Longitudinal linear-epitope profiling of SLE identifies novel antigens associated with disease and disease activity

Abstract Systemic lupus erythematosus (SLE) is characterized by diverse autoantibodies, yet established biomarkers capture only a fraction of autoreactivity. We profiled 1,268 plasma samples from 695 patients with SLE and 301 healthy controls using phage immunoprecipitation sequencing (PhIP-seq). Machine learning identified a reproducible autoreactivity signature that discriminated SLE from controls across two validation cohorts. SLE autoreactivities were enriched for RNA-binding proteins, while targets containing Basic Acidic Dipeptide repeat domains were associated with high disease activity. A YTHDF2-derived peptide was the most prevalent shared autoreactivity, occurring in ∼25% of patients and exhibiting sequence similarity to Epstein-Barr virus (EBV) lytic switch protein BZLF1. Competition assays demonstrated higher-affinity binding to the viral peptide, consistent with EBV-associated molecular mimicry. Orthogonal validation using a 10-peptide Luminex panel strongly discriminated SLE from healthy controls. These findings expand the SLE autoantibody landscape, identify disease activity associated autoreactivities and insight linking EBV-driven immunity to epitope-specific autoimmunity in lupus.