A Brief Review of Cryogenic Electron Microscopy

Introduction

Cryogenic electron microscopy (cryo-EM) has revolutionized biological imaging by allowing researchers to generate three-dimensional (3D) models of biological specimens at atomic resolution such as the one shown in Figure 1. This invention was built on decades of development of electron microscopy techniques, which began with the invention of the transmission electron microscope (TEM) in 1933 by Ernst Ruska and Max Knoll. The impetus for cryo-EM came from the field of microbiology searching for a way to look deeper and deeper into biological specimens, but this revolutionary technique finds many applications biology thanks to its ability to protect specimens from the deleterious effects of electron microscopy and provide extremely high resolution images.

Figure 1. “Uncleavable Spike Protein of SARS-CoV-2 in Closed Conformation” mapped by Wrobel AG, Benton DJ, Rosenthal PB, Gamblin SJ and deposited for use in EMDataResource

Biological samples cannot withstand the harsh environment required for standard electron microscopy techniques. The vacuum level required to maintain the electron source and beam coherence evaporates the water from hydrated specimens, and the intensity of the electron beam will burn or destroy organic molecules and cells. The latter issue can be mitigated by reducing the intensity of the electron beam, but the resulting increase in signal noise, a result of the low atomic number elemental makeup of biological samples, significantly reduces the image resolution.

Before the invention of cryo-EM, researchers could work around these limitations by using a technique called negative staining. A high atomic number reagent, such as lead citrate or uranyl acetate, is applied to the sample, and this staining protects the sample by scattering the incident electrons thus protecting the biological materials underneath. This method enabled imaging resolution on the order of 4 nm to 2 nm and is still in use today when cryo-EM is not available or is not needed. However, this method leaves much to be desired. Staining of the sample makes the interior of the sample opaque to the electron beam because the stain scatters the electrons before they penetrate to the inside of the specimen. Furthermore, the staining process can alter the structure of the specimen. For example, Scarff et al. showed a vast difference in the appearance of C-proteins under transmission electron microscopy depending on the method used to apply the stain to the sample [1]. These limitations are overcome by cryo-EM, which has allowed researchers to see irradiation sensitive materials in greater detail than ever before. In this report, I will give a brief overview of the principles of the technology with historical context, the two main modes of cryo-EM, a use case in microbiology as well as uses in other fields, and the limitations of the technique.

Principles of Cryo-EM

Cryo-EM combines a cryogenic sample freezing technique, called vitrification, with a noise-reducing image processing method to enable methods of producing high resolution images and 3D models of biological specimens. These two techniques are framed by the work of Richard Henderson. In 1975, he published the first 3D model of a protein by using TEM crystallography in air, which ignited the pursuit of a higher resolution and more flexible imaging technique. By 1990, Henderson was able to use cryo-EM to produce the first atomic-resolution 3D reconstruction of a protein [2]. This work would not be possible without the two techniques described below.

Figure 2. Graphical representation of the sample vitrification process invented by Dubochet and McDowall and implemented broadly for cryo-EM today. Illustration: © Johan Jarnestad/The Royal Swedish Academy of Sciences

Sample Vitrification

In the 1970’s, researchers identified through a series of publications that cryogenic temperatures protect hydrated organic molecules from the damaging effects of electron beams. Specimens at cryogenic temperatures could withstand as much as ten times the dose of radiation as at room temperature [3]. Glaeser and Taylor recognized in their review of this effect that it could be leveraged to increase the resolution of electron microscopy of biological specimens [4]. In 1982, Dubochet and McDowall published a method of freezing specimens in amorphous ice, called vitrification, that allowed for the preparation, handling, and microscopy of biological samples at cryogenic temperature [5]. In this process, the specimen is suspended in pure water, which is deposited on a TEM mounting grid. The grid is then blotted to reduce the thickness of the water film on the grid to a thin meniscus. The thickness must be on the order of hundreds of nanometers to allow the electron beam to transmit to the detector. The grid is then rapidly immersed into liquid ethane or helium that is kept at liquid nitrogen temperature (-196 ℃, 77 K) such that the water freezes at a rate faster than the rate of ice crystal formation; Any crystals in the ice would diffract some portion of the electron beam and induce artifacts in the images. The sample can then be mounted to a cold stage in the electron microscope for imaging. The specimen can then endure much greater irradiation dosage than it would at room temperature, which allows for a stronger electron beam to be used and consequently results in higher resolution [6].

