β₂-Glycoprotein 1 Antibodies Directly Activate the Platelet IgG Receptor, FcγRIIa, and Cause Thrombosis in FCGR2A Transgenic but Not in Wild Type Mice

Abstract 106

Antibodies targeting β₂-glycoprotein 1 (β₂-GP1; β₂-Abs) are of primary importance in antiphospholipid syndrome (APS), a thrombotic autoimmune disorder. The predominance of the IgG antibody isotype in APS is conspicuously associated with increased risk of thrombosis, raising the question whether the platelet IgG receptor, FcγRIIa, may play a role in thrombosis in APS, as is the case in heparin-induced thrombocytopenia (HIT). The hypothesis that platelet FcγRIIa may contribute to thrombosis in APS has received little attention, with research emphasis instead placed on several proposed alternative mechanisms of action. We have shown that, like HIT antibodies, antibodies targeting VEGF or CD154 are also potently thrombotic in mice transgenic for human FcγRIIa but have no activity in mice lacking FcγRIIa (i.e., wild type mice). We therefore asked whether antiphospholipid antibodies can activate platelets and cause thrombosis via FcγRIIa. To this end, we tested mouse monoclonal (mAb) and goat polyclonal anti-human β₂-Abs alone or complexed to human β₂-glycoprotein 1 (i.e., to form immune complexes, or ICs) by the serotonin release assay (SRA) and platelet aggregation methods using washed human platelets. We found that two of three commercially available polyclonal anti-β₂-Abs (pAb1 and pAb3), both alone or in IC form, induced platelet granule release and aggregation, and that this activity was abolished by anti-FcγRIIa mAb, IV.3. pAb2 and mAb were inactive. Activity analysis (SRA) of preformed ICs using constant pAb1 concentration with varied β₂-GP1 stoichiometries revealed a zone-of-equivalence pattern, with maximal activity near balanced stoichiometry (1:1). Because pAb2 did not activate platelets, we sought to determine its capacity to form higher order ICs (which are known to be required for FcγRIIa activation) by size exclusion chromatography (SEC). All antibodies tested in isolation were shown by HPLC-SEC to be free of aggregates or degradation products. Antibody-antigen complex size analysis revealed that, as expected, mAb+β₂-GP1 in 1:1 stoichiometry failed to form higher order ICs (i.e., complexes having ≥2 IgGs/complex). pAb2 (inactive) did form higher order ICs at 1:1 and 4:1 (IgG:Ag) stoichiometries, but less extensively than pAb1 (active), which efficiently formed higher order ICs at both 1:1 and 4:1 stoichiometries. pAb1 alone (i.e., not in IC form) caused aggregation, while pAb2 did not. Furthermore, preincubation of washed platelets with pAb2 prevented pAb1-induced aggregation, suggesting that pAb1 activity is β₂-GP1-specific. A single intravenous injection of anti-β₂-GP1 ICs (20 μg β₂-GP1 plus 120 μg pAb1, a 1:2 molar stoichiometry) induced severe thrombocytopenia (>90%) and caused thrombotic shock in FcγRIIa-transgenic mice but not in wild type mice. Symptoms of shock occurred within 10 minutes. Pervasive occlusive thrombi were observed in the lungs of all FcγRIIa-transgenic but not in any wild type mice (H&E microscopy). Injection of pAb2 ICs produced none of these effects in transgenic or wild type mice. Finally, injection of pAb1 or pAb2 alone (neither of which bind mouse β₂-GP1) had no effects in transgenic mice. In summary, these findings confirm previous in vitro studies that β₂-Abs can directly activate platelets in a manner wholly dependent on the platelet IgG receptor, FcγRIIa. Additionally, we have shown for the first time in vivo that β₂-Abs can also cause FcγRIIa-dependent thrombosis. This mechanism may contribute to thrombosis in APS, suggesting that further studies on the importance of FcγRIIa in APS are warranted.