Get 20M+ Full-Text Papers For Less Than $1.50/day. Subscribe now for You or Your Team.

Learn More →

Illuminating the secrets of crystals: microcrystal electron diffraction in structural biology

Illuminating the secrets of crystals: microcrystal electron diffraction in structural biology Downloaded from https://academic.oup.com/biohorizons/article/doi/10.1093/biohorizons/hzy013/5256436 by DeepDyve user on 16 October 2022 BioscienceHorizons Volume 11 2018 10.1093/biohorizons/hzy013 ............................................................................................ ..................................................................... Review article Illuminating the secrets of crystals: microcrystal electron diffraction in structural biology Rob Barringer C.101, Biomedical Sciences, University walk, University of Bristol, United Kingdom, BS8 1TD *Corresponding author: Centre for structure Biology, Sir Ernst Chain Building, Imperial College London, United Kingdom, SW7 2AZ Email: r.barringer@bristol.ac.uk Supervisor: Professor Thomas Meier, Room 501, Sir Ernst Chain building, imperial college London, South Kensington, London, United Kingdom, SW7 2AZ. ............................................................................................ ..................................................................... X-ray crystallography (XRC) has visualised biological macromolecules in exquisite detail for over 50 years, relying on a com- bination of mathematical principles to offer insight into atomic structures. Crystals can diffract various electromagnetic waves aside from the conventional X-ray, offering an alternative approach to crystallographic structural analysis. Microcrystal electron diffraction (MicroED) illuminates crystals with electron waves instead of X-rays. Two specialised groups have demonstrated that MicroED can give high-resolution (often atomic) data, and now appears to be developing into a powerful alternative method to XRC or electron microscopy of macromolecules. How MicroED compares to XRC will be key to assessing it as a stand-alone crystallographic technique. This review presents a critical analysis of MicroED, with comments on theoretical and practical aspects and suggestions of further work and development. Key words: MicroED, crystallography, structural biology, method development Submitted on 23 July 2018; editorial decision on 23 November 2018 ............................................................................................ ..................................................................... the use of complex experimental methods, denoting the field Introduction of ‘structural biology’. There are three main methods used to assess protein structure: Electron microscopy (EM), Nuclear Life is dependent on the ability of cells to perform a myriad of magnetic resonance, and X-ray crystallography (XRC) functions alone or in communities as tissues. When cellular (Curry, 2015), with XRC having the richest history of the processes falter, diseases can arise depending on the aberrant three (Wilkins, 2013). This review focusses on a new crystal- process. Understanding these processes is therefore essential lographic method (microcrystal electron diffraction; to understanding the cellular basis of disease pathology. MicroED) that uses electrons instead of X-rays, outlining a Cellular functions depend on proteins, of which there are brief history of both and then presenting a critical comparison likely over 19 000 in humans (Ezkurdia et al., 2014), each of theoretical and practical aspects. with unique functions and interacting with various biomole- cules (Rolland et al., 2014). Proteins act as nano-scale cellular A short history of crystallography ‘tools’ with functions that are intimately linked to their unique structure; the role that protein architecture plays in XRC has been the primary macromolecular structural meth- biomolecular interactions has a long history, most commonly od for over 50 years (Jaskolski, Dauter and Wlodawer, typified in the mind of the public by the ‘lock-and-key’ 2014). The theory of the method is simple: a pure crystal con- hypothesis (Koshland, 1994). Since proteins are incredibly taining exquisitely ordered repeating units of identical mole- small (for example haemoglobin has a diameter of ~1/200 cules are placed in the path of an X-ray beam, which is 000th of a mm) (Erickson, 2009), structural studies require scattered by the molecules within the crystal to produce a ............................................................................................... .................................................................. © The Author(s) 2019. Published by Oxford University Press. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited. Downloaded from https://academic.oup.com/biohorizons/article/doi/10.1093/biohorizons/hzy013/5256436 by DeepDyve user on 16 October 2022 Review article Bioscience Horizons � Volume 11 2018 ............................................................................................... .................................................................. the molecules within the crystal to solve the atomic structure of the molecules. Electron crystallography X-rays interfere weakly with matter compared to electrons, so XRC requires large volumes of matter create sufficient X-ray dif- fraction before the destructive effects of the radiation degrade the crystal (Henderson, 1995). The damaging effects are because X-rays deposit 1333× more energy than electrons into the crys- tal per elastic scattering (ES) event (single scattering of a wave without loss of amplitude or phase) (80 keV vs 60 eV per ES respectively) but electrons interact more frequently, degrading crystals very quickly (Glaeser, 1971). To accommodate for this challenge, 2-dimensional electron crystallography (2DEC) was developed in the 1970s, using thin crystals that reduce electron interaction, thereby reducing noise and enabling crystal illumin- ation with electrons. The technique successfully determined the Figure 1. The X-ray diffraction pattern of glycine modified mono- structure of purple membrane protein (today known as the acetoacetyl insulin. Protein crystals give unique diffraction patterns light-driven proton pump ‘bacteriorhodopsin’)and theenzyme due to the atomic arrangement of the protein (i.e. the structure of the macromolecule). Each crystal has a characteristic diffraction pattern catalase (Henderson and Unwin, 1975; Unwin and Henderson, that can be analysed to determine the atomic structure of the 1975) and had the unique ability to form 2-dimensional crystals crystallised molecule. Above is the diffraction pattern of glycine of proteins within a lipid bilayer (an important property of modified mono-acetoacetyl insulin, which presents a symmetrical six- native membrane proteins, Gonen et al.,2005; Andrews, sided pattern. This research was originally published in the Biochemical Reichow and Gonen, 2008; Wisedchaisri, Reichow and Gonen, Journal (Lindsay and Shall 1969), reprinted with permission. 2011). During the late 2000s, Jan Pieter Abrahams applied elec- tron crystallography to cryogenic 3-dimensional macromolecu- lar crystals to create a method termed 3-dimensional electron constellation-style pattern (Fig. 1). While beautiful to the crystallography (3DEC), gaining single diffraction patterns from naked eye, the symmetry and ‘brightness’ of each spot of the microcrystals of Lysozyme which had previously proved difficult ‘diffraction pattern’ contains crucial structural information to achieve (Georgieva et al., 2007b). As it turned out, the indi- about the molecules that scattered the X-rays (Sweet, 1985). vidual diffraction patterns provided a challenge as the relative Crystallographers use the information from diffraction pat- orientation of the crystal lattice for each pattern was unknown, terns to elucidate the atomic structures of biomolecules, so merging diffraction data from multiple crystals was difficult. allowing us to gain an atomistic perspective of nature’s tools. Data processing methods were then developed (Jiang et al., X-rays were first used in crystallography by Max von Laue, 2009)(Abrahams, 2010), before improved detectors yielded bet- who hypothesised that X-ray wavelengths are short enough ter signal-to-noise ratios (SNR) (Nederlof et al., 2011). to be diffracted by atoms within a 3-dimensional crystal − −2 −1 Reducing electron dosage (to 0.1 e Å s ) prevented degrad- (Eckert, 2012). Incoming X-rays excite the electrons sur- ation and acquired more diffraction patterns per crystal rounding atoms of the protein molecules in the crystal, and (Nederlof et al.,2013), allowing patterns to be orientated. the X-ray energy is subsequently redistributed in all directions as a wave. In certain directions, the diffracted X-rays of each identical molecule interfere coherently (in phase) to produce The history of MicroED observable diffraction ‘spots’. The Braggs showed that these angles of observation depend on the X-ray wavelength and Low dose electron beams using 3D the spacing between molecules in the crystal (Bragg and microcrystals Bragg, 1913) and that the characteristic ‘spotting’ of diffrac- tion patterns results from these two properties. The spots are MicroED (hereon used as a catch-all term for MicroED/ observed because the scattered X-rays of the repeating mole- 3DEC) was developed by Tamir Gonen, using cryogenically cules only amplify when they overlap with each other in treated 3D Lysozyme microcrystals of 2 μm × 2 μm × 0.5 μm phase, which occurs at discrete angles relative to the crystal to to demonstrate proof of concept (Fig. 2)(Shi et al., 2013). As − 2 create a ‘reciprocal lattice’ of spotting. The brightness of each with Abrahams, Gonen found that after 9e /Å the diffrac- spot results from the cumulative X-ray waves scattered from tion patterns deteriorated, thus by using ~0.5 μM microcrys- − −2 −1 every atom of every molecule, overlapping in unique phases tals and a 0.01e Å s dosage over 10 second exposures at each angle to give an average phase that produces a charac- per angle, they collected 90 diffraction patterns per crystal, teristic spot intensity (i.e. spot ‘brightness’). Each crystal three times more than (Nederlof et al., 2013). A wavelength therefore gives a unique diffraction pattern (Fig. 1), which is of 0.025 Å (with a 200 kV acceleration current) was used, sig- measured and used to determine the exact atomic positions of nificantly shorter than typical XRC wavelengths (~1 Å). ............................................................................................... .................................................................. 