397 Uncovering causative, noncoding genetic variants in cutaneous diseases

Abstract

Noncoding regions of the human genome are frequently and strongly linked with genetically complex human diseases by genome wide association studies (GWAS). These regions are thought to harbor single nucleotide polymorphisms (SNPs) with gene regulatory (enhancer or repressor) activity that modulate expression of disease mediating genes. Given the large number of SNPs in a linked genomic region, identification of causative variants has been hampered by a lack of methods to rapidly measure the gene regulatory function of numerous SNPs at once. Here we apply a combination of epigenomic filtering and a massively parallel reporter assay (MPRA) to identify causative, noncoding SNPs underlying several cutaneous diseases. Integrating published GWAS data from 8 polygenic skin diseases, including cutaneous cancers and inflammatory diseases with epigenetic profiling of human primary keratinocytes (ATAC-Seq and ChIP-Seq) identified numerous candidate disease-linked SNPs in regions of transcriptionally active chromatin. To functionally characterize the filtered SNPs, we measured allele-specific gene regulatory activity using MPRA – a multiplexed, next-gen sequencing method based on the canonical luciferase-based reporter assay. We performed MPRA with 516 SNP pairs using primary human keratinocytes in distinct states of differentiation and identified disease-linked SNPs with allele- and differentiation state-specific gene regulatory activity. We further linked active SNPs to putative disease-mediating target genes using the GTEx eQTL database and chromatin looping analysis (Hi-ChIP) in primary human keratinocytes. Our results not only help further mechanistic studies for understanding the basis of cutaneous disorders, but also provide a framework for identifying causative, noncoding SNPs in complex genetic diseases.