Binding is not enough; autoantibodies do more to garner complements

See allHide authors and affiliations

Science Immunology  05 Jul 2019:
Vol. 4, Issue 37, eaax9510
DOI: 10.1126/sciimmunol.aax9510


Pathogenic autoantibodies in neuromyelitis optica require antigen array assembly and a specific epitope to initiate robust complement activation.

Autoantibodies can serve as valuable biomarkers for the diagnosis of autoimmune diseases. Those that are pathogenic also serve this purpose, and, given their direct role in disease, they are presumed to offer a further association with disease severity. Yet such associations are not invariably observed. Furthermore, autoantibodies can present another curious phenomenon: they effect damage to a restricted set of tissues despite the presence of target antigens on tissues that remain unharmed. Such is the case in neuromyelitis optica (NMO), where anti-aquaporin 4 (AQP4-IgG) autoantibodies are both biomarkers and pathogenic but do not correlate with disease status, and autoantibody-mediated tissue damage is restricted to the central nervous system (CNS) despite abundant expression of AQP4 in the kidneys.

The study by Soltys and colleagues provides new data that help explain these counterintuitive characteristics. They approached the questions by seeking to understand how autoantibody epitope specificity influences Fc-effector functions, namely complement-dependent cytotoxicity (CDC), a well-established mechanism of pathogenicity in autoimmune disease. Specifically, they investigated epitope specificity of AQP4-IgGs and showed that those targeting a distinct extracellular loop epitope, regardless of affinity, enhanced CDC. The study further showed that the M23 AQP4 isoform, which forms supramolecular orthogonal arrays, arrange in a manner that benefits autoantibody multimeric complexes supporting Fc-Fc interactions that are critical for CDC. The M1 isoform does not assemble into such an arrangement and consequently does not support the robust activation of complement seen with M23 isoforms.

This study helps to answer two fundamental questions in NMO, and more broadly in human autoantibody-mediated disease. First, AQP4-IgG serum titers fail to correlate with disease severity because antigen binding alone (the sole property measured by existing diagnostic assays) does not account for the multiple arrangement steps necessary to initiate CDC and subsequent tissue damage. Second, pathogenic AQP4-IgG are epitope specific, limiting CDC to the CNS where the M23—not M1 isoform—is abundant. These findings, along with their model of CDC initiation, are likely to extrapolate to other autoantibody-mediated diseases, offering better understanding of disease progression, novel diagnostics, and targeted therapeutic interventions.

Highlighted Article

View Abstract

Navigate This Article