Abstract 5239: Integrative analysis of 3D chromatin organization at GWAS loci identifies RAPGEF1 as a melanoma susceptibility gene

• Published in: Cancer research (Chicago, Ill.) Ill.), 2023-04, Vol.83 (7_Supplement), p.5239-5239
• Main Authors: Thakur, Rohit, Xu, Mai, Thornock, Alexandra, Sowards, Hayley, Long, Epring, Rheling, Thomas, Funderburk, Karen, Yin, Jinhu, Hennessey, Rebecca, Chari, Raj, Zhang, Tongwu, Jessop, Lea, Myers, Timothy, Johnson, Matthew E., Wells, Andrew D., Chesi, Alessandra, Grant, Struan F., Iles, Mark I., Landi, Maria T., Law, Matthew, Consortium, Melanoma Meta-Analysis, Machiela, Mitchell, Choi, Jiyeon, Zon, Leonard I., Brown, Kevin M.
• Format: Article
• Language: English
• ISSN: 1538-7445; 1538-7445
• DOI: 10.1158/1538-7445.AM2023-5239

Many GWAS loci occur in non-coding regions and often overlap with gene regulatory elements such as distant enhancers, making functional interpretation and target gene discovery challenging. The 3D chromatin organization brings enhancers in spatial proximity with a promoter to regulate target gene expression. Therefore, to map chromatin interactions between GWAS variants and target gene promoters we performed region-focused chromatin conformation capture assay (Capture-C) in primary human melanocytes. We baited the entire region of association for all 68 independent signals from the recent melanoma GWAS, and Capture-C interactions were called using CHiCAGO tool. Integrative analysis of Capture-C interactions with melanocyte- and melanoma-specific ATAC-sequencing, massive-parallel reporter assay (MPRA), ROADMAP chromatin imputed state model, and gene expression datasets helped prioritize target genes for functional follow-up. Capture-C assays identified physical chromatin interactions between fine-mapped risk variants and candidate causal gene (CCG) promoters at 90% of the GWAS loci; For 84% of the 68 loci, we observed at least one variant-to-gene promoter interaction longer than 100 kb, and for 20% of loci we found interactions beyond 1 Mb. For 76% of the 68 loci, the CCG-interacting variant was in annotated melanocyte or melanoma enhancer regions consistent with CCG regulation via an enhancer-promoter interaction. We observed at least one CCG-interacting variant in 63% and 51% of the 68 loci with distinct allele specific transcriptional activity in melanocyte and melanoma MPRA datasets respectively. A majority of the loci (60%) harbored CCG-linked risk variants in accessible chromatin regions in melanocytes and melanoma. Pathway enrichment analyses of Capture-C-nominated CCGs identified embryonic development, aryl hydrocarbon receptor signaling, and DNA repair pathways. Notably, we observed chromatin interactions between risk variants located near the 3’ of Rap Guanine Nucleotide Exchange Factor 1 (RAPGEF1) to the RAPGEF1 and UCK1 promoter regions. The risk allele of the lead variant (rs3780269) at this locus was associated with higher RAPGEF1 mRNA expression and was not associated with UCK1 expression in melanocytes. We performed a CRISPR knockout proliferation screen in immortalized melanocytes and identified RAPGEF1 as an essential gene for melanocyte growth or survival. Further, we validated the results of this screen by overexpressing RAPGEF1 in immor