TFPIα Has Reduced Ability to Inhibit FV Leiden

Introduction: Tissue factor pathway inhibitor (TFPI)α dampens the initiation phase of coagulation by inhibiting the tissue factor/factor VIIa complex and early thrombin generation. Activated protein C (APC) dampens the propagation phase of coagulation by degrading factor Va (FVa) and factor VIIIa. Kinetic assays using purified TFPIα, protein C, and thrombomodulin have shown that TFPIα and APC act synergistically to inhibit thrombin production. In these assays, thrombin generation is markedly altered by factor V Leiden (FVL), a prothrombotic mutation that reduces the ability of APC to degrade FVa. A 50% reduction in TFPIα (2.5nM to 1.25nM) completely abolishes the anticoagulant effect of APC in reactions performed with FVL, but has no effect in reactions performed with FV. The physiological importance of these in vitro results was subsequently supported by the severe perinatal thrombosis observed in mice with homozygous FVL and heterozygous TFPI deficiency. TFPIα inhibits thrombin generation via a high affinity exosite interaction between its basic C-terminus and an acidic region in early forms of FVa. We hypothesized that the Leiden mutation alters this interaction in a manner that reduces TFPIα inhibitory activity during the initiation of coagulation.

Methods: Plasma and platelet-rich plasma (PRP) were collected from homozygous FVL, heterozygous FVL, and control donors. Thrombin generation assays were performed in the presence or absence of recombinant TFPIα, a peptide mimicking the TFPIα basic C-terminus (LIKTKRKRKK), or anti-TFPI antibodies. Purified prothrombinase assays were performed using FV810, a B-domainless form of FV(a) that binds and is inhibited by TFPIα, and FVL810.

Results: Thrombin generation assays were initiated with limiting FXa to determine an activation threshold. The threshold for thrombin generation was between 10 and 20 pM FXa in all 10 control plasmas, but was reduced in homozygous FVL plasmas, with 4 of 10 responding to only 5 pM FXa. Additional thrombin generation assays were performed using 100 pM FXa to determine the effect of FVL on TFPI activity, as there were no baseline differences between control and FVL plasmas in these conditions. When substances that alter TFPI function were added, the lag time for initiation of thrombin generation was primarily affected. Addition of 1 nM recombinant TFPIα and 1 μM LIKTKRKRKK prolonged the lag time in control plasma by 58% and 28%, respectively, but had progressively less effect in FVL heterozygous (45% and 19%) and homozygous (39% and 9.6%) plasmas (p<0.03 for all comparisons except addition of TFPIα to heterozygous plasma). Conversely, anti-TFPI antibodies directed against the second (anti-K2) and third (anti-K3) Kunitz domains decreased the lag time 32% and 23% in control plasmas, respectively, but only 25% and 17% in FVL homozygous plasma (p=0.03 for both). Thrombin generation assays were repeated using PRP. There were minimal and equivalent effects of recombinant TFPIα or LIKTKRKRKK in control and FVL PRP, possibly due to the presence of additional exogenous TFPIα in platelets. However, the anti-TFPI antibodies decreased the lag time more in control plasma than in FVL homozygous plasma, with anti-K2 decreasing the lag time of control PRP by 28% and FVL homozygous PRP by 20% (p=0.009). In purified prothrombinase assays, TFPIα was a weaker inhibitor of prothrombinase containing FV810 (IC50 = 4.8nM; C.I.= 3.8-6.1nM) compared to FVL810 (IC50 = 7.3nM; C.I. = 6.9-7.9nM). A similar effect of FVL was observed when inhibiting with LIKTKRKRKK.

Conclusion: The biochemical mechanism underlying the procoagulant phenotype of FVL is well-described as a resistance to APC, which dampens the propagation phase of clot formation. The data described here suggest a potential second biochemical mechanism underlying this procoagulant phenotype, a reduced ability of TFPIα to inhibit FVL. This effect is evidenced by alterations in the threshold for initiation of thrombin generation in FVL homozygous plasma and PRP, as well as the reduced IC50 for inhibition of thrombin production by prothrombinase assembled with FVL. This effect is enhanced by reducing TFPIα concentration, offering an explanation for the dramatic procoagulant effect of 50% reduced TFPI in assays using human proteins and mice with FVL. The findings could have clinical relevance in patients with FVL who use oral contraceptives that can reduce the plasma TFPIα concentration.