Unravelling the invisible complexities of the genome

罗莎琳德·富兰克林为建立DNA原子结构所做的开创性工作，在很大程度上巩固了我们对“生命分子”的理解，然而，在细胞中，DNA是缠绕和扭曲的，采用复杂的拓扑结构，并且经常保持在超螺旋压力下。缠绕、扭曲和打结对复杂基因组DNA的影响影响到它的功能以及它与分子结构的相互作用。然而，这种分子的复杂性和灵活性意味着，关于缠绕和扭曲的DNA的结构和相互作用的许多方面仍然没有明确的定义。至关重要的是，我们必须提高这种认识，因为构成基因组大部分的复杂DNA结构对我们的健康有巨大的影响：在衰老、癌症和对抗传染病方面。原子力显微镜在其他以亚纳米分辨率工作的结构工具中是独一无二的，其定义特征是在生理温度下对液体中的单分子进行成像的能力，在生理温度下，生物分子是活跃的(并且可以自由探索其天然构象空间)，尽管它被绑在表面上。原子力显微镜使用锐利的探头，在液体中以纳米级的精度‘感觉’吸附在平板上的分子的表面。我的高分辨率原子力显微镜(AFM)方法的独特之处在于，它们能够提供关于DNA结构、功能和动力学的定量信息，而无需标记或平均，我的工作表明DNA的双螺旋结构沿单个分子发生变化。尽管AFM具有在流体中以高分辨率观察单个分子的独特能力，但由于技术的复杂性和对产生的强大数据的有限分析，AFM的广泛采用受到了限制。传统上，大多数AFM分析都是手工进行的，依赖于训练有素和经验丰富的研究人员。当再加上高度依赖操作员专业知识的数据获取时，这意味着原子力显微镜还没有被用作解决目前其他结构生物学工具无法解决的问题的工具，这些工具在这种长度范围内运行。我将率先使用高分辨率原子力显微镜和自动分析来克服这些限制，并揭示DNA结构和构象对DNA-蛋白质相互作用的影响。为了实现这一目标，我将与业界合作，将最先进的原子力显微镜技术发展与新的自动化分析工具结合起来，利用多种机器学习方法，促进对DNA底物的拓扑、结构和构象的跟踪和量化。我将与AFM社区合作，确保所有级别的研究人员都可以使用这些工具，提高他们分析的重复性，并降低这种成像方法的激活能，目前这是一个相当大的进入门槛。使用这些工具，我将确定DNA的结构异质性如何影响其与关键抗生素和抗癌靶点的相互作用：拓扑异构酶；基因编辑工具CRISPR-Cas9；以及G-四链，这是一种具有潜在抗癌新靶点的替代DNA结构。这项计划的重点是具有翻译潜力的系统，使我能够影响制药发展。这一雄心勃勃的计划以我在高分辨率AFM方面的专业知识为基础，并得到包括一家AFM制造商在内的各种跨学科团队的支持，以及互补性单分子生物物理技术、机器学习、分子动力学模拟和生物化学方面的专家。我们将一起揭示基因组中的拓扑应力对整个构象图景中相互作用的隐藏影响。这些知识可以通过帮助合理设计新的治疗方法来与疾病作斗争。

Rosalind Franklin's pioneering work to establish the atomic structure of DNA has underpinned much of our understanding of the 'molecule of life', however in the cell, DNA is tangled and twisted, adopts complex topologies and is frequently maintained under superhelical stress. The effect of coiling, twisting and knotting on complex genomic DNA affects its function and how it interacts with molecular machinery. However the complexity and flexibility of this molecule means that much about the structure and interactions of tangled and twisted DNA remains poorly defined. It is critical that we improve this understanding as complex DNA structures, which make up the majority of the genome, have a huge impact on our health: in aging, cancer and fighting infectious disease. The defining feature of AFM, unique among other structural tools operating at sub-nanometre resolution, is its capacity for imaging single molecules in liquid at physiological temperatures, where biomolecules are active (and free to explore their native conformational space), albeit tethered to a surface. AFM uses a sharp probe to 'feel' the surface of molecules adsorbed on a flat substrate, with nanometre precision in liquid. My high-resolution atomic force microscopy (AFM) methods are unique in their ability to provide quantitative information on DNA structure, function and kinetics without labelling or averaging, demonstrated by my work showing variation in the double-helical structure of DNA along a single molecule. Despite its unique capabilities for observing individual molecules at high resolution in fluid, the widespread adoption of AFM has been limited by the complexity of the technique, and the limited analysis of the powerful data produced. Traditionally, the majority of AFM analysis has been carried out by hand, relying on a highly trained and experienced researcher. When coupled with data acquisition that is highly dependent on the expertise of the operator, this has meant that AFM has not been adopted as the tool that can solve problems currently inaccessible to other tools of structural biology, which operate at this length scale. I will pioneer the use of high-resolution AFM and automated analysis to overcome these limitations and uncover the effect of DNA structure and conformation on DNA-protein interactions. To achieve this I will work in collaboration with industry to combine state-of-the-art atomic force microscopy developments, with new automated analysis tools that facilitate tracing and quantification of the topology, structure and conformation of DNA substrates, using multiple machine learning approaches. I will work with the AFM community to ensure that these tools are available to researchers at all levels, improving the reproducibility of their analysis, and lowering the activation energy for this method of imaging, currently a considerable barrier to entry. Using these tools I will determine how the structural heterogeneity of DNA impacts its interactions with key antibiotic and anti-cancer targets: topoisomerases; the gene editing tool CRISPR-Cas9; and G-quadruplexes, alternative DNA structures with potential as new anti-cancer targets. This programme is focussed on systems with translational potential, to enable me to impact pharmaceutical development. This ambitious programme is underpinned by my expertise in high-resolution AFM and supported by a diverse interdisciplinary team including an AFM manufacturer, and experts in complementary single-molecule biophysics techniques, machine learning, molecular dynamics simulations and biochemistry. Together we will uncover the hidden effect of topological stress in the genome on interactions across its entire conformational landscape. This knowledge can be harnessed to combat disease by aiding in the rational design of novel therapeutics.