Cell type-specific enhancer-promoter connectivity maps in the human brain and disease risk association

Abstract

Identification of cell type-specific regulatory elements in the human brain enables interpretation of non-coding GWAS risk variants. AbstractUnique cell type-specific patterns of activated enhancers can be leveraged to interpret non-coding genetic variation associated with complex traits and diseases such as neurological and psychiatric disorders. Here, we have defined active promoters and enhancers for major cell types of the human brain. Whereas psychiatric disorders were primarily associated with regulatory regions in neurons, idiopathic Alzheimer's disease (AD) variants were largely confined to microglia enhancers. Interactome maps connecting GWAS variants in cell type-specific enhancers to gene promoters revealed an extended microglia gene network in AD. Deletion of a microglia-specific enhancer harboring ADrisk variants ablated BIN1 expression in microglia but not in neurons or astrocytes. These findings revise and expand the genes likely to be influenced by non-coding variants in AD and suggest the probable brain cell types in which they function. of neurons, microglia, astrocytes, oligodendrocytes and other cell types that reside within the brain (1-4). In contrast, transcriptional mechanisms that control their developmental and functional properties in health and disease remain less well understood. Studies in animal models have provided deep insights into conserved processes, but significant differences between species often limit the translatability of findings in model systems to humans (5).Genome-wide association studies (GWASs) provide a genetic approach to identifying molecular pathways involved in complex traits and diseases by defining associations between genetic variants and phenotypes of interest (6, 7). Large-scale GWASs have discovered hundreds of single nucleotide polymorphisms (SNPs) that are associated with the risk of neurological and psychiatric disorders. The vast majority of disease-risk genetic variants are located in non-coding regions of the genome (7) and the specific cell type (s) in which the disease-risk variants are active is often unclear. Furthermore, linkagedisequilibrium between disease-risk variants at GWAS significant loci complicates the identification of causal variants. Collectively, these limitations have hindered the assignment and interpretation of disease-risk genes.GWAS-identified risk variants that are located in non-coding regions of the genome can exert phenotypic effects through several mechanisms, including perturbation of transcriptional enhancers and promoters (6). While promoters provide the essential sites of transcriptional initiation of mRNAs, they are frequently not sufficient to direct appropriate developmental and signal-dependent levels of gene expression (8,9). This additional information is provided by enhancers, short regions of DNA that, when bound by transcription factors, enhance mRNA expression from target promoters. Enhancers can reside hundreds of thousands of base pairs away from their target gene and their function is generally consid...

Genome-wide meta-analysis, fine-mapping, and integrative prioritization identify new Alzheimer’s disease risk genes

“…Across the 36 loci, we found 391 colocalisations with at least 80% probability of a shared causal variant between AD and eQTL, representing 80 distinct genes at 27 loci (Supplementary Tables 3-4). The genes implicated by colocalisation include many which have alternative lines of evidence for roles in AD, such as PTK2B 37,38 , BIN1 39,40 , PILRA 41 , CD33 42,43 , and TREM2 44,45 , as well as novel candidates including FCER1G, TSPAN14, APH1B, and ACE. However, the presence of multiple genes with colocalisation evidence within individual loci suggests that additional lines of evidence are important for prioritizing relevant genes.…”

“…There were 21 variants with a mean causal probability above 50% across the fine-mapping methods, and 79 further variants with probabilities from 10 -50% (Table 1 and Supplementary Table 7). These include SNPs near established AD risk genes, such as rs6733839 ~20 kb upstream of BIN1, which has recently been shown to alter a microglial MEF2C binding site 14 and to regulate BIN1 expression specifically in microglia 40 . High- is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint…”

“…Column 'SNP prob' indicates the mean fine-mapping probability for the SNP; the SpliceAI score is the maximum splicing probability for donor gain/loss or acceptor gain/loss, with nonzero values highly enriched for splicing effects; the DeepSEA functional significance score represents the significance above expectation for chromatin feature changes, as well as evolutionary conservation, with lower values more significant. References for specific SNPs are shown 2,7,9,14,40,41,[60][61][62][63][64][65][66][67][68][69] . confidence variants also include a well-known missense SNP in PILRA 41 , and a splicealtering missense SNP in CD33 7 .…”

“…We downloaded bed files based on imputed data for Roadmap Epigenomics DNase, histone peaks, and 25state genome segmentations for 127 epigenomes 53 . We grouped these into groups "all", "brain" (epigenomes 7,9,10,53,54,67,68,69,70,71,72,73,74,81,82,125), and "blood & immune" (epigenomes 33,34,37,38,39,40,41,42,43,44,45,47,48,62,29,30,31,32,35,36,46,50,51,116). For genome segmentations, we considered 9 states to represent enhancers: TxReg, TxEnh5, TxEnh3, TxEnhW, EnhA1, EnhA2, EnhAF, EnhW1, EnhW2.…”

Interleukin-33 coordinates a microglial phagocytic response and limits corticothalamic excitability and seizure susceptibility

“…The background sequences were from the comparing condition as indicated in the figure legends. Chromatin immunoprecipitation-sequencing (ChIP-seq): ChIP for H3K27ac was performed essentially as described previously 82 . In brief, FACS purified cells were fixed with 1% formaldehyde for 10 min at room temperature.…”