Polymorphisms of the Tissue Factor Pathway Inhibitor (TFPI) Gene in Patients With Acute Coronary Syndromes and in Healthy Subjects
Impact of the V264M Substitution on Plasma Levels of TFPI
Abstract—Mutations of the gene encoding tissue factor pathway inhibitor (TFPI), an inhibitor of TF-induced activation of the coagulation cascade, were screened for in 130 patients and 142 healthy controls to determine whether these variants contribute to acute coronary syndromes or modify plasma TFPI levels. The following 3 new polymorphisms were identified: 384T→C in exon IV, which does not change the corresponding amino acid (tyrosine 57); −33C→T in intron 7 (the T/T, C/T, and C/C genotypes were found in ≈50%, 40%, and 10% of subjects in both groups); and 874G→A in exon IX (GTG→ATG), which predicts a valine to methionine change (V264M) in the carboxy-terminus tail of TFPI. The V264M polymorphism was found in 9.2% of the cases and 4.9% of the controls; the associated odds ratio (OR) for acute coronary syndromes was 2.0 (95% confidence interval [CI], 0.7 to 5.1). The OR increased to 3.6 (95% CI, 0.8 to 15.7) and 3.2 (95% CI, 0.9 to 11.8) in nonsmokers and patients without other risk factors, respectively. The possible link between the V264M polymorphism and coronary heart disease was checked in a large case-control study of myocardial infarction (Etude Cas-Témoins de l’Infarctus du Myocarde [the ECTIM Study]). The results showed no link between the V264M polymorphism and coronary syndromes. Interestingly, however, 5 patients heterozygous for the V264M polymorphism had significantly lower plasma TFPI levels than did 13 patients with the most common genotype. Although our present results do not support an association between TFPI polymorphisms and acute coronary syndromes, the possibility that 1 of them, especially the exon IX polymorphism, is associated with subtypes of myocardial infarction or to evolutive particularities that were not assessed in this study, cannot be excluded and is currently being evaluated.
- Received June 8, 1998.
- Accepted August 6, 1998.
Thrombus formation on disrupted atherosclerotic lesions plays a fundamental role in the onset of acute coronary syndromes.1 Tissue factor (TF), an integral membrane glycoprotein, is the main initiator of the coagulation cascade.2 Recently, TF activity and antigen were detected in directional coronary atherectomy specimens.3 Moreover, circulating monocytes express TF in unstable angina.4 Finally, TF pathway inhibition prevents arterial thrombosis in several animal models, underlining the major role of TF in arterial thrombosis.5
Tissue factor pathway inhibitor (TFPI), a circulating, multidomain, Kunitz-type protease inhibitor, is thought to play a major role in the inhibition of TF–factor VIIa proteolytic activity in vivo. The TFPI molecule consists of a negatively charged NH2 terminus, 3 tandem Kunitz-type inhibitor domains, and a positively charged COOH-terminal tail.6 The DNA sequence, including the promoter region, has been elucidated7 8 9 ; the TFPI gene consists of 9 exons separated by 8 introns. Kunitz-type domains 1, 2, and 3 are encoded by exons IV, VI, and VIII, respectively. The NH2 terminus and COOH-terminal tail are encoded by exons III and IX, respectively.
Experimental data strongly suggest that TFPI plays an important role as a natural anticoagulant. Immunodepletion of TFPI lowers the threshold at which TF induces disseminated intravascular coagulation.10 Conversely, infusion of recombinant TFPI protects against thrombosis in numerous experimental models.5 Despite extensive screening of plasma from patients with arterial or venous thrombosis, no quantitative or qualitative hereditary TFPI abnormalities have so far been detected.
The aim of this study was to screen the TFPI gene for point sequence variations in selected subjects with a history of myocardial infarction (MI) or unstable angina in comparison with healthy subjects. Because the Kunitz-type domains and the carboxy-terminal tail of TFPI appear to have major functional importance, the exons coding for these domains were analyzed first. Plasma TFPI was assayed in a small subgroup of patients.