Noise Reduction Via Image Processing

The vitrification sample preparation process enabled much higher resolution imaging of biological samples than could be previously achieved, but limitations remained. Vacuum pumps powerful enough to prevent the deposition of crystalline ice on the sample during microscopy were not invented until the 1980’s and modern field emission guns enabling illumination coherence 1000 times greater than thermionic sources wouldn’t be available until the 1990’s [2], so researchers employed image processing techniques to clarify their TEM images and extract higher resolution micrographs from the inherent noise.

In 1970, Joachim Frank and colleagues introduced the cross-correlation function as a method of aligning two noisy electron microscopy images [7,8], and in 1976 Frank and lead author Saxton established an empirical relationship between the imaging contrast (c), resolution (d), and the critical irradiation dosage of the specimen (p_crit) to determine the particle size that could resolved, as shown in the equation below.

$D\; =\; 3/c2dp$_crit

This formula related the signal-to-noise ratio to critical aspects of the imaging process and demonstrated that it would be possible to resolve molecular systems on the order of 15-150 nm, which ignited the pursuit of computational methods of image alignment, averaging and reconstruction. By 1981, Frank and his colleagues had developed a custom software suite, called “System for Processing of Image Data from Electron microscopy and Related fields” (SPIDER), for processing electron microscopy images and reconstructing them into 3D models. SPIDER has been upgraded, maintained, and widely distributed to this day [8,9].

Operating Modes of Cryo-EM

Cryo-EM techniques can be divided into two main categories: cryo-electron tomography, and single-particle electron microscopy. Researchers have developed specific variations of these techniques, which are mentioned in the reviews cited in this paper, and I will describe the basics of each technique for brevity and refer you to those reviews for more information.

Cryo-Electron Tomography

Tomography is the process of taking images of several cross sections of an object and reconstructing them into a 3D model. This method can be employed in a TEM to generate 3D reconstructions of biological specimens via cryo-EM. After a sample is properly prepared for cryo-EM, it can be mounted on a rotating stage inside the specimen chamber of the TEM. A series of images at different rotational angles, called a tilt-series, are created, and then computationally combined into a 3D image, called a tomogram [10]. Morphologically heterogeneous specimens can be imaged at resolutions between 5 – 10 nm, and resolution as high as 0.8 nm can be achieved by coupling these results with X-ray diffraction or nuclear magnetic resonance spectroscopy and a specialized computational approach called “constrained single-particle tomography” [11]. For further reading on the development of this method, recent advances, and associated modes, I recommend Koning et al.’s review on the subject [10].

Single Particle Electron Tomography

Single-particle electron microscopy leverages the preponderance of similar or identical specimens usually found in a sample to provide thousands of rotated views of a particle that can be used to generate 3D reconstructions. These individual projections can be extracted from the overall images and then sorted by shape to cluster like orientations. Image processing is then employed to average these like images and remove noise. Then, central projection theorem can be used to relate common features in the averaged projections and assemble a 3D reconstruction of the particle. Highly homogeneous specimens, and specimens with high symmetry, are especially well suited for this method and can be imaged at sub-nanometer resolution [11]. Advances to this method continue thanks to developments in electron detectors and image processing algorithms; In 2020, Nakane et al. demonstrated 0.122 nm resolution single-particle electron microscopy of a mouse apoferritin protein molecule [12]. Joachim Frank’s 30-year retrospective on this method provides a comprehensive history on the evolution of the image processing algorithms enabling single-particle electron microscopy up to 2009 [8], and Cheng et al. provided a comprehensive overview of the technique in 2015 [13].

Example Applications of Cryo-EM

Applications of cryo-EM in microbiology abound and are well reviewed. For example, Milne et al. provide a comprehensive overview for non-microscopists [11] and Luque and Castón reviewed the use of cryo-EM in virus assembly in 2020 [14]. I encourage you to seek out reviews like these for a deeper perspective of cryo-EM applications in microbiology. Here, I will present an extremely timely application of cryo-EM and explore non-biological uses for cryo-EM.

A Timely Application: Imaging SARS-CoV-2 Virus

Cryo-EM has been a crucial tool for learning more about the virus behind the COVID-19 pandemic that spread across the globe in 2020, SARS-CoV-2. Various cryo-EM modes have been used to elucidate the structure of the virus down to the atomic level to discover its functions and vulnerabilities. Klein et al. used cryo-electron tomography to image the RNA of the virus and the compartment in which it is replicated, which offered insights about the budding mechanism of the virus [15]. Wrapp et al. employed single-particle cryo-EM to characterize the spike protein of the virus at the atomic level. The spike protein attaches the virus to human cells and transfers the virus’s genetic material to the host, which causes disease. Understanding its structure has guided the development of vaccines that can block the function of this protein [16]. Cryo-EM has enabled atomic-resolution reconstruction of flexible and heterogeneous protein complexes like these that would have been difficult or impossible via X-ray crystallography [17].

