Modelling Carbohydrate Solution Structure Using a Novel Combined Experimental-Computational Strategy

我们在生物化学方面取得新发现以及开发新药物、先进材料、能源和食品的能力越来越依赖于我们对生物分子在分子水平上如何工作的理解。这反过来又取决于我们对这些生物分子的结构的理解，以及它们的结构如何控制它们的功能。在过去的几十年里，这一直是导致医学、生物学、化学和材料科学取得许多进步的指导原则。这些进展在很大程度上归功于通过X射线结晶学和核磁共振光谱学获得的蛋白质和核酸的结构信息，并彻底改变了我们对生命运行方式的看法。不幸的是，这些技术更难应用于主要的生物分子--碳水化合物。因此，尽管碳水化合物约占地球生物量的99%，并在从藻类到植物再到人类的生命系统中发挥着几乎无限的作用，但我们并不真正理解它们是如何工作的，就像我们对蛋白质和DNA所做的那样。缺乏明确的数据意味着，对于我们如何定义碳水化合物聚合物的结构，仍然存在相当大的争论。为了更好地利用碳水化合物在科学和工业上的巨大潜力，我们必须更详细地了解指导它们组装、组织和与其他分子相互作用的分子原理。这将需要新的方法来研究碳水化合物的结构。为了满足这一迫切的要求，我们将开发一种计算机模拟和光谱实验室相结合的方法来表征碳水化合物的结构，从单糖到已知参与调节生物功能的关键碳水化合物聚合物。该项目的三个组成部分将结合在一起，产生一个独特的、尖锐的糖生物学新工具。首先，高水平的量子化学计算将提供非常详细的光谱，这些光谱对碳水化合物构象的所有方面都很敏感，从而使我们能够识别细微的结构差异。其次，由于最近的工作表明水化作用在碳水化合物构象控制中起着重要作用，我们将使用分子动力学模拟来确定每个碳水化合物在自然存在的溶剂化环境中形成哪些结构，以及它们如何与周围的水分子相互作用。第三，我们将测量非常详细的拉曼光谱，以提供证明我们的计算和建模正确所需的黄金标准基准。这也将为我们提供一个严格的标准来验证新的计算机建模。虽然这种新的计算工具的开发将集中在碳水化合物的结构和行为上，但最终产品也将广泛适用于所有其他生物分子，特别是蛋白质和核酸。我们将克服的挑战是研究人员试图对其他分子的行为进行建模所面临的挑战，例如结构有多稳定？它的组件是如何交互的？与溶剂水或其他分子的相互作用如何影响其形状？由于我们的新型计算工具是通用的，它们将能够为许多其他研究领域提供新的见解，例如蛋白质-配体相互作用和DNA-药物分子结合。

Our ability to make new discoveries in biochemistry and to develop new pharmaceuticals, advanced materials, energy sources and foodstuffs is increasingly dependent on our understanding of how biomolecules work at the molecular level. This in turn depends on our understanding of the structure of these biomolecules, and how their structures control their functions. This has been the guiding principle that has resulted in many of the advances in medicine, biology, chemistry and materials science of the last few decades. These advances have owed much to the structural information that has been obtained for proteins and nucleic acids by X-ray crystallography and NMR spectroscopy, and which has revolutionised our view of how life works. Unfortunately, these techniques are far more difficult to apply to the main class of biomolecules, carbohydrates. As a result, even though carbohydrates constitute around 99% of the biomass of our planet and perform an almost limitless number of roles in living systems, from algae to plants to humans, we don't really understand how they work in the same way that we do for proteins and DNA. The lack of definitive data means there is still considerable debate as to how we even define structure in carbohydrate polymers. In order to best capitalise on the great potential of carbohydrates in both science and industry we will have to understand to a much greater level of detail the molecular principles that govern their assembly, organisation and interactions with other molecules. This will require new approaches to studying carbohydrate structure.In order to meet this urgent requirement we will develop a combined computer modelling and spectroscopic lab-based approach to characterising the structures of carbohydrates, from simple sugars to key carbohydrate polymers known to be involved in regulating biological functions. Three components to the project will combine to generate a uniquely incisive new tool for glycobiology. First, high level quantum chemistry calculations will provide highly detailed spectra that are sensitive to all aspects of carbohydrate conformation, so allowing us to identify subtle structural differences. Secondly, as recent work shows that hydration plays an important role in controlling carbohydrate conformation, we will use molecular dynamics simulations to identify which structures are formed by each carbohydrate in the solvated environments in which they are found naturally and how they interact with the water molecules around them. Thirdly, we will measure highly detailed Raman spectra to provide the gold standard benchmarks required to prove that our calculations and modelling are correct. This will also provide us with a rigorous standard against which to validate the novel computer modelling. Although this development of new computational tools will be focused on the structures and behaviour of carbohydrates, the end product will also be widely applicable to all other biomolecules, particularly proteins and nucleic acids. The challenges we will overcome are those faced by researchers attempting to model how other molecules behave, e.g. how stable is the structure? how do its components interact? how do interactions with solvent water or other molecules affect its shape? Because our novel computational tools are generic they will be able to provide new insights into many other areas of research, such as protein-ligand interactions and DNA-drug molecule binding.