Vilela Fernando, Bezault Armel, Rodriguez de Francisco Borja, Sauvanet Cécile, Xu Xiao-Ping, Swift Mark F, Yao Yong, Marrasi Francesca M, Hanein Dorit, Volkmann Niels
Institut Pasteur, Université Paris Cité, CNRS UMR3528, Structural Studies of Macromolecular Machines in Cellulo Unit, F-75015 Paris, France; Institut Pasteur, Université Paris Cité, CNRS UMR3528, Structural Image Analysis Unit, F-75015 Paris, France.
Scintillon Institute, San Diego, CA, USA.
J Struct Biol. 2022 Dec;214(4):107916. doi: 10.1016/j.jsb.2022.107916. Epub 2022 Nov 1.
Nanodiscs have become a popular tool in structure determination of membrane proteins using cryogenic electron microscopy and single particle analysis. However, the structure determination of small membrane proteins remains challenging. When the embedded protein is in the same size range as the nanodisc, the nanodisc can significantly contribute to the alignment and classification during the structure determination process. In those cases, it is crucial to minimize the heterogeneity in the nanodisc preparations to assure maximum accuracy in the classification and alignment steps of single particle analysis. Here, we introduce a new in-silico method for the characterization of nanodisc samples that is based on analyzing the Feret diameter distribution of their particle projection as imaged in the electron microscope. We validated the method with comprehensive simulation studies and show that Feret signatures can detect subtle differences in nanodisc morphologies and composition that might otherwise go unnoticed. We used the method to identify a specific biochemical nanodisc preparation with low size variations, allowing us to obtain a structure of the 23-kDa single-span membrane protein Bcl-xL while embedded in a nanodisc. Feret signature analysis can steer experimental data collection strategies, allowing more efficient use of high-end data collection hardware, as well as image analysis investments in studies where nanodiscs significantly contribute to the total volume of the full molecular species.
纳米圆盘已成为使用低温电子显微镜和单颗粒分析来确定膜蛋白结构的一种常用工具。然而，小膜蛋白的结构确定仍然具有挑战性。当嵌入的蛋白质与纳米圆盘大小相当时，纳米圆盘在结构确定过程中会对对齐和分类产生显著影响。在这些情况下，尽量减少纳米圆盘制剂中的异质性对于确保单颗粒分析的分类和对齐步骤的最大准确性至关重要。在此，我们介绍一种基于分析电子显微镜成像的纳米圆盘颗粒投影的费雷特直径分布来表征纳米圆盘样品的新计算机模拟方法。我们通过全面的模拟研究验证了该方法，并表明费雷特特征可以检测到纳米圆盘形态和组成中的细微差异，否则这些差异可能会被忽视。我们使用该方法鉴定出一种具有低尺寸变化的特定生化纳米圆盘制剂，这使我们能够在纳米圆盘嵌入的情况下获得23 kDa单跨膜蛋白Bcl-xL的结构。费雷特特征分析可以指导实验数据收集策略，从而在纳米圆盘对全分子物种总体积有显著贡献的研究中更有效地使用高端数据收集硬件以及图像分析资源。
