DNA Folding in Chromatin at the Supra-nucleosome Level

基因组功能的变化与染色质空间组织的变化是耦合的。在过去的一年中，我们在核小体水平上获得了关于乳腺癌染色质重组的详细全基因组信息。真核生物的DNA被紧密地包裹在细胞核内，然而它的序列却被许多重要的蛋白质因子有效地识别。为了阐明这种识别的结构机制，人们需要对DNA组织的第二级或30纳米纤维有详细的了解。为了解决这个问题，之前我们计算了具有DNA连接体L = 10 - 70 bp（核小体重复长度，NRL = 157 - 217 bp）的双起始染色质纤维的所有可能构型。因此，我们观察到两种不同的构象家族（即拓扑异构体）以不同的DNA拓扑为特征。纤维的最佳几何形状取决于连接子的长度：连接子L = 10n和10n+ 5bp的纤维每个核小体的DNA连接数分别为-1.5和-1.0。换句话说，DNA超卷绕的水平与染色质中的核小体间距直接相关。这使我们能够解决长期存在的被称为“连接数悖论”的矛盾。我们假设上述染色质纤维的拓扑多态性可能在转录过程中发挥作用，已知转录过程在RNA聚合酶的上游和下游产生不同水平的DNA超卷曲。对酵母基因中NRL分布的全基因组分析证实了这一假设。我们还发现两种纤维拓扑异构体（L = 10n和10n+ 5bp）不仅在平衡构型和平均DNA连接数上存在差异，而且在动力学上也存在差异。特别是，新的10n+5拓扑异构体的特点是可塑性增加，这使得染色质更容易被转录因子（tf）接近。此外，基因组DNA可以通过核小体的部分解包裹而获得。有趣的是，染色质结构、拓扑结构和基因调控之间的这种结构-功能关系是否也存在于人类中。为此，我们彻底研究了乳腺癌（BRC）细胞中的核小体重定位（与英国埃塞克斯大学的V. Teif合作）。我们在来自同一BRC患者的配对肿瘤和正常组织中生成了高分辨率的核小体图谱，并将其与相应的无细胞DNA （cfDNA）进行了比较。肿瘤组织的特征是：(1)关键调控区域的单核小体重定位，(2)全基因组部分核小体解包裹增加，(3)NRL减少5-10 bp。这些影响受核苷酸含量和DNA序列重复的存在所调节，并与不同的DNA甲基化和连接体组蛋白变体H1.4和H1X的结合有关，后者能稳定染色质。对肿瘤组织中NRL缩短的观察揭示了许多关于癌症中核小体到细胞水平重组的重要问题，这些问题需要在未来进行额外的分析。在实践方面，我们首次分析了来自同一患者的配对正常和肿瘤组织中的核小体定位，并为之前在cfDNA中观察到的效果提供了可能的解释，并在临床中进行了经验应用。此外，我们的发现为nrl相关的液体活检分析开辟了新的领域。总之，这项研究使我们能够为基于cfdna的核体分析提供坚实的理论基础，用于患者诊断、监测和分层。描述这些结果的论文已提交给《癌症研究》；参见bioRxiv预印本：https://doi.org/10.1101/2023.04.17.537031。未来的研究需要将这种方法应用于更大的患者群体和更多的癌症类型。因此，我们与W.M. Linehan （NCI泌尿肿瘤学分会）建立了合作，比较几种类型的肾细胞癌（透明细胞癌、肉瘤样肾细胞癌和憎色肾细胞癌）的核小体结构。初步结果表明，与健康组织相比，肿瘤组织中的核小体有显著的重排。重要的是，我们观察到肿瘤细胞的侵袭性水平与核小体定位的变化之间存在很强的相关性，其数值特征是核小体重复长度的减少（在进行中）。

Changes in genome functioning are coupled with the changes in spatial organization of chromatin. In the past year, we obtained a detailed genome-wide information about chromatin reorganization in breast cancer at the nucleosome level. Eukaryotic DNA is tightly packed in the nucleus, nevertheless its sequence is effectively recognized by numerous protein factors essential for regulation. To elucidate structural mechanisms of this recognition one needs to have a detailed information about the second level of DNA organization, or 30-nm fibers. To tackle this problem, earlier we computed all possible configurations of the two-start chromatin fibers with DNA linkers L = 10 - 70 bp (nucleosome repeat length, NRL = 157 - 217 bp). As a result, we observed two different families of conformations (i.e., topoisomers) characterized by different DNA topologies. The optimal geometry of a fiber depends on the linker length: the fibers with linkers L = 10n and 10n+5 bp have DNA linking numbers per nucleosome delta(Lk) = -1.5 and -1.0, respectively. In other words, the level of DNA supercoiling is directly related to the nucleosome spacing in chromatin. This allowed us to resolve the long-standing discrepancy known as the linking-number paradox. We hypothesize that topological polymorphism of chromatin fibers described above may play a role in the process of transcription, which is known to generate different levels of DNA supercoiling upstream and downstream from RNA polymerase. A genome-wide analysis of the NRL distribution in yeast genes confirmed this assumption. We also found that the two fiber topoisomers (with L = 10n and 10n+5 bp) differ not only in their equilibrium configuration and average DNA linking number, but also in their dynamics. In particular, the novel 10n+5 topoisomer is characterized by an increased plasticity, which makes chromatin more accessible to transcription factors (TFs). In addition, genomic DNA is made accessible by partial unwrapping of nucleosomes. It is interesting to see if this structure-function relationship between chromatin structure, topology and gene regulation also exists in humans. To this aim, we thoroughly investigated nucleosome repositioning in breast cancer (BRC) cells (in collaboration with V. Teif, Essex University, UK). We generated high-resolution nucleosome maps in paired tumor and normal tissues from the same BRC patients and compared these with the corresponding cell-free DNA (cfDNA). Tumor tissues were characterized by (1) single-nucleosome repositioning at key regulatory regions, (2) genome-wide increase in partial nucleosome unwrapping and (3) decrease in NRL by 5-10 bp. These effects were modulated by the nucleotide content and the presence of DNA sequence repeats and linked to differential DNA methylation and binding of linker histone variants H1.4 and H1X which stabilize chromatosomes. The observation of NRL shortening in tumor tissues opens a number of important questions about the nucleosome-to-cell level reorganization in cancer that require additional analyses in the future. On the practical side, we analyzed for the first time nucleosome positioning in paired normal and tumor tissues from the same patients and provided a possible explanation for the effects previously observed in cfDNA and used empirically in the clinic. In addition, our findings opened a new venue for NRL-related analyses of liquid biopsies. Taken together, this study allows us to develop solid theoretical foundations for cfDNA-based nucleosomics analysis for patient diagnostics, monitoring and stratification. The paper describing these results has been submitted to Cancer Research; see also bioRxiv preprint: https://doi.org/10.1101/2023.04.17.537031. Future studies are needed to apply this methodology to larger patient cohorts and more cancer types. Therefore, we established a collaboration with W.M. Linehan (Urologic Oncology Branch, NCI) to compare nucleosome landscapes in several types of renal cell carcinoma (clear cell, sarcomatoid and chromophobe renal cell carcinoma). The initial results suggest that there are significant rearrangements of nucleosomes in the tumor tissues (compared to the healthy ones). Importantly, we observe a strong correlation between the level of aggressiveness of the tumor cells and the changes in nucleosome positioning, characterized numerically as a decrease in nucleosome repeat length (in progress).