LAUSR

Abstract Most disease variants lie within poorly-annotated noncoding parts of the genome, and their functional interpretation is challenging. To functionally assess the relevance of noncoding sequences in neuropsychiatric disorders, we hypothesized that disease-relevant noncoding sequences likely overlap with cell-specific open chromatin regions (OCRs). Neuropsychiatric disorders, including schizophrenia (SZ), afflict a significant fraction of the population. Because SZ GWAS suggests a possible pathogenic role of impaired excitatory neurotransmission, we carried out a genome-wide OCR profiling of excitatory neuronal differentiation from human iPSCs using an Assay for Transposase-Accessible Chromatin by sequencing (ATAC-seq). We found that OCRs were highly dynamic, affecting both core cellular and neuronal-specific genomic locus during iPSC differentiation. Importantly, we identified a genome-wide enrichment of SZ risk variants in neural OCRs and their flanking transcriptional factor binding sites (TFBS), and prioritized putatively functional SZ risk variants that may impact OCRs and consequently, cellular development. At a leading SZ-risk locus flanking MIR137, we narrowed down putatively functional risk variants to a common SNP rs1198588 in neuronal OCR. Excitatory neurons of isogenic iPSCs with CRISPR/Cas9-edited rs1198588 showed altered MIR137 expression, OCR dynamics and dendrite arborization/synapse maturation, which is consistent with the effects of a known functional rare SZ-risk variant at the same MIR137 locus. Our study suggests that OCR profiling in a human iPSC model of neuron differentiation can predict functional noncoding sequences that regulate neurodevelopment, creating a useful resource that can help prioritize functional risk variants of SZ and other mental disorders.