Prevalence, Domain Specificity, and Residue Specificity of IgG, IgM, and IgA Anti-β2GPI Antibodies in Tetra-Positive Antiphospholipid Syndrome Patients: Identification of Two Non-Overlapping Epitopes in Domain I of β2GPI

Background. Risk stratification of antiphospholipid syndrome (APS) patients involves several laboratory tests, including lupus anticoagulant (LAC), anti-cardiolipin antibodies (anti-CL), anti-β2GPI antibodies (anti-β2GPI), and, more recently, anti-phosphatidylserine/prothrombin antibodies (anti-PS/PT). Patients with positive results for all four tests, known as tetra-positive APS patients, are considered at high risk for thrombotic recurrence. β2GPI is a protein composed of five domains: DI, DII, DIII, DIV, and DV, arranged in a flexible elongated structure. Previous studies have demonstrated that DI is frequently targeted by IgG in thrombotic APS patients. However, the prevalence, domain specificity, and residue specificity of the three different isotypes of anti-β2GPI antibodies in tetra-positive APS patients remain unknown. Objective. To establish the prevalence, domain specificity, and residue specificity of IgG, IgM, and IgA anti-β2GPI antibodies in tetra-positive APS patients. Methods. Plasma samples from 32 clinically confirmed tetra-positive APS patients and 8 healthy controls were analyzed for levels of IgG, IgM, and IgA using an enzyme-linked immunosorbent assay (ELISA) with immobilized β2GPI as the antigen. Domain and residue mapping were investigated using five deleted domain variants and eight single-point mutations. Mutants lacking specific disulfide bonds were used to assess whether the epitopes targeted by the antibodies are linear or conformational. Results. Among the 32 patients tested, 31 (96%) had IgG, 18 (56%) had IgM, and 18 (56%) had IgA anti-β2GPI antibodies. Ten patients (31%) had all three isotypes, 7 (22%) had IgG and IgM, and 8 (25%) had IgG and IgA. Six patients (19%) had only IgG, 1 had only IgM, and none had only IgA. The removal of DI from the β2GPI molecule led to an almost complete loss of binding (~80%) for the majority of IgG antibodies. There was a lesser but still significant reduction in binding when DII and DV were removed. However, no loss of binding was observed with variants lacking DIII and DIV. Notably, IgA antibodies behaved similarly to IgG but differently from IgM, as their reactivity towards DI and DV was significantly attenuated. To investigate the contribution of specific residues to binding, charged residues K19, E23, E26, E27, R39, R43, K44, and K59 in DI were individually replaced with alanine (A). The most important residue for IgG and IgA antibodies was R43, followed by R39 and K44. In contrast, most residues contributed to optimal binding for IgM antibodies, suggesting a less specific epitope definition. Interestingly, the mutations R39A and K44A often led to loss of binding in conjunction with the mutation R43A, but rarely simultaneously, suggesting the existence of two distinct epitopes for anti-DI antibodies-one located north and one south of R43. Finally, IgG, IgM, and IgA antibodies did not react against the mutant C32S/C60S but showed good reactivity against the mutant C288S/C326S, indicating that epitopes in DI are conformational and require the integrity of the disulfide bond 32-60. Conclusions. In tetra-positive APS patients, IgG anti-β2GPI antibodies are the most prevalent isotype, followed by IgA and IgM. The IgG and IgA isotypes exhibit a pronounced preference for DI, while the IgM isotype demonstrates lower specificity, which aligns with IgM's role as initial responders in infections. Notably, our research revealed that anti-DI antibodies are diverse, as they recognize epitopes within DI that are closely related but physically distinct. Ongoing studies aim to further characterize these antibody subtypes and understand their significance in APS. This knowledge may inform improved diagnostic and therapeutic approaches for APS patients.