Selection of Cases and Controls
The patients and controls gave their written, informed consent for the study, which was approved by the Pitié Salpétrière ethics committee. One hundred thirty unrelated patients were genotyped; all were under 65 years of age and had been admitted to Bichat Hospital coronary care unit with a diagnosis of MI or unstable angina. Acute MI was diagnosed on the basis of any 2 of the following criteria: typical chest pain lasting ≥30 minutes, unequivocal ECG changes, or an increase in at least 2 of the 3 serum cardiac enzymes to >2× the upper limit of normal, or an increase in creatine kinase and MB isoenzyme. Unstable angina was diagnosed on the basis of typical chest pain lasting >5 minutes (class I to III of Braunwald’s classification11 ) with or without ST-T abnormalities and with no elevation of the cardiac enzymes above the upper limit defined for MI. Among patients with unstable angina, those with angiography-proven ≥50% diameter stenosis were selected. Blood was taken for genotyping either at the time of the acute coronary event (67% of patients) or 3 to 9 months later. Similar genotyping analyses were performed on 142 unrelated control subjects matched with the patients for age and sex and who had no documented personal history of stable or unstable angina, MI, peripheral arterial thrombosis, or stroke. The control subjects had no family history of arterial thrombosis in first-degree relatives before the age of 60 years. The control subjects were recruited almost equally among hospital employees and blood donors. Subjects (cases and controls) and their parents and grandparents were all white; 30 cases (23%) and 13 controls (9%) originated from North Africa (Maghreb).
All subjects completed a questionnaire including age; smoking habits; and a history of hypertension, hypercholesterolemia, or diabetes. Smokers were defined as all subjects who reported current smoking or who had stopped <5 years previously. Subjects were classified as diabetic, hypertensive, or hypercholesterolemic if they reported that they were currently taking prescribed drugs for these conditions and as obese when their body mass index (BMI) was ≥27.3 kg/m2. These latter 4 variables were grouped together as metabolic risk factors. Heredity was considered positive when subjects had a family history of arterial thrombosis.
The polymorphism identified in exon IX was also investigated in a large case-control study of MI, the ECTIM Study (Etude Cas-Témoins de l’Infarctus du Myocarde). The inclusion criteria for this study have been described in detail elsewhere.12 13 Men aged 25 to 64 years were recruited between 1988 and 1991 in regions covered by the WHO MONICA (World Health Organization MONItoring trends and determinants in CArdiovascular disease) registers in Northern Ireland and France.14 Cases were recruited to the study 3 to 9 months after the event, and they had to satisfy WHO criteria for definite acute MI (category I). Controls were randomly recruited from the same geographic areas, and age stratification was used to obtain an approximate match between the age distribution of the controls and cases. DNA extracted from 509 patients and 562 controls was screened for the polymorphism in exon IX. Coronary angiograms were available for 314 cases. The number of arteries with >50% stenosis was used to assess the degree of stenosis.
The screening strategy used a method comprising polymerase chain reaction (PCR) amplification of exons IV, VI, VIII, and IX followed by denaturing gradient gel electrophoresis (DGGE) of the amplified fragments. Mutations were identified by direct sequencing of fragments with abnormal electrophoretic behavior. The first 6 patients in whom we detected a sequence variation in exon IX were screened for other mutations in the remaining DNA coding sequences (exons III, V, and VII) by means of direct sequencing; none were found. The cDNA sequence reported in this article has been published by Wun et al.15
PCR and DGGE
DNA was extracted from blood leukocytes by the method of Bell et al.16 Enzymatic amplification of exons IV, VI, VIII, and IX was performed with a set of oligonucleotide primers from GIBCO Life Technologies designed for DGGE analysis (Table 1⇓). Because it is composed of 2 domains with different melting temperatures, exon VIII was studied with 2 sets of primers. A standard PCR was performed in a 50-μL reaction mixture containing 50 ng of genomic DNA, 0.4 pmol/μL of each primer, 200 μmol/L of each deoxynucleotide 5′-triphosphate (dNTPs from GIBCO), and 0.25 U of Taq polymerase (Super Taq from ATGC) in buffer containing 10 mmol/L Tris-HCl, 5 mmol/L KCl, 1.5 mmol/L MgCl2, and 0.01% gelatin, pH 9. The reaction mixture was subjected to 35 cycles of PCR with denaturation at 94°C for 20 seconds, annealing for 30 seconds at temperatures depending on the primer set, and extension at 70°C for 45 seconds in a Trio-Thermoblock thermal cycler (Biometra). The reaction mixture was then held at 70°C for 7 minutes and 94°C for 5 minutes. The annealing temperatures and the size of amplification products are shown in Table 1⇓.
