsic coagulation system may have important implications in the pathophysiology of thrombosis and atherosclerosis.
TFPI activity or antigen concentration may not necessarily reflect its anticoagulant potency.18 Thus, determination of the anticoagulant potential of TFPI in an extrinsic coagulation assay may provide important information about the functional properties of TFPI from different pools. The aim of the present study was to determine how cholesterol lowering in patients with type II hyperbetalipoproteinemia (HLP) influenced the distribution of TFPI in different intravascular pools and how these changes affected the anticoagulant potency of plasma- and endothelium-associated pools.
Patients and Study Design
Twenty-five patients referred to the Department of Medicine at the University Hospital of Tromsø for phenotypic type II hyperlipoproteinemia were recruited for the study. Fourteen patients had type IIa HLP (total cholesterol above 8.0 mmol/L and triglycerides less than 2.0 mmol/L) and 11 had type IIb HLP (total cholesterol above 8.0 mmol/L and triglycerides between 2.0 and 5.0 mmol/L). These patients had been on a diet low in cholesterol and saturated fat (American Heart Association Step I diet) for at least 3 months. All participants had normal thyroid, renal, and hepatic function, and none had diabetes or other chronic illnesses. This study group included 20 men and 5 women between 27 and 55 years of age. Seven patients had tendon xanthomas and 20 had family histories of premature cardiovascular disease. Four patients had manifest cardiovascular disease (1 had type IIa HLP and 3 had type IIb HLP). These patients were using aspirin and β-blockers, whereas the rest of the patients did not use any drugs.
The patients were asked before the dietary intervention to participate in a clinical trial that included additional blood sampling and heparin injection after the dietary intervention (ie, at visit 1) and again after cholesterol-lowering treatment by lovastatin (Merck Sharp and Dohme), a hydroxymethylglutaryl–coenzyme A reductase inhibitor (at visit 2). Patients with peptic ulcers, gastrointestinal disorders likely to influence drug absorption, alcoholism, drug abuse, mental illness, or contraindications to heparin injection were not asked to participate in the study. Height was measured at visit 1, and body weight was recorded at both visits with the same digital scale while the subjects had only light clothes on. Informed consent was obtained from all participants. The protocol was approved by the Regional Ethical Committee on Human Research.
After at least 3 months of dietary intervention (ie, at visit 1), the patients were considered for additional drug treatment. Eight of 25 patients continued dietary intervention (n=5) or were given 1200 mg/d gemfibrozil (Lopid, Parke-Davis) (n=3), whereas the rest of the patients were treated with 40 mg/d lovastatin (n=17). Patients considered for lovastatin treatment (11 with type IIa HLP and 6 with type IIb HLP) were included in an open observational study. These patients were asked to maintain their usual lifestyle and dietary restrictions for 8 to 10 more weeks with 40 mg/day lovastatin treatment until a final blood sampling and heparin injection.
Blood Sampling and Heparin Injection
After the subjects fasted overnight, blood was drawn with minimal stasis from an antecubital vein on the right arm by a 19-gauge needle in a Vacutainer system (Becton Dickinson). Immediately afterwards, 1.0 mL unfractionated heparin (5000 IU/mL, Nycomed Pharma AS) was injected into an antecubital vein on the left arm. Exactly 5 minutes after the heparin injection, another blood sample was collected from the contralateral arm. Blood was collected into Vacutainers containing 0.129 mol/L sodium citrate (blood:anticoagulant, 10:1) for plasma preparation and into glass tubes for serum preparation. Plasma was prepared by centrifugation at 2000g for 15 minutes at 22°C and stored at −70°C until further testing. Blood for reference plasma was obtained from 32 healthy blood donors (16 of each sex) 22 to 50 years of age at the Blood Bank of the Tromsø University Hospital. Serum was prepared by clotting of whole blood in a glass tube at room temperature for 1 hour followed by centrifugation at 1200g for 10 minutes.
