Domain and residue mapping of autoantibodies to β2GPI reveals differences among antiphospholipid syndrome phenotypes

Abstract

Abstract Antiphospholipid antibodies targeting β2-glycoprotein I (β2GPI) are a hallmark of antiphospholipid syndrome (APS), associated with an increased risk of thrombosis and pregnancy morbidity. Among these, antibodies targeting domain I (DI) are common in individuals at higher risk; however, their epitopes and prevalence among APS phenotypes remain unclear. Here, we use a large collection of 29 structurally and functionally validated β2GPI variants to provide new insights into the molecular mechanisms of autoantibody recognition in APS. Using the prototypic human-derived monoclonal anti-DI antibody MBB2, we identified positively charged residue R39 as the key driver of MBB2 binding, followed by residues R43, N56, and T57. Structural analyses revealed that although R39 is solvent exposed, R43 is not, because it is caged by residues N56 and T57. The narrow epitope footprint explains why MBB2 exhibits a modest affinity for soluble β2GPI. The cage structure accounts for the epitope being conformational rather than linear. Mutational analyses of immunoglobulin G anti-β2GPI antibodies from 52 patients with triple-positive APS, 37 with a history of thrombosis and 15 nonvascular obstetric patients, confirmed significant reactivity against DI and showed signatures of 2 conformational epitopes: one similar to MBB2 (epitope I), in which the presence of R39 is essential, and another that does not require R39 (epitope II). Although less frequent than epitope II in our cohort, epitope I reactivity was notably enriched in patients with vascular-obstetric APS. Varying epitope specificities for DI may therefore aid in identifying different APS phenotypes and predicting clinical outcomes.