Differentiating naturally occurring and disease associated autoantibodies in MPO-ANCA vasculitis through immunogenomic and epitope comparisons.

Introduction: ANCAs are diagnostic and pathogenic in ANCA-associated vasculitis however ANCAs can persist or reappear in clinical remission. It is a known phenomenon that naturally occurring autoantibodies, including ANCAs, are also detectable in healthy individuals. That autoantibodies do not directly correlate with clinical disease suggests that characteristics of the autoantibody response, rather than its mere presence, may determine pathogenicity. We aimed to define immunogenomic and epitope-level differences between naturally occurring and disease-associated MPO-ANCA. Methods: Patients with MPO-ANCA vasculitis (active disease and long-term remission off therapy) and healthy individuals were enrolled at a single center. Peripheral blood mononuclear cells were screened using highly sensitive methods to isolate MPO specific B cells from healthy individuals and patients with MPO-ANCA. Sorted cells were expanded in vitro with antigen stimulation, and supernatants were tested for MPO-ANCA IgG. Bulk B cell receptor (BCR) sequencing of heavy and light chains was performed on autoantigen-specific and control B cells. Serum IgG from the same cohort was used to map MPO epitopes using a yeast surface display library of overlapping MPO-derived peptides. Results: We found that autoantigen-specific B cells were present at similar frequencies in patients with active disease, patients in long-term remission off therapy, and healthy individuals. However, only MPO specific B cells from MPO-ANCA patients produced high levels of MPO-ANCA in vitro, whereas MPO specific B cells from healthy individuals did not. MPO specific B cells from patients exhibited shared immunogenomic features such as reduced clonotype diversity and clonal dominance compared with non-MPO B cells, and there was greater overlap in CDR3 sequences. Epitope mapping revealed that healthy individuals had naturally occurring MPO-reactive IgG, however patient IgG conferred distinct MPO epitope signatures. Discussion: Our findings demonstrate that naturally occurring and disease-associated MPO-ANCA differ in both BCR repertoire features and epitope specificity. These data support the concept that immunological remission in MPO-ANCA vasculitis cannot be defined solely by the presence or absence of ANCAs, but depends on qualitative differences in the autoantibody response, including BCR sequence and epitope targeting.