2 Downloaded from https://academic.oup.com/biohorizons/article/doi/10.1093/biohorizons/hzy013/5256436 by DeepDyve user on 16 October 2022 Bioscience Horizons � Volume 11 2018 Review article ............................................................................................... .................................................................. Figure 2. A comparison of Lysozyme microcrystals for MicroED (black arrows) and large crystals for XRC (white arrows). Microcrystals are ~6 orders of magnitude smaller than XRC crystals. Microcrystals are ~0.5–1 μM thick with ~55 × 106 unit cells. Crystals are screened by initial illumination to assess diffraction quality. For discussions on Figure 3. Illumination using electrons creates slices (left) through volume limitations, see optimal crystal sizes. Scale bar is 50 μM. This reciprocal space instead of lunes typically observed in XRC (right). research was originally published in eLIFE (Shi et al., 2013). Image has Diffraction patterns of lysozyme are shown. The images show that at undergone a minor edit to include white arrows. an angle perpendicular to a crystal axis, every reciprocal lattice point within the plane was observed using electron sources (with a complete Consequently, the Bragg angles are more acute and show absence of lune formation), while the XRC experiment demonstrated more diffraction spots in a single illumination; the Ewald circular lunes. This research was originally published in Current Opinion sphere (a mathematical tool to determine accessible reciprocal in Structural Biology (Nannenga and Gonen, 2014), published under the lattice points from a single illumination) is so large that it’s CC-BY-NC-ND license. surface is virtually flat, producing slices through reciprocal space instead of the lunes seen in XRC, sampling more lattice matched previous observations that rotation reduces IES/ points (Fig. 3, Nannenga and Gonen, 2014). Crystals were MES contribution to diffraction data (Georgieva et al., rotated 1° between illuminations by tilting the specimen stage, 2007a; Gjonnes et al., 1998; Gemmi et al., 2003), resulting in with each spot averaging 34 observations from multiple continuous rotation becoming standard in MicroED. A sug- angles. Since the slices through reciprocal space do not inter- gestion that larger unit cells (the repeating unit of molecules sect spots perfectly, true spot intensities were difficult to throughout the crystal) and lower symmetry (the orientation derive, leading to loss of intensity data. Crude intensity of proteins within the unit cell) of the static vs CR Lysozyme approximations were made by assuming that the highest spot structure may have increased noise due to lower protein/solv- intensity represented the ‘true’ value. Using this initial ent ratios, increasing solvent contribution to the diffraction approach, the Lysozyme structure was solved to 2.9 Å, using pattern and lowering resolution (see ’Optimal crystal sizes’ molecular replacement (MR, a technique that derives phase section). Further MicroED work used XRC software to gener- information from prior experiments of the same molecule) ate a Catalase structure to 3.2 Å (Nannenga et al., 2014b). (Cipriani et al., 2012). The resolution breakthrough Continuous crystal rotation in MicroED More MicroED macromolecular structures were solved, lead- The Gonen group introduced ‘continuous rotation’ MicroED; ing to protocols outlining data collection and analysis uninterrupted observation of diffraction data during contin- (Hattne et al., 2015, 2016; Shi et al., 2016) including micro- ual crystal rotation, with frame rates of 4 s/frame with 0.09°/s crystal acquisition from macro-crystals (de la Cruz et al., rotation (Nannenga et al., 2014b). The subsequent lysozyme 2017). Two breakthrough studies using MicroED without pattern compared better with XRC diffraction data than complementary X-ray MR data were published: the structure static-MicroED (2.5 vs 2.9 Å respectively, Fig. 4), likely of α-synuclein gained by MR that utilised β-strand motifs as a because (a) rotation allows proper intersection of reciprocal search model for this simple protein (Rodriguez et al., 2015) lattice points with the Ewald sphere and (b) rotation reduces and an Sup35 amyloid core component (GNNQQNY, a hep- beam contact time, lessening inelastic-scattering (IES, where ta-peptide) (Sawaya et al., 2016). This marked a turning point incident electrons are deflected by atomic electrons and for MicroED; solving protein structures without prior XRC impart some energy into the crystal) and multiple elastic- data (ab initio), an achievement previously applied only to scattering (MES, where electrons elastically scatter multiple organic compounds (van Genderen et al., 2016). The first times), both of which interfere with diffracting waves to cre- novel macromolecule structure, TGF-βm; TβRII, was published ate noise in data (see ’MicroED and XRC’ section). This soon after (de la Cruz et al.,2017). More than 17 structures ............................................................................................... .................................................................. 3 Downloaded from https://academic.oup.com/biohorizons/article/doi/10.1093/biohorizons/hzy013/5256436 by DeepDyve user on 16 October 2022 Review article Bioscience Horizons � Volume 11 2018 ............................................................................................... .................................................................. Figure 4. Lysozyme diffraction data of CR MicroED, static-frame MicroED and XRC experiments. The (001) diffraction plane of lysozyme using continuous rotation MicroED (c), static-MicroED (a) and XRC (b) are shown. Continuous rotation MicroED shows the best agreement with the XRC diffraction pattern. The Pearson correlation coefficients shown in (d) were 0.76 between continuous rotation MicroED (c) and X-ray data (b), vs 0.56 for static-frame MicroED (a) and X-ray data (b). Reprinted by permission from Springer Nature: Springer Nature Methods (Nannenga et al., 2014b). have been solved by MicroED (Rodriguez, Eisenberg and suites (EM, tilting stage, carbon grids and liquid ethane vitrifi- Gonen, 2017), with subsequent structures appearing in litera- cation) are sufficient for MicroED. MicroED requires a high ture on a semi-regular basis (Krotee et al.,2017; Gallagher- frame rate camera fast enough to capture individual reflec- Jones et al.,2018; Guenther et al.,2018), many within the sub- tions and prevent spot overlap, and high-speed detectors to angstrom range such as the fused in sarcoma (FUS) protein minimise between-frame readout times, such as direct electron amyloid forming core to 0.73 Å (Luo et al.,2018). Recent detectors (as used previously by the Abrahams group, van reviews highlight MicroED’s history and future improvements Genderen et al., 2016). A previous guide was also outlined (Rodriguez, Eisenberg and Gonen, 2017) and mathematical explaining how to make a device to control continuous crystal and theoretical principles (Clabbers and Abrahams, 2018). tilt (Shi et al., 2016) and which equipment, data acquisition and processing methods are necessary. Briefly, the process involves crystallisation and microcrystal identification, EM/ stage height/tilt calibration and setting up data collection pro- Technical requirements and cesses. Data can now be processed and refined using standard method procedure XRC software, and structural models can be built using COOT, as outlined by Hattne et al., 2015. Thus, aside from Can the typical EM department implement MicroED? A com- an appropriate camera, MicroED is readily available to any prehensive guide with a trouble-shooting section was recently department equipped for single particle analysis, with sup- outlined (Nannenga and Gonen, 2018), explaining that stand- porting protocols and guides to aid researchers. ard equipment used in single particle analysis/typical EM ............................................................................................... .................................................................. 