Trans-Ancestral Fine-Mapping and Epigenetic Annotation as Tools to Delineate Functionally Relevant Risk Alleles at <i>IKZF1</i> and <i>IKZF3</i> in Systemic Lupus Erythematosus

• Main Authors: Timothy J. Vyse, Deborah S. Cunninghame Graham
• Format: Article
• Language: English
• Published: MDPI AG 2020-11-01
• Series: International Journal of Molecular Sciences
• Subjects: trans-ancestral fine-mapping; Systemic Lupus Erythematosus; epigenetics; functional annotation
• Online Access: https://www.mdpi.com/1422-0067/21/21/8383

<b>Background:</b> Prioritizing tag-SNPs carried on extended risk haplotypes at susceptibility loci for common disease is a challenge. <b>Methods:</b> We utilized trans-ancestral exclusion mapping to reduce risk haplotypes at <i>IKZF1</i> and <i>IKZF3</i> identified in multiple ancestries from SLE GWAS and ImmunoChip datasets. We characterized functional annotation data across each risk haplotype from publicly available datasets including ENCODE, RoadMap Consortium, PC Hi-C data from 3D genome browser, NESDR NTR conditional eQTL database, GeneCards Genehancers and TF (transcription factor) binding sites from Haploregv4. <b>Results:</b> We refined the 60 kb associated haplotype upstream of <i>IKZF1</i> to just 12 tag-SNPs tagging a 47.7 kb core risk haplotype. There was preferential enrichment of DNAse I hypersensitivity and H3K27ac modification across the 3′ end of the risk haplotype, with four tag-SNPs sharing allele-specific TF binding sites with promoter variants, which are eQTLs for <i>IKZF1</i> in whole blood. At <i>IKZF3</i>, we refined a core risk haplotype of 101 kb (27 tag-SNPs) from an initial extended haplotype of 194 kb (282 tag-SNPs), which had widespread DNAse I hypersensitivity, H3K27ac modification and multiple allele-specific TF binding sites. Dimerization of Fox family TFs bound at the 3′ and promoter of <i>IKZF3</i> may stabilize chromatin looping across the locus. <b>Conclusions:</b> We combined trans-ancestral exclusion mapping and epigenetic annotation to identify variants at both <i>IKZF1</i> and <i>IKZF3</i> with the highest likelihood of biological relevance. The approach will be of strong interest to other complex trait geneticists seeking to attribute biological relevance to risk alleles on extended risk haplotypes in their disease of interest.