Regulatory genomic profiling in pure iPSC-derived human cortical neurons

For decades, technical and cost hurdles have prevented the systematic investigation of non-coding sequences in complex human diseases, and thus our knowledge about autism spectrum disorders (ASD) has been primarily obtained from analysis of protein-coding sequences. We have combined the analysis of whole genome sequencing with global studies of regulatory sequences of human cortical neurons to reveal the regulatory architecture of ASD. Analysis of de novo mutations from whole genome sequencing of 261 autism families revealed the physical proximity of ASD de novo mutations specifically to the cortical expression quantitative loci (eQTLs) of synaptic genes. We performed ATAC-Seq, ChIP-Seq, RNA-Seq and Hi-C experiments on human cortical neurons, which for the first time provided a paranormal view of the regulatory landscape in these cells. We found that ASD de novo mutations preferentially affect regulatory elements, and the associated genes are shared targets of two ASD syndromic factors, CHD8 and PTEN. Analyzing 15 chromatin states across 127 human tissue/cell types revealed a significant enrichment of ASD de novo mutations in active transcription start sites and the perturbed genes implicated in neuron functions; this distribution enabled us to develop a machine-learning algorithm to assess potential ASD risk for a given individual. Taken together, our study for the first time revealed the regulatory landscape in human neurons, demonstrated the importance of the non-coding genome in ASD and provides a general framework for analyzing regulatory mutations for other complex human diseases. We performed ATAC-SEQ, RNA-SEQ, in situ HI-C, ChIP-SEQ (H3K27AC and CHD8) in this neuron line, and also CHIP-SEQ for CHD8 in a postmortem brain sample (Brodmann Area 44)

数十年来，技术与成本壁垒长期阻碍了复杂人类疾病中非编码序列的系统性研究，因此我们对自闭症谱系障碍 (autism spectrum disorders, ASD) 的认知主要源自蛋白质编码序列的分析。本研究将全基因组测序分析与人类皮层神经元调控序列的全局研究相结合，揭示了自闭症谱系障碍的调控架构。对261个自闭症家庭的全基因组测序所得新发突变 (de novo mutations) 进行分析后发现，自闭症谱系障碍的新发突变与突触基因的皮层表达数量性状基因座 (expression quantitative loci, eQTLs) 存在显著的物理邻近性。我们在人类皮层神经元中开展了ATAC-Seq、ChIP-Seq、RNA-Seq及Hi-C实验，首次为这些细胞的调控景观提供了全景视角。我们发现，自闭症谱系障碍的新发突变优先靶向调控元件，其关联基因为两种自闭症综合征相关因子CHD8与PTEN的共同靶标。通过对127种人类组织/细胞类型的15种染色质状态进行分析，我们发现自闭症谱系障碍的新发突变在活跃转录起始位点及与神经元功能相关的受扰动基因中显著富集；这一分布特征使我们得以开发机器学习算法，以评估个体潜在的自闭症谱系障碍风险。综上，本研究首次揭示了人类神经元的调控景观，证实了非编码基因组在自闭症谱系障碍发病机制中的重要性，并为分析其他复杂人类疾病的调控突变提供了通用研究框架。我们在该神经元细胞系中开展了ATAC-SEQ、RNA-SEQ、原位Hi-C及针对H3K27AC与CHD8的ChIP-SEQ实验，同时还在一例死后大脑样本 (布罗德曼44区, Brodmann Area 44) 中开展了针对CHD8的ChIP-SEQ实验。