HIGH-RESOLUTION GENOMEWIDE PROMOTER-FOCUSED CONNECTOME IMPLICATES MICROGLIA CAUSAL GENES FOR ALZHEIMER’S DISEASE

Abstract

Microglia are implicated as playing a central role in the pathogenesis of Alzheimer's Disease (AD). Indeed, genome-wide association studies (GWAS) have identified multiple genomic variants associated with inflammation and microglial immune response. That said, GWAS reports genomic signals associated with a given trait and not necessarily the precise localization of culprit effector genes. High resolution chromatin conformation capture-based techniques that detect contacts between distant regions of the genome offer a valuable tool to understand GWAS signals that reside almost exclusively in non-coding regions, principally through their presumed influence on regulatory elements. To improve on the low resolution of available Hi-C data, we expanded a high resolution Capture-C based method to simultaneously characterize the physical genome-wide interactions of all human promoters - which we applied to the human microglial cell line, HMC3. A custom Agilent SureSelect library was designed to target both ends of DpnII restriction fragments that overlap promoters of both protein-coding and noncoding transcripts, totaling 36,691 RNA baited fragments. Following sequencing, we investigated significant interactions at different resolutions. In parallel, we generated ATAC-seq open chromatin maps to filter for informative proxy (using a liberal r2>0.4) single nucleotide polymorphisms (SNPs) for each of 38 AD common independent sentinel SNPs reported to date beyond the APOE locus. ATAC-seq fine-mapping yielded 43 candidate SNPs in open chromatin for 14 of these loci. By further constraining on our promoter connectome data in HMC3 cells, at both one and four DpnII fragment resolution (median distance between interacting regions 24kb and 122kb, respectively; median region size = 265bp and 1,441bp, respectively), we observed contacts to “open” promoters for 12 putative target genes, relevant to four of the original GWAS loci. These included RTFDC1 at the ‘CASS4’ locus and MADD and PACSIN3 at the ‘CELF1’ locus. We observed informative contacts between proxy SNPs and putative effector genes for ∼11% of AD GWAS loci in the human microglia-specific context. Further efforts in other relevant cell types, where many of the other loci may in fact be principally operating, should shed light on additional signals. Follow-up functional studies are warranted to validate these findings.