Total cholesterol and triglycerides in serum were measured on a Boehringer Mannheim/Hitachi 737 analyzer with reagents from the manufacturer. HDL cholesterol was measured in the supernatant after precipitation of VLDL and LDL by use of the manganese-heparin procedure of Burstein et al.19 LDL cholesterol was calculated by the formula of Friedewald et al20 : LDL cholesterol=total cholesterol−HDL cholesterol−(triglyceride/2.2) (all values in mmol/L). Persons involved in analysis were unaware of the subject group from which each sample came.
A prepacked gel filtration column (Superose 6, 10 mm×30 cm) and high-molecular-weight standard proteins were obtained from Pharmacia LKB Biotechnology. Recombinant factor VIIa was a gift from Dr U. Hedner (Novo Nordisk A/S, Biopharmaceutical Division, Research). Bovine factors X and Xa were purchased from Chromogenix AB. TF was a crude extract of human brain prepared by T. Janson (Nycomed), essentially as described by Hjort.21 Freeze-dried vials of TF were diluted in distilled water to the optimal concentration at which the clotting time of normal plasma was 16 seconds. This stock solution of TF was frozen and stored until further dilution for use in the assays described below. The chromogenic substrate S-2222, used to quantitate factor Xa formation, was purchased from Chromogenix AB. Polyclonal rabbit anti-human TFPI IgG was a gift from Dr A.R. Hubbard and Dr T. Barrowcliffe, National Institute for Biological Standards and Control, Hertfordshire, UK.
TFPI Activity Assay
The TFPI assay is based on the ability of a test sample containing TFPI to inhibit TF/factor VIIa catalytic activity in the presence of factor Xa. A two-stage amidolytic assay with minor modifications was used to quantitate TFPI as described by Sandset et al.13 In our assay, we used recombinant factor VIIa (Novo Nordisk A/S) at a final concentration of 15.0 ng/mL instead of purified factor VIIa. Incubation 1 was 20 minutes long and incubation 2 was 30 minutes long, and both were at 37°C. Plasma and gel-filtered samples were prepared by being heated at 56°C for 15 minutes (to remove factor VII activity and fibrinogen) in vials containing 0.25 mL of either type of liquid. The samples were immediately put on ice for 2 minutes and the heat precipitate was removed by centrifugation at 13 500g for 3 minutes. The TFPI activity in the reference plasma was defined as 1 U/mL. The intra-assay coefficient of variation varied from 3% to 5%.
TFPI Antigen Determination
TFPI antigen was determined by use of a solid-phase two-site enzyme immunoassay with reagents kindly donated by P. Østergaard, PhD, Novo Nordisk. Microtiter wells were first coated with a monoclonal anti-TFPI antibody with specificity for the third Kunitz domain (P. Østergaard, PhD, written communication, 1994). The microtiter wells were then incubated with at least two dilutions of each plasma sample to allow free TFPI molecules in plasma to bind to the monoclonal antibody. Bound TFPI was detected by a peroxidase-labeled monospecific polyclonal anti-TFPI, and standard curves were constructed by use of serial dilutions of a recombinant, full-length TFPI preparation.
Heparin concentration in plasma was measured by its anti–factor Xa activity with a commercial chromogenic assay (COATEST Heparin, Chromogenix AB). Reference curves were prepared by addition of dilutions of the heparin used for the human studies to pooled human plasma for final concentrations of 0.0 to 1.0 IU/mL.