4 Downloaded from https://academic.oup.com/biohorizons/article/doi/10.1093/biohorizons/hzy013/5256436 by DeepDyve user on 16 October 2022 Bioscience Horizons � Volume 11 2018 Review article ............................................................................................... .................................................................. limit for XRC to be 0.5 Å (Jelsch et al., 2000). With more MicroED and XRC development, one might consider it plausible that MicroED To critically assess the efficacy and technical requirements of will surpass XRC. XRC currently achieves higher resolution MicroED, it is useful to compare it with the current leading than MicroED for macromolecules indicating that more methods in XRC. Synchrotron facilities and X-ray free-elec- refinement is required, but understanding how variables affect tron lasers (XFELs) generate continuous or pulsed X-rays use- resolution may significantly close this gap (see ‘Improvements ful for the determination of high-resolution XRC structures. for MicroED’ section). One notable difference between XRC Fortunately, MicroED and XFELs have acquired multiple and MicroED lies in the ‘phase problem’; while XRC rou- lysozyme structures (a molecule often used when developing tinely uses experimental methods to acquire phase informa- and validating crystallographic methods), which offers a use- tion, such methods are challenging in electron ful means of comparison. XFELs have yielded a 0.65 Å lyso- crystallography, requiring a dependence on XRC MR data or zyme structure (Wang et al., 2007) compared to the 1.8 Å ab initio phasing (see ‘Overcoming the phase problem’). A structure from MicroED (de la Cruz et al., 2017), demonstrat- better understanding of the critical MicroED parameters that ing that XRC techniques are currently still superior to affect resolution and development of phasing methods is MicroED for macromolecular crystallography. The difference required to improve resolution. in resolution may be partially due to MicroED’s unique chal- lenges with relation to the manner in which electrons/matter Challenges facing MicroED interactions give rise to IES and MES. Complications arising from IES for XRC/MicroED result from the addition of noise Nannenga et al. (2014b) noted that large crystals may con- (as lower-energy deflected electrons interfere with diffracted tribute more MES events, while separate work predicted that waves) and crystal degradation. It is interesting to note that ≥0.1 μm thickness would give unusable data (Subramanian the ratio of IES/ES is much lower for electron interactions et al., 2015). Surprisingly, proteinase K crystals of 0.1–1 μm than X-ray interactions, implying that with further refine- thickness acquired good resolution when studied by whereby ment, MicroED might have a lower IES-induced noise base- crystal disorder was suggested to explain the disparity line than XRC. Conversely, the fact that electrons interact between the experiment and predictions. Nonetheless, a nega- strongly with matter compared to X-rays creates significantly tive correlation between crystal volume and resolution was more scattering and more MES effects for MicroED, due to observed; thick crystals absorbed electrons more frequently. the weak scattering of X-rays by matter. In conventional This presents a problem; crystals must be large enough for XRC multiple scattering events are rare and presumed negli- observable ES, while minimising IES/MES. IES occurs three gible, with individual diffraction spots derived mostly from times more than ES to add noise, presenting an apparent chal- single ES events (aka ‘kinematical scattering’) rather than lenge (Henderson, 1995). Phasing methods present another more complicated multiple scattering events (aka ‘dynamical challenge; acquiring phases is challenging for X-ray (Taylor, scattering’). While dynamic scattering in rotational electron 2003), electron (Dorset, 1997) and neutron crystallography crystallography can be modelled (Oleynikov and Hovmöller, (Hauptman and Langs, 2003). XRC overcomes this using 2007), the complexity often obfuscates correct data interpret- Isomorphous Replacement (IR: heavy metal soaking of crys- ation, and should be reduced as much as possible. While tals to off-set phases and infer the original phases) and anom- MicroED might present with unique challenges, there are alous scattering (AS: whereby X-ray wavelength is altered to potentially some advantages that MicroED has over XRC. deposit energy into heavy metals which off-set phases to infer MicroED undoubtedly offers a cheaper and more accessible original phases) (Hendrickson and Ogata, 1997), but alternative to high-quality X-ray sources; many research insti- MicroED has not implemented IR/AS so far, an area of tutes have an EM suite on site with short wait times, whereas research that requires further investigation. synchrotron/XFEL facilities tend to only be found in regional hubs that typically require long wait times (Shi et al., 2016). MicroED also requires a single or a few crystals compared to Improvements for MicroED XFELs that require many, often taking time to produce. Crystallisation of macromolecules is often the critical time- Rotation scope consuming bottleneck in XRC studies. When crystallisation MicroED typically uses a ±70° crystal tilt, which becomes fails to produce large crystals (for synchrotron sources), such problematic depending on crystal symmetry; most crystals ‘failed’ conditions often produce MicroED-suitable micro- present with space groups (notations denoting the repeating crystals (Stevenson et al., 2014, 2016). While previously symmetry of proteins within unit cells) of P2 2 2 and P2 1 1 1 1 regarded as a by-product of improper crystallisation, (Wukovitz and Yeates, 1995), giving unique diffraction spots MicroED (and XFEL analysis) allows microcrystals to be use- over 180°.A ±70° tilt limits the accessible data to 140° from ful for structural studies, allowing crystallographers to make a single crystal (commonly known as the ‘missing wedge’ of use of more crystallisation conditions. MicroED methodology EM, Bartesaghi et al., 2008). A larger tilt might improve sam- currently uses wavelengths of ~0.025 Å compared to ~0.55 Å pling, and efforts to create innovative solutions (such as those in the highest resolution XRC structure, which is Cambrin attempted by Barnard et al., 1992) could be explored in the (Schmidt et al., 2011), with some considering the wavelength future. ............................................................................................... .................................................................. 5 Downloaded from https://academic.oup.com/biohorizons/article/doi/10.1093/biohorizons/hzy013/5256436 by DeepDyve user on 16 October 2022 Review article Bioscience Horizons � Volume 11 2018 ............................................................................................... .................................................................. Optimal crystal sizes Reducing electron dosage – electron wavelength Crystals require sufficient volume to provide signal, while excessively large crystals generate significant MES/IES, The incident beam might be adjusted to reduce electron dos- prompting to suggest an upper limit of 500 μm for optimal age; electrons with shorter wavelengths degrade crystals diffraction. Microcrystals can be acquired from larger slower and improve longevity (Glaeser, 1971). While (>500 μm) crystals (de la Cruz et al., 2017), but understand- MicroED uses an electron wavelength of 0.025 Å, no work ing the volume/resolution relationship is necessary. assessing wavelength effect on ES/IES/MES has been pub- Interestingly, the highest resolution data currently derives lished. Electron exposure was assessed in proteinase K and from small unit cell crystals with complex symmetry (see GSNQNNF crystals; high-resolution spots degraded at the Table 1), suggesting that perhaps the protein atomicity of a 0.025 Å wavelength (likely due to their typical low signal) crystal (i.e. protein molecules per Å ) or high solvent content with high-resolution data (<2.0 Å) being severely degraded at − −2 (e.g. when protein atomicity is low) contributes to volume/ ≥3e Å (Hattne et al., 2018). At low doses electrons were resolution relationships. It is already known that solvents gen- absorbed and shielded atoms from proper ES, e.g. at ≥0.9 e −2 − −2 erate noise (Bragg and Perutz, 1952)(Fraser, MacRae and Å disulphide bridges deteriorated and broke at 5.78 e Å , Suzuki, 1978), and are typically disordered and not uniformly while glutamate/aspartate residues lost carboxyl density at − −2 − −2 oriented in crystals (Weichenberger et al., 2015), which could ≥2e Å , which was absent at 5 e Å , meaning electron increase noise when protein atomicity is low (since solvents dosage contributes to structural quality and is an important occupy more volume). When reviewing MicroED structures element to control. Higher frequency electrons interact with (Table 1), an observable trend between protein atomicity and matter less frequently (Birkhoff, 1958); further research could resolution appears (Fig. 5), however this compares structures build on the work of Hattne et al. (2018) to determine over years from many laboratories, so variability is likely sig- whether shorter wavelengths might reduce interactions and nificant. Interestingly, Sawaya et al. (2016) analysed IES/MES to improve resolution. GNNQQNY in two different symmetries; the higher sym- metry crystal gave better resolution, providing insight into Overcoming the phase problem atomicity/resolution relationships. Further research is essen- tial to characterise the relationship between ES/MES/IES and 2DEC IR techniques have been attempted, but shielding of crystal volume and atomicity. metal nuclei by large electron clouds weakened IR interfer- ence (Ceska and Henderson, 1990), however some argued Reducing electron dosage – crystal rotation that heavy metal phase contributions can theoretically give sufficient information to solve phases (Burmester and speed Schröder, 1997). While MicroED has successfully used direct Electrons deposit less IES per ES than X-rays and less energy per methods to process data (Sawaya et al., 2016; de la Cruz IES, but interact more frequently with matter leading to quicker et al., 2017; Vergara et al., 2017), these computational meth- crystal degradation (Henderson, 1995). Electron dosage is kept ods require resolution to be 1.2 Å or better, (aka ‘Sheldricks below a critical threshold (Shi et al.,2013) and continuous rota- rule’, Sheldrick, 1990; Morris and Bricogne, 2003). tion reduces electron exposure to different crystal