Platelets as pivot in the antiphospholipid syndrome

In this issue of Blood, Proulle et al demonstrate that platelets are essential for the prothrombotic effects of anti–β₂-glycoprotein I autoantibodies in a mouse model of the antiphospholipid syndrome (APS).¹ 

APS is characterized by the presence of antiphospholipid antibodies in patients with thrombosis or pregnancy morbidity.²  Although its name suggests otherwise, antiphospholipid antibodies are not directed against phospholipids but toward plasma proteins with affinity for anionic phospholipids. Experiments in a mouse model for APS have shown that autoantibodies against the plasma protein β₂-glycoprotein I are responsible for the increased risk of thrombosis. A complication in our understanding of the syndrome is that no physiological function has been assigned to β₂-glycoprotein I and that men and mice without β₂-glycoprotein I seem to be healthy.³  How antibodies directed against β₂-glycoprotein I can lead to an increased risk of thrombotic complications is unsolved and the cause of vivacious debates. A consensus now is that the autoantibodies induce a new activity for the protein. It has been shown that β₂-glycoprotein I antibody complexes, but not β₂-glycoprotein I alone, can activate different cell types that are involved in the regulation of hemostatic response, resulting in a prothrombotic state. A major question that remains is: “What are the target cells in vivo for the antibody-β₂-glycoprotein I complexes?”

β₂-glycoprotein I can exist in 2 completely different conformations.⁴  In plasma, it is present as a circular protein in which its N-terminal domain interacts with its C-terminal domain. After interaction with antibodies, the protein opens up and forms a hockey stick–like conformation. This stretched conformation now expresses a hidden epitope in its C-terminal domain that is involved in the binding of the antibody–β₂-glycoprotein I complex to cells. To complicate matters, this complex is very sticky and binds in vitro to many different proteins and cellular receptors, questioning the value of the results of these experiments. Animal experiments are an essential tool to understand the consequences of cellular interactions of the autoantibodies–β₂-glycoprotein I complexes. However, mouse models of APS have shown that endothelial cells, monocytes, and platelets all became activated when anti–β₂-glycoprotein I antibodies were infused.⁵  The question remains whether 1 specific cell is the prime target or that the complexes can activate different cell types independently of each other. The experiments of Proulle et al showed that inhibition of platelet activation prevents the activation of endothelial cells. Thus platelets are the first target for the complexes, and products released from activated platelets are responsible for the local activation of endothelial cells.

Some key issues remain to be answered. How do activated platelets mediate the activation of coagulation? The authors showed that products released from platelets activate endothelial cells and activated endothelial cells can express tissue factor. No information is given by the authors whether tissue factor is indeed induced in endothelial cells. In their experiments, the authors did not visualize monocytes. It is conceivable that the antibody complexes induce the formation of platelet monocyte complexes. Platelets mediate extravascularization of monocytes with subsequent tissue factor expression on the macrophage.⁵  Inhibition of platelet activation also inhibits platelet–monocyte formation. Infusion of autoantibodies against β₂-glycoprotein I in mice resulted in the expression of tissue factor on macrophages in the vessel wall.⁶  Monocytes/macrophages are excellent candidates as source for tissue factor. It is even possible that monocytes act as “trait-d’union” between platelets and endothelial cells. Experiments in which monocytes are depleted with gadolinium chloride could easily answer this question. However, in a model in which monocytes also participate, platelets seem to be the first target of the antibody–β₂-glycoprotein I complexes.

A second unknown is the receptor on platelets that mediates the binding of the antibody–β₂- glycoprotein I complexes. Many different receptors have been described that bind β₂-GPI-antibody complexes,³  such as Toll-like receptors 2 and 4, annexin A2, glycoprotein Ibα, and LRP8. Platelets express most of these receptors; only annexin A2 has been shown to be absent in platelets.⁷  Thus, a consequence of the observation that platelets are the first target of antibody–β₂-glycoprotein I complexes does not solve the question which cellular receptors are important to explain the APS pathophysiology.

The observations were made in a model using the cremaster muscle microcirculation, and both arterioles and venules were tested. APS is an autoimmune disease characterized by thrombotic complications in arteries and veins as well as in smaller vessels. However, the majority of the thrombotic complications are in large vessels: deep venous thrombosis and stroke. It is very important to know whether these observations made in the microcirculation could be translated to the larger arterial and venous vascular beds. If so, it would be interesting to see whether we can treat anti–β₂-glycoprotein I antibody–mediated venous thrombosis with aspirin or clopidogrel.