TFPI Anticoagulant Assay
The in vitro inhibitory activity of plasma TFPI was tested in a modified diluted thromboplastin time (dTP) assay with diluted TF. Citrated plasma (40 μL) from each patient, collected before and after heparin injection, was incubated with 10 μL Tris-buffered saline (TBS) (0.05 mol/L Tris, 0.15 mol/L NaCl, and pH adjusted to 7.5 in distilled water by 1N HCl) and 50 μL TF diluted in TBS for 3 minutes at 37°C. After the preincubation, 50 μL 35 mmol/L CaCl2 was added and the clotting times were determined automatically (Amelung KC1A, Heinrich Amelung GmbH Medizinische Laboragräte). TF was diluted from the stock solution to yield a clotting time of approximately 50 seconds in the assay (final dilution, 1:300). The coefficient of variation was 1.4% (n=20; mean, 50.2 seconds; range, 49.3 to 51.7) for aliquots from the reference plasma. All clotting times are given as the mean of two measurements. To abolish the effect of heparin-mediated activation of antithrombin, a 40-μL plasma sample was preincubated with 10 μL hexadimethrine bromide (HDB; Sigma Chemical Co) (final concentration, 5 μg/mL) or TBS before the dTP assay. The HDB concentration chosen (5 μg/mL) completely inhibited the heparin-mediated activation of antithrombin in the postheparin plasmas collected from the patients. The relative prolongation of dTP in postheparin plasma treated with HDB compared with preheparin plasma reflected the anticoagulant property of heparin-releasable TFPI. For assessment of the contribution of TFPI to clotting times in preheparin plasma, 40 μL of plasma was preincubated with 10 μL of anti-TFPI (final dilution, 1:30) or TBS for 5 minutes at room temperature before the dTP assay. The relative shortening of dTP by anti-TFPI is assumed to reflect the contribution of TFPI to clotting times in normal plasma. In control experiments, a 1:30 final dilution of anti-TFPI abolished more than 98% of plasma TFPI activity.
Gel filtration was performed with the fast protein liquid chromatography system from Pharmacia LKB Biotechnology. A sample volume of 0.2 mL was loaded onto the Superose 6 column, which was run in a buffer of 0.05 mol/L Tris-HCl, pH 7.40, with 0.3 mol/L NaCl, 0.01 mol/L sodium citrate, and 2.0 μg/mL HDB. The flow rate was 0.4 mL/min. After the first 6.0 mL was discarded, fractions of 0.25 mL were collected. The areas under the curve for TFPI activity during gel filtration were calculated as the deviation from the basal value integrated over the sampling fractions. Area under the curve was estimated separately for peaks I and II (fractions 6 through 32) and peak III (fractions 32 through 42).
The sas statistical software package was used.22 Descriptive statistics on variables from the clinical trial revealed normal distribution of data. Student’s paired t test was used for comparison of treatment effects on variables. Correlations were tested by Pearson’s r. Statistical significance was set at P<.05.
No adverse effects due to heparin administration or drug intervention were recorded. The two groups of patients were comparable with regard to age, sex, and body mass index (Table 1⇓). The patients with type IIa HLP had significantly higher HDL (P<.01) and lower serum triglycerides (P<.001) than the patients with type IIb HLP, whereas no statistical differences were observed for total and LDL cholesterol.
TFPI and Lipoproteins
Patients with type IIa and type IIb HLP had similar total TFPI activity in plasma, but type IIb patients had slightly higher TFPI antigen than type IIa patients (P<.05) (Table 2⇓). When data from all patients with type II HLP were included in the analysis (n=25), total plasma TFPI activity was strongly correlated to total (r=.64, P<.001) and LDL (r=.68, P<.001) cholesterol. However, no significant associations were found with HDL cholesterol (r=.28, P=.18) or serum triglycerides (r=−.00, P=.99). In contrast, TFPI antigen was not found to be associated with total (r=−.11, P=.64) or LDL (r=−.20, P=.38) cholesterol, but it had a positive correlation with serum triglycerides (r=.52, P<.01; n=25). The unexpected but fairly strong correlation between serum triglycerides and TFPI antigen needs confirmation before an interpretation of physiological importance is appropriate.
