Alzheimer?s disease (AD) is a devastating neurodegenerative disorder without an effective cure. Recent genome- wide association studies (GWAS) on AD have identified >20 reproducible risk loci, providing an opportunity for understanding novel aspects of AD biology and developing effective interventions. However, each GWAS locus typically spans several genes and many equally associated genome-wide significant index/proxy single- nucleotide polymorphisms (SNPs) that are in strong linkage disequilibrium; it thus remains challenging to identify which genes and risk SNPs are causally involved in AD pathogenesis. Most GWAS risk variants are noncoding and likely regulate gene expression. Because accessible or open chromatin overlaps with cis-regulatory sequences, we hypothesize that many causal AD risk variants modulate chromatin accessibility to transcription factors, thereby altering molecular and cellular phenotypes relevant to AD. We have recently shown that open chromatin profiles in neurons derived from human induced pluripotent stem cells (hiPSCs) can help prioritize regulatory GWAS risk variants of schizophrenia. By directly comparing the quantitative measurements of open chromatin between the two alleles of a heterozygous SNP within the same sample, i.e., allele-specific open chromatin (ASoC) assay, we further showed that schizophrenia GWAS risk variants frequently exhibit ASoC in hiPSC-neurons. More importantly, we found that the neuronal ASoC variants are highly enriched for AD GWAS risk variants and can readily inform putatively regulatory risk variants at three leading AD risk loci (BIN1, CD2AP and CLU). Here, we will extend the ASoC approach to other hiPSC-derived AD-relevant cell types and also harness our expertise in CRISPR genome/epigenome editing and AD cell biology to address three specific questions: (1) Which AD GWAS risk variants are functional? For this, we will map open chromatin by ATAC-seq in glutamatergic and GABAergic neurons, astrocytes and microglia, and search for regulatory AD-risk variants that present ASoC; (2) Which genes are regulated by the putatively functional AD GWAS variants and are thus likely to be causal? For this, we will cost-effectively assay the regulatory effects of all putative AD-risk variants in hiPSC-derived neurons, astrocytes and microglia by combining multiplexed CRISPR/cas9 epigenome editing with single-cell RNA-seq; (3) What are the cellular phenotypic changes caused by the putatively causal AD variants/genes? For this, we will study the cis-regulatory effects of regulatory variants in BIN1, CD2AP and CLU (and other prioritized genes) on AD-relevant biochemical and cellular phenotypes in CRISPR-engineered hiPSC models. This project would impact the field by moving beyond GWAS to decipher causal mechanisms and develop effective treatments.
We will systematically identify functional noncoding Alzheimer?s disease (AD) GWAS risk variants that affect chromatin accessibility and gene expression in order to characterize their functional effects on cellular phenotypes of isogenic hiPSC models. Identifying the functional noncoding AD risk variants/genes and their associated cellular biology in hiPSC-derived neurons/glia will help illuminate disease causal mechanisms, facilitating the clinical translation of AD genetic findings into risk prediction and treatment.