The J-shape of β2GPI reveals a cryptic discontinuous epitope across domains I and II

Abstract

Beta-2-Glycoprotein I is the main target for pathogenic antiphospholipid syndrome autoantibodies. It can adopt several conformations, including an O-shape and two more linear J- and S-shapes. The in vivo existence of the O-shape is debated, and doubt remains pertaining to the pathogenic impact of each shape. Studies have shown that APS antibodies react weakly with the O-shape and bind to the linear shapes due to the exposure of a cryptic epitope in the 1st domain. How the protein transitions from O-shape to the linear shapes remains unknown. While the main epitope is widely recognised as the R39-R43 peptide, there is evidence pointing to a discontinuous epitope across domains I and II (DI & DII). We used molecular dynamics simulations to examine the potential pathways of conformational shift from the O-shape to the open forms, and the impact of plasmin clipping on these pathways. Through these studies, starting in a theorised O-shape, we identified that peptides R39-R43, T50-N56 and R63-F67 become more exposed and have increased stability in the J- and S-shapes relative to the O-shape. These changes are likely due to a shift in DII of the T106-G109 loop, which twists to form contacts with the DI K33-Y36 loop. The R39-R43 peptide is brought closer to R63-F67 suggesting a more complex DI epitope than previously theorised. These effects were observed in the wild type and plasmin clipped model, with the effect being larger in the latter. These results are in good agreement with the increased antibody binding observed experimentally for the clipped protein. We therefore suggest that we have been able to identify the structural mechanism at the residue level which results in increased antibody binding in the J-Shape, and specifically in the clipped protein.