SEARCH

SEARCH BY CITATION

Keywords:

  • anticoagulation;
  • coagulation;
  • contact activation system;
  • factor XI;
  • factor XII;
  • genetically altered mice;
  • intrinsic pathway

Abstract

  1. Top of page
  2. Abstract
  3. Factor XI and the intrinsic pathway of coagulation
  4. Factor XI deficient mice
  5. Thrombosis models in FXI deficient mice
  6. Activation of FXI in arterial thrombosis
  7. Deficiency in FXI in combinations with other clotting proteins
  8. Evolutionary aspects of FXI deficiency
  9. Acknowledgments
  10. Disclosure of Conflict of Interests
  11. References

Summary.  The blood coagulation system forms fibrin to limit blood loss from sites of injury, but also contributes to occlusive diseases such as deep vein thrombosis, myocardial infarction, and stroke. In the current model of a coagulation balance, normal hemostasis and thrombosis represent two sides of the same coin; however, data from coagulation factor XI-deficient animal models have challenged this dogma. Gene targeting of factor XI, a serine protease of the intrinsic pathway of coagulation, severely impairs arterial thrombus formation but is not associated with excessive bleeding. Mechanistically, factor XI may be activated by factor XII following contact activation or by thrombin in a feedback activation loop. This review focuses on the role of factor XI, and its deficiency states as novel target for prevention of thrombosis with low bleeding risk in animal models.


Factor XI and the intrinsic pathway of coagulation

  1. Top of page
  2. Abstract
  3. Factor XI and the intrinsic pathway of coagulation
  4. Factor XI deficient mice
  5. Thrombosis models in FXI deficient mice
  6. Activation of FXI in arterial thrombosis
  7. Deficiency in FXI in combinations with other clotting proteins
  8. Evolutionary aspects of FXI deficiency
  9. Acknowledgments
  10. Disclosure of Conflict of Interests
  11. References

In the classic waterfall scheme of the coagulation cascade, thrombin and fibrin formation proceed through a series of sequential reactions involving activation of plasma proteases by limited proteolysis [1–3]. The cascade/waterfall model describes two converging pathways that are initiated either by exposure of blood to a damaged vessel wall (the extrinsic pathway) or by blood-borne components of the vascular system (the intrinsic pathway). The extrinsic pathway, which is essential for fibrin formation at a site of injury following vascular damage, is initiated when the plasma protease factor VIIa forms a complex with tissue factor (TF), which is ubiquitously expressed on subendothelial vascular cells. As deficiency of factor VII in humans and mice, or low TF levels in mice is associated with severe hemorrhage, the extrinsic pathway of coagulation is essential for coagulation activity at a wound site [4,5]. Activity of factor VIIa/TF complexes is tightly regulated by the TF pathway inhibitor (TFPI), which rapidly inactivates factor VIIa/TF enzyme complexes on membranes [6,7]. Therefore, additional stimuli for propagating fibrin formation to support the formation of a three-dimensional thrombus are required [8,9].

The intrinsic or ‘contact’ pathway is initiated when factor XII comes into contact with negatively charged surfaces, in a reaction involving high molecular weight kininogen and plasma kallikrein. Factor XII activation is triggered by negatively charged polyanions, and one of the most potent activators, kaolin, is used in the clinical activated partial thromboplastin time (aPTT) assay to assess the integrity of the intrinsic pathway [10,11]. In reactions that are propagated by platelets, activated factor XII activates its substrate factor XI (FXI), which in turn activates factor IX [12]. FXI deficient humans suffer from mild hemorrhage (hemophilia C), which is characterized by trauma or soft tissue-related hemorrhage, primarily involving tissues with high fibrinolytic activity [13,14]. In contrast, factor XII deficiency is not associated with increased bleeding [15,16], demonstrating the existence of factor XII-independent mechanisms for FXI activation. Indeed, in the revised model of coagulation, thrombin was demonstrated to convert FXI to the active protease FXIa in the presence of dextran sulfate [17]. It was postulated that thrombin generated early in coagulation by the tissue factor pathway activates FXI creating a positive feedback loop that sustains coagulation following inactivation of the factor VIIa/TF complex by TFPI [18–20]. The relative importance of these two FXI activation modes in vivo is not entirely understood and so far it remains unclear why the bleeding severity in FXI deficiency in humans is not associated with FXI plasma levels. Therefore, the role of FXI in fibrin formation has been extensively studied in mammalian animal models of hemostasis and thrombosis.

