Mapping DNA interaction landscapes in psoriasis susceptibility loci highlights KLF4 as a target gene in 9q31

Abstract

Background Genome-wide association studies (GWAS) have uncovered many genetic risk loci for psoriasis, yet many remain uncharacterised in terms of the causal gene and their biological mechanism in disease. This is largely a result of the findings that over 90% of GWAS variants map outside of protein-coding DNA and instead are enriched in cell type- and stimulation-specific gene regulatory regions. Results Here, we use a disease-focused Capture Hi-C (CHi-C) experiment to link psoriasis-associated variants with their target genes in psoriasis-relevant cell lines (HaCaT keratinocytes and My-La CD8+ T cells). We confirm previously assigned genes, suggest novel candidates and provide evidence for complexity at psoriasis GWAS loci. For one locus, uniquely, we combine further epigenomic evidence to demonstrate how a psoriasis-associated region forms a functional interaction with the distant (> 500 kb) KLF4 gene. This interaction occurs between the gene and active enhancers in HaCaT cells, but not in My-La cells. We go on to investigate this long-distance interaction further with Cas9 fusion protein-mediated chromatin modification (CRISPR activation) coupled with RNA-seq, demonstrating how activation of the psoriasis-associated enhancer upregulates KLF4 and its downstream targets, relevant to skin cells and apoptosis. Conclusions This approach utilises multiple functional genomic techniques to follow up GWAS-associated variants implicating relevant cell types and causal genes in each locus; these are vital next steps for the translation of genetic findings into clinical benefit.

Chromatin Looping Links Target Genes with Genetic Risk Loci for Dermatological Traits

“…We found significant overlap between capture Hi-C and HiChIP, although HiChIP identifies significantly more genes that are specifically active in these cell types ( Figure 1 c and d, Supplementary Figures S4 b and S5 a and b, and Supplementary Materials and Methods ). Moreover, we recovered most of the interacting genes highlighted in the previous work, such as IL23A , STAT3 , B3GNT2 , COMMD1 , ERRFI1, and SOX4 ( Ray-Jones et al., 2020 ).…”

“…This is consistent with recent reports about topologically associating domains determining the scope of gene regulation ( Rao et al., 2014 ; Rowley and Corces, 2018 ). We compared this new dataset with our previous study that used capture Hi-C in the same cell lines and targeted GWAS loci from a number of autoimmune disorders ( Ray-Jones et al., 2020 ). We found significant overlap between capture Hi-C and HiChIP, although HiChIP identifies significantly more genes that are specifically active in these cell types ( Figure 1 c and d, Supplementary Figures S4 b and S5 a and b, and Supplementary Materials and Methods ).…”

Transcriptional and signalling regulation of skin epithelial stem cells in homeostasis, wounds and cancer

“…ZNF750 promotes epidermal differentiation by closely associating with Krüppel‐like factor 4 (KLF4), 93,95,96 which is critical for skin barrier formation 97,98 . Of note, mutations of ZNF750 99 and KLF4 100 have been linked to psoriasis, an inflammatory skin disease strongly associated with defects in innate immunity and skin barrier function. These studies share a common theme in which germline mutations of squamous lineage TFs are frequently found in an overlapping spectrum of human ectodermal diseases, suggesting that these TFs are instrumental for early ectoderm specification and subsequently are repurposed to regulate squamous differentiation.…”

Functional genomics in autoimmune diseases

“…HiChIP involves performing chromatin immunoprecipitation (ChIP) on a Hi-C library, using a protein of interest such as H3K27ac, such that the resulting data contain information regarding protein occupancy and interaction frequency ( 48 ). CHi-C and H3K27ac HiChIP have recently been applied to keratinocyte and skin resident cytotoxic T-cell lines, leading to the validation of existing potentially causal genes for Ps, as well as the identification of novel candidates ( 49 ). A more extensive analysis of this data, including other publicly available Hi-C, CHi-C and HiChIP data have identified relevant genes for Ps, PsA and SSc and highlighted the enhanced tissue-specificity of H3K27ac HiChIP data ( 50 ).…”