Computer analysis was performed with the melt 87 (Bio-Rad) program.17 Because the sequence of interest must be located within the first melting domain of the fragment, a psoralen derivative (ChemiClamp) was attached to the 5′ extremity of 1 of the amplification primers18 (Table 1⇑). Psoralen-modified oligonucleotides were purchased from Eurogentec. Before DGGE, the PCR products were exposed to UV radiation (365 nm) for 15 minutes. Amplification products were subjected to electrophoresis at 160 V in 6.5% polyacrylamide gel containing a gradient of denaturing agents. The 90% solution of denaturing agent contained 6.3 mol/L urea and 36% formamide in 1× TAE buffer (40 mmol/L Tris, 20 mmol/L acetic acid, and 1 mmol/L EDTA, pH 8.3). The gradients used for each DGGE run are shown in Table 1⇑. The bands corresponding to the normal and mutant alleles were excised from the gel, reamplified, purified, and automatically sequenced (Genome Express) by cycle fluorescent sequencing.
Restriction Analysis of PCR-Amplified Products
Exon IX Polymorphism
The restriction sites of the sequence were not altered by the G-to-A substitution. To confirm the exon IX variation, a modified 30-mer oligonucleotide with an I-to-G substitution was synthesized (5′-TAA CAA AAA TTT CTT CAT ATG CTA TTG TCA-3′; the substituted base is underlined). This creates a restriction site for the enzyme MaeIII (Boehringer Mannheim; ↓GTNAC/CANTG↑). After amplification, the PCR products were digested by the enzyme overnight at 55°C and checked on a 2.5% agarose gel.
Intron 7 Polymorphism
Because the intron 7 polymorphism creates a restriction site for the enzyme NdeI (CA↓TATG/GTAT↑AC), this enzyme was used to confirm the intron 7 variation. The PCR products were digested overnight at 37°C by NdeI (BioLabsinc) and checked on a 2.5% agarose gel.
Allele-Specific Oligonucleotide Hybridization
ECTIM DNAs were genotyped by using allele-specific oligonucleotides (ASOs), as described.19 The following ASOs were synthesized by Eurogentec: 5′ TAT TTT CAC TCT CTG CT 3′ and 5′ AGC AGA GAA TGA AAA TA 3′, and were used to detect the CTC (valine) and GAA (methionine) codons, respectively.
Blood was collected from 12-hour–fasted patients 3 to 9 months after the acute coronary event. Blood was taken before and 90 minutes after a subcutaneous injection of 0.66 mg/kg body weight enoxaparin.
TFPI Antigen Assay
Total and free TFPI contents were measured by using the Asserachrom total and Asserachrom free TFPI ELISA kits kindly donated by Diagnostica Stago. In both assays the capture antibody is a monoclonal F(ab′)2 fragment against the second Kunitz domain of TFPI. The detecting monoclonal antibodies are conjugated to peroxidase. The tag monoclonal antibody used for the free TFPI assay is specific for the Kunitz 3 domain; the tag antibody used for total TFPI assay is specific for total TFPI, as shown using recombinant full-length and C-terminal–truncated forms. The Asserachrom total TFPI assay measures free and bound, as well as intact and truncated, forms of TFPI, whereas the Asserachrom free TFPI assay measures only TFPI free of lipoprotein (personal communication from J. Amiral, Serbio, Gennevilliers, France.) The within-run coefficients of variation were 6.0% and 6.8% with the total ELISA and free ELISA, respectively (n=10).
Chromogenic Assay of TFPI Activity
TFPI activity was measured in microplates with a 2-stage amidolytic assay according to Sandset et al,20 with minor modifications. TF (Neoplastine), bovine factor Xa, human factor X, and the chromogenic substrate CBS 5244 were from Diagnostica Stago; factor VIIa was ACSET kindly provided by LFB Laboratories. Absorbance was read at 405 nm in a microplate reader (Molecular Devices).
