Pathological Anti-β2-Glycoprotein I Antibodies Disrupt the Anticoagulant Activity of Annexin A5: A Possible Explanation for the Lupus Anticoagulant Paradox.

Abstract

One of the most enigmatic features of the antiphospholipid (aPL) syndrome is the paradoxical correlation between thrombosis and the in vitro lupus anticoagulant (LAC) effect. It has recently been shown that only LAC that are due to β2-glycoprotein I (β2GPI) antibodies correlate with thrombosis. It has also been reported that aPL antibodies can disrupt the formation of an anticoagulant crystal shield that is formed by annexin A5 (AnxA5) over phospholipids bilayers. This disruption can be assayed by measuring resistance to AnxA5 anticoagulant activity. We therefore investigated whether the presence of these β2GPI-dependent LACs might correlate with the disruption of AnxA5 anticoagulant activity. We performed a double blind study with 30 patients divided into three groups; Group A: plasma of patients that show a β2GPI-dependent LAC; Group B: plasma of patients that show a β2GPI-independent LAC; Group C: plasma of SLE patients without a LAC. As previously demonstrated, we were able to discriminate between a β2GPI-dependent and a β2GPI-independent LAC by titrating cardiolipin into the plasma sample. A β2GPI-dependent LAC could be normalized by the addition of cardiolipin in contrast to a β2GPI-independent LAC which was prolonged by cardiolipin. To investigate the effects on AnxA5, we performed an APPT-based coagulation assay. In this assay we added AnxA5 to the plasma of patients and measured the prolongation in clotting time. We calculated the ratio of the clotting time divided by the clotting time with AnxA5 added to the plasma (=AnxA5 ratio). This assay represents the resistance against AnxA5 and gives an indication of its effect on the anticoagulant shield formed by AnxA5. Eleven patients displayed a β2GPI-dependent LAC (group A, all thrombosis), that gave an AnxA5 ratio of 151.3% (± SD 14.5). For 9 patients that displayed a β2GPI-independent LAC (group B, thrombosis n=3) we found a ratio of 225.4% (± SD 29.4). Ten patients did not have LAC (group C, thrombosis n=2) that gave a ratio of 219.2% (± SD 31.0). The annexin A5 ratio of group A was significantly lower than the ratio of group B (P=0.0002) and group C (P=0.0001). There was no difference in AnxA5 ratio between group B and group C (P=0.7802). β2GPI-dependent LAC correlate strongly with both thrombosis and AnxA5 resistance. The results therefore suggest that anti-β2GPI antibodies that cause LAC, in contrast to aPL antibodies with other specificities, are capable of disrupting the anticoagulant effect of AnxA5 on phospholipids. The disruption of the AnxA5 anticoagulant shield by LAC-causing anti-β2GPI antibodies may be a mechanism for thrombosis in APS and offers a potential explanation for the LAC paradox. [Display omitted]