Non-Coding Functional SNPs Within the Arthritis-Associated TRAF1-C5 Locus

Abstract

The TRAF1-C5 locus is associated by genome-wide association studies (GWAS) with susceptibility to rheumatoid arthritis and juvenile idiopathic arthritis. Monocytes from healthy individuals with the arthritis-associated risk variant rs3761847 express lower intracellular TRAF1 protein in response to LPS and have greater LPS-induced production of IL-6 and TNF, consistent with a role in inflammatory disease. However, the functional interpretation of this finding remains challenging. Tagging SNPs identified by GWAS are often not causal themselves, but rather simply reside in close association with true functional variants. Further, most GWAS-defined risk loci – including TRAF1-C5 – contain no candidate exonic variants, such that most causal SNPs are believed to operate by modulating the binding of regulatory proteins. This thesis is focused on discovering the causal variant(s) at TRAF1-C5 that modulate TRAF1 expression, and to define the protein-DNA association that drives this mechanism. We screened a library of 132 TRAF1-C5 SNPs in linkage disequilibrium with rs7039505 using SNP-seq, a new technique developed in the mentor’s lab that employs type IIS enzyme restriction and next generation sequencing to identify SNPs that bind proteins from nuclear extract. The 11 candidate functional SNPs identified via this method were tested via electrophoretic mobility shift and luciferase assays in THP1 monocytic cells, revealing allele-specific differences in protein binding capacity and transcription activity at rs7034653, rs10760129 and rs1609810. To further investigate the regulatory mechanism by which these SNPs execute their function, we aim to identify the associated regulatory proteins using Flanking Restriction Enhanced Pulldown (FREP), which takes advantage of flanking restriction to eliminate the non-specific binding protein at both end of the SNP bait sequence, as well as supershift assays with antibodies against transcription factors that recognize consensus sequences potentially altered by the candidate variants. Together, these studies will define the mechanism through which variation at TRAF1-C5 promotes susceptibility to human inflammatory disease.