Effect of Heparin Injection on TFPI Levels
The intravenous injection of 5000 IU unfractionated heparin caused a prominent but variable enhancement in the plasma heparin concentration (which varied from 0.8 to 2.4 IU/mL) (Table 2⇑). The increase in heparin concentration was inversely correlated with the body weight (r=−.58, P=.002; n=25). Heparin induced a prominent and similar rise in total plasma TFPI activity and TFPI antigen in patients with type IIa and type IIb HLP. However, the magnitude of the increase was strikingly different between total plasma TFPI activity (2.2-fold) and TFPI antigen (7.5-fold) when all patients were included in the analysis. The rise in TFPI activity and antigen induced by heparin was not correlated with the heparin concentration in plasma (r=−.03, P=.90 and r=.31, P=.17 for TFPI activity and antigen, respectively).
Heparin and the Anticoagulant Potential of TFPI
Heparinization of plasma in vitro (1.5 IU/mL) prolonged dTP, and this effect was neutralized by HDB (Table 3⇓). Intravenous injection of heparin in vivo causing similar plasma concentrations of heparin prolonged dTP, but in this situation HDB did not completely inhibit the heparin-induced prolongation of dTP. The remaining prolongation of dTP from 18% to 46% above basal levels could be abolished by addition of polyclonal anti-TFPI antibodies (final dilution, 1:30). Thus, the prolongation of dTP after addition of HDB reflected the anticoagulant potential of the heparin-releasable pool of TFPI. Anti-TFPI antibodies added to plasma (final dilution, 1:30) without any heparin present shortened dTP by 4% to 11%.
The heparin injection caused a marked prolongation of dTP that was strongly correlated with the heparin concentration in plasma (r=.84, P<.001). By addition of HDB to plasma collected after heparin injection, dTP was shortened, but it was still 31±8% longer for all patients with type II HLP (n=25) than dTP measured in plasma collected before heparin injection. No statistical differences were observed between patients with type IIa and type IIb HLP (Table 2⇑).
TFPI Activity in Eluates From Human Plasma
Gel filtration of plasma from seven randomly selected patients revealed that 83.4±3.2% of the TFPI activity appeared in peaks I and II, representing LDL-bound TFPI, and 16.6±3.2% of the activity was retained in peak III (mostly carrier-free TFPI) (Fig 1⇓). The heparin injection caused a selective twofold to fourfold increase in peak III. A similar increase in peak III TFPI activity was noted after cholesterol-lowering treatment (data not shown). However, caution should be taken in interpreting the results because more than 95% of TFPI activity was recovered before heparin injection, whereas only 40% to 50% of TFPI activity was recovered in eluates from postheparin plasma.
Cholesterol Lowering and Intravascular Pools of TFPI
Body mass index remained stable but total and LDL cholesterol levels were markedly reduced during the cholesterol-lowering treatment with lovastatin (Table 4⇓). HDL cholesterol was slightly increased and the triglyceride level showed an insignificant decrease. The reduction in total and LDL cholesterol was paralleled by a decrease in total plasma TFPI activity. A modest lowering in TFPI antigen was also observed after lovastatin treatment. Gel filtration of plasma collected from the seven randomly selected subjects before and after lovastatin treatment (Fig 2⇓) showed a reduction in peaks I and II (23±15%, P=.025, estimated from calculations of area under the curve), whereas peak III was unchanged.
Intravenous injection of 5000 IU heparin induced a similar rise in plasma heparin concentration before and after cholesterol lowering (from baseline, 1.76±0.49 and 1.91±0.49 IU/mL, respectively, P=.38). Similar increases in both TFPI antigen (Fig 3A⇓) and total plasma TFPI activity (Fig 3B⇓) resulted from the heparin injection before and after lovastatin treatment. Addition of blocking anti-TFPI antibodies shortened dTP similarly before and after cholesterol lowering (6.7±1.7% and 7.1±1.8%, respectively, P=.45) (Fig 4⇓). Heparin injection caused a marked and similar prolongation of dTP before and after lovastatin treatment. The heparin-releasable pool of TFPI had the same anticoagulant potential, reflected by percent prolongation of dTP after neutralization with HDB, before and after cholesterol lowering (32±8% and 37±11%, respectively, P=.09; n=17) (Fig 4⇓).
