Center for Neurodegenerative Science, Van Andel Research Institute, Grand Rapids, MI, United States.
Center for Neurodegenerative Science, Van Andel Research Institute, Grand Rapids, MI, United States.
Neurobiol Dis. 2018 Jun;114:53-64. doi: 10.1016/j.nbd.2018.02.007. Epub 2018 Feb 24.
In genome-wide association studies of complex diseases, many risk polymorphisms are found to lie in non-coding DNA and likely confer risk through allele-dependent differences in gene regulatory elements. However, because distal regulatory elements can alter gene expression at various distances on linear DNA, the identity of relevant genes is unknown for most risk loci. In Parkinson's disease, at least some genetic risk is likely intrinsic to a neuronal subpopulation of cells in the brain regions affected. In order to compare neuron-relevant methods of pairing risk polymorphisms to target genes as well as to further characterize a single-cell model of a neurodegenerative disease, we used the portionally-dopaminergic, neuronal, mesencephalic-derived cell line LUHMES to dissect differentiation-specific mechanisms of gene expression. We compared genome-wide gene expression in undifferentiated and differentiated cells with genome-wide histone H3K27ac and CTCF-bound regions. Whereas promoters and CTCF binding were largely consistent between differentiated and undifferentiated cells, enhancers were mostly unique. We matched the differentiation-specific appearance or disappearance of enhancers with changes in gene expression and identified 22,057 enhancers paired with 6388 differentially expressed genes by proximity. These enhancers are enriched with at least 13 transcription factor response elements, driving a cluster of genes involved in neurogenesis. We show that differentiated LUHMES cells, but not undifferentiated cells, show enrichment for PD-risk SNPs. Candidate genes for these loci are largely unrelated, though a subset is linked to synaptic vesicle cycling and transport, implying that PD-related disruption of these pathways is intrinsic to dopaminergic neurons.
在复杂疾病的全基因组关联研究中,许多风险多态性被发现在非编码 DNA 中,并且可能通过等位基因依赖性的基因调控元件差异来赋予风险。然而,由于远端调控元件可以在线性 DNA 上的各种距离改变基因表达,因此大多数风险位点的相关基因身份未知。在帕金森病中,至少一些遗传风险可能是大脑受影响区域中神经元亚群固有的。为了比较将风险多态性与靶基因配对的与神经元相关的方法,并进一步表征神经退行性疾病的单细胞模型,我们使用部分多巴胺能、神经元、中脑衍生细胞系 LUHMES 来剖析基因表达的分化特异性机制。我们比较了未分化和分化细胞的全基因组基因表达与全基因组组蛋白 H3K27ac 和 CTCF 结合区域。虽然启动子和 CTCF 结合在分化和未分化细胞之间基本一致,但增强子大多是独特的。我们将分化特异性增强子的出现或消失与基因表达的变化进行匹配,并通过接近性确定了 22057 个增强子与 6388 个差异表达基因配对。这些增强子至少富集了 13 个转录因子反应元件,驱动了一组参与神经发生的基因簇。我们表明,分化的 LUHMES 细胞而非未分化细胞显示出 PD 风险 SNP 的富集。这些位点的候选基因大多不相关,但一部分与突触小泡循环和运输有关,这意味着 PD 相关的这些途径的中断是多巴胺能神经元固有的。