Total cholesterol and triglyceride levels in plasma were measured by using commercial enzymatic kits (Kone Instruments SA). Plasma anti–factor Xa activity was measured using the Rotachrom heparin kit from Diagnostica Stago.
Links between each polymorphism and MI or unstable angina were examined by simple cross-tabulation and by calculating the odds ratio (OR) as an approximation of the relative risk. The extent to which the link between the TFPI V264M variation and disease onset was modified by other characteristics was assessed by means of stratified analyses. Confidence intervals (95% CIs) were calculated using Woolf’s method. An OR was considered statistically significant when the lower limit of the 95% CI was >1.0. The χ2 test (or Fisher’s exact test, when sample numbers were too small) was used to test for deviation of the genotype distribution from Hardy-Weinberg equilibrium and to identify significant differences in allele or genotype frequencies between the cases and controls. Two-tailed probability values of 0.05 or less were considered statistically significant. TFPI levels in patients and controls were compared using the Mann-Whitney test for unpaired data.
Characteristics of the Study Population
One hundred forty-two controls and 130 cases (15 with unstable angina and 115 with MI) were studied. Table 2⇓ shows the characteristics of the subjects and the prevalence of selected risk factors for coronary heart disease among them. The 2 groups were matched for age and sex, and there were, therefore, no significant differences in these variables. All of the subjects were white, but 23% of the cases were from North Africa, compared with only 9% of the controls. As expected, risk factors for coronary heart disease (including current smoking, hypercholesterolemia, diabetes mellitus, hypertension, and obesity) were more frequent in the cases.
DGGE Migration and Restriction Patterns
No abnormal DGGE patterns were detected in exon VI, which encodes the second Kunitz domain (data not shown). We found an abnormal migration pattern for exon IV (Figure 1A⇓) in 4 patients and 2 controls. The 6 subjects were heterozygous for a thermostabilizing variant, which was further characterized by sequencing. The variation was located at position 384 (TAT→TAC) in exon IV (384T→C); it did not change the corresponding amino acid (tyrosine 57) in the Kunitz 1 domain of TFPI.
Exon VIII (coding for the third Kunitz domain) was studied by using 2 different DGGEs corresponding to the beginning and end of the exon. No abnormality was detected in the end of exon VIII, but an abnormal migration pattern was detected at a high frequency in amplified fragments including the end of intron 7 and the beginning of exon VIII (Figure 1B⇑). Sequence analysis showed that the variation was a C-to-T substitution at position −33 in intron 7 (intron 7 −33C→T).7 This was confirmed by restriction site analysis with NdeI (Figure 1D⇑). The restriction site for this enzyme was absent from the normal strand, which remained undigested (198 bp); the homozygous mutated strand, in contrast, was completely digested into 2 fragments of 167 and 31 bp; in heterozygous subjects, both patterns, corresponding to the undigested (198 bp) and digested (167 bp) fragments, were present.
The DGGE profiles obtained for exon IX are shown in Figure 1C⇑. A thermodestabilizing heterozygous variation was found in some patients and controls. These subjects were shown to be heterozygous for a missense mutation in nucleotide 1006 (GTG→ATG) in exon IX (nucleotides numbered according to Wun et al).15 This variation predicts a valine-to-methionine change in the sequence (V264M), which is located in the carboxy-terminus tail of the molecule. Restriction analysis with MaeIII was used to confirm the sequence variation (Figure 1E⇑). When the normal strand was amplified and digested, the fragment was split into 2 fragments of 100 and 31 bp (the latter was not visible on the gel). The restriction site was absent from the mutant fragment, which remained undigested (131 bp). Both patterns were observed in heterozygotes. As shown in Figure 2⇓, the predicted valine residue in position 264 of the carboxy-terminal tail of human TFPI is highly conserved relative to 3 other species.21 22 23
Distribution of TFPI Genotypes in Patients With MI or Unstable Angina and in Control Subjects
The distribution of TFPI genotypes for exon IV and intron 7 among patients and controls is shown in Table 3⇓. The neutral 384T→C variant in exon IV was found in 4 cases (3.1%) and 2 controls (1.4%). Five of these subjects originated from North Africa and the other was French. No statistical case-control difference in its frequency was observed. The variant of intron 7 was frequent in both groups. The OR associated with the T/T genotype was 1.2 (95% CI, 0.7 to 1.9) relative to the C/T+C/C genotypes, a difference that was not statistically significant. C/C homozygotes represented ≈10% of the subjects in each group. The genotype distributions in cases and controls conformed to Hardy-Weinberg equilibrium.
