Factor XI (fXI) is the zymogen of a plasma protease (fXIa) that contributes to coagulation by activating factor IX. The mechanism by which fXI is converted to fXIa in plasma has been a topic of recent debate. When plasma is exposed to a charged surface, factor XII (fXII) is converted to fXIIa, which then activates fXI. The importance of this reaction to hemostasis in vivo is questionable, as fXII deficiency does not cause abnormal bleeding. This suggests that fXI can be activated by other proteases, with α-thrombin receiving considerable attention in this regard. Results from several laboratories support a model in which α-thrombin activates fXI to propagate coagulation. This notion has been challenged by a recent study that found

  1. no direct evidence of fXI activation by α-thrombin in plasma and

  2. that fXI activation during plasma preparation can give the false impression that fXI is activated independent of fXIIa.

We developed two plasma systems to examine thrombin generation (measured by calibrated automated thrombography) in the absence of fXII, with due consideration to the possibility that traces of fXIa can affect results. In the first system, fXI deficient plasma is initially treated with corn trypsin inhibitor to neutralize fXIIa, and then supplemented with fXI treated previously with DFP to neutralize contaminating fXIa. The second system uses fXII deficient plasma, and endogenous fXI is neutralized with an antibody if a fXI deficient state is required. Coagulation is initiated in both systems by addition of Ca2+ with or without tissue factor (TF - <10 pM), α-thrombin (5 nM), or factor Xa (6 pM). In both systems, significant thrombin generation was detected only in the presence of fXI, and required TF, α-thrombin, or factor Xa. Ca2+ alone did not stimulate thrombin generation. Thrombin generation was detected in fXI deficient plasma stimulated with as little as 3.0 pM fXIa. However, only 0.3 pM fXIa was required to induce thrombin generation if fXI was present, indicating additional fXIa is generated after addition of the fXIa trigger. The fXI deficient plasma system was not reconstituted by fXI variants defective in factor IX activation, nor by a fXI variant that is activated poorly by α-thrombin but normally by fXIIa. The results support a model in which fXI is activated in plasma by thrombin, with fXIa subsequently contributing to additional thrombin generation through factor IX activation. α-thrombin generated early in these reactions could promote subsequent thrombin generation through activation of factors V and VIII, as well as conversion of fibrinogen to fibrin. These reactions involve interactions with anion binding exosite I (ABE-I) on α-thrombin. When thrombin with a dysfunctional ABE-I (β-thrombin or α-thrombin with ABE-I mutations) were tested in the plasma systems, fXI-dependent thrombin generation was actually greater, and occurred earlier, than in the same system stimulated with α-thrombin. Studies with purified proteins and SDS-PAGE showed that β-thrombin and the ABE-I mutants convert fXI to fXIa similarly to α-thrombin. α-thrombin was also able to activate fXI in the presence of the ABE-I blocking peptide hirugen. β-thrombin and the exosite I mutants may promote fXI-dependent thrombin generation in plasma better than α-thrombin because there is no competition from fibrinogen. The different behavior of α-thrombin compared to β-thrombin and the ABE-I mutants supports the broader concept that thrombin activates fXI in plasma, and indicates that fXI activation by thrombin does not require ABE-I. Natural products of prothrombin activation lacking ABE-I, such as β-thrombin, therefore, may contribute to factor XI activation in plasma.

Disclosures: No relevant conflicts of interest to declare.

Author notes

Corresponding author

Sign in via your Institution