locations Consequently, this method is currently limited to small mole- (Nannenga et al., 2014a). Rotation speeds and frame rates are a cules, which tend to give the highest resolution data (Taylor, compromise between adequate sampling of diffraction spots 2010) as seen in the Sawaya et al. (2016) publication. Direct while preventing spot overlap (Hattne et al.,2015). methods have been used on macromolecular data of ~2 Å in XRC by prospective MR using archetypal α-helices/β-sheets A faster rotation with a shorter frame rate might reduce as search models, solving the structure of a previously electron dosage while preventing spot overlap. Rotation unknown 111-residue protein (Rodriguez et al., 2009). While speeds varied over experiments, from 0.1°/s (Vergara et al., useful, this may not be appropriate for macromolecules lack- 2017) to 0.29 /s (de la Cruz et al., 2017), but Table 1 shows ing sufficient α-helix and β-sheet structures. XRC also over- that resolution vs rotation speed (Fig. 6) and frame scope (the comes the phase problem using AS, requiring wave angle covered by a single frame, Fig. 7) appears to show no absorption by heavy atoms (Hendrickson and Ogata, 1997) apparent trend; different rotation speeds and frame scopes an effect that electrons are believed to be capable of achieving did not correlate with resolution; however, the data likely (Burmester and Schröder, 1997). A breakthrough is needed in include significant inter-laboratory and inter-assay variability MicroED analogous to IR/AS to find dependable phasing which may contribute to resolution quality. Indeed, the methods for macromolecular crystals. Gonen group attribute increased resolution over their three lysozyme structures to improved data collection and process- ing (Nannenga and Gonen, 2018). To date, no published Improvements to equipment and data work specifically dedicated to the effect of rotation speeds is processing available. Further research is required to probe rotation speed/frame rate contribution to diffraction patterns using The electron detector used in an EM plays a key part in detec- identical crystals to control for other variables. tion and analysis of diffracted waves. Traditionally, EMs ............................................................................................... .................................................................. 6 Downloaded from https://academic.oup.com/biohorizons/article/doi/10.1093/biohorizons/hzy013/5256436 by DeepDyve user on 16 October 2022 Bioscience Horizons � Volume 11 2018 Review article ............................................................................................... .................................................................. ............................................................................................... .................................................................. Table 1. crystal rotation speed, frame rate and resolution Molecule Crystal rotation Frame rate Rotation per Space Molecules Unit cell Volume per Resolution Reference 3 3 speed (°/s) (s/frame) frame (°) group per unit cell volume (Å ) molecule (Å ) (Å) Proteinase 0.090 4.0 0.360 P4 2 2 8 452 897 56 612 1.60 de la Cruz et al. 3 1 K (2017) 0.089 5.1 0.454 P4 2 2 8 452 267 56 533 1.71 Hattne et al. (2018) 3 1 0.090 4.0 0.360 P43212 8 457 776 57 222 1.75 Hattne et al. (2016) Lysozyme 0.090 4.0 0.360 P4 2 2 8 215 821 26 978 1.80 de la Cruz et al. 3 1 (2017) 0.152 0.5 0.076 P2 2 2 4 228 046 57 012 2.11 Clabbers et al. (2017) 1 1 0.450 2.0 0.900 P2 2 2 4 222 650 55 663 2.20 Xu et al. (2018) 1 1 0.090 4.0 0.360 P4 2 2 8 212 691 26 586 2.50 Nannenga et al. 3 1 (2014b) N/A 10.0 N/A P4 2 2 8 219 373 27 422 2.90 Shi et al. (2013) 3 1 Catalase 0.09 6.0 0.540 P2 2 2 4 2 125 468 531 367 3.20 Nannenga et al. 1 1 1 (2014a) a a a 0.75 2.0 1.500 P2 2 2 4 2466 129 616 532 3.20 Yonekura and Maki- 1 1 1 Yonekura (2016) GNNQQNY 0.30 2.0 0.600 P2 2 2 4 4625 1156 1.05 Sawaya et al. (2016) 1 1 1 0.30 2.0 0.600 P2 2 2726 1363 1.10 Sawaya et al. (2016) The table shows MicroED experiments that used crystals of four proteins (proteinase K, lysozyme, catalase and the hepta-peptide GNNQQNY) under different crystal rotation speeds, frame rates, crystal unit cells and volumes, along with the resultant resolution of the final structures. References to relevant publications are given in the table. The relationship of resolution plotted against protein atomicity (density within the unit cell, defined as the volume per molecule), crystal rotation speed and frame scope (the total angle sampled by a single frame) are shown in Figs 5– 7, respectively. Note: for this catalase experiment, various frame rates and rotation speeds were used from 0.5 to 1.0° over 1–3 s per frame median values of these ranges are displayed in the table. Downloaded from https://academic.oup.com/biohorizons/article/doi/10.1093/biohorizons/hzy013/5256436 by DeepDyve user on 16 October 2022 Review article Bioscience Horizons � Volume 11 2018 ............................................................................................... .................................................................. Resolution plotted against Resolution vs Frame atomicity scope 3.5 3.5 Very high atomicity Proteinase K 2.5 2.5 High Lysozyme atomicity 1.5 Catalase 1.5 Medium 1 atomicity GNNQQNY 0.5 0.5 Very low atomicity 0 200000 400000 600000 800000 01 2 Volume per protein molecule (Å3) Frame scope ( /frame) Figure 5. Resolution vs molecular crystal atomicity from MicroED Figure 7. Resolution vs frame scope from MicroED structures of common structures of common proteins. MicroED structures of proteinase K proteins. The relationship between resolution and frame scope (angle (blue), lysozyme (red), catalase (green) and GNNQQNY hepta-peptide sampled per camera frame) of all Table 1 molecules is shown above, (purple) are plotted to correlate structure resolution to unit cell colour coded for atomicity. Very high atomicity = 1000–1500 A /molecule, atomicity, i.e. volume per protein molecule. Crystals with higher high atomicity = 25 000–30 000 A /molecule, medium atomicity = 3 3 symmetry and tighter packing have less volume per molecule and 55 000–60 000 A /molecule, very low atomicity = 50 0000–650 000 A / higher atomicity. More densely packed cells with higher atomicity give molecule. higher final structure resolution. All experimental information is derived from Table 1. provided that electron dosage was not sufficiently high to dam- age the detector (Clough, Moldovan and Kirkland, 2014). Tim Resolution vs Rotation Gruene’s group is actively engaging in detector development speed (Nederlof et al., 2013; van Genderen et al.,2016; Clabbers 3.5 Very high et al.,2017; Matheson et al., 2017), recently demonstrating a atomicity novel ultrathin hybrid detector that samples smaller pixel areas 2.5 and has shorter dead time between frames, reducing data loss High from pixel-related overlap of intensities (Tinti et al., 2018). The atomicity detector circuitry was specifically designed with beam resistance 1.5 Medium to reduce damage at energies of 100, 200 and 300 keV, while 1 atomicity the Abrahams group applied new DIALS integration software 0.5 Very low (Clabbers et al.,2018). The Gonen group noted that IES elec- atomicity 0 trons may be filtered out of the diffraction data by using energy 0 0.5 1 filters to remove IES (Nannenga et al., 2014b) and improve Frame scope ( /s) SNR in line with other work (Yonekura, Maki-Yonekura and Namba, 2002; Leis et al., 2009), but have seemingly not yet uti- Figure 6. Resolution vs crystal rotation speed from MicroED structures of lised this approach. Future innovation and development of common proteins. The relationship between resolution and crystal equipment and data processing to improve SNR will greatly rotation speed of all Table 1 molecules are shown above, colour coded for 3 improve resolution, and energy filters may be an attractive atomicity. Very high atomicity = 1000–1500 A /molecule, high atomicity = 3 3 option to remove IES from diffracting waves. 25 000–30 000 A /molecule, medium atomicity = 55 000–60 000 A / molecule, Very low atomicity = 500 000–650 000 A /molecule. MicroED in the wider community have used two types of indirect detectors, Charge-Coupled Devices and Complementary Metal-Oxide-Semiconductors The worldwide structural biology community is increasingly (CMOS). Both convert electrons to photons using a scintillator using MicroED in structural studies and are taking an interest which are then detected by the sensor (a comparison is outlined in the technical and theoretical aspects. One group published in Allé et al., 2016). The scintillation inherently introduces scat- Lysozyme structures to 2.2 Å and commented on MicroED’s ter, and bulky detectors can scatter photons laterally into neigh- main challenges, stating that crystal degradation and gonio- bouring pixels, spreading the peak, making indirect detectors metric imperfections in crystal rotation are key challenges to sub-optimal for electron crystallography (Faruqi, 2001). Hybrid overcome (Xu et al., 2018). The first use of MicroED to study detectors detect charge directly from electrons to prevent amyloid fibrils outside of the Gonen group was published scintillator-associated scatter but due to the strong electron/mat- recently, solving the (FUS) amyloid core to 0.73 Å (Luo et al., ter interactions, they are easily damaged by the electron beam. 2018). Other groups are attempting to improve the technique Initial studies suggested that hybrid detectors may nonetheless and contribute to MicroED theory; some are researching and perform better than indirect detectors to generate signal, commenting on the relationship between microcrystal size ............................................................................................... .................................................................. Resolutiopn (Å) Resolution (Å) Resolution (Å) Downloaded from https://academic.oup.com/biohorizons/article/doi/10.1093/biohorizons/hzy013/5256436 by DeepDyve user on 16 October 2022 Bioscience Horizons � Volume 11 2018 Review article ............................................................................................... .................................................................. and diffraction spots (Williams et al., 2017), while others sug- a keen interest in crystallography and wrote this article during gest that maps generated from MicroED might be utilising his M.Res dissertation in Structural Molecular Biology at sub-optimal MR methods as structure factors fail to take Imperial College London, after a BSc in Biomedical Science, negative contributions from IES into account (Wang, 2017). MSc in Biomedical Blood Science at Keele University, and a While the technique is still very young, it is likely that the PGCert in Protein Crystallography at Birkbeck. While work- future will see a growing rate of MicroED structures being ing in clinical laboratories (transfusion and emergency medi- solved and published as the community begins to delve deeper cine in Salisbury and Addenbrookes), Rob undertook the into the intricacies of the method and develop it further. PGCert and decided to pursue a career in basic research. He While the work of the Gonen and Abrahams group have then worked as a Tissue Acquisition Officer for the BCI, undoubtedly contributed significantly to the technique, the before working at UCB Celltech as an Associate Scientist in wider community is now beginning to realise the potential biopharmaceutical method development. In September 2017 that this novel method has within structural biology. he began his MRes, with the PhD beginning in September 2018. Post-PhD Rob would like to contribute to MicroED development within structural biology. Conclusion MicroED is an exciting crystallographic method utilising short Acknowledgements wavelengths to offer better theoretical resolution limits than XRC, while simultaneously sampling more reciprocal space per I would like to thank my supervisor, Thomas Meier, for his illumination. While modern macromolecular EM work typically guidance and feedback. I would like to thank Morgan Beeby relies on single particle analysis without phasing methods or crys- for his invaluable insight concerning electron microscopy and tals (Orlova and Saibil, 2004), MicroED offers significantly high- electron/matter interactions. I would also like to thank Jose er resolution; a characteristic that makes crystallographic Rodriguez for his fruitful correspondence concerning methods so popular (Wilkins, 2013). MicroED offers an afford- MicroED, current research and future directions for the able and accessible XRC alternative by utilising EMs commonly method. used in many structural laboratories with minor adjustments. Continuous rotation MicroED and XRC diffraction software have improved the method, but more work is required to charac- References terise the effect of crystal volume, unit cell packing, electron wave- length, crystal rotation speed, frame rate capture, detector type Abrahams, J. P. (2010) The strong phase object approximation may and the use of energy filters on resolution. While MicroED works allow extending crystallographic phases of dynamical electron dif- very well with small molecule crystals, its use on larger molecules fraction patterns of 3D protein nano-crystals, Zeitschrift fur lacks comparable resolution, likely due to solvent/protein ratios Kristallographie-Crystalline Materials, 225, 67–76. of large unit cells with low symmetry crystals creating noise. For Allé, P., Wenger, E., Dahaoui, S. et al. (2016) Comparison of CCD, CMOS high macromolecular resolution, MicroED depends on MR from and hybrid pixel X-ray detectors: detection principle and data qual- XRC studies. To be a stand-alone technique, MicroED must ity, Physica Scripta, 91, 063001. develop phasing methods for macromolecular datasets. Computational methods have somewhat addressed this challenge Andrews, S., Reichow, S. L. and Gonen, T. (2008) Electron crystallog- at resolutions of 1.2 Å and 2.0 Å, but macromolecules typically raphy of aquaporins, IUBMB Life, 60, 430–436. give lower resolutions than 1.2 Å and even at 2.0 Å may not pre- Barnard, D. P., Turner, J. N., Frank, J. et al. (1992) A 360° single-axis tilt sent sufficient secondary structures to utilise the phasing method, stage for the high-voltage electron microscope, Journal of therefore phasing is particularly troublesome and must be Microscopy, 167, 39–48. addressed by developing innovative phasing methods. As it stands, MicroED is a powerful technique for small molecules, elu- Bartesaghi, A., Sprechmann, P., Liu, J. et al. (2008) Classification and 3D cidating these structures in unprecedented detail and might per- averaging with missing wedge correction in biological electron haps be considered better than XRC in terms of resolution and tomography, Journal of Structural Biology, 162 (3), 436–450. practicality for this class of molecules. Further work and innov- Birkhoff, R. D. (1958) The passage of fast electrons through matter, in ation will be essential for solving the MicroED phase problem, Flügge S. (ed), Corpuscles and Radiation in Matter II, 1st edition, and given that the wider community is beginning to contribute to Springer, Berlin/Heidelberg, pp. 53–138. the technique, we may very well be on the cusp of producing the next series of breakthroughs for this impressive technique. Bragg, W. H. and Bragg, W. L. (1913) The reflection of X-rays by crystals, Proceedings of the Royal Society London, 88, 428–438. Author biography Bragg, W. H. and Perutz, M. F. (1952) The external form of the haemo- globin molecule. I, Acta Crystallographica, 5, 277–283. Rob is now a BBSRC PhD candidate of Structural Biology at the University of Bristol, working with Dr Paul Race studying Burmester, C. and Schröder, R. R. (1997) Solving the phase problem in pro- the structure and function of streptococcal adhesins. Rob has tein electron crystallography: multiple isomorphous replacement and ............................................................................................... .................................................................. 9 Downloaded from https://academic.oup.com/biohorizons/article/doi/10.1093/biohorizons/hzy013/5256436 by DeepDyve user on 16 October 2022 Review article Bioscience Horizons � Volume 11 2018 ............................................................................................... .................................................................. anomalous dispersion as alternatives to imaging, Scanning Microscopy, precession technique and high-resolution electron microscopy, 11, 323–334. Acta Crystallographica A—Foundations and Advances, 59, 117–126. Ceska, T. A. and Henderson, R. (1990) Analysis of high-resolution elec- Georgieva, D., Abrahams, J. P., Zandbergen, H. et al. (2007a) Solving ab- tron diffraction patterns from purple membrane labelled with initio protein and nanostructures with precession electron diffrac- heavy-atoms, Journal of Molecular Biology, 213, 539–560. tion, Microscopy and Microanalysis, 13, 952–953. Cipriani, F., Roewer, M., Landret, C. et al. (2012) CrystalDirect: a new Georgieva, D. G., Kuil, M. E., Oosterkamp, T. H. et al. (2007b) Heterogeneous method for automated crystal harvesting based on laser-induced nucleation of three-dimensional protein nanocrystals, Acta photoablation of thin films, Acta Crystallographica Section D— Crystallographica Section D—Biological Crystallography, 63, 564–570. Biological Crystallography, 68, 1393–1399. Gjonnes, J., Hansen, V., Berg, B. S. et al. (1998) Structure model for the Clabbers, M. T. B. and Abrahams, J. P. (2018) Electron diffraction and phase almfe derived from three-dimensional electron diffraction three-dimensional crystallography for structural biology, intensity data collected by a precession technique. comparison Crystallography Reviews, 24 (3), 176–204. with convergent-beam diffraction, Acta Crystallographica Section A- Foundations and Advances, 54, 306–319. Clabbers, M. T. B., Gruene, T., Parkhurst, J. M. et al. (2018) Electron dif- fraction data processing with DIALS, Acta Crystallographica section Glaeser, R. M. (1971) Limitations to significant information in biological D—Structural Biology, 74, 506–518. electron microscopy as a result of radiation damage, Journal of Ultrastructure Research, 36, 466–482. Clabbers, M. T. B., van Genderen, E., Wan, W. et al. (2017) Protein struc- ture determination by electron diffraction using a single three- Gonen, T., Cheng, Y. F., Sliz, P. et al. (2005) Lipid-protein interactions in dimensional nanocrystal, Acta Crystallographica Section D— double-layered two-dimensional AQP0 crystals, Nature, 438, 633–638. Structural Biology, 73, 738–748. Guenther, E. L., Ge, P., Trinh, H. et al. (2018) Atomic-level evidence for Clough, R. N., Moldovan, G. and Kirkland, A. I. (2014) Direct detectors for packing and positional amyloid polymorphism by segment from electron microscopy, Journal of Physics Conference Series, 522, 012046. TDP-43 RRM2, Nature Structural & Molecular Biology, 25, 311–319. Curry, S. (2015) Structural biology: a century-long journey into an Hattne, J., Reyes, F. E., Nannenga, B. L. et al. (2015) MicroED data collec- unseen world, Interdisciplinary Science Reviews, 40, 308–328. tion and processing, Acta Crystallographica A-Foundation and Advances, 71, 353–360. de la Cruz, M. J., Hattne, J., Shi, D. et al. (2017) Atomic-resolution struc- Hattne, J., Shi, D., de la Cruz, M. J. et al. (2016) Modeling truncated pixel tures from fragmented protein crystals with the cryoEM method values of faint reflections in MicroED images, Journal of Applied MicroED, Nature Methods, 14 (4), 399–402. Crystallography, 49, 1029–1034. Dorset, D. L. (1997) Direct phase determination in protein electron crystal- Hattne, J., Shi, D., Glynn, C. et al. (2018) Analysis of global and site- lography: the pseudo-atom approximation, Proceedings of the National specific radiation damage in cryo-EM, Structure (London, England : Academy of Sciences of the United States of America, 94, 1791–1794. 