In all of the subjects the TFPI V264M variation in exon IX was identified by DGGE screening and confirmed by restriction analysis. As shown in Table 4⇓⇓, the frequency of the TFPI V264M variation was 9.2% (12/130) in the cases (of whom 11 had MI and 1 had unstable angina) and 4.9% (7/142) of the controls. No homozygotes were found. The OR for MI or unstable angina associated with the TFPI V264M variation was 2.0 (95% CI, 0.7 to 5.1). The link was unaffected when cases and controls originating from Maghreb were excluded from the analysis (OR=2; 95% CI, 0.7 to 5.9; P=0.18). In contrast, the risk increased to 2.6 (95% CI, 0.9 to 7.3, P=0.06) in patients over 45 years old. We then explored whether the risk associated with the V264M variation was different in subjects with other major risk factors or specific combinations of risk factors. ORs were calculated for smokers and for subjects with 1 or more metabolic risk factor. The proportion of smokers was 79% among the patients and 38% among the controls (Table 2⇑). As shown in Table 5⇓, current or recent smoking was associated with an OR for acute coronary events of 6.8 in comparison with nonsmokers with the V/V genotype (OR, 6.8; 95% CI, 3.8 to 12.1; P<0.0001). One or more metabolic risk factors (hypertension, diabetes mellitus, hypercholesterolemia, or obesity) was present in 59% of the patients and 30% of the controls, yielding a 3.5-fold increase in the OR (OR, 3.5; 95% CI, 2.1 to 6.0; P<0.0001) among patients with 1 or more of these risk factors compared with patients with none. The V264M variation was associated with an increase of 3.6 and 3.2 times the relative risk among nonsmokers (95% CI, 0.8 to 15.7; P=0.09) and patients without metabolic risk factors (95% CI, 0.9 to 11.8; P=0.09), respectively. In smokers and subjects with risk factors, the polymorphism increased the relative risk from 6.8 to 9.7 and from 3.5 to 4.2, respectively.
Frequency of the V264M Genotype in the ECTIM Study Population
Because our patient and control groups were relatively small, the V264M polymorphism was screened for in 509 cases and 562 controls from the ECTIM Study. This was done by using ASO hybridization. To simplify the data presentation, the results from the 3 French centers were pooled after checking that the results were not statistically different across the centers. Mean age was similar in the cases and controls (54±8.2 and 53±8.5 years, respectively). As shown in Table 6⇓, the overall V264M genotype and allele frequencies did not differ significantly between the cases and controls. Two homozygous subjects were identified in the control group and 1 in the patient group. Genotype frequencies conformed to Hardy-Weinberg equilibrium in all subgroups. Similar results were obtained in subjects under 45 years of age. The genotype and allele frequencies did not differ significantly between the Northern Irish and French subjects. In the 314 French cases for whom coronary angiographic data were available, 10 were heterozygous and 1 was homozygous. No link was found with the degree of stenosis (data not shown).
TFPI Levels According to Genotype
To assess the effect of the V264M variation on plasma TFPI levels, the latter were measured in 5 V/M and 13 V/V patients (Table 7⇓). The 5 patients with the V264M mutation were being treated with statins. Because the cholesterol level is known to affect TFPI levels,24 the 13 patients with the wild-type variant were chosen for also being on statins. As shown in Table 7⇓, the 2 groups of patients had similar ages and total cholesterol, LDL cholesterol, triglyceride, and glycemia levels. Total and free TFPI antigen levels were lower in the patients heterozygous for the V264M mutation than in patients with the wild-type variant (P<0.05). TFPI activity levels measured by a chromogenic assay were also lower, but the difference did not reach significance.
