Proteome-Wide Analysis of Disease-Associated SNPs That Show Allele-Specific Transcription Factor Binding
A causative role for single nucleotide polymorphisms (SNPs) in many genetic disorders has become evident through numerous genome-wide association studies. However, identification of these common causal variants and the molecular mechanisms underlying these associations remains a major challenge. Differential transcription factor binding at a SNP resulting in altered gene expression is one possible mechanism. Here we apply PWAS (“proteome-wide analysis of SNPs”), a methodology based on quantitative mass spectrometry that enables rapid screening of SNPs for differential transcription factor binding, to 12 SNPs that are highly associated with type 1 diabetes at the IL2RA locus, encoding the interleukin-2 receptor CD25. We report differential, allele-specific binding of the transcription factors RUNX1, LEF1, CREB, and TFAP4 to IL2RA SNPs rs12722508*A, rs12722522*C, rs41295061*A, and rs2104286*A and demonstrate the functional influence of RUNX1 at rs12722508 by reporter gene assay. Thus, PWAS may be able to contribute to our understanding of the molecular consequences of human genetic variability underpinning susceptibility to multi-factorial disease. Genome-wide association studies (GWAS) are a powerful approach to identifying genes contributing to risk of disease. However, individual mapped single nucleotide polymorphisms (SNPs) may not map close to a gene, and it can be difficult to distinguish marker SNPs from causal SNPs. Furthermore, the molecular mechanism responsible for disease association is usually not clear. Here we develop a method termed “proteome-wide analysis of SNPs” (PWAS) that identifies differentially binding transcription factors (TFs) and thereby helps to unravel the molecular mechanisms by which the SNPs may exert their effect on gene regulation. We use quantitative interaction proteomics to identify proteins with allele-specific binding. Applied to fine-mapped SNPs conferring risk in type 1 diabetes, PWAS revealed preferential binding of common transcription factors to certain disease-associated SNPs, suggesting they could be causal. In general, a proportion of causal SNPs are likely to function by mimicking binding motifs for transcription factors, increasing their occupancy and modulating gene expression. In addition, PWAS is streamlined and can be used as an informative follow-up approach to GWAS results.