1993), 26, 759–766. Eckert, M. (2012) Max von Laue and the discovery of X-ray diffraction in Hauptman, H. A. and Langs, D. A. (2003) The phase problem in neutron 1912, Annalen Der Physik, 524 (5), A85. crystallography, Acta Crystallographica Section A, 59, 250–254. Erickson, H. P. (2009) size and shape of protein molecules at the nano- meter level determined by sedimentation, gel filtration, and elec- Henderson, R. (1995) The potential and limitations of neutrons, elec- tron microscopy, Biological Procedures Online, 11, 32–51. trons and X-rays for atomic-resolution microscopy of unstained bio- logical molecules, Quarterly Reviews of Biophysics, 28, 171–193. Ezkurdia, I., Juan, D., Manuel Rodriguez, J. et al. (2014) Multiple evi- dence strands suggest that there may be as few as 19 000 human Henderson, R. and Unwin, P. (1975) 3-Dimensional model of purple protein-coding genes, Human Molecular Genetics, 23, 5866–5878. membrane obtained by electron-microscopy, Nature, 257, 28–32. Faruqi, A. R. (2001) Prospects for hybrid pixel detectors in electron crys- Hendrickson, W. A. and Ogata, C. M. (1997) Phase determination from tallography, Nuclear Instruments and Methods in Physics Research A, multiwavelength anomalous diffraction measurements, Methods in 466, 146–154. Enzymology, 276, 494–523. Fraser,R.D.B., MacRae,T.P.and Suzuki, E.(1978) An improved method for Hughes, M. P., Sawaya, M. R., Goldschmidt, L. et al. (2017) Low- calculating the contribution of solvent to the X-ray diffraction pattern complexity domains adhere by reversible amyloid-like interactions of biological molecules, Journal of Applied Crystallography, 11, 693–694. between kinked β-sheets. bioRxiv (in press) doi: 10:10:/153817 Gallagher-Jones, M., Glynn, C., Boyer, D. R. et al. (2018) Sub-angstrom Jaskolski, M., Dauter, Z. and Wlodawer, A. (2014) A brief history of cryo-EM structure of a prion protofibril reveals a polar clasp, Nature macromolecular crystallography, illustrated by a family tree and its Structural & Molecular Biology, 25 (2), 131–134. Nobel fruits, FEBS, 281, 3985–4009. Gemmi, M., Zou, X., Hovmöller, S. et al. (2003) Structure of Ti2P solved Jelsch, C., Teeter, M. M., Lamzin, V. et al. (2000) Accurate protein crystal- by three-dimensional electron diffraction data collected with the lography at ultra-high resolution: valence electron distribution in ............................................................................................... .................................................................. 10 Downloaded from https://academic.oup.com/biohorizons/article/doi/10.1093/biohorizons/hzy013/5256436 by DeepDyve user on 16 October 2022 Bioscience Horizons � Volume 11 2018 Review article ............................................................................................... .................................................................. crambin, Proceedings of the National Academy of Sciences of the from submicrometre three-dimensional protein crystals, Acta United States of America, 97, 3171–3176. Crystallographica Section D—Biological Crystallography, 69, 1223–1230. Jiang, L., Georgieva, D., Zandbergen, H. W. et al. (2009) Unit-cell determin- Oleynikov, P. and Hovmöller, S. (2007) Precession electron diffraction: ation from randomly oriented electron-diffraction patterns, Acta observed and calculated intensities, Ultramicroscopy, 107, 523–533. Crystallographica Section D—Biological Crystallography, 65, 625–632. Orlova, E. V. and Saibil, H. R. (2004) Structure determination of macro- Koshland, D. E. (1994) The key-lock theory and the induced fit theory, molecular assemblies by single-particle analysis of cryo-electron Angewandte Chemie-International Edition, 33, 2375–2378. micrographs, Current Opinion in Structural Biology, 14, 584–590. Krotee, P., Rodriguez, J. A., Sawaya, M. R. et al. (2017) Atomic structures Rodriguez, J. A., Eisenberg, D. S. and Gonen, T. (2017) Taking the meas- of fibrillar segments of hIAPP suggest tightly mated beta-sheets are ure of MicroED, Current Opinion in Structural Biology, 46, 79–86. important or cytotoxicity, Elife, 6, e19273. Rodriguez, D. D., Grosse, C., Himmel, S. et al. (2009) Crystallographic ab Leis, A., Rockel, B., Andrees, L. et al. (2009) Visualizing cells at the nano- initio protein structure solution below atomic resolution, Nature scale, Trends in Biochemical Sciences, 34, 60–70. Methods, 6, U39. Lindsay, D. G. and Shall, S. (1969) Acetoacetylation of insulin, Rodriguez, J. A., Ivanova, M. I., Sawaya, M. R. et al. (2015) Structure of Biochemical Journal, 115, 587–595. doi:10.1042/bj1150587 the fig- the toxic core of alpha-synuclein from invisible crystals, Nature, 525, ure used in Figure 1 was original published by Portland Press under 486–490. the STM permission guidelines (https://www.stm-assoc.org/ Rolland, T., Tasan, M., Charloteaux, B. et al. (2014) A proteome-scale copyright-legal-affairs/permissions/permissions-guidelines/), and map of the human interactome network, Cell, 159, 1212–1226. request has been granted to use the figure. Sawaya, M. R., Rodriguez, J., Cascio, D. et al. (2016) Ab initio structure Luo, F., Gui, X., Zhou, H. et al. (2018) Atomic structures of FUS LC determination from prion nanocrystals at atomic resolution by domain segments reveal bases for reversible amyloid fibril forma- MicroED, Proceedings of the National Academy of Sciences of the tion, Nature Structural & Molecular Biology, 25, 341–346. United States of America, 113, 11232–11236. Matheson, J., Moldovan, G., Kirkland, A. et al. (2017) Testing and com- Schmidt, A., Teeter, M., Weckert, E. et al. (2011) Crystal structure of small parison of imaging detectors for electrons in the energy range protein crambin at 0.48 Å resolution, Acta Crystallographica 10–20 keV, Journal of Instrumentation, 12, C11016. International Section F- Structural Biology Communications, 67, 424–428. Conference on Position Sensitive Detectors. Sheldrick, G. M. (1990) Phase annealing in Shelx-90—direct methods Morris, R. J. and Bricogne, G. (2003) Sheldrick’s 1.2 angstrom rule and for larger structures, Acta Crystallographica Section A, 46, 467–473. beyond, Acta Crystallographica Section D—Biological Crystallography, 59, 615–617. Shi, D., Nannenga, B. L., de la Cruz, M. J. et al. (2016) The collection of MicroED data for macromolecular crystallography, Nature Protocols,11, 895–904. Nannenga, B. L. and Gonen, T. (2014) Protein structure determination by MicroED, Current Opinion in Structural Biology, 27, 24–31. The fig- Shi, D., Nannenga, B. L., Iadanza, M. G. et al. (2013) Three-dimensional ure used in Figure 3 was originally published by Elsevier and electron crystallography of protein microcrystals, Elife, 2, e01345. request has been granted to use the figure by the Copyright The figure used in Figure 2 was original published under the Clearance Centre (license 4367131293431). Creative Commons Attribution 3.0 Unported (CC BY 3.0) license (https://creativecommons.org/licenses/by/3.0/), and request has Nannenga, B. L. and Gonen, T. (2018) MicroED: a versatile cryoEM meth- been granted to use the figure. od for structure determination, Emerging Topics in Life Sciences,2, 1–8. Stevenson, H. P., Lin, G., Barnes, C. O. et al. (2016) Transmission electron microscopy for the evaluation and optimization of crystal growth, Nannenga, B. L., Shi, D., Hattne, J. et al. (2014a) Structure of catalase Acta Crystallographica Section D—Structural Biology, 72, 603–615. determined by MicroED, Elife, 3, e03600. Stevenson, H. P., Makhov, A. M., Calero, M. et al. (2014) Use of transmis- Nannenga, B. L., Shi, D., Leslie, A. G. W. et al. (2014b) High-resolution sion electron microscopy to identify nanocrystals of challenging structure determination by continuous-rotation data collection in protein targets, Proceedings of the National Academy of Sciences of MicroED, Nature Methods, 11, 927–930. The figure used in Figure 4 the United States of America, 111, 8470–8475. was original published by Springer Nature and request has been granted to use the figure by the Copyright Clearance Centre Subramanian, G., Basu, S., Liu, H. et al. (2015) Solving protein nanocrys- (license 4367140499674). tals by cryo-EM diffraction: Multiple scattering artifacts, Ultramicroscopy, 148, 87–93. Nederlof, I., Georgieva, D. and Abrahams, J. P. (2011) Electron diffrac- tion of submicron 3D protein crystals, Acta Crystallographica Section Sweet, R. M. (1985) Introduction to Crystallography, Methods in A-Foundation and Advances, 67, C228. Enzymology, 114, 19–46. Nederlof, I., van Genderen, E., Li, Y. et al. (2013) A Medipix quantum Taylor, G. (2003) The phase problem, Acta Crystallographica Section D— area detector allows rotation electron diffraction data collection Biological Crystallography, 59, 1881–1890. ............................................................................................... .................................................................. 