To evaluate the impact of the V264M variation on endothelial cell–bound TFPI, blood was drawn 90 minutes after a subcutaneous enoxaparin injection, because Bara et al25 have shown that TFPI levels peak at this time. In the 2 groups of subjects, the mean (±SD) anti–factor Xa activity level at 90 minutes was 0.4 IU/mL. In both groups, total TFPI levels increased 1.5-fold while free TFPI increased ≈3-fold. The mean TFPI level in the patients with the genetic variation remained 15% lower than in the other patients, although the difference did not reach significance.
In patients with MI or unstable angina, TF appears to play a determining role in triggering thrombosis in contact with the atherosclerotic plaque.1 The search for genetic variations in TFPI thus appeared important. In this study, we identified 3 new polymorphisms in the TFPI gene. The polymorphism of exon IX, predicting a methionine instead of a valine in position 264 of the protein, emerged as the most interesting change for several reasons: (1) in the first group of subjects studied, its frequency was 4.9% in the controls and 9.2% in the cases; (2) the valine residue in position 264 of TFPI is highly conserved among species and is located within the heparin-binding site; and (3) it is this polymorphism that was associated with low plasma levels of TFPI.
The exon IV variation (384T→C) identified herein did not modify the protein sequence, because TAT and TAC both code for tyrosine; 5 of the 6 subjects carrying this polymorphism were from North Africa, suggesting that the C allele frequency is higher in Maghreb (12%) than in Europe (0.4%; P<0.0001). A common C/T polymorphism in intron 7 was located 33 bp upstream from the beginning of exon VIII. In the sequence described by van der Logt et al,7 the base identified at this site is a C. Our results clearly show, however, that the most frequent allele, encountered in ≈70% of the patients and controls, is a T. No significant difference in the frequencies of the 3 genotypes was observed between the cases and controls.
In the first group of subjects studied, the frequency of the polymorphism in exon IX (V264M) was higher in the cases (9.2%) than in the controls (4.9%) (OR, 2; 95% CI, 0.7 to 5.1; P=0.16). This trend toward an increased relative risk associated with the mutation was stronger in patients who did not smoke (OR, 3.6; 95% CI, 0.8 to 15.7; P=0.09) and in those without other risk factors (OR, 3.2; 95% CI, 0.9 to 11.8; P=0.09). Since these results showed a trend toward an increased OR and given the relatively small size of the study group, we investigated the V264M polymorphism in ECTIM, a large, multicenter, case-control study. The frequency of heterozygotes (2.9% to 5.3%, according to the geographic area) and the M264 allele frequency (0.017 to 0.028) were similar in the cases and controls (Table 6⇑); moreover, the values were very similar to those found in the controls of our initial study (V/M genotype frequency, 4.9%; M264 allele frequency, 0.025). There was no difference in the distribution of either genotype or allele frequencies in the high-risk populations of Northern Ireland or France. Among the possible reasons why the initial trend toward a link between the TFPI V264M polymorphism and the occurrence of acute coronary events was not confirmed in the ECTIM Study is the geographic origin of the subjects. In ECTIM, it was limited to 3 French regions and North Ireland, whereas our initial study included subjects originating from all over Europe and North Africa. In addition, the initial study included subjects of both sexes, whereas only men were included in ECTIM. However, restricting the analysis to the male population or to subjects originating from Europe (ie, excluding those from Maghreb) and to the patients with MI (ie, excluding those with unstable angina) did not change the results. Although our results for the ECTIM Study population do not indicate a link between MI and the V264M polymorphism, certain baseline characteristics (including the severity of coronary disease) of our initial group of patients may have differed from those of the ECTIM population. Our results for the former subjects suggest that this polymorphism might increase the severity of coronary thrombosis. In the small subgroup of patients with unstable angina (n=15), only 1 patient was heterozygous for the V264M polymorphism. However, by study design, patients with unstable angina had a ≥50% stenosis. Therefore, our results do not rule out the role of the polymorphism in patients with less severe angiographic lesions who may be prone to plaque rupture as well.26
