MicroED methodology and procedures

In this section, we present an overview of the MicroED procedures—from sample preparation to structure determination. Because the descriptions are only briefly addressed, we encourage readers to refer to more detailed previously published literature.^{Reference Shi, Nannenga, de la Cruz, Liu, Sawtelle, Calero, Reyes, Hattne and Gonen4,Reference Hattne, Reyes, Nannenga, Shi, de la Cruz, Leslie and Gonen10–Reference de la Cruz, Martynowycz, Hattne and Gonen12}

The procedure for MicroED analysis begins with the preparation of the crystalline material on a carbon-coated electron microscopy (EM) grid. For many samples, especially biological samples, it is important that the hydration of the sample be preserved; therefore, samples are applied to the carbon-coated EM grid, blotted to remove excess solution, and immediately plunged into liquid ethane to vitrify the sample. These procedures are similar to what is commonly used in single particle cryo-EM.^{Reference Passmore and Russo13} For samples where maintaining hydration is not important (e.g., crystals of inorganic materials or some small organic molecules), the vitrification process can be bypassed. In cases where the samples are in the form of dry powders, the sample can be directly added to the grid, shaken to remove excess material, and directly used for MicroED.^{Reference Jones, Martynowycz, Hattne, Fulton, Stoltz, Rodriguez, Nelson and Gonen2,Reference Gruene, Wennmacher, Zaubitzer, Holstein, Heidler, Fecteau-Lefebvre, De Carlo, Muller, Goldie, Regeni, Li, Santiso-Quinones, Steinfeld, Handschin, van Genderen, van Bokhoven, Clever and Pantelic3}

Once the samples have been loaded into the cryo-TEM, they are screened at low magnification to assess the initial quality of the prepared sample. Grids that have many areas with unbroken carbon and microcrystals on the surface are considered good for further screening. The crystals are then screened using a higher magnification where the field of view consists of approximately one grid square (∼50 μm), and promising crystals (well separated from other crystals, with sharp edges) are then used to collect initial diffraction patterns by switching the microscope into diffraction mode. If a crystal shows high-resolution and high-quality diffraction, a full data set is collected from this crystal. The data are collected by continuously rotating the crystal in the electron beam as the high-speed detector records the diffraction data.^{Reference Nannenga, Shi, Leslie and Gonen5} It is important that the search and data collection steps be performed under a low electron dose in the TEM. Doses for data collection are typically on the order of <0.01 e^–/Ų/s, and the dose during searching is even lower than that used during exposure.

Because the diffraction data collected from continuous rotation resemble the data collected by the rotation method in x-ray crystallography, the same data processing programs can be used for MicroED data.^{Reference Nannenga and Gonen1,Reference Nannenga, Shi, Leslie and Gonen5,Reference Clabbers, Gruene, Parkhurst, Abrahams and Waterman9,Reference Hattne, Reyes, Nannenga, Shi, de la Cruz, Leslie and Gonen10} Typically, only a single crystal or a small number of crystals are needed to generate a data set sufficient for structure determination in MicroED. For protein samples, structure solution consists of phasing the data using molecular replacement followed by refinement using electron-scattering factors in standard crystallographic refinement packages. For peptides, small molecules, or materials samples where resolution is generally much higher, phasing can be performed using direct methods followed by structure refinement. The final structures determined by MicroED are of similar quality to those determined by x-ray methods, with the important difference being the size of crystals used, which are orders of magnitude smaller for MicroED compared with what is required by x-ray crystallography.

Hardware and software for MicroED

Microscope choice and alignment are critical for a successful MicroED experiment. For proteins, we recommend the use of an electron microscope with a field-emission electron source and a cryogenic stage. For small molecules and organic and inorganic materials, a field-emission electron source is not imperative as demonstrated by recent MicroED studies.^{Reference Lanza, Margheritis, Mugnaioli, Cappello, Garau and Gemmi14,Reference Zhou, Luo, Luo, Li, Liu and Li15} Additionally, camera choice is important. The dynamic range and speed should be high. The physical size of the chip and pixels are also important. Hybrid pixel detectors such as TimePix and EIGER have a small sensor size and large pixels, which can be detrimental for diffraction from samples with large unit cells, however, these cameras have high speeds and high sensitivity.^{Reference van Genderen, Clabbers, Das, Stewart, Nederlof, Barentsen, Portillo, Pannu, Nicolopoulos, Gruene and Abrahams16,Reference Tinti, Frojdh, van Genderen, Gruene, Schmitt, de Winter, Weckhuysen and Abrahams17} The FalconIII direct electron detector^{Reference Hattne, Martynowycz, Penczek and Gonen18} and Direct Electron DE16 have been used successfully in MicroED experiments and appear promising. They can be shared between different cryo-EM modalities (e.g., electron tomography or single particle cryo-EM) where imaging is used.

