Genome-wide Analysis of Chromatin Interactions in Human Cells

Millions of cis-regulatory sequences have recently been found in the human genome, but the function of most cis-elements are not yet clear, in part due to the difficulty in determining their regulatory targets, which are often located millions of base pairs away and separated by one or more unrelated genes. To address this problem, the Hi-C method has been developed to identify long-range looping interactions in a genome-wide, unbiased fashion. However, current data analysis of Hi-C datasets cannot fully resolve regulatory interactions between enhancers and promoters due to the low resolution. Here, we generated a high-depth Hi-C dataset and applied a new analysis method that offers improved resolution permitting genome-wide identification of nearly one million chromatin interactions. We demonstrated the use of Hi-C to identify target promoters of enhancers regulated by NF-κB signaling and signal-dependent dynamic chromatin interaction at these enhancers in human cells. Surprisingly, our results showed that most NF-κB binding sites are looped to their regulatory targets prior to activation of the signaling pathway, and appear to undergo little change during signaling. This observation suggests that the chromatin organization landscape, once established in a cell type, is rather stable and may influence the selection and activation of target genes by a transcription factor. We performed Hi-C analysis using a human fibroblast cell line IMR90 before and after NF-κB activation. In the meantime, we also performed ChIP-seq experiments to map the location of NF-κB p65 subunit, RNA polymerase II, p300, and several histone modifications (including H3K4me1, H3K4me3, H3K27ac and H3K36me3) in IMR90 cells before and after transient TNF-α stimulation. Additionally, to monitor the dynamic transcription profiles, we also performed Global Run-On sequencing (GRO-seq).

近年来，人类基因组中已发现数百万个顺式调控序列（cis-regulatory sequences），但多数顺式作用元件（cis-elements）的功能仍未明确，究其原因，部分在于确定其调控靶标存在较大难度——这些靶标往往相距数百万碱基对，且中间被一个或多个无关基因所分隔。为解决这一难题，学界开发出Hi-C技术（Hi-C），可在全基因组范围内以无偏倚的方式识别远程染色质环相互作用。然而，受限于较低的分辨率，当前Hi-C数据集的数据分析无法完全解析增强子（enhancers）与启动子（promoters）之间的调控相互作用。本研究构建了高深度Hi-C数据集，并应用了一种新型分析方法，该方法可提升分辨率，从而实现全基因组范围内近百万个染色质相互作用的鉴定。本研究证实，可利用Hi-C技术在人类细胞中鉴定受核因子κB（NF-κB）信号通路调控的增强子的靶启动子，以及这些增强子处的信号依赖性动态染色质相互作用。令人意外的是，本研究结果显示，大多数核因子κB结合位点在信号通路激活前就已与其调控靶标形成环化结构，且在信号传导过程中几乎未发生变化。这一观察结果表明，细胞类型一旦确立，其染色质组织景观便相当稳定，且可能影响转录因子（transcription factor）对靶基因的选择与激活。本研究使用人成纤维细胞系IMR90（IMR90），在核因子κB激活前后分别开展了Hi-C分析。与此同时，本研究还进行了染色质免疫共沉淀测序（ChIP-seq）实验，以定位IMR90细胞在瞬时肿瘤坏死因子α（TNF-α）刺激前后，核因子κB p65亚基、RNA聚合酶II（RNA polymerase II）、p300以及多种组蛋白修饰（包括H3K4me1、H3K4me3、H3K27ac和H3K36me3）的结合位点。此外，为监测动态转录谱，本研究还开展了全局转录运行测序（GRO-seq）实验。