Prekallikrein, factor XII, and high-molecular-weight kininogen are members of the contact activation system of coagulation. Despite decades of research, the contact activation system has been enigmatic. Although deficiencies of the contact activation factors result in prolongation of the activated partial thromboplastin time, they are not associated with a bleeding diathesis. However, thrombosis models using factor XII-deficient and high-molecularkininogen-deficient mice have suggested that these two proteins contribute to thrombosis.1 Additionally, there has been a resurgence of interest in the study of the contact activation system because of its possible roles in a variety of other processes, including blood pressure regulation, cell proliferation, angiogenesis, apoptosis, and inflammation.2 Now Liu et al. from the laboratory of Edward Feener at Harvard Medical School have reported results using rat and murine models that suggest that kallikrein may play a pathogenic role in stroke. The path that led to this observation is an example of how a logical direction can lead to unexpected turns in the laboratory. In an earlier study of carbonic anhydrase-dependent edema following intracerebral hemorrhage (ICH) in rats, the Feener laboratory became interested in kallikrein because of the association between deficiency of C1 esterase inhibitor, which inhibits kallikrein, and hereditary angioedema.3 They found that antibodies to kallikrein inhibited retinal edema and proposed that the effect was due to the well-known production of the proedematous peptide bradykinin following proteolytic cleavage of high-molecular-weight kininogen by kallikrein.
In the present study, Liu et al. addressed the clinical observation that diabetes mellitus and hyperglycemia are associated with ICH in stroke patients. They found that ICH expansion following intracerebral infusion of autologous blood was increased in diabetic rats and non-diabetic hyperglycemic rats. Intracerebral injection of purified kallikrein increased ICH expansion, which was blocked by antibodies to kallikrein. In an analogous murine model, ICH expansion was decreased in prekallikrein-deficient mice. Surprisingly, bradykinin receptor antagonists had no effect on ICH expansion in rats. Searching for a possible anti-hemostatic role of kallikrein, they found that it inhibited collagen-induced rat platelet aggregation in vitro, but not ADP- or thrombin-induced platelet aggregation. Continuing to search for an association between hyperglycemia and ICH, they found that glucose enhanced the inhibitory effect of kallikrein on collagen-stimulated platelet aggregation. Additionally, using surface plasmon resonance spectroscopy, they found that kallikrein binds collagen, suggesting that kallikrein binds to exposed subendothelium and prevents platelet adhesion and activation. Consistent with this finding, they observed that kallikrein bound to de-endothelialized rat aorta, and that binding increased with increasing concentrations of glucose.
On the platelet membrane surface, collagen binds to glycoprotein (GP) VI, whose expression depends on its association with the FcRγ chain. Liu et al. found that administration of a neutralizing monoclonal anti-GPVI antibody to mice increased ICH expansion. Additionally, ICH expansion was increased in FcRγ-deficient mice. They also observed that hyperosmolar mannitol and sodium chloride produced effects similar to glucose on the binding of kallikrein to collagen, inhibition by kallikrein of collagen-stimulated platelet aggregation, and kallikreindependent ICH expansion. These findings suggest that mannitol, which is used to reduce intra-cerebral pressure and edema in individuals with stroke, may also have adverse effects.
In Brief
Liu et al. propose a novel mechanism in which inhibition by kallikrein of collagen-induced platelet aggregation contributes to ICH following stroke, thus implicating another member of the contact activation system in thrombotic pathology. They additionally propose that kallikrein is a potential pharmacologic target. As the authors note, their findings are limited to rodent ICH models that may have important differences from stroke settings in humans. Nonetheless, their findings should stimulate further investigation into the possible roles of the contact activation system in thrombosis, including hemorrhagic stroke, and in platelet function.
References
Competing Interests
Dr. Lollar indicated no relevant conflicts of interest.
