Methods:  We devised an unbiased high-throughput screen that employs enzymatic restriction to identify genetic variants that allelically modulate the binding of regulatory proteins to DNA. In this method, termed Single Nucleotide Polymorphism-next generation sequencing (SNP-seq), each potentially regulatory variant is engineered into an individual DNA construct. A pool of constructs is then incubated with nuclear extract containing regulatory proteins such as transcription factors. PCR and next generation sequencing are then employed to amplify and quantitate sequences that bind nuclear proteins and are therefore spared from enzymatic degradation. To identify these bound proteins, we applied Flanking Restriction Enhanced Pulldown (FREP, Li…Nigrovic PLOS Genetics 2016;e1006292), an efficient method to employ SNP-containing sequences as bait constructs for protein identification by mass spectrometry.

Results:  We piloted SNP-seq + FREP on the CD40 locus, associated with rheumatoid arthritis, identifying three SNPs that determined expression of this co-stimulator via a previously unrecognized protein complex. We then performed a high-throughput screen across 1223 alleles of 608 SNPs in LD R²>0.8 with 27 JIA loci, using peripheral blood mononuclear cells as protein source, resulting in identification of 148 candidate regulatory variants. Confirmatory testing at the STAT4 locus established two novel regulatory DNA-protein interactions that regulate this T cell-associated gene in a manner implicated in JIA risk.

Conclusion:  We have developed a novel experimental strategy to identify genetic variants that modulate disease risk by altering the binding of transcription factors and other regulatory proteins. Application of this strategy to JIA has allowed us to begin to bridge the gap between GWAS and mechanism in this important but poorly-understood disease.