In this project we set out to investigate several leads that could identify which patients with Human Platelet Antigen-1a (HPA-1a) antibodies are at high risk of developing Fetal and neonatal alloimmune thrombocytopenia (FNAIT). Previously our work had suggested afucosylation of these IgG antibodies are preferentially formed in patients at risk, which in turns enhances their elimination through IgG-Fc receptors on myeloid cells and/or NK cells. Other groups had at the start of the project shown that pathogenic HPA-1a antibodies bind stronger to the HPA-1a antigen when complexed with the αV integrin found on endothelial cells, suggesting FNAIT is mainly caused by immune recognition of endothelial cells of the blood vessels, not by directly degrading platelets where HPA-1a associates with the αIIb integrin. To study this, we generated cells that express the HPA-1a antigen in either αV or αIIb context and studied their interaction with patient antibodies. We purified these bound antibodies, to study their glycosylation. We found no evidence of the existence of antibodies that preferentially recognize either form. Antibodies bound to these cells also showed identical glycosylation per patient, also supporting the notion that anti-HPA-1a antibodies do not differentially recognize HPA-1a depending on their α-partner.
We did however, clone and functionally characterize antibodies from one patient who showed that HPA-1a antibodies can come in different flavours, most which do not activate platelet directly, while others seem to trigger platelet aggregation through the platelet FcγRIIa. This is worthy of future study, as this might explain different pathologies between patients with HPA-1a antibodies.
Lastly, we investigated the glycosylation of HPA-1a antibodies in the recent HIP cohort consisting of a prospective patient with HPA-1a antibodies, most of which had no pathologies. Unlike patients with severe FNAIT, this cohort mostly consisted of patients with highly fucosylated antibodies with lower effector function potential. All in all, this study shows that the quality and quantity of HPA-1a matters, which is independent of the integrin α partner. This quality determined by the antibody fucosylation, but also the intrinsic epitope and stoichiometry which sometimes can enable the antibody to activate platelets through their IgG-Fc-receptor.

Summarizing Figure: In this project we characterized functional properties of alloantibodies to human platelets contributing to disease in Fetal and neonatal alloimmune thrombocytopenia (FNAIT). A) Antibodies to the most common FNAIT inducing alloantibodies to Human Platelet Antigen 1a (HPA-1a) were cloned from patients. We discovered that not all HPA-1a are equally pathogenic: some activate platelets through the IgG Fc Receptor (FcγRIIa), while most HPA-1a antibodies require effector functions from myeloid or NK cells. B) The effector functions can be categorized in antigen-binding (Fab, above) and Fc-mediated effector functions (bottom). Top: Without requiring effector functions, these antibodies can block binding of the receptor to their ligands (e.g. by locking them in their inactive function, right) or induce cellular death through apoptosis of either megakaryocytes (MK) – the platelet progenitors – or endothelial cells. Together this can contribute to bleeding tendencies seen in FNAIT. Bottom: With effector functions, antibodies in FNAIT tend to be afucosylated, in particular in severe FNAIT patients. This strongly elevates their potential to induce effector functions through FcγRIII on myeloid- an NK-cells. In addition, these antibodies all have very high galactosylation, which can elevate their potential to activate complement, especially in the presence of other co-occurring alloantibodies (e.g. anti-HLA). This also elevates their pathogenicity. Finally – we found that antibodies recognize HPA-1a epitope (found on the β3 integrin) indiscriminately weather it is found associated with αII on platelets or αV on endothelial cells. In addition, HPA-1a antibodies affinity purified from either molecular form of HPA-1a have identical glycosylation. Together this strongly indicates that HPA-1a antibodies do not have different origin and recognize both endothelial cells and platelets identically, simplifying future diagnostic screening and future efforts aimed at unravelling the disease mechanism behind FNAIT.

THESIS 2026: Maternal antibodies to fetal platelets: a complex-related story - Janita J. Oosterhoff