Interferometric 3D Super-Resolution Imaging and Structure and Stoichiometry Mapping in Living Cells

摘要 我们正处于一个令人兴奋的生物学时代，细胞的内部工作可以通过快速发展来探索 成像方法。荧光显微镜有两大优势：标记的特异性和活细胞 兼容性。然而，它受到衍射的限制，分辨率约为250 nm。最近出现的 单分子开关纳米显微镜(SMSN，也称为Palm/Storm/FPALM)已经克服了这一点 通过随机开启和关闭单一染料的基本限制，使得它们的发射事件 在时间上是分开的。这允许他们的中心位置在空间上以高精度定位，从而导致 重建的超分辨率图像分辨率降至~25 nm。然而，它的生物应用是 受限制的原因有两个：(1)由于时间性差，SMSN应用通常限于固定样本 分辨率和(2)应用仅限于薄样品中接近盖层滑动的结构，因为它 深度方向(Z)分辨率较低，厚样品分辨率迅速下降。此外，SMSN 每个数据集生成数千到数百万个精确的单分子位置-大量 由于缺乏数据量化方法，信息很少被挖掘。克服这些障碍将使 活细胞中纳米结构的可视化和量化，确定荧光的化学计量 标记蛋白质，从而极大地扩展了SMSN的应用范围。 我们建议：(1)发展用于活细胞和厚的超高分辨率成像的干涉型SMSN 通过成像深度高达50微米、各向同性5-10 nm的样品捕捉3D活细胞动力学 (2)在空间、时间和多种颜色上进行结构和化学计量作图，建立高分辨率 活细胞中大分子复合体和大蛋白质组合的三维分辨率模型；以及(3)进一步 将重建模型的分辨率提高了另一个数量级(~1 nm精度) 允许对数千个细胞(每小时约3000个细胞)进行统计量化的内容系统。应用这些 发展，我们将研究三个不同的肌球蛋白在不同的分子组织和功能 活分裂酵母中的胞质分裂和神经元运动集中在活神经元的生长锥上。 这项拟议的研究将首次使细胞的超高分辨率可视化成为可能 厚的和活的样本，允许建立高度分辨的和演变的结构和化学计量模型 体内的大分子组装和蛋白质簇，并根据它们的活细胞进一步分类 背景。这使我们能够确定肌球蛋白分子在体内的组织结构，可视化它们的相互作用 与肌动蛋白网络结合，研究它们在细胞分裂过程中细胞动环内张力产生中的作用。 提议的研究得到了我与自适应公司Martin Booth的密切合作的热情支持 牛津大学光学专家，普渡大学和托马斯·波拉德的丹尼尔·苏特神经元生物学家， 他的研究重点是细胞运动和胞质分裂的分子基础。

Abstract We are in an exciting era of biology where the inner workings of cells can be explored by rapidly developing imaging methods. Fluorescence microscopy has two major advantages: labeling specificity and live cell compatibility. However, it is limited by diffraction to approximately 250 nm resolution. The recent advent of single molecule switching nanoscopy (SMSN, also known as PALM/STORM/FPALM) has overcome this fundamental limit by stochastically switching single dyes on and off such that their emission events are separated in time. This allows their center positions to be localized with high precision in space leading to a reconstructed super resolved image with a resolution down to ~25 nm. However, its biological application is limited for two reasons: (1) SMSN applications are typically limited to fixed samples due to the poor temporal resolution and (2) the application been limited to structures close to the coverslip in thin samples because of its inferior resolution in the depth direction (z) and rapidly deteriorating resolution in thick samples. Further, SMSN generates thousands to millions of precise single molecule positions per dataset - a large amount of information rarely explored due to the lack of data quantification methods. Overcoming these hurdles will allow visualization and quantification of nanostructures in living cells, determine the stoichiometry of fluorescently tagged proteins and thus drastically expand the breadth of SMSN applications. We propose to (1) develop interferometric SMSN for ultra-high resolution imaging in live cells and thick samples capturing 3D live cell dynamics through an imaging depth up to 50 µm with isotropic 5-10 nm resolution; (2) develop structure and stoichiometry mapping in space, time and multiple color to build high- resolution 3D models of macromolecular complexes and large protein assemblies in live cell; and (3) further improve the resolution by another order of magnitude (~1 nm precision) of the reconstructed model by a high- content system allowing statistical quantification over thousands of cells (~3000 cells per hour). Applying these developments, we will study the distinct molecular organization and function of three different myosins during cytokinesis in live fission yeast and neuronal motility focusing on the growth cones in live neuron. The proposed research will, for the first time, make ultra-high resolution visualization of cells possible in thick and live samples, allow building highly-resolved and evolving structure and stoichiometry models of macromolecular assemblies and protein clusters in vivo and further categorizing them based on their live-cell context. This allows us to determine the  organization of myosin molecules in vivo, visualize their interaction with actin network and study their function in tension generation within the cytokinetic ring during cell division. The proposed research is enthusiastically supported by my close collaboration with Martin Booth, adaptive optics expert from Oxford University, Daniel Suter neuron biologist from Purdue University and Thomas Pollard, whose research focuses on molecular basis of cellular motility and cytokinesis from Yale University.