11 Downloaded from https://academic.oup.com/biohorizons/article/doi/10.1093/biohorizons/hzy013/5256436 by DeepDyve user on 16 October 2022 Review article Bioscience Horizons � Volume 11 2018 ............................................................................................... .................................................................. Taylor, G. L. (2010) Introduction to phasing, Acta Crystallographica Wilkins, S. W. (2013) Celebrating 100 years of X-ray crystallography, Section D—Biological Crystallography, 66, 325–338. Acta Crystallographica Section A, 69, 1–4. Tinti, G., Frojdh, E., van Genderen, E. et al. (2018) Electron crystallog- Williams, S. R., Dilanian, R. A., Quiney, H. M. et al. (2017) Analysis of dif- raphy with the EIGER detector, IUCRJ, 5, 190–199. fracted intensities from finite protein crystals with incomplete unit cells, Crystals, 7, 220. Unwin, P. and Henderson, R. (1975) Molecular-structure determination by electron-microscopy of unstained crystalline specimens, Journal Wisedchaisri,G., Reichow, S. L. andGonen,T.(2011) Advances in of Molecular Biology, 94, 425–440. structural and functional analysis of membrane proteins by elec- tron crystallography, Structure (London, England : 1993),19, van Genderen, E., Clabbers, M. T. B., Das, P. P. et al. (2016) Ab initio struc- 1381–1393. ture determination of nanocrystals of organic pharmaceutical com- pounds by electron diffraction at room temperature using a Timepix Wukovitz, S. W. and Yeates, T. O. (1995) Why protein crystals favor quantum area direct electron detector, Acta Crystallographica some space-groups over others, Nature Structural Biology,2, A-Foundation and Advances,72, 236–242. 1062–1067. Vergara, S., Lukes, D. A., Martynowycz, M. W. et al. (2017) MicroED struc- Xu, H., Lebrette, H., Yang, T. et al. (2018) A rare lysozyme crystal form ture of Au-146(p-MBA)(57) at subatomic resolution reveals a twinned solved using highly redundant multiple electron diffraction data- FCC cluster , Journal of Physical Chemistry Letters,8,5523–5530. sets from micron-sized crystals, Structure (London, England : 1993), 26, 667–675. Wang, J. (2017) On the appearance of carboxylates in electrostatic potential maps, Protein Science, 26, 396–402. Yonekura, K. and Maki-Yonekura, M. (2016) Refinement of cryo-EM structures using scattering factors of charged atoms, Journal of Wang, J.,Dauter, M.,Alkire, R. et al.(2007) Tricliniclysozymeat0.65 Å reso- Applied Crystallography, 49, 1517–1523. lution, Acta Crystallographica Section D—Biological Crystallography, 63, 1254–1268. Yonekura, K., Maki-Yonekura, S. and Namba, K. (2002) Quantitative comparison of zero-loss and conventional electron diffraction from Weichenberger, C. X., Afonine, P. V., Kantardjieff, K. et al. (2015) The solv- two-dimensional and thin three-dimensional protein crystals, ent component of macromolecular crystals, Acta Crystallographica Biophysical Journal, 82, 2784–2797. Section D—Biological Crystallography, 71, 1023–1038. ............................................................................................... .................................................................. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png BioScience Horizons Oxford University Press

Illuminating the secrets of crystals: microcrystal electron diffraction in structural biology

BioScience Horizons , Volume 11 – Jan 1, 2018

Loading next page...
 
/lp/oxford-university-press/illuminating-the-secrets-of-crystals-microcrystal-electron-diffraction-8ZSKL8hUFS

References (164)

Publisher
Oxford University Press
Copyright
Copyright © 2022 Oxford University Press
eISSN
1754-7431
DOI
10.1093/biohorizons/hzy013
Publisher site
See Article on Publisher Site

Abstract

Downloaded from https://academic.oup.com/biohorizons/article/doi/10.1093/biohorizons/hzy013/5256436 by DeepDyve user on 16 October 2022 BioscienceHorizons Volume 11 2018 10.1093/biohorizons/hzy013 ............................................................................................ ..................................................................... Review article Illuminating the secrets of crystals: microcrystal electron diffraction in structural biology Rob Barringer C.101, Biomedical Sciences, University walk, University of Bristol, United Kingdom, BS8 1TD *Corresponding author: Centre for structure Biology, Sir Ernst Chain Building, Imperial College London, United Kingdom, SW7 2AZ Email: r.barringer@bristol.ac.uk Supervisor: Professor Thomas Meier, Room 501, Sir Ernst Chain building, imperial college London, South Kensington, London, United Kingdom, SW7 2AZ. ............................................................................................ ..................................................................... X-ray crystallography (XRC) has visualised biological macromolecules in exquisite detail for over 50 years, relying on a com- bination of mathematical principles to offer insight into atomic structures. Crystals can diffract various electromagnetic waves aside from the conventional X-ray, offering an alternative approach to crystallographic structural analysis. Microcrystal electron diffraction (MicroED) illuminates crystals with electron waves instead of X-rays. Two specialised groups have demonstrated that MicroED can give high-resolution (often atomic) data, and now appears to be developing into a powerful alternative method to XRC or electron microscopy of macromolecules. How MicroED compares to XRC will be key to assessing it as a stand-alone crystallographic technique. This review presents a critical analysis of MicroED, with comments on theoretical and practical aspects and suggestions of further work and development. Key words: MicroED, crystallography, structural biology, method development Submitted on 23 July 2018; editorial decision on 23 November 2018 ............................................................................................ ..................................................................... the use of complex experimental methods, denoting the field Introduction of ‘structural biology’. There are three main methods used to assess protein structure: Electron microscopy (EM), Nuclear Life is dependent on the ability of cells to perform a myriad of magnetic resonance, and X-ray crystallography (XRC) functions alone or in communities as tissues. When cellular (Curry, 2015), with XRC having the richest history of the processes falter, diseases can arise depending on the aberrant three (Wilkins, 2013). This review focusses on a new crystal- process. Understanding these processes is therefore essential lographic method (microcrystal electron diffraction; to understanding the cellular basis of disease pathology. MicroED) that uses electrons instead of X-rays, outlining a Cellular functions depend on proteins, of which there are brief history of both and then presenting a critical comparison likely over 19 000 in humans (Ezkurdia et al., 2014), each of theoretical and practical aspects. with unique functions and interacting with various biomole- cules (Rolland et al., 2014). Proteins act as nano-scale cellular A short history of crystallography ‘tools’ with functions that are intimately linked to their unique structure; the role that protein architecture plays in XRC has been the primary macromolecular structural meth- biomolecular interactions has a long history, most commonly od for over 50 years (Jaskolski, Dauter and Wlodawer, typified in the mind of the public by the ‘lock-and-key’ 2014). The theory of the method is simple: a pure crystal con- hypothesis (Koshland, 1994). Since proteins are incredibly taining exquisitely ordered repeating units of identical mole- small (for example haemoglobin has a diameter of ~1/200 cules are placed in the path of an X-ray beam, which is 000th of a mm) (Erickson, 2009), structural studies require scattered by the molecules within the crystal to produce a ............................................................................................... .................................................................. © The Author(s) 2019. Published by Oxford University Press. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited. Downloaded from https://academic.oup.com/biohorizons/article/doi/10.1093/biohorizons/hzy013/5256436 by DeepDyve user on 16 October 2022 Review article Bioscience Horizons � Volume 11 2018 ............................................................................................... .................................................................. the molecules within the crystal to solve the atomic structure of the molecules. Electron crystallography X-rays interfere weakly with matter compared to electrons, so XRC requires large volumes of matter create sufficient X-ray dif- fraction before the destructive effects of the radiation degrade the crystal (Henderson, 1995). The damaging effects are because X-rays deposit 1333× more energy than electrons into the crys- tal per elastic scattering (ES) event (single scattering of a wave without loss of amplitude or phase) (80 keV vs 60 eV per ES respectively) but electrons interact more frequently, degrading crystals very quickly (Glaeser, 1971). To accommodate for this challenge, 2-dimensional electron crystallography (2DEC) was developed in the 1970s, using thin crystals that reduce electron interaction, thereby reducing noise and enabling crystal illumin- ation with electrons. The technique successfully determined the Figure 1. The X-ray diffraction pattern of glycine modified mono- structure of purple membrane protein (today known as the acetoacetyl insulin. Protein crystals give unique diffraction patterns light-driven proton pump ‘bacteriorhodopsin’)and theenzyme due to the atomic arrangement of the protein (i.e. the structure of the macromolecule). Each crystal has a characteristic diffraction pattern catalase (Henderson and Unwin, 1975; Unwin and Henderson, that can be analysed to determine the atomic structure of the 1975) and had the unique ability to form 2-dimensional crystals crystallised molecule. Above is the diffraction pattern of glycine of proteins within a lipid bilayer (an important property of modified mono-acetoacetyl insulin, which presents a symmetrical six- native membrane proteins, Gonen et al.,2005; Andrews, sided pattern. This research was originally published in the Biochemical Reichow and Gonen, 2008; Wisedchaisri, Reichow and Gonen, Journal (Lindsay and Shall 1969), reprinted with permission. 