An active chromatin interactome elucidates the biological mechanisms underlying genetic risk factors of dermatological conditions in disease relevant cell lines

In the last 15 years Genome wide association studies (GWAS) have uncovered the genetic factors that contribute to disease risk for many disorders, yet the biological significance of many of these associations remain unclear. This is because about 90% of these variants are predicted to affect regulatory elements that are highly cell-type specific and can affect distal genes through chromatin interaction mechanisms and, therefore, understanding their role in disease is challenging. Previous studies have attempted to use chromatin conformation methods to describe the functional mechanism in disease associated loci. However, the majority of these studies have focused on cell lines derived from blood, which might not be the most relevant cell lines for dermatological conditions. Here, we generated HiChIP for H3K27ac, Hi-C and RNA-seq datasets from a keratinocyte and a skin-plaque derived CD8+ T cell line. We studied their active chromatin conformation with the aim to identify biological mechanisms that could be affected by variants associated with skin related diseases: psoriasis (Ps), psoriatic arthritis (PsA), systemic sclerosis (SSc), melanoma and atopic dermatitis. In our analysis we show that HiChIP interactions provide functional evidence for gene regulation by showing that enhancers linked to genes that respond to IFN-γ stimulation are strongly enriched for motifs for TFs related to the stimulation. We also show that by using chromatin conformation we can recall 53% of eQTLs identified by GTEx in skin. In addition to our datasets, we integrated public datasets for a B cell-like lymphoblastoid cell line and primary CD4+ T cells to expand our analysis to include GWAS variants associated with the aforementioned diseases likely to be mediated specifically by B cells and CD4+ T cells (such as in systemic sclerosis and melanoma). We then identified the numerous genes that interact with GWAS variants associated and show that these genes strongly enrich for pathways that are related to the trait studied. For example, the top enriched pathways for autoimmune conditions such as PsA and psoriasis included responses to cytokines and T cell regulation, whereas for melanoma they were related to replicative senescence. We show examples of how our analysis can inform changes in the current description of a few psoriasis associated risk loci. For example, the variant rs10794648, which had previously been assigned to IFNLR1, was linked to GRHL3 in our dataset, a gene essential in skin repair and development. This can indicate a renewed importance of skin related factors in the risk of disease. In conclusion, this study allows us to further characterize disease risk loci by using novel chromatin conformation techniques, which can help identify the genes that are putatively involved in disease and has the potential to identify therapeutic targets.