The influence of the V264M genotype on TFPI levels was assessed in a small subgroup of patients, all of whom were on statin therapy, before and after an injection of low-molecular-weight heparin. Interestingly, total and free TFPI antigen levels were significantly lower before heparin in the heterozygous patients (Table 7⇑). TFPI activity and TFPI levels measured after heparin administration were still lower in these patients, but the difference did not reach significance. Val264 is located in the carboxy-terminus tail of TFPI, within the positively charged basic sequence (KTKRKRKKQRVK) involving residues 254 to 265. This sequence has been shown to mediate TFPI binding to heparin.27 Because infusion of heparin increases levels of circulating TFPI,28 it has been postulated that TFPI, via its basic region, binds to glycosaminoglycans on the endothelial surface and that heparin, by competing for this basic region, releases bound TFPI.28 Many other functions of TFPI have been reported to be dependent on the presence of the carboxy-terminus tail, including optimal inhibition of activated factor X,27 binding to fibrin and subsequent degradation by thrombin,29 and interaction with hepatoma cells, which is followed by LDL-related protein–mediated internalization and degradation of TFPI.30 A number of these functions, including those involving TFPI clearance and degradation, appear to depend on the interaction of the positively charged residues of the carboxy terminus of TFPI with the negative charges carried by glycosaminoglycans of the vascular wall or by fibrin. Although the replacement of a valine by a methionine is not expected to strongly modify the charge of the basic sequence, it might increase the TFPI interaction with glycosaminoglycans, thereby reducing circulating free TFPI levels. The measurement of free TFPI antigen after heparin injection is of particular interest in estimating TFPI bound to the vessel wall, which is released by heparin.31 In the patients carrying the V264M polymorphism, the decrease in total and free TFPI levels observed before heparin injection was attenuated after heparin, suggesting that the release process and the amount of endothelial cell–bound TFPI were not strongly affected. Free TFPI is thought to be the active part of TFPI in inhibiting coagulation, and a fall in its concentration might thus affect thrombotic potential. Moreover, evidence that the valine in position 264 varies little across species (Figure 2⇑) supports a potential role in TFPI function.
In conclusion, we identified 3 new polymorphisms in the TFPI gene. We found no evidence supporting a link between TFPI polymorphisms and the risk of coronary heart disease, but further studies are underway to determine whether the exon IX polymorphism is associated with particular subtypes of MI or certain outcomes.
This study was supported in part by Rhône-Poulenc Rorer Bellon Laboratories (to D.d.P.). and by a grant from Assistance Publique-Hôpitaux de Paris (CRC 95135) (to D.d.P., M.C.A., L.S., M.G., D.D.). We thank C. Lacombe and L. Sustendal for their excellent technical assistance; J. Amiral and M. Grosley (Serbio Laboratory, Gennevilliers, France) for providing the TFPI ELISA kits; and Drs A.M. Becker, S. Menasie, and R. Babou (E.T.S. de l’AP-HP, Site Transfusionnel, Hôpital Bichat-Claude Bernard) for collecting blood samples from the control subjects. We also thank Professor F. Cambien (INSERM SC7) for stimulating discussions and critical reading of the manuscript.
Fuster V. Mechanisms leading to myocardial infarction: insights from studies of vascular biology. Circulation. 1994;90:2126–2146.
Nemerson Y. Tissue factor and hemostasis. Blood. 1988;71:1–8.
Marmur JD, Thiruvikraman SV, Fyfe BS, Guhu A, Sharma SK, Ambrose JA, Fallon JT, Nemerson Y, Taubman MB. Identification of active tissue factor in human coronary atheroma. Circulation. 1996;94:1226–1232.
Jude B, Agraou B, Mcfadden EP, Susen S, Bauters C, Lepelley P, Vanhaesbroucke C, Devos P, Cosson A, Asseman P. Evidence for time-dependent activation of monocytes in the systemic circulation in unstable angina but not in acute myocardial infarction or in stable angina. Circulation. 1994;90:1662–1668.
Girard TJ, Eddy R, Wesselschmidt RL, Macphail LA, Likert KM, Byers MG, Shows TB, Broze GJ Jr. Structure of the human lipoprotein-associated coagulation inhibitor gene: intron/exon gene organization and localization of the gene to chromosome 2. J Biol Chem. 1991;266:5036–5041.
Sandset PM, Warn-Cramer BJ, Maki SL, Rapaport SI. Immunodepletion of extrinsic pathway inhibitor sensitizes rabbits to endotoxin-induced intravascular coagulation and the generalized Schwartzman reaction. Blood. 1991;78:1496–1502.