Factor XI deficient mice

  1. Top of page
  2. Abstract
  3. Factor XI and the intrinsic pathway of coagulation
  4. Factor XI deficient mice
  5. Thrombosis models in FXI deficient mice
  6. Activation of FXI in arterial thrombosis
  7. Deficiency in FXI in combinations with other clotting proteins
  8. Evolutionary aspects of FXI deficiency
  9. Acknowledgments
  10. Disclosure of Conflict of Interests
  11. References

The plasmatic coagulation systems of humans and mice share high similarity [21,22]. All coagulation factors and inhibitors are well conserved [23]. Both human and murine FXI are 160 kDa homodimers comprised of two 80 kDa polypeptides connected by a disulfide bond in the apple 4 domain. Addition of murine FXI to human FXI deficient plasma ‘rescues’ the prolonged aPTT, however with a slightly reduced activity compared with the human protein. Murine FXI is susceptible to activation by both human factor XII and thrombin supporting the similarity between the species, which is also reflected by its 78% homology to human FXI at the protein level. Of note, some differences between human and mouse FXI exist. The murine protein is predominantly, if not exclusively, expressed in the liver, whereas human FXI can also be detected in kidney and pancreas. Additionally, murine FXI appears to have slightly reduced enzymatic activity although it is difficult to precisely assess the enzymatic properties, as a FXI-specific chromogenic substrate is missing [24]. Using a classical homologous recombination-based approach, Gailani et al. [25] targeted the FXI gene for deletion in mice. FXI−/− mice are healthy, fertile, and phenotypically indistinguishable from wild-type animals. Mating between heterozygous FXI+/− mice followed the expected Mendelian ratio arguing against an association of FXI deficiency and increased risk of abortion. Consistent with this observation, factor XII levels are not associated with aborting [26]. Similar to its human counterparts, plasma from FXI null mice showed a severely prolonged aPTT compared with wild-type animals, whereas the aPTT of plasma from heterozygous animals, having 50% of the wild-type FXI level, was somewhere in between. Importantly, bleeding times in adult FXI−/− mice, as assessed by a tail bleeding assay, were indistinguishable from wild-type animals, suggesting that FXI does not significantly contribute to fibrin formation in this kind of injury model. To further support, large surgical procedures did not reveal abnormal hemostasis following injury of various vascular beds including the brain in FXI null mice [27].

Thrombosis models in FXI deficient mice

  1. Top of page
  2. Abstract
  3. Factor XI and the intrinsic pathway of coagulation
  4. Factor XI deficient mice
  5. Thrombosis models in FXI deficient mice
  6. Activation of FXI in arterial thrombosis
  7. Deficiency in FXI in combinations with other clotting proteins
  8. Evolutionary aspects of FXI deficiency
  9. Acknowledgments
  10. Disclosure of Conflict of Interests
  11. References

Initially, Rosen et al. [28] analyzed arterial thrombus formation in FXI−/− mice, using a FeCl3-induced (20% FeCl3) carotid artery injury model. Formation of thrombi was severely defective in the absence of FXI, and intravenous infusion of human FXI restored formation of vessel occlusive thrombi indicating that FXI operates similarly in these species in vivo. The importance of FXI for arterial thrombus formation was confirmed by studies that compared factor IX and FXI deficient mice in the carotid artery thrombosis model, in which occlusive thrombus formation was induced by application of varying concentrations of FeCl3 to the exterior of the vessel [29]. All wild-type animals had vessel occlusion within 5–10 min after exposure to 3.5% FeCl3, while factor IX null mice were completely protected from occlusion at 5% FeCl3 and partially protected at 7.5% FeCl3. Remarkably, FXI−/− mice responded to FeCl3-induced carotid artery injury in a manner similar to factor IX null mice, showing that the contribution of the two proteases to thrombus formation in this model does not correlate with their importance in hemostasis. Using the same system, Wang et al. [30] showed that FXI null mice are also protected from occlusive thrombosis in the vena cava. Although the data suggest that FXI deficiency confers thromboprotection in veins, the importance of FXI for venous thrombosis in humans is difficult to assess. FeCl3 causes disruption of the vascular endothelial layer and is probably associated with significant exposure of collagen to the flowing blood, which is of minor importance for the initiation of venous thrombosis in large vessels. It is not clear if activated FXI functions to simply increase the total factor IX activity at the wound site, or acts to generate factor IXa at strategic locations within the clot or clot–vessel interface [31]. In vitro, FXI activity is important for maintaining clot integrity over time when plasma or blood is induced to clot with TF or thrombin in the presence of an activator of fibrinolysis. This effect is mediated, at least in part, through FXIa-dependent thrombin generation leading to the activation of the plasma metalloproteinase thrombin-activatable fibrinolysis inhibitor (TAFI) [32–34]. TAFI suppresses fibrinolysis by fibrin proteolysis.

