Systematic Functional Interpretation of Regulatory Variants in Neuropsychiatric Disorders

ABSTRACT Despite the mounting risk loci in genome-wide association studies (GWAS) of neuropsychiatric disorders, identifying the causal variants/genes has been challenging, which hinders the translation of GWAS findings into novel disease biology. A major hurdle is that most risk variants lie in noncoding regions of DNA without easily interpretable function. Noncoding regulatory sequences often reside in open chromatin regions (OCRs). With human induced pluripotent stem cell (hiPSC) neurons as a model in our initial productive R01 period, we have identified abundant regulatory variants in OCRs that affect chromatin accessibility, exhibiting allele- specific open chromatin (ASoC). ASoC SNPs frequently affect gene expression and are strongly enriched for schizophrenia risk variants. However, most causal variants/genes of schizophrenia and other neuropsychiatric disorders remain unknown. Because regulatory variants often act in specific biological context, e.g., cellular stimulation or perturbation, we hypothesize that many neuropsychiatric disease variants may alter chromatin accessibility and gene expression only in activated neurons. Various neuronal stimuli cause membrane depolarization, resulting in robust activity-dependent chromatin and expression changes in mouse neurons. Our pilot data in KCl-depolarized human neurons also showed substantial activity-dependent chromatin and expression changes, notably, with hundreds of activity-dependent ASoC SNPs some of which are schizophrenia risk variants. Leveraging the tractable hiPSC model that can be perturbed in the context of genetic variation, this competitive renewal application will address three specific questions: (1) To what extent genetic variation influences neural activity-dependent chromatin accessibility and gene expression? For this, we will assay cell type-specific chromatin and expression at single-cell resolution in both baseline and activated neurons of a well-powered hiPSC cohort, and perform quantitative trait loci (QTL) mapping to identify activity-dependent ASoC and expression QTL (eQTL). (2) What is the contribution of activity-dependent regulatory variants to neuropsychiatric disorders? For this, we will jointly analyze ASoC and eQTL SNPs with neuropsychiatric GWAS datasets to fine-map causal disease variants that affect activity-dependent chromatin and expression, followed by a multiplex CRISPR base editing to validate their function and cis-target genes. (3) What is the mechanism of activity-dependent chromatin changes? For this, we will examine whether activity- dependent chromatin regions are enriched for specific transcriptional factors (TFs), and explore the effects of CRISPR-editing of TFs on chromatin accessibility, expression, and cellular phenotypes in human neurons. This study will yield novel mechanistic insights into the contribution of neural activity-dependent chromatin and expression changes to neuropsychiatric disorders.