Prioritizing autoimmune-associated genetic variants that alter regulatory element activity in B cells

SUMMARY Genome-wide association studies have discovered hundreds of genetic loci throughout the genome that associate with autoimmune diseases. However, the causal genetic variants that drive autoimmune disease in each locus are for the most part unknown. It is difficult to identify the causal variant(s) in each locus because 1)

there is often tight linkage disequilibrium between causal variants and tens to hundreds of non-causal variants, making it difficult to determine which variant is indeed causal, and 2) because most genetic associations occur in non-coding regions, where their actions on disease-relevant genes and cell types are hard to determine.

However, variants are enriched in non-coding candidate cis-regulatory regions that may modulate gene expression of disease-relevant genes. Autoimmune genetic variants enrich highly in B cell accessible chromatin, indicating that variant modulation of cis-regulatory activity in this cell types may be important for disease

progression. One way to determine the variants that can alter cis-regulatory activity is through testing them in massively parallel reporter assays, which can test thousands of variants in one assay for their ability to alter reporter expression in an allele-specific fashion. This approach is highly effective, as we have found variants that

altered reporter expression enrich up to 58-fold for causal variants (according to statistical fine-mapping). In this proposal, we will apply MPRA in B cells to test ~18,000 variants associated with type 1 diabetes, multiple sclerosis, inflammatory bowel disease, rheumatoid arthritis, and psoriasis for their ability to alter reporter

expression. To ensure B cell MPRA hits are relevant for autoimmune disease, we will assess whether they enrich for causal variants according to statistical fine-mapping. We will rank variants according to their allelic bias effect size and choose 3 high effect variants for further mechanistic follow-up. For each of the 3 variants, we will base

edit variant alleles into primary naïve B cells from healthy donors and observe effects on local B cell gene expression (within 1 Mb of variant), cytokine secretion, and markers of B cell differentiation. For 1-2 variants where we find base editing to alter B cell phenotypes, we will identify homozygous risk and non-risk allele carriers

for both healthy subjects and subjects with autoimmune disease within the Benaroya Research Institute biorepository. We will assess whether variants correlate with altered B cell phenotypes, testing whether risk vs. non-risk variant carriers in health and disease have differences in the percentage of naïve, memory, plasma-

blasts, and plasma cell B cell populations, and corresponding differences in gene expression that may highlight disease-relevant pathways downstream of the variant. Upon completion of this study, we will have exhaustively tested variants for their ability to alter cis-regulatory activity in MPRA in B cells and followed up on high effect

variants to determine their biological relevance to disease. Mechanistic dissection of these prioritized risk variants could provide a wealth of targetable pathways for which to design new therapeutics for autoimmunity.