The importance of FXI for thrombus formation is not restricted to the FeCl3 model. Intravital microscopy and laser injury confirmed the importance of FXI in thrombus formation in the arterioles of the cremaster muscles in FXI null mice [35]. Following laser injury, both fibrin deposition and platelet accumulation were reduced in FXI null mice compared with a wild-type control. While the reduction in fibrin deposition was relatively modest (∼50%), the defect in platelet accumulation was significantly greater (>90%), supporting classical observations, which showed that the intrinsic pathway is operating on platelet surfaces (reviewed by [36]). Platelet-rich thrombi developed at the site of vessel injury in FXI deficient mice, but the thrombi were unstable and rapidly embolized, restoring vessel patency. The antithrombotic effect of FXI deficiency does not appear to be related to reduced TAFI activation, as mice lacking TAFI are not protected from arterial occlusion after FeCl3-induced injury (unpublished data). In rabbits, Minnema and colleagues showed that inhibition of FXI by a function-neutralizing antibody enhances tPA-induced clot lysis in an ear vein thrombus model [37], and anti-FXI antibodies interfered with thrombus propagation on injured neointima of the rabbit iliac artery [38]. FXI is also essential for thrombosis in primates, supported by studies using FXI blocking antibodies, which demonstrated that FXI inhibition prevented growth of occlusive platelet rich thrombi in an arterio-venous shunt model [39] and vascular graft occlusion [40] in baboons. Thus, FXI appears to have prothrombotic properties across a wide range of mammalian species.

Activation of FXI in arterial thrombosis

  1. Top of page
  2. Abstract
  3. Factor XI and the intrinsic pathway of coagulation
  4. Factor XI deficient mice
  5. Thrombosis models in FXI deficient mice
  6. Activation of FXI in arterial thrombosis
  7. Deficiency in FXI in combinations with other clotting proteins
  8. Evolutionary aspects of FXI deficiency
  9. Acknowledgments
  10. Disclosure of Conflict of Interests
  11. References

Based on the premise that the fibrin forming mechanisms in thrombus formation and hemostasis are identical but differ in intensity or localization, it was assumed that FXI activation in the arterial injury models was a factor XII-independent process, as suggested by the revised model of coagulation. However, work comparing thrombus formation in mice lacking factor XII and FXI raises serious questions about this premise. Compared with wild-type animals, both FXI and factor XII null animals were largely protected from arterial occlusions induced by FeCl3, mechanical damage, or ligation types of injury, and intravital microscopy clearly demonstrates the instability of platelet-rich thrombi in the vessels of both gene knockout strains [27]. Subsequent work using a middle cerebral artery model of transient ischemia-reperfusion injury demonstrated a significant thromboprotective effect for both factor XII and FXI deficiency, with substantially lower volumes of infarcted tissue and less fibrin deposition in postischemic brain microvessels [41]. Together the studies demonstrate that, at least in mice, FXI is activated by factor XII during ischemia reperfusion injury and within the growing thrombus according to the classical intrinsic pathway cascade. Furthermore, these novel observations with factor XII−/− and FXI−/− mice suggest that the concept, in which pathologic thrombus formation represents a disequilibrium of the hemostatic process active at a wound site, is not entirely accurate [42]. However, whether this premise can be transferred to thromboembolism in humans needs to be investigated [43,44]. While certainly sharing common features, physiological hemostasis and pathological thrombosis appear to employ some different mechanisms [41]. The extrinsic pathway of coagulation is important for both hemostasis and thrombosis, and deficiency or increased activity of this pathway may result in bleeding or pathological thrombosis, respectively [45]. However, to stabilize occlusive thrombi in the arterial system, additional fibrin or platelet activation is necessary, and this process is critically dependent on FXI, whereas FXI-driven fibrin formation has minor importance in preventing blood loss at a site of vessel injury. This raises the possibility that therapeutically targeting FXI may offer a strategy for preventing or treating arterial thrombosis that is not associated with the high rate of hemorrhage that accompanies many currently used anticoagulants [46].