In two independent investigations, patients with familial hypercholesterolemia were reported to have higher plasma TFPI activity than age- and sex-matched control subjects.14 15 Therapeutic intervention with hydroxymethylglutaryl–coenzyme A reductase inhibitors decreased the TFPI activity in parallel with the reduction in LDL cholesterol. Gel filtration of plasma from these patients demonstrated a specific drop in LDL-TFPI complexes, whereas carrier-free TFPI was unchanged.15 The present investigation in patients with type IIa or type IIb HLP confirmed that cholesterol-lowering treatment results in a decrease in TFPI activity, which is a drop specific for the LDL-TFPI complexes. Recently, Abumiya et al16 demonstrated that a high-cholesterol diet was associated in monkeys with parallel increases in TFPI and LDL cholesterol. These observations indicate a strong biological association between LDL and TFPI, in which the LDL cholesterol level may predict the total plasma TFPI activity.
The extrinsic coagulation system has been suggested to play a crucial role in the initiation of blood coagulation in atherosclerotic disease.1 23 TF is synthesized in perturbed endothelial cells, which may render them thrombogenic, and is present in the cores of atherosclerotic plaques.1 2 Thus, transient exposure of TF at the surface of atherosclerotic plaque or perturbed endothelial cells may cause low-grade triggering of blood coagulation. In epidemiological studies, factor VII has been found to be associated with cholesterol24 25 and to predict coronary heart disease.26 Given this background, it may be suggested that the elevated TFPI levels seen in hypercholesterolemia may represent a compensatory mechanism to prevent activation of blood coagulation.
The mechanisms underlying the high total plasma TFPI activity in hypercholesterolemia are unknown. LDL may potentiate the synthesis of TFPI in, or release of TFPI from, the endothelium; inhibit clearance; or alter distribution between different intravascular pools. It has recently been suggested that endothelium associated TFPI is the principal source of increased TFPI in hypercholesterolemia because of transfer of TFPI from the endothelium to LDL in concert with raised lipoprotein levels.16 This hypothesis has been supported by animal studies by Novotny et al,12 who demonstrated that exogenously added 125I-labeled heparin-releasable TFPI can bind LDL/VLDL and HDL in vivo. The latter mechanism may imply that high circulating levels of LDL would attract TFPI from the endothelium, thereby making the endothelial surface more susceptible to thrombosis. However, even though diet-induced hypercholesterolemia in monkeys causes increased plasma levels of TFPI-LDL complexes, the magnitude of the heparin-releasable pool from vascular endothelium remains unchanged.16
In our study with human subjects, neither TFPI levels (antigen determination) nor TFPI activity released from the vascular endothelium induced by heparin administration changed during cholesterol lowering. Moreover, the anticoagulant effect of the heparin-releasable pool of TFPI remained unaltered during the lovastatin treatment. Thus, our data demonstrate that neither the size nor the anticoagulant potency of the endothelial pool of TFPI was affected by a decrease in plasma TFPI due to a specific drop in LDL-TFPI complexes. These findings may indicate that the endothelial pool of TFPI and the LDL-TFPI complexes in plasma are regulated independently.
A dTP assay has been used to assess the anticoagulant potency of different forms of TFPI.17 Several years ago it was demonstrated that neutralization of heparin by HDB did not completely abolish the anticoagulant effect in plasma collected from persons receiving heparin.27 This phenomenon was referred to as the postheparin effect, and it is attributed to TFPI, because addition of blocking anti-TFPI antibodies together with heparin neutralization did abolish the anticoagulant effect of heparin administration completely.28 During the progression of disease in patients with cancer, plasma TFPI is gradually increased29 because of a specific rise in carrier-free TFPI that possesses anticoagulant activity similar to that of postheparin TFPI.17 TFPI from postheparin plasma has also been shown to be a more powerful inhibitor of coagulation than recombinant TFPI.30 Addition of blocking anti-TFPI antibodies to normal plasma caused a shortening in dTP,30 which implies that TFPI also contributes to regulation of coagulation under normal conditions. Therefore, the anticoagulant potency of TFPI in postheparin plasma could be assessed after neutralization of heparin by HDB, whereas shortening of dTP by blocking anti-TFPI antibodies in normal plasma would provide information about the inhibitory potential under normal circumstances.
