Developing a one-tube circularized ligation product sequencing (CLP-seq) method for the mapping of 3D genome architecture in small cell populations or single cells.

哺乳动物基因组的三维结构在转录调控中起着关键作用。通过DNA 成环的非编码顺式调节元件可以调节数百个酶以外的靶基因。 由于这种复杂性，生成一个全面的长距离DNA循环相互作用图将有助于 极大地促进了我们对基因组功能的理解。我们以前的工作第一次 证明了在哺乳动物细胞中用30亿~ 50亿Hi-C读段绘制3D基因组是可行的， 分辨率为5- 10 kb。在此分辨率下，单个顺式调节元件之间的相互作用可以被 揭示最近，还测试了单细胞Hi-C方法，以揭示细胞间的变异性， 染色体结构快速发展的3D基因组研究领域要求以 在不同的生理或病原扰动下的多种细胞或组织类型。为了实现 3D基因组作图技术具有广泛的适用性，必须解决以下挑战：（i）能够 分析稀有生物样品;（ii）产生用于数十亿水平的深度测序的高质量文库 （iii）分析大量单细胞以分析复杂组织的能力，或 细胞异质性然而，当使用高分子材料时，来自Hi-C及其衍生物的文库质量通常较差。 起始材料的量小。本提案的总体目标是制定一个简单而有效的 3C-seq方法（环化连接产物测序，或CLP-seq），以生成高质量的 适合于来自少量细胞的超深度测序的文库。与Hi-C及其 CLP-seq是独特的，因为它通过一系列的酶促反应来富集连接连接产物。 酶促反应，而不需要生物素标记和下拉。在一个试点实验中，我们 据估计，这种新方法只需要不到1%的细胞作为起始材料就可以完成测序 在十亿次读取的水平上的深度（比Hi-C提高100倍以上）。此外，由于中电- seq避免了生物素标记和下拉，它适合于单管单细胞CLP的开发。 seq协议（scCLP-seq）用于大规模可扩展的单细胞分析。在这个项目中，我们将建立和 优化这些新技术，并作为原理证明，还产生了大量有价值的 数据资源与这些方法在以下三个目标。在目标1中，我们将优化CLP-seq 产生高复杂性文库以用于从小细胞群体超深度测序的方案，或 罕见的人体组织在目标2中，我们将开发一个完整的CLP-seq数据分析管道来检测 并在双酶分辨率下可视化DNA成环相互作用。我们将生成高分辨率的3D 在一些困难的人类组织中绘制基因组图谱，并对非- 在相关人类疾病中编码GWAS SNP。在目标3中，我们将进一步开发单管SCCLP- 用于同时分析数百个单细胞的SEQ方案。作为测试用例，我们将生成3D 人类胰岛组织中数百个单细胞的基因组数据，并探索执行 使用聚类方法进行亚群分析。我们相信这项技术的进步将扩大 三维基因组研究领域，并最终有利于我们了解基因组功能和人类 疾病

The 3D architecture of mammalian genome plays a key role in transcription regulation. Through DNA looping, non-coding cis-regulatory elements may regulate target genes from hundreds of kilobases away. Because of this complexity, generating a comprehensive map of long-range DNA looping interactions will greatly facilitate our understanding of genome functions. Our previous work for the first time demonstrated that it is feasible to map the 3D genome in mammalian cells with 3~5 billion Hi-C reads, at a resolution of 5-10kb. At this resolution, interactions between individual cis-regulatory elements can be revealed. Recently, single cell Hi-C approach has also been tested to reveal cell-to-cell variability of chromosome structures. The fast growing field of 3D genome research calls for 3D genome maps in a variety of cell or tissue types under different physiological or pathogenic perturbations. In order to achieve broad applicability, 3D genome mapping technology must address the following challenges: (i) Ability to assay rare bio-samples; (ii) Generating high-quality library for deep sequencing at a level of several billion reads; (iii) The ability to analyze a large number of single cells for the analyses of complex tissue or cellular heterogeneity. However, the library quality from Hi-C and its derivatives is usually poor when the amount of starting material is small. The overall goal of this proposal is to develop a simple and efficient 3C-seq method (Circularized Ligation Products sequencing, or CLP-seq) to generate high-quality libraries suitable for ultra-deep sequencing from a small number of cells. In contrast to Hi-C and its derivatives, CLP-seq is unique because it enriches ligation junction products through a series of enzymatic reactions without the need for biotin labeling and pull-down. From a pilot experiment, we estimate that this new method requires less than 1% of cells as starting material to reach sequencing depth at that level of a billion reads (over 100-fold improvement over Hi-C). Furthermore, because CLP- seq avoids biotin labeling and pull-down, it is amenable to the development of a one-tube single cell CLP- seq protocol (scCLP-seq) for massive scalable single cell analysis. In this project, we will establish and optimize these new technologies, and as proof-of-principle, also produce a significant amount of valuable data resources with these methods in the following three aims. In aim 1, we will optimize CLP-seq protocol to generate high-complexity libraries for ultra-deep sequencing from small cell populations or rare human tissues. In aim 2, we will develop a full-package CLP-seq data analysis pipeline to detect and visualize DNA looping interactions at kilobase resolution. We will generate kilobase-resolution 3D genome maps in a few difficult human tissues and perform preliminary functional annotation of non- coding GWAS SNPs in relevant human diseases. In aim 3, we will further develop a one-tube scCLP- seq protocol for simultaneous analysis of hundreds of single cells. As test cases, we will generate 3D genome data in hundreds of single cells from human islet tissues, and explore strategies to perform subpopulation analysis using clustering methods. We believe this technology advance will expand the field of 3D genome study and eventually benefit our understanding of genome functions and human diseases.