Deficiency in FXI in combinations with other clotting proteins

  1. Top of page
  2. Abstract
  3. Factor XI and the intrinsic pathway of coagulation
  4. Factor XI deficient mice
  5. Thrombosis models in FXI deficient mice
  6. Activation of FXI in arterial thrombosis
  7. Deficiency in FXI in combinations with other clotting proteins
  8. Evolutionary aspects of FXI deficiency
  9. Acknowledgments
  10. Disclosure of Conflict of Interests
  11. References

As FXI contributes to pathological fibrin formation in mice, FXI null animals appeared as an attractive system to ‘rescue’ the prothrombotic phenotype of plasminogen and Protein C deficient mice. Targeted deficiency in Protein C leads to early neonatal death due to lethal coagulopathy in various tissues. In contrast, mice with combined homozygous deficiency in FXI and Protein C (FXI−/−/PC−/−) survived the early lethality observed in Protein C null animals and reached up to 3 months of age indicating that they, even though lacking intrinsic pathway-mediated fibrin generation, succumbed to overwhelming thromboembolic events later in life [47]. The data indicate that FXI is critical for fibrin formation in vivo and support the hypothesis that blocking FXI activity provides a new strategy to treat severe thromboembolic disorders. In contrast, FXI deficiency worsens the excessive disseminated fibrin depositions that underlie the wasting phenotype with impaired wound healing of plasminogen (Plg) null mice. Plg is the precursor of the serine protease plasmin, which plays a major role in regulating hemostasis by proteolytically degrading fibrin. FXI−/−/Plg−/− mice have a reduced live span, and increased leukocyte infiltration into the lung compared with Plg−/− animals [48]. Surprisingly, deficiency in factor IX rescues the wasting phenotype of Plg null mice and increases survival suggesting that FXI may have additional functions in regulation of inflammation or tissue repair distinct from its role in coagulation. Indeed, FXI may contribute to host defense systems. Tucker et al. [49] challenged FXI null mice in a bacterial model of lethal sepsis induced by cecal ligation and puncture. FXI deficiency increased survival and reduced leukocyte infiltration and coagulopathy, suggesting that FXI contributes to inflammation and pharmacological FXI inhibitors might be beneficial in septic disease.

Evolutionary aspects of FXI deficiency

  1. Top of page
  2. Abstract
  3. Factor XI and the intrinsic pathway of coagulation
  4. Factor XI deficient mice
  5. Thrombosis models in FXI deficient mice
  6. Activation of FXI in arterial thrombosis
  7. Deficiency in FXI in combinations with other clotting proteins
  8. Evolutionary aspects of FXI deficiency
  9. Acknowledgments
  10. Disclosure of Conflict of Interests
  11. References

Genetic analyses have shown that FXI is a relatively ‘modern’ protein. FXI and all other contact system proteins such as factor XII, plasma kallikrein or high molecular weight kininogen are absent from inframammalian vertebrates. As genes of contact system proteases and FXI are absent in chicken, vultures, and ostriches, bird blood clots slowly upon exposure to foreign surfaces. FXI shares high homology with plasma kallikrein both in sequence and domain arrangement [50], indicating that a gene duplication event gave rise to these paralogs [51]. Indeed, a single precursor of FXI and plasma kallikrein is found in frog, chicken, and platypus, suggesting its first appearance among early tetrapods. In contrast, the opossum genome has two distinct genes for both FXI and plasma kallikrein, indicating that the gene duplication event leading to separate factors occurred early in mammalian evolution [22]. Most probably due to inbreeding, FXI deficiency is found in some cattle [52–54], dogs [55], and cats [56,57]. Surprisingly, and in contrast to mice and most FXI deficient humans, some of these animals suffer from abnormal bleeding showing that the precise role of FXI for hemostasis and thrombosis in vivo still needs to be investigated further.

Acknowledgments

  1. Top of page
  2. Abstract
  3. Factor XI and the intrinsic pathway of coagulation
  4. Factor XI deficient mice
  5. Thrombosis models in FXI deficient mice
  6. Activation of FXI in arterial thrombosis
  7. Deficiency in FXI in combinations with other clotting proteins
  8. Evolutionary aspects of FXI deficiency
  9. Acknowledgments
  10. Disclosure of Conflict of Interests
  11. References

This work was supported in part by grants from the Deutsche Forschungsgemeinschaft (SFB 688) and the EU-funded ERARE program. We apologize to all such colleagues whose works could not be cited due to space limitations.

References

  1. Top of page
  2. Abstract
  3. Factor XI and the intrinsic pathway of coagulation
  4. Factor XI deficient mice
  5. Thrombosis models in FXI deficient mice
  6. Activation of FXI in arterial thrombosis
  7. Deficiency in FXI in combinations with other clotting proteins
  8. Evolutionary aspects of FXI deficiency
  9. Acknowledgments
  10. Disclosure of Conflict of Interests
  11. References