A Computational Biology Approach to Mapping Nucleosomes in Stem Cells

描述(由申请人提供)： 多水平的表观遗传调控对于维持胚胎干细胞的多能性状态是必不可少的。组蛋白修饰和DNA甲基化已被证明可以控制ES细胞的干性。尽管核小体定位在表观遗传调控中的重要性，但核小体是否以及如何调控干细胞功能仍然缺乏明确的定义，部分原因是在高等生物中获得高分辨率核小体图谱的技术障碍。最近，我们结合了一种新的化学作图方法和贝叶斯去卷积算法，获得了碱基对分辨率的酵母核小体图谱。这个目前的项目试图扩展这一新的方法，以绘制小鼠基因组的核小体。使用表达工程化组蛋白H4的ES细胞，我们的初步结果证明了为高等生物构建高分辨率核小体图谱的可行性。化学图谱需要在组蛋白H4的第47位引入一种独特的半胱氨酸，以共价方式连接一个硫羟基反应性铜螯合标记。由于组蛋白H4基因在小鼠基因组中存在多个拷贝，这一过程变得复杂起来。因此，我们的首要目标是开发一种化学作图方案，并在培养的哺乳动物细胞中进行全基因组化学作图。根据化学数据确定核小体的中心位置需要对来自局部重叠的核小体的切割信号进行去卷积。我们之前为酵母开发的贝叶斯算法对于鼠标数据来说计算效率很低。因此，我们的第二个目标是开发更高效的计算算法和软件工具。为了实现上述目标，我们将生成多能ES细胞和分化成纤维细胞的核小体化学图谱，并通过RNA-SEQ实验并行量化基因表达。我们将对小鼠基因组核小体定位的全球特征进行高分辨率分析，并研究核小体如何与其他表观调控因子协同调控基因表达。最后，利用化学核小体图谱，我们旨在确定抑制染色质标记H3K27me3对整个基因组中核小体定位的影响。综上所述，这项拟议的工作将以前所未有的细节和准确性描绘胚胎干细胞的核小体图景，为多能细胞状态的表观遗传调控提供新的视角。

DESCRIPTION (provided by applicant): Multiple levels of epigenetic regulation are essential to maintain the pluripotent state of embryonic stem (ES) cells. Histone modification and DNA methylation have been shown to control the stemness of ES cells. Despite the importance of nucleosome positioning in epigenetic regulation, whether and how nucleosomes regulates stem cell functions remains poorly defined, in part due to technical obstacles to obtain high-resolution nucleosome maps in higher organisms. Recently we obtained a yeast nucleosome map at base-pair resolution by combining a novel chemical mapping approach and a Bayesian deconvolution algorithm. This current project seeks to extend this new approach to map nucleosomes for the mouse genome. Using ES cells expressing an engineered histone H4, our preliminary results have demonstrated the feasibility of constructing high-resolution nucleosome maps for higher organisms. The chemical mapping requires introducing a unique cysteine into histone H4 at position 47 to covalently attach a sulfhydryl-reactive copper-chelating label. This procedure is complicated by existence of multiple copies of histone H4 genes in the mouse genome. Thus our first aim is to develop a chemical mapping protocol, and carry out genome-wide chemical mapping in cultured mammalian cells. To define the center positions of nucleosomes based on chemical data requires deconvolution of cleavage signals from locally overlapping nucleosomes. The Bayesian algorithm we developed previously for yeast is computationally inefficient for the mouse data. Thus our second aim is to develop a more efficient computing algorithm and software tools. With above aims achieved, we will generate chemical maps of nucleosomes for both pluripotent ES cells and differentiated fibroblast cells, and quantify the gene expression in parallel by RNA-seq experiments. We will perform high-resolution analysis on global features of nucleosome positioning for the mouse genome, and investigate how nucleosomes regulate gene expression in coordination with other epi-regulators. Lastly, using chemical nucleosome maps we aim to determine the impact of the repressive chromatin mark H3K27me3 on nucleosome positioning throughout the genome. Taken together, the proposed work will delineate the nucleosome landscape of embryonic stem cells in unprecedented details and accuracy, providing insight into a new aspect of epigenetic regulation of the pluripotent cellular state.