In the present study, TFPI activity increased 2.2-fold, whereas TFPI antigen rose 7.5-fold after heparin injection before drug intervention. TFPI antigen was determined by a sandwich enzyme immunoassay method with monoclonal antibodies raised against the third Kunitz domain in TFPI, which is probably responsible for the association between TFPI and lipoproteins.31 Thus, theoretically, the enzyme immunoassay method used should reflect the level of carrier-free TFPI in plasma. This assumption is supported by the fact that TFPI antigen, in contrast to TFPI activity, was not correlated with total or LDL cholesterol. Although gel filtration of postheparin plasma has a low recovery for TFPI activity, the specific rise in carrier-free TFPI indicates that the main heparin-releasable pool of TFPI actually is carrier-free TFPI. Thus, we claim that the enzyme immunoassay method almost exclusively measures the carrier-free fraction of TFPI, whereas the activity assay reflects the global TFPI in plasma. Gel filtration of plasma before heparin injection revealed that 17% of the TFPI activity was within the carrier-free fraction (peak III). If this assumption holds true, almost 91% of the increase in total TFPI activity could be explained by the 7.5-fold increase in TFPI antigen during heparin injection, based on calculations from the whole group (n=25). These estimations lead us to propose that the increase in plasma TFPI induced by heparin administration is almost entirely caused by enhancement in carrier-free TFPI.
In an umbilical vein model, Almus et al32 showed that reducing plasma TFPI by 50% causes a similar increase of the factor VIIa–TF catalytic activity toward factor IX in a reaction mixture. They suggest that physiological variations in plasma TFPI might influence the regulation of TF-dependent coagulation during hemostasis, an assumption that may hold true for therapeutically induced changes in plasma TFPI. But does a drop in total TFPI—more specifically, a 30% drop in LDL-TFPI complexes—induced by cholesterol-lowering therapy reduce the anticoagulant potential of TFPI in plasma? Lindahl et al18 have proposed that the anticoagulant potency of TFPI is restricted to carrier-free TFPI, which would imply that specific changes in LDL-TFPI complexes did not influence the anticoagulant potency of TFPI. In fact, we have shown for the first time that a specific and prominent drop in LDL-TFPI complexes by lovastatin did not affect anticoagulant potency in human plasma.
In conclusion, the present investigation confirmed that total and LDL cholesterol are correlated with TFPI activity in plasma and that therapeutic lowering of LDL cholesterol in type II hyperlipoproteinemia is paralleled by a similar decrease in TFPI activity. The drop in TFPI activity is mainly due to a specific lowering of circulating LDL-TFPI complexes. Downregulation of TFPI in plasma by lovastatin did not affect the magnitude of the vascular pool of TFPI mobilized into the circulation by heparin administration. Moreover, the drug-induced lowering of plasma LDL-TFPI complexes was not found to influence the anticoagulant potency of TFPI either in plasma or in the vascular endothelial pool.
This study was supported in part by grants from Merck Sharp and Dohme (MSD-Norway) and the Norwegian Council on Cardiovascular Diseases.
Nemerson Y. Tissue factor and haemostasis. Blood. 1988;71:1-8.
Wilcox JN, Smith KM, Schwartz SM, Gordon D. Localization of tissue factor in the normal vessel and in the atherosclerotic plaque. Proc Natl Acad Sci U S A. 1989;86:2839-2843.
Broze GJ, Miletich JP. Characterization of the inhibition of tissue factor in serum. Blood. 1987;69:150-155.