Beyond Biological Samples

Cryo-EM is more than a microscopy technique for microbiologists; It has found applications across scientific disciplines. For example, cryo-EM has enabled high resolution imaging of polymeric materials, small molecules, and other soft materials that are sensitive to radiation damage. The use of cryo-EM to unveil the morphology of block copolymers was reviewed by Zhong and Pochan in 2010 [18], and Newcomb et al. extended this work by reviewing methods of observing block copolymer self-assembly in response to various stimuli, such as Ph or temperature changes [19]. In a 2017 review, Patterson et al. discussed the ability of cryo-EM to shed light on the complexities of crystal nucleation. This method allows researchers to observe transient phenomena in hydrated specimens at atomic resolution, which could not be achieved previously. The authors also noted the ability of the technique to investigate the interfaces of organic and inorganic materials and how the morphology of inorganic materials growing from organic sites can be induced by this interface. This can be used to better understand the growth of bio-mineral materials for biomimetic applications [20].

Limitations of Cryo-EM

The revolutionary capabilities of cryo-EM do not come without some limitations. As mentioned earlier, cryogenic temperatures reduce radiation damage to the samples, but damage is not entirely prevented; Exposure time must still be considered when using cryo-EM to minimize damage. During the vitrification process, the specimen particles tend to be attracted to the water-air interface, which can cause the particles to locate at and align relative to the top and bottom of the water film, which causes resolution anisotropy in the scanned images [21]. The sample thickness is limited to hundreds of nanometers thick so that the electron beam can be transmitted to the receiver, which puts an upper cap on the types of samples that can be imaged via cryo-EM [19]. However, cryo-sectioning techniques, such as diamond knife cutting have been developed to allow for slicing large specimens into thinner sections [10].

Summary

Cryo-EM enables atomic-resolution imaging of radiation sensitive materials by combining the protective properties of cryogenic temperatures with image processing techniques that can identify and extract high resolution images from the signal noise caused by scanning low-atomic number specimens. This technology, for which three scientists earned the Nobel Prize in 2017, has revolutionized microbiological research and has found many applications across materials science and chemistry. Cryo-EM has elucidated the structure, function, and potential vulnerabilities of the SARS-CoV-2 virus among many others and has been used to explore how crystals and inorganic materials nucleate and grow. Advancements in electron microscope technology, such as improved vacuum pumps, electron guns, and electron detectors, growing computational power, and improvements in image processing algorithms have advanced the capabilities of cryo-EM to atomic resolution. This technology has had a profound effect on science since it was developed by Joachim Frank, Jacques Dubochet, and Richard Henderson with the help of innumerable colleagues and predecessors, and their 2017 Nobel Prize is well deserved.

References

[1]        C.A. Scarff, M.J.G. Fuller, R.F. Thompson, M.G. Iadaza, Variations on negative stain electron microscopy methods: Tools for tackling challenging systems, Journal of Visualized Experiments. 2018 (2018) 57199. https://doi.org/10.3791/57199.

[2]        R. Henderson, M. Alsari, History of Cryo-EM, Scientific Video Protocols. 1 (2020) 1–4. https://doi.org/10.32386/scivpro.000022.

[3]        K.A. Taylor, R.M. Glaeser, Electron microscopy of frozen hydrated biological specimens, Journal of Ultrasructure Research. 55 (1976) 448–456. https://doi.org/10.1016/S0022-5320(76)80099-8.

[4]        R.M. Glaeser, K.A. Taylor, Radiation damage relative to transmission electron microscopy of biological specimens at low temperature: a review, Journal of Microscopy. 112 (1978) 127–138. https://doi.org/10.1111/j.1365-2818.1978.tb01160.x.

[5]        J. Dubochet, A.W. McDowall, Vitrification of Pure Water for Electron Microscopy, Journal of Microscopy. 124 (1981) 3–4. https://doi.org/10.1111/j.1365-2818.1981.tb02483.x.

[6]        P. Broadwith, Explainer: What is cryo-electron microscopy, Chemistry World. (2017). https://www.chemistryworld.com/news/explainer-what-is-cryo-electron-microscopy/3008091.article (accessed December 3, 2020).

[7]        J. Frank, Untersuchungen von elektronen-microskopischen Aufnahmen hoher Auflosung mit Bilddefferenz- und Rekonstruktionsverfahren, 1970.

