Plasma Kallikrein Contributes to Ellagic Acid-Induced Coagulation in the Absence of FXI By Activating Factor IX

Background - Factor XI (FXI) plays a critical role in thrombosis, as its inhibition reduces clot formation in animal models, and human congenital FXI deficiency appears to protect against venous thromboembolism and stroke. In contrast, in both settings, the impact on hemostasis is rather low, especially compared to FIX deficiency. Little is known about the cause of this difference. As thrombin generation can be observed in FXI-deficient plasma, we speculated that intrinsic activation of FIX is not solely dependent on FXI(a).

Aim - Evaluate whether plasma kallikrein contributes to FIX activation and subsequent thrombin generation in FXI deficiency.

Methods - Kaolin-initiated coagulation and fibrin formation was assessed in different plasma types in the presence or absence of an active site inhibitor of human plasma kallikrein (PKa) using the activated partial thromboplastin time (aPTT). Thrombin generation was measured in ellagic acid-induced contact activation in human FXI-deficient plasma in the presence or absence of the PKa inhibitor. Levels of FIXa-antithrombin (FIXa-AT) or FXa-AT were determined at defined time points using in-house developed ELISAs. The role of thrombin and its substrate FVIII in ellagic acid-induced contact activation in the absence of FXI was investigated using dabigatran or a human monoclonal FVIII antibody and assessing the FIXa-AT and FXa-AT plasma levels. In in vivo experiments, F11^-/- mice were treated with ellagic acid in the presence or absence of the PKa inhibitor and FXIa-AT levels were assessed 30 min after administration.

Results - Plasma kallikrein inhibition dose-dependently prolonged kaolin-triggered clotting time in human FXI-deficient plasma but not in normal pooled or plasma prekallikrein depleted plasma, indicating a role for PKa in coagulation in the absence of FXI. In FXI-deficient plasma, ellagic acid activated the coagulation system, as could be demonstrated by measurement of thrombin generation, and FIXa-AT and FXa-AT levels, which immediately reached the upper detection limit of 1.6 nM suggesting a FXI bypass route towards FIX activation. Addition of a PKa inhibitor caused a delayed and diminished FXa-AT and FIXa-AT complex formation and led to a reduction in thrombin peak height (EC₅₀: 3 µM). This indicates that PKa might contribute to FXI-independent coagulation by activating FIX. In the presence of dabigatran FIXa-AT complex formation in FXI-deficient plasma was completely inhibited, which could be reversed by addition of purified thrombin. In the presence of purified thrombin, a FVIII antibody inhibited thrombin-dependent FIXa-AT complex formation suggesting that the activation of FVIII by thrombin is essential for FIX activation by PKa or to support the activity of FIXa. In line with this, thrombin generation in FXI-defcient plasma was inhibited by the FVIII antibody. In ellagic acid-treated F11^-/- mice, FIXa-AT complex formation was significantly increased compared to controls (612 pM vs. 121 pM; P < 0.001), and was attenuated upon inhibition of PKa.

Conclusions - We demonstrate that, activation of FXII leads to thrombin generation via FIX activation by plasma kallikrein in the absence of FXI. In addition, the activation of FVIII by thrombin is essential for the FXI bypass route towards FIX activation. These findings support the idea that FXI and FIX have distinctly different roles within the intrinsic coagulation cascade upon FXII activation and may contribute to variations in bleeding in FXI-deficient subjects due to different prekallikrein plasma levels.