Applications of MicroED for high-resolution structure determination

MicroED has been successfully used to study a wide variety of microcrystalline samples (Figure 2). In this section, we present and highlight some key applications of MicroED for protein, peptide, materials, and small molecule research.

Figure 2.

Representative structures determined by microcrystal electron diffraction (MicroED). The MicroED method has been used on a variety of targets, including (a) soluble proteins such as lysozyme^{Reference Nannenga, Shi, Leslie and Gonen5} and (b) membrane proteins such as the NaK ion channel. Asn68 and Gly67 are amino acids where Asn68 = asparagine 68 and Gly67 is glycine 67;³⁴ (c) peptides such as the 11 residue fragment of α-synuclein;^{Reference Rodriguez, Ivanova, Sawaya, Cascio, Reyes, Shi, Sangwan, Guenther, Johnson, Zhang, Jiang, Arbing, Nannenga, Hattne, Whitelegge, Brewster, Messerschmidt, Boutet, Sauter, Gonen and Eisenberg53} (d) small molecules such as carbamazepine;^{Reference Gallagher-Jones, Glynn, Boyer, Martynowycz, Hernandez, Miao, Zee, Novikova, Goldschmidt, McFarlane, Helguera, Evans, Sawaya, Cascio, Eisenberg, Gonen and Rodriguez40} (e) natural products such as VFA-thiaGlu (valine-phenylalanine-alanine-thiaglutamate);^{Reference Ting, Funk, Halaby, Zhang, Gonen and van der Donk50} and (f) nanomaterials such as gold nanoparticles. Representative gold and hydrogen atoms in (f) are indicated by yellow and gray arrows, respectively. The “+” shows the location of the atoms, and the dashed lines indicate the distances between the atoms in the unit cell.^{Reference Vergara, Lukes, Martynowycz, Santiago, Plascencia-Villa, Weiss, de la Cruz, Black, Alvarez, Lopez-Lozano, Barnes, Lin, Weissker, Whetten, Gonen, Yacaman and Calero43}

Summary and future view

MicroED is a relatively new method that is still in its infancy, and yet, in the short six years since the initial demonstration,^{Reference Shi, Nannenga, Iadanza and Gonen27} it has allowed the determination of several samples that were intractable by other methods,^{Reference Nannenga and Gonen1,Reference Xu, Lebrette, Clabbers, Zhao, Griese, Zou and Högbom33,Reference Krotee, Griner, Sawaya, Cascio, Rodriguez, Shi, Philipp, Murray, Saelices, Lee, Seidler, Glabe, Jiang, Gonen and Eisenberg38,Reference Krotee, Rodriguez, Sawaya, Cascio, Reyes, Shi, Hattne, Nannenga, Oskarsson, Philipp, Griner, Jiang, Glabe, Westermark, Gonen and Eisenberg41,Reference Zatsepin, Li, Colasurd and Nannenga51–Reference Guenther, Cao, Trinh, Lu, Sawaya, Cascio, Boyer, Rodriguez, Hughes and Eisenberg54} and is quickly gaining momentum in the scientific community. More than 75 structures have been deposited to the protein data bank and other depositories, and several peer-reviewed publications have resulted. Now is the time to invest in MicroED infrastructure and expertise. Service centers and user facilities similar to those at synchrotrons should be established to provide the scientific community easy, quick, and efficient access to this technology.

The first MicroED dedicated facility has opened at the University of California, Los Angeles; the MicroED Imaging Center. The center is open to both academic labs and for-profit organizations and is operated for MicroED studies similar to a synchrotron facility. Users send their samples to the center and dedicated experienced personnel prepare the cryoEM grids, collect MicroED data, and send the data back to users. Center personnel can also solve structures if needed and provide MicroED training.

As the significance of MicroED grows, we predict that several such dedicated centers will be established worldwide. A good starting point is synchrotrons and National Institute of Health funded National and regional centers that already invest in cryoEM equipment. The same equipment and cameras can be used for MicroED and other cryoEM modalities. All that remains is training, which can be achieved through workshops.