Single-molecule chromatin architectures of human centromeres and telomeres

Centromeres and telomeres are essential structural components of eukaryotic genomes, and alterations in their structure and function contribute to various human diseases. Currently, our understanding of the chromatin architecture underlying centromeres and telomeres is limited owing to their highly repetitive DNA content. To overcome this, we applied long-read single-molecule chromatin fiber-sequencing (Fiber-seq) to a human cell line with a complete telomere-to-telomere (T2T) reference genome, resulting in single-nucleotide precise mapping of chromatin accessibility, nucleosome positioning, and transcription factor occupancy across 1.7 million intact ~16kb chromatin fibers - resolving the chromatin architecture for ~97% of human telomeric and centromeric DNA. We identify that the centromeric kinetochore binding region contains highly atypical chromatin that juxtaposes highly accessible and tightly compacted DNA - a feature unique to the kinetochore. This structure mirrors the alpha-satellite DNA repeat and appears driven via clustered CENP-B occupancy at hypo-CpG methylated repeats. In addition, telomeres often contain several kilobases of non-nucleosomal chromatin, and the transition between nucleosome-bound chromatin and the telomere cap is highly heterogeneous and often punctuated by CTCF-bound elements. Our findings uncover non-standard chromatin architectures as the predominant organizational structure of telomeres and the kinetochore binding region, and provide a framework for future studies of complete human epigenomes. Overall design: Single-molecule chromatin fiber-sequencing (Fiber-seq) of CHM13 cells using a complete reference genome. Single-molecule chromatin fiber-sequencing (Fiber-seq) of GM12878 cells using a complete reference genome.

着丝粒（centromeres）与端粒（telomeres）是真核基因组不可或缺的结构性组分，二者结构与功能的异常会引发多种人类疾病。当前，由于着丝粒与端粒携带高度重复的DNA序列，学界对其背后的染色质结构（chromatin architecture）的认知仍十分有限。
为突破这一研究瓶颈，我们针对搭载端粒到端粒（T2T）完整参考基因组的人类细胞系，应用长读长单分子染色质纤维测序（Fiber-seq）技术，对总计170万条完整的约16kb染色质纤维实现了染色质开放性（chromatin accessibility）、核小体定位（nucleosome positioning）及转录因子结合占据情况的单核苷酸精度精准定位，解析了约97%人类端粒与着丝粒DNA的染色质结构。
我们发现，着丝粒动粒（kinetochore）结合区域存在高度非典型的染色质结构：该区域同时兼具高度开放与高度致密的DNA特征，这一特性为动粒所独有。该结构与α卫星DNA（alpha-satellite DNA）重复序列特征相符，其形成似乎依赖于低CpG甲基化重复序列上成簇分布的CENP-B结合占据。
此外，端粒通常包含数千碱基的非核小体染色质，而核小体结合染色质与端粒帽（telomere cap）之间的过渡区域具有高度异质性，且常被CTCF结合元件所打断。
本研究揭示，非典型染色质结构是端粒与动粒结合区域的主要组织形式，同时为完整人类表观基因组（epigenomes）的后续研究提供了系统性研究框架。
实验总体设计：针对搭载完整参考基因组的CHM13细胞与GM12878细胞，分别开展单分子染色质纤维测序（Fiber-seq）实验。