[8]        J. Frank, Single-particle reconstruction of biological macromolecules in electron microscopy-30 years, Quarterly Reviews of Biophysics. 42 (2009) 139–158. https://doi.org/10.1017/S0033583509990059.

[9]        W.O. Saxton, J. Frank, Motif detection in quantum noise-limited electron micrographs by cross-correlation, Ultramicroscopy. 2 (1976) 219–227. https://doi.org/10.1016/S0304-3991(76)91385-1.

[10]       R.I. Koning, A.J. Koster, T.H. Sharp, Advances in cryo-electron tomography for biology and medicine, Annals of Anatomy. 217 (2018) 82–96. https://doi.org/10.1016/j.aanat.2018.02.004.

[11]       J.L.S. Milne, M.J. Borgnia, A. Bartesaghi, E.E.H. Tran, L.A. Earl, D.M. Schauder, J. Lengyel, J. Pierson, A. Patwardhan, S. Subramaniam, Cryo-electron microscopy - A primer for the non-microscopist, FEBS Journal. 280 (2013) 28–45. https://doi.org/10.1111/febs.12078.

[12]       T. Nakane, A. Kotecha, A. Sente, G. McMullan, S. Masiulis, P.M.G.E. Brown, I.T. Grigoras, L. Malinauskaite, T. Malinauskas, J. Miehling, T. Uchański, L. Yu, D. Karia, E. v. Pechnikova, E. de Jong, J. Keizer, M. Bischoff, J. McCormack, P. Tiemeijer, S.W. Hardwick, D.Y. Chirgadze, G. Murshudov, A.R. Aricescu, S.H.W. Scheres, Single-particle cryo-EM at atomic resolution, Nature. 587 (2020) 152–156. https://doi.org/10.1038/s41586-020-2829-0.

[13]       Y. Cheng, N. Grigorieff, P.A. Penczek, T. Walz, A primer to single-particle cryo-electron microscopy, Cell. 161 (2015) 438–449. https://doi.org/10.1016/j.cell.2015.03.050.

[14]       D. Luque, J.R. Castón, Cryo-electron microscopy for the study of virus assembly, Nature Chemical Biology. 16 (2020) 231–239. https://doi.org/10.1038/s41589-020-0477-1.

[15]       S. Klein, M. Cortese, S.L. Winter, M. Wachsmuth-Melm, C.J. Neufeldt, B. Cerikan, M.L. Stanifer, S. Boulant, R. Bartenschlager, P. Chlanda, SARS-CoV-2 structure and replication characterized by in situ cryo-electron tomography, Nature Communications. 11 (2020) 1–10. https://doi.org/10.1038/s41467-020-19619-7.

[16]       D. Wrapp, N. Wang, K.S. Corbett, J.A. Goldsmith, C.-L. Hsieh, O. Abiona, B.S. Graham, J.S. McLellan, Cryo-EM structure of the 2019-nCoV spike in the prefusion conformation, Science. 367 (2020) 1255–1260. https://doi.org/10.1126/science.abb2507.

[17]       S. Subramaniam, COVID-19 and cryo-EM, IUCrJ. 7 (2020) 575–576. https://doi.org/10.1107/S2052252520008799.

[18]       S. Zhong, D.J. Pochan, Cryogenic transmission electron microscopy for direct observation of polymer and small-molecule materials and structures in solution, Polymer Reviews. 50 (2010) 287–320. https://doi.org/10.1080/15583724.2010.493254.

[19]       C.J. Newcomb, T.J. Moyer, S.S. Lee, S.I. Stupp, Advances in cryogenic transmission electron microscopy for the characterization of dynamic self-assembling nanostructures, Current Opinion in Colloid and Interface Science. 17 (2012) 350–359. https://doi.org/10.1016/j.cocis.2012.09.004.

[20]       J.P. Patterson, Y. Xu, M.A. Moradi, N.A.J.M. Sommerdijk, H. Friedrich, CryoTEM as an Advanced Analytical Tool for Materials Chemists, Accounts of Chemical Research. 50 (2017) 1495–1501. https://doi.org/10.1021/acs.accounts.7b00107.

[21]       D. Lyumkis, Challenges and opportunities in cryo-EM single-particle analysis, Journal of Biological Chemistry. 294 (2019) 5181–5197. https://doi.org/10.1074/jbc.REV118.005602.

Acknowledgements

This work is a reproduction, with some grammatical edits and removal of copyright protected images, of a report originally submitted as a portion of the final project for professor Tao Sun’s Fall 2020 Materials Science and Engineering course, Characterization of Materials (MSE 6120) at the University of Virginia. I want to thank professor Tao for inspiring this work and for his excellent lectures throughout the semester.