Physical Modelling of Chromatin at Individual Nucleosome Scale

在真核细胞中，遗传信息存储在染色体中，每条染色体都由一种名为染色质的复合体组成，该复合体由DNA和各种结构蛋白组成。这种包装的第一层是核小体，它由八个组蛋白八聚体组成，周围包裹着大约147个碱基对的DNA，大致排列成直径约10纳米的圆盘。在这些核小体之间是游离链DNA的片段，通常长度在20到90个碱基对之间，导致一种类似于串上珠子的结构。然而，人们对下一层包装知之甚少；典型的图景是30 nm的光纤，其中串上的珠子折叠成螺线形或锯齿形，但这些几何模型存在问题，因为连接件长度不一致，因此这种规则有序的排列可能过于简化。事实上，尽管这种结构已经在实验室中被证明是在受控条件下进行的，但在细胞条件更加复杂的体内是否存在这种30 nm的纤维仍存在争议，来自细胞机械的物理和热力学压力以及组蛋白的生化修饰都会导致这些纤维的破坏。DNA包装方式的最大影响之一是它对基因表达的影响。由于真核细胞变得高度专业化，只利用其DNA中包含的一小部分遗传信息，它们需要一种机制来控制哪些片段被转录，哪些片段不被转录。包裹在核小体中的DNA片段不受转录机器的影响，而自由片段则暴露在外面，很容易接触到。此外，一些转录因子更容易与负超螺旋(缠绕下)而不是正超螺旋(缠绕)DNA结合，因此由核小体和蛋白质位置施加的拓扑限制也可能具有重要意义。此外，启动子和增强子与基因相互作用的确切机制还不是很清楚。虽然最近已经建立了核苷酸水平的DNA(分辨双螺旋)和染色质折叠水平的染色体(分辨染色质环和结构域)的模型，但需要一个适用于启动子和增强子大小的中间模型。因此，本项目的目标是建立一个新的粗粒模型，用于单个核小体的染色质纤维的分辨，并使用分子动力学模拟来了解这些核小体在DNA上的位置变化如何在更高水平上改变结构。这将遵循之前不太现实的染色质串珠模型，但增加了复杂性，同时仍保持足够高的计算效率，以模拟足够长的基因组片段，使其具有生物相关性。该模型将考虑自然的核小体进入和离开角度，考虑沿聚合物链轮廓的旋转力，并包括基于现有更高分辨率模型的结果的核小体势的唯象表示。与生物学家合作并使用他们通过实验获得的关于核小体位置的数据，意在能够更准确地预测这些染色质纤维在体内的几何形状和行为，因此应该能够更好地理解基因表达是如何受到增强子和启动子区域DNA包装变化的影响的。

In eukaryotic cells, genetic information is stored in chromosomes, each made from a complex called chromatin comprising of DNA and various structural proteins. The first level of this packaging is the nucleosome, which consists of eight histone octamers around which around 147 base pairs of DNA are wrapped, roughly arranged as a disk of about 10nm in diameter. Between these nucleosomes are sections of free linker DNA, usually varying between 20 and 90 base pairs in length, resulting in a structure analogous to beads on a string. However, less is known about the next level of packaging; the canonical picture is the 30nm fibre, where these beads on a string are folded into either a solenoidal or a zigzagged form, but there are issues with these geometrical models, since linker length is not uniform, so such regularly ordered arrangements may be an oversimplification. Indeed, although this structure has been demonstrated in laboratories under controlled conditions, it is debated whether the 30nm fibre even exists in vivo where cell conditions are much more complex, with physical and thermodynamic stresses from cell machinery and biochemical modifications of the histone proteins both leading to disruption of these fibres.One of the biggest implications of how DNA is packaged is the effect it has on gene expression. Since eukaryotic cells become highly specialised, only making use of a small fraction of the genetic information contained in their DNA, they require a mechanism to control which sections are transcribed and which are not. Sections of DNA wrapped up into nucleosomes are shielded from transcription machinery, whilst free sections are exposed so easily accessible. Adding to this, some transcription factors bind more readily to negatively supercoiled (under-wound) than to positively supercoiled (over-wound) DNA, so topological constraints enforced by nucleosome and protein locations may also have importance. Further, the exact mechanisms of promoter and enhancer interactions with genes are not well understood. Whilst recent models have been developed for DNA at the nucleotide scale (resolving the double helix) and for chromosomes at the scale of chromatin folding (resolving chromatin loops and domains) an intermediate model that is applicable to the sizes of promoters and enhancers is needed.Hence, the aim of this project is to develop a new coarse grained model for chromatin fibres at the resolution of individual nucleosomes, and to use molecular dynamics simulations to understand how varying the positions of these nucleosomes along the DNA changes the structure at higher levels. This will follow from previous, less realistic, beads-on-a-string models of chromatin at this scale, but increase the complexity whilst still keeping the computational efficiency high enough to simulate long enough sections of the genome to be biologically relevant. The model will consider natural nucleosome entry and exit angles, account for rotational forces along the contour of the polymer chain, and include phenomenological representations of the nucleosome potentials based on results from existing higher resolution models. Collaboration with biologists and the use of their experimentally obtained data on nucleosome positions is intended to allow for more accurate predictions of the geometry and behaviour of these chromatin fibres in vivo, and therefore should enable greater understanding of how gene expression is affected by changes to the underlying packaging of DNA at enhancer and promoter regions.