Braunwald E. Unstable angina: a classification. Circulation. 1989;80:410–414.
Cambien F, Poirier O, Lecerf L, Evans A, Cambou J, Arveiler D, Luc G, Bard J, Bara L, Ricard S, Tiret L, Amouyel P, Alhenc-Gelas F, Soubrier F. Deletion polymorphism in the gene for angiotensin-converting enzyme is a potent risk factor for myocardial infarction. Nature. 1992;359:641–644.
Parra HJ, Arveiler D, Evans AE, Cambou JP, Amouyel P, McMaster D, Schaffer P, Douste-Blazy P, Luc G, Richard JL, Ducimetière P, Fruchart JC, Cambien F. A case-control study of lipoprotein particles in two populations at contrasting risk of coronary heart disease: the ECTIM study. Arterioscler Thromb. 1992;12:701–707.
Wun TC, Kretzmer KK, Girard TJ, Miletich JP, Broze GJ. Cloning and characterization of a cDNA coding for the lipoprotein-associated coagulation inhibitor shows that it consists of three tandem Kunitz-type inhibitory domains. J Biol Chem. 1988;263:6001–6004.
Bell G, Karam J, Rutter W. Polymorphic DNA region adjacent to the 5′ end of the human insulin gene. Proc Natl Acad Sci U S A. 1981;78:5759–5763.
Costes B, Girodon E, Ghanem N, Chassignol M, Thuong NT, Dupret D, Goossens M. Psoralen-modified oligonucleotide primers improve detection of mutations by denaturing gradient gel electrophoresis and provide an alternative to GC-clamping. Hum Mol Genet. 1993;2:393–397.
Belaaouaj A, Kuppuswamy MN, Birktoft JJ, Bajaj SP. Revised cDNA sequence of rabbit tissue factor pathway inhibitor. Thromb Res. 1998;69:547–553.
Girard TJ, Gailani D, Broze GJ. Complementary DNA sequencing of canine tissue factor pathway inhibitor reveals a unique nanomeric repetitive sequence between the second and third Kunitz domains. Biochem J. 1994;303:923–928.
Kamei S, Kamikubo Y, Fujimoto H, Ishihara M, Yonemura H, Miyamoto S, Funatsu A, Enjyoji K, Amubiya T, et al. Amino acid sequence and inhibitor activity of rhesus monkey tissue factor pathway inhibitor (TFPI): comparison with human TFPI. J Biochem. 1994;115:708–714.
Kokawa T, Enjyoji K, Kumeda K, Kamikubo Y, Haradashiba M, Koh H, Tsushima M, Yamamoto A, Kato H. Measurement of the free form of TFPI antigen in hyperlipidemia: relationship between free and endothelial cell–associated forms of TFPI. Arterioscler Thromb Vasc Biol. 1996;16:802–808.
Bara L, Bloch ML, Zitoun D, Samama M, Collignon F, Frydman A, Uzan A, Bouthier J. Comparative effects of enoxaparin and unfractionated heparin in healthy volunteers on prothrombin consumption in whole blood during coagulation, and release of tissue factor pathway inhibitor. Thromb Res. 1993;69:443–452.
Wesselschmidt R, Likert K, Girard T, Wun TC, Broze GJ. Tissue factor pathway inhibitor: the carboxy-terminus is required for optimal inhibition of factor-Xa. Blood. 1992;79:2004–2010.
Novotny WF, Palmier M, Wun TC, Broze GJ, Miletich JP. Purification and properties of heparin-releasable lipoprotein-associated coagulation inhibitor. Blood. . 1991;78:394–400.
Ohkura N, Enjyoji K, Kamikubo Y, Kato H. A novel degradation pathway of tissue factor pathway inhibitor: incorporation into fibrin clot and degradation by thrombin. Blood. 1997;90:1883–1892.
Warshawsky I, Bu GJ, Mast A, Saffitz JE, Broze GJ, Schwartz AL. The carboxy terminus of tissue factor pathway inhibitor is required for interacting with hepatoma cells in vitro and in vivo. J Clin Invest. 1995;95:1773–1781.