Time-resolved methodologies to provide both spatial and temporal resolution in Electron Microscopy

结构生物学一直是解开蛋白质突变变化秘密的关键，并为治疗干预提供了平台。传统上，x射线晶体学提供了所需的高分辨率信息，但它依赖于有序的晶体，限制了它的范围，特别是膜蛋白和大蛋白复合物。然而，这些系统需要进行结构性调查。另外，电子显微镜（EM）非常适合研究蛋白质复合物和膜蛋白，但传统上受到较低分辨率的阻碍。然而，EM经历了一场“分辨率革命”，由更稳定的显微镜、先进的处理算法、快速的自动数据收集和新的直接电子探测器驱动，这些探测器可以提供更高质量的图像。因此，非对称蛋白质的原子分辨率正在成为标准，以前棘手但重要的医学靶标现在可以使用EM衍生结构进行基于结构的药物设计。EM的一个强大但未被充分利用的优势是能够捕获样品中存在的不同构象态。然而，这目前依赖于在“基态”的样品中存在的不同构象态的计算排序。为了克服这个问题，可以使用时间解决方法。该方法依赖于在液态冷冻剂中冷冻之前快速混合蛋白质和底物，以在ms时间框架内捕获反应中间体。重要的挑战是，要用电子显微镜对蛋白质成像，蛋白质必须被冷冻在玻璃冰中的一个专门的网格上，这种网格薄到足以让电子通过，但不能比被研究的样品更薄。这通常是使用自动“吸光”装置（如Vitrobot）来实现的，该装置从网格中吸去多余的溶液，留下适当的量，以便在快速冻结时形成适当厚度的玻璃状冰层。我们进行时间分辨研究的主要途径是在微流体室中快速混合蛋白质和底物，然后直接将其喷洒到网格上并进行冷冻。我们有一个基本的设置，能够达到50毫秒的时间分辨率，但方法仍处于起步阶段，尚未充分发挥其潜力。这受到两个主要因素的限制，首先是在直接喷涂方法不一致的情况下获得合适的冰厚。通过筛选一系列网格类型，包括最近开发的微纤维网格，我们相信我们可以找到一个有效的解决方案。第二个限制是在获得一个可靠的，可重复的和快速的快速混合装置。通过与汉堡大学的新合作，我们将把为尖端衍射技术（x射线自由电子激光和SAXS）设计的技术转化为开发快速混合单元。通过与领先的网格准备系统（Vitrobot）集成，我们可以开发一个通用的系统，任何感兴趣的新兴市场团体都可以访问，将这项技术从利基推向主流。这笔资金将使我们能够通过产生能够以更可控和可靠的方式产生50毫秒时间分辨率的原型系统，从而迈出重要的第一步。展望未来，结构生物学的下一个重要里程碑将是捕获时间和空间分辨率，极大地增加我们对蛋白质和蛋白质复合物的理解，并使我们摆脱目前在该领域占主导地位的“静态”观点。

Structural biology has been key to unlocking the secrets of mutational change within proteins and provides the platform for therapeutic intervention. Traditionally, X-ray crystallography has provided the high resolution information required, but it is dependent on well-ordered crystals, limiting its scope, especially for membrane proteins and large protein complexes. Yet these systems demand structural investigation. Alternatively, electron microscopy (EM) is ideally suited to studying protein complexes and membrane proteins but has traditionally been hindered by a more modest resolution. However, EM has undergone a "resolution revolution" driven by more stable microscopes, advanced processing algorithms, rapid automatic data collection and new direct electron detectors which offer greater quality images. As a consequence atomic resolution of non-symmetric proteins is becoming the norm and previously intractable but important medical targets are now amenable to structure based drug design using EM derived structures. A powerful but under exploited advantage of EM is the ability to trap different conformational states that exist within the sample. However, this is currently reliant on computational sorting of different conformational states that exist within the sample in the "ground state". To overcome this, time-resolved methodologies can be used. This methodology relies on mixing a protein and substrate rapidly before freezing in a liquid cryogen to trap the reaction intermediate in the ms time frame. The significant challenge is that to image a protein with the electron microscope it must be frozen on a specialised grid in vitreous ice which is thin enough for the electrons to pass through but not thinner than the sample being studied. This is commonly achieved using an automated "blotting" device (such as the Vitrobot), which blots away excess solution from the grid leaving a suitable amount such that upon rapid freezing a vitreous ice layer is formed of an appropriate thickness. Our primary route for time-resolved studies will be to rapidly mix protein and substrate in a microfluidic chamber before directly spraying it onto a grid and plunge freezing. We have a basic setup that is capable of a time resolution >50ms but methodologies are still in their infancy and have yet to realise their full potential. This is limited by two major factors, the first is in obtaining a suitable ice thickness with inconsistencies of the direct spraying approach. By screening a range of grid types including the recently developed micro-fibre grids that have been designed to negate this problem we are confident we can find a working solution. The second limitation is in obtaining a reliable, reproducible and fast rapid mixing device. Through a new collaboration with Hamburg University we will translate the technology designed for cutting edge diffraction technologies (x-ray free electron laser and SAXS) to develop a rapid mixing unit. By integrating with the leading grid preparation system (Vitrobot) we can develop a universal system accessible by any interested EM group, moving this technology from niche to mainstream. This funding will allow us to take the first significant step towards this by generating a prototype system capable of producing time resolution >5ms in a more controlled and reliable manner. Looking forward the next significant milestone in structural biology will be the capture of temporal as well as spatial resolution, greatly increasing our understanding of proteins and protein complexes and moving us away from the "static" view that currently predominates in the field.