Comment on Cho and Mosher, page 3555
Thrombi in conditional fibronectin knock-out mice continuously shed platelets and exhibit slowed growth and decreased stability. Cho and Mosher report results from in vitro studies that identify the molecular mechanisms for these effects and define the role of fibronectin in thrombus growth and stability.
Nearly 4 decades ago, the investigation of a patient presenting with chronic intravascular coagulation secondary to an occult neoplasm that was characterized by cold-induced plasma precipitation led to the isolation of plasma fibronectin.1 Since then, many roles have been proposed for fibronectin, and there have been associations between fibronectin and disease.2 The concentrations of fibronectin are generally reduced in disseminated intravascular coagulation, liver disease, sepsis, and trauma or surgery, but are increased in some kinds of cancer. In one family with a congenital deficiency of fibronectin, there was abnormal wound healing but no bleeding.3 There is an association between moderately elevated levels of fibronectin and coronary artery disease, and large quantities of fibronectin have been found in developing fibrous atherosclerotic plaques.4
What is the function of plasma fibronectin? There has been no shortage of studies of fibronectin, and there is an extensive list of this protein's effects on coagulation, thrombus formation, and many other physiologic and pathologic processes.2 Most of these functions have revolved around fibronectin's role as an extracellular matrix protein, its binding and covalent attachment to fibrin by factor XIII, and its specific binding to several integrins present on platelets and other cells.
The relative significance of these various proposed functions in vivo can be evaluated in several ways. For many proteins, the use of knock-out mice has provided such a role. Animals deficient in fibrinogen showed no defect in thrombogenesis, except in the stability of larger thrombi. Animals deficient in fibrinogen and von Willebrand factor showed delayed thrombogenesis, but occlusive thrombi still formed. On the other hand, animals deficient in β3 integrins showed a complete absence of thrombus formation. These results demonstrate that another protein in addition to fibrinogen and von Willebrand factor must bind to β3 integrins to support thrombus formation.FIG1
The fundamental role of fibronectin was not appreciated at first, because of the prominence of these other proteins and because many studies were carried out under static or low shear rates. It is significant that fibronectin is essential for embryonic development, so its function in blood coagulation could not be studied by the use of knock-out mice with complete absence of fibronectin.
Inactivation of the fibronectin gene in conditional knock-out mice reduced plasma fibronectin levels to less than 2% and platelet fibronectin to about 20% of that in control mice.5 In a model of arterial injury in these mice, it was observed that platelet adhesion was normal initially, but there was a delay of several minutes in thrombus formation and the thrombi continuously shed platelets. These studies showed that fibronectin plays a significant role in thrombus growth and stability at arterial shear rates.
The mechanisms that account for these results are analyzed with in vitro experiments reported in a paper published in this issue of Blood. Thrombus formation was studied in parallel plate flow chambers. Thrombi were larger on surfaces with fibronectin cross-linked to fibrin. If plasma fibronectin was perfused with platelets, the thrombi were even larger. These effects were blocked by addition of a 70-kDa N-terminal fragment of fibronectin or a bacterial peptide that inhibits fibronectin binding to fibroblasts. Furthermore, platelet activation with lysophosphatidic acid was necessary, and all of these effects were concentration dependent. Finally, there was a synergism of the 2 mechanisms of thrombus enhancement by fibronectin: cross-linking to fibrin and assembly by adherent and aggregating platelets (see figure).
It is not yet entirely clear which platelet receptors are responsible for these effects of fibronectin. It seems that binding of the type III modules 9 to 10 of fibronectin to β3 integrins is responsible for enhancing platelet adherence to fibronectin-fibrin versus fibrin, and β3 integrins likely mediate adhesive interactions of platelets to assembled fibronectin. However, the binding of the N-terminal 70-kDa region of fibronectin may be mediated by nonintegrin receptors.
These results can be put into the context of other adhesive proteins that play important roles in thrombogenesis. Evidence from a variety of sources now suggests that von Willebrand factor, fibrinogen, and fibronectin all have separate roles in thrombus formation. Von Willebrand factor is important in initiating and supporting platelet–vessel wall and platelet-platelet interactions at high shear rates. Fibrin(ogen) is present at high concentrations in plasma, and its high affinity for activated αIIbβ3 means it will have a fundamental role in platelet aggregation under normal conditions. Because platelet adhesion to fibrin and platelet-platelet interactions are both affected by fibronectin, it seems that fibronectin is involved at all stages.
In conclusion, these results provide a potential explanation for the association of elevated levels of plasma fibronectin with coronary artery disease. The finding, reported in this paper, that thrombus formation in fibrin matrices is enhanced as fibronectin increases suggests that higher levels of plasma fibronectin may predispose subjects to larger and more stable arterial thrombi. Further research will be necessary to determine if fibronectin could be a potential risk factor and a target for reduction. ▪