Abstract Text: In autoimmune disease, an individual’s immune system destroys host tissues1-3. Developing a precise understanding of the fine-grained cell states that mediate the genetics of autoimmunity is critical to uncover causal disease mechanisms and develop potentially curative therapies. Here, we leveraged multimodal single-nucleus (sn) snRNA-seq and snATAC-seq data across 28,678 cells from the inflamed synovium of 12 donors with arthritis to identify accessible regions of the genome that covary with gene expression patterns across cells. For 12 immune-mediated and seven control traits, we discovered that cell-state-dependent (“dynamic”) peaks disproportionately captured autoimmune heritability in immune cell types, compared to cell-state-invariant (“invariant”) peaks. These dynamic peaks implicated peripheral helper and regulatory T cell states in rheumatoid arthritis (RA), and dendritic and STAT1+CXCL10+ myeloid cell states in inflammatory bowel disease (IBD). Our study shows that leveraging dynamic regulatory measurements in single cells can pinpoint cell states enriched for disease-critical genetic variation.
