Longitudinal linear-epitope profiling of SLE identifies novel antigens associated with disease and disease activity
Hurabielle, C.; Wang, Y. V.; Castillo-Rojas, B.; Wang, C.-Y.; Caspar, C.; Asaki, J.; Dayao, M.; Calcoen, M.; Merrill, E. D.; Bodansky, A.; Klepper, A.; Chung, E. N.; Vissers, C.; Zorn, K. C.; Anderson, M. S.; Wilson, M.; Izmirly, P.; Buyon, J.; Yazdany, J.; Dall'Era, M.; Ye, C. J.; DeRisi, J. L.
Show abstract
Systemic lupus erythematosus (SLE) is the prototypical systemic autoimmune disease defined by the production of autoantibodies directed against diverse self-antigens. Antinuclear antibodies (ANA) are central to SLE diagnosis, and anti-double-stranded DNA (anti-dsDNA) antibodies are clinically used to track disease activity. However, these biomarkers provide only partial insight into the full spectrum of autoimmune changes in patients. To expand the known landscape of SLE autoreactivity shared among patients and associated with disease activity, we performed an unbiased characterization of 1,268 SLE plasma samples from 695 SLE patients and 301 healthy controls using phage immunoprecipitation and sequencing (PhIP-seq). This technique preferentially displays linear epitopes, and thus provides a complementary set of autoreactivities to previously discovered autoreactive targets. Despite the marked clinical heterogeneity of SLE, machine learning applied to this data surfaced a shared, reproducible signature that robustly discriminates between SLE and healthy controls in two validation cohorts. Autoreactivities in SLE were significantly enriched for RNA-binding protein targets. Autoreactive proteins containing Basic Acidic Dipeptides (BAD) repeat domains, features associated with biomolecular condensate formation, distinguished patients with high disease activity (by SLEDAI score) from patients with no disease activity. Among all shared autoreactivities, a peptide derived from YTHDF2 emerged as the most prevalent target across SLE cohorts occurring in ~25% of SLE patients and displayed striking sequence similarity to a peptide derived from the Epstein Barr virus (EBV) lytic switch protein BZLF1. Competition assays demonstrated higher-affinity binding to the viral peptide, supporting a model in which antiviral humoral responses to EBV may drive secondary epitope spreading toward molecularly mimicked self-antigens. Remarkably, orthogonal validation using the Luminex bead platform containing a panel of only 10 peptides, including BZLF1, demonstrated strong discrimination between SLE and healthy controls (AUC = 0.95). Together, these findings identify a rich, greatly expanded landscape of autoreactivities in SLE, candidate biomarkers associated with high disease activity, and provide mechanistic insight linking EBV-driven immunity to epitope-specific autoimmunity in lupus.
Matching journals
The top 7 journals account for 50% of the predicted probability mass.
Similar papers in this journal
- Mapping the plasma proteomic architecture of systemic lupus erythematosus 96%
- Single cell transcriptomics reveals distinct effector profiles of infiltrating T cells in lupus skin and kidney 94%
- Patient ancestry significantly contributes to molecular heterogeneity of systemic lupus erythematosus. 94%
Similar papers in this journal
- SAP expressing T peripheral helper cells identify systemic lupus erythematosus patients with lupus nephritis 93%
- Immune responses and disease biomarker long-term changes following COVID-19 mRNA vaccination in a cohort of rheumatic disease patients 93%
- High throughput profiling of the B cell repertoire identifies systematic changes in the repertoire of individuals with Crohn's disease 92%
Similar papers in this journal
- Multimodal repertoire analysis unveils B cell biology in immune-mediated diseases 95%
- Meta-analysis of 208,370 East Asians identifies 113 genomic loci and yields new non-immune cell relevant biological insights for systemic lupus erythematosus 94%
- BNT162b2 vaccine-induced humoral and cellular responses against SARS-CoV-2 variants in Systemic Lupus Erythematosus 93%
Similar papers in this journal
"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.