Rao LVM, Rapaport SI. Studies of a mechanism inhibiting the initiation of the extrinsic pathway of coagulation. Blood. 1987;69:645-651.
Lindhout T, Blezer R, Schoen P, Nordfang O, Reutelingsperger C, Hemker HC. Activation of factor X and its regulation by tissue factor pathway inhibitor in small-diameter capillaries lined with human endothelial cells. Blood. 1992;79:2909-2916.
Sandset PM, Warn-Cramer BJ, Rao LVM, Maki SL, Rapaport SI. Depletion of extrinsic pathway inhibitor (EPI) sensitizes rabbits to disseminated intravascular coagulation induced with tissue factor: evidence supporting a physiological role for EPI as a natural anticoagulant. Proc Natl Acad Sci U S A. 1991;88:708-712.
Bajaj MS, Kuppuswamy MN, Saito H, Spitzer SG, Bajaj SP. Cultured normal human hepatocytes do not synthesize lipoprotein-associated coagulation inhibitor: evidence that endothelium is the principal site of its synthesis. Proc Natl Acad Sci U S A. 1990;87:8869-8873.
Østergaard P, Nordfang O, Petersen LC, Valentin S, Kristensen H. Is tissue factor pathway inhibitor involved in the antithrombotic effect of heparins? Haemostasis. 1993;23(suppl 1):107-111.
Novotny WF, Palmier M, Wun TC, Broze GJ, Miletich JP. Purification and properties of heparin releasable lipoprotein-associated inhibitor. Blood. 1991;78:394-400.
Sandset PM, Lund H, Norseth J, Abildgaard U, Ose L. Treatment with hydroxymethylglutaryl-coenzyme A reductase inhibitors in hypercholesterolemia induces changes in the components of the extrinsic coagulation system. Arterioscler Thromb. 1991;11:138-145.
Hansen JB, Huseby NE, Sandset PM, Svensson B, Lyngmo V, Nordøy A. Tissue factor pathway inhibitor and lipoproteins: evidence for association with and regulation by LDL in human plasma. Arterioscler Thromb. 1994;14:223-229.
Abumiya T, Nakamura S, Takenaka A, Takenaka O, Yoshikuni Y, Miyamoto S, Kimura T, Enjyoji KI, Kato H. Response of plasma tissue factor pathway inhibitor to diet-induced hypercholesterolemia in crab-eating monkeys. Arterioscler Thromb. 1994;14:483-488.
Burstein M, Scholnick HR, Mofin R. Rapid method for the isolation of lipoproteins from human serum by precipitation with polyamines. J Lipid Res. 1970;11:583-595.
Friedewald WT, Levy RI, Fredrickson DS. Estimation of the concentration of low-density lipoprotein cholesterol in plasma without use of the preparative ultracentrifuge. Clin Chem. 1972;18:499-502.
Hjort PF. Intermediate reactions in the coagulation of blood with tissue thromboplastin. Scand J Clin Lab Invest. 1957;9(suppl 27):1-173.
SAS Institute Inc. SAS/STAT Guide for Personal Computers, Version 6. Cary, NC: SAS Institute; 1987.
Prydz H. The tissue factor pathway and cardiovascular disease. In: Francis RB Jr, ed. Atherosclerotic Cardiovascular Disease, Hemostasis, and Endothelial Function: Fundamental and Clinical Cardiology, Volume 8. New York, NY: Marcel Dekker, Inc; 1992:35-52.
Lindahl AK, Abildgaard U, Larsen ML, Aamodt LM, Nordfang O, Beck TC. Extrinsic pathway inhibitor (EPI) and the post-heparin anticoagulant effect in tissue thromboplastin induced coagulation. Thromb Res. 1991;62(suppl 14):39-48.
Almus FE, Rao LVM, Rapaport SI. Regulation of factor VIIa/tissue factor functional activity in an umbilical vein model. Arterioscler Thromb. 1993;13:105-111.