2011). During the late 2000s, Jan Pieter Abrahams applied elec- tron crystallography to cryogenic 3-dimensional macromolecu- lar crystals to create a method termed 3-dimensional electron constellation-style pattern (Fig. 1). While beautiful to the crystallography (3DEC), gaining single diffraction patterns from naked eye, the symmetry and ‘brightness’ of each spot of the microcrystals of Lysozyme which had previously proved difficult ‘diffraction pattern’ contains crucial structural information to achieve (Georgieva et al., 2007b). As it turned out, the indi- about the molecules that scattered the X-rays (Sweet, 1985). vidual diffraction patterns provided a challenge as the relative Crystallographers use the information from diffraction pat- orientation of the crystal lattice for each pattern was unknown, terns to elucidate the atomic structures of biomolecules, so merging diffraction data from multiple crystals was difficult. allowing us to gain an atomistic perspective of nature’s tools. Data processing methods were then developed (Jiang et al., X-rays were first used in crystallography by Max von Laue, 2009)(Abrahams, 2010), before improved detectors yielded bet- who hypothesised that X-ray wavelengths are short enough ter signal-to-noise ratios (SNR) (Nederlof et al., 2011). to be diffracted by atoms within a 3-dimensional crystal − −2 −1 Reducing electron dosage (to 0.1 e Å s ) prevented degrad- (Eckert, 2012). Incoming X-rays excite the electrons sur- ation and acquired more diffraction patterns per crystal rounding atoms of the protein molecules in the crystal, and (Nederlof et al.,2013), allowing patterns to be orientated. the X-ray energy is subsequently redistributed in all directions as a wave. In certain directions, the diffracted X-rays of each identical molecule interfere coherently (in phase) to produce The history of MicroED observable diffraction ‘spots’. The Braggs showed that these angles of observation depend on the X-ray wavelength and Low dose electron beams using 3D the spacing between molecules in the crystal (Bragg and microcrystals Bragg, 1913) and that the characteristic ‘spotting’ of diffrac- tion patterns results from these two properties. The spots are MicroED (hereon used as a catch-all term for MicroED/ observed because the scattered X-rays of the repeating mole- 3DEC) was developed by Tamir Gonen, using cryogenically cules only amplify when they overlap with each other in treated 3D Lysozyme microcrystals of 2 μm × 2 μm × 0.5 μm phase, which occurs at discrete angles relative to the crystal to to demonstrate proof of concept (Fig. 2)(Shi et al., 2013). As − 2 create a ‘reciprocal lattice’ of spotting. The brightness of each with Abrahams, Gonen found that after 9e /Å the diffrac- spot results from the cumulative X-ray waves scattered from tion patterns deteriorated, thus by using ~0.5 μM microcrys- − −2 −1 every atom of every molecule, overlapping in unique phases tals and a 0.01e Å s dosage over 10 second exposures at each angle to give an average phase that produces a charac- per angle, they collected 90 diffraction patterns per crystal, teristic spot intensity (i.e. spot ‘brightness’). Each crystal three times more than (Nederlof et al., 2013). A wavelength therefore gives a unique diffraction pattern (Fig. 1), which is of 0.025 Å (with a 200 kV acceleration current) was used, sig- measured and used to determine the exact atomic positions of nificantly shorter than typical XRC wavelengths (~1 Å). ............................................................................................... .................................................................. 2 Downloaded from https://academic.oup.com/biohorizons/article/doi/10.1093/biohorizons/hzy013/5256436 by DeepDyve user on 16 October 2022 Bioscience Horizons � Volume 11 2018 Review article ............................................................................................... .................................................................. Figure 2. A comparison of Lysozyme microcrystals for MicroED (black arrows) and large crystals for XRC (white arrows). Microcrystals are ~6 orders of magnitude smaller than XRC crystals. Microcrystals are ~0.5–1 μM thick with ~55 × 106 unit cells. Crystals are screened by initial illumination to assess diffraction quality. For discussions on Figure 3. Illumination using electrons creates slices (left) through volume limitations, see optimal crystal sizes. Scale bar is 50 μM. This reciprocal space instead of lunes typically observed in XRC (right). research was originally published in eLIFE (Shi et al., 2013). Image has Diffraction patterns of lysozyme are shown. The images show that at undergone a minor edit to include white arrows. an angle perpendicular to a crystal axis, every reciprocal lattice point within the plane was observed using electron sources (with a complete Consequently, the Bragg angles are more acute and show absence of lune formation), while the XRC experiment demonstrated more diffraction spots in a single illumination; the Ewald circular lunes. This research was originally published in Current Opinion sphere (a mathematical tool to determine accessible reciprocal in Structural Biology (Nannenga and Gonen, 2014), published under the lattice points from a single illumination) is so large that it’s CC-BY-NC-ND license. surface is virtually flat, producing slices through reciprocal space instead of the lunes seen in XRC, sampling more lattice matched previous observations that rotation reduces IES/ points (Fig. 3, Nannenga and Gonen, 2014). Crystals were MES contribution to diffraction data (Georgieva et al., rotated 1° between illuminations by tilting the specimen stage, 2007a; Gjonnes et al., 1998; Gemmi et al., 2003), resulting in with each spot averaging 34 observations from multiple continuous rotation becoming standard in MicroED. A sug- angles. Since the slices through reciprocal space do not inter- gestion that larger unit cells (the repeating unit of molecules sect spots perfectly, true spot intensities were difficult to throughout the crystal) and lower symmetry (the orientation derive, leading to loss of intensity data. Crude intensity of proteins within the unit cell) of the static vs CR Lysozyme approximations were made by assuming that the highest spot structure may have increased noise due to lower protein/solv- intensity represented the ‘true’ value. Using this initial ent ratios, increasing solvent contribution to the diffraction approach, the Lysozyme structure was solved to 2.9 Å, using pattern and lowering resolution (see ’Optimal crystal sizes’ molecular replacement (MR, a technique that derives phase section). Further MicroED work used XRC software to gener- information from prior experiments of the same molecule) ate a Catalase structure to 3.2 Å (Nannenga et al., 2014b). (Cipriani et al., 2012). The resolution breakthrough Continuous crystal rotation in MicroED More MicroED macromolecular structures were solved, lead- The Gonen group introduced ‘continuous rotation’ MicroED; ing to protocols outlining data collection and analysis uninterrupted observation of diffraction data during contin- (Hattne et al., 2015, 2016; Shi et al., 2016) including micro- ual crystal rotation, with frame rates of 4 s/frame with 0.09°/s crystal acquisition from macro-crystals (de la Cruz et al., rotation (Nannenga et al., 2014b). The subsequent lysozyme 2017). Two breakthrough studies using MicroED without pattern compared better with XRC diffraction data than complementary X-ray MR data were published: the structure static-MicroED (2.5 vs 2.9 Å respectively, Fig. 4), likely of α-synuclein gained by MR that utilised β-strand motifs as a because (a) rotation allows proper intersection of reciprocal search model for this simple protein (Rodriguez et al., 2015) lattice points with the Ewald sphere and (b) rotation reduces and an Sup35 amyloid core component (GNNQQNY, a hep- beam contact time, lessening inelastic-scattering (IES, where ta-peptide) (Sawaya et al., 2016). This marked a turning point incident electrons are deflected by atomic electrons and for MicroED; solving protein structures without prior XRC impart some energy into the crystal) and multiple elastic- data (ab initio), an achievement previously applied only to scattering (MES, where electrons elastically scatter multiple organic compounds (van Genderen et al., 2016). The first times), both of which interfere with diffracting waves to cre- novel macromolecule structure, TGF-βm; TβRII, was published ate noise in data (see ’MicroED and XRC’ section). This soon after (de la Cruz et al.,2017). More than 17 structures ............................................................................................... .................................................................. 3 Downloaded from https://academic.oup.com/biohorizons/article/doi/10.1093/biohorizons/hzy013/5256436 by DeepDyve user on 16 October 2022 Review article Bioscience Horizons � Volume 11 2018 ............................................................................................... .................................................................. Figure 4. Lysozyme diffraction data of CR MicroED, static-frame MicroED and XRC experiments. The (001) diffraction plane of lysozyme using continuous rotation MicroED (c), static-MicroED (a) and XRC (b) are shown. Continuous rotation MicroED shows the best agreement with the XRC diffraction pattern. The Pearson correlation coefficients shown in (d) were 0.76 between continuous rotation MicroED (c) and X-ray data (b), vs 0.56 for static-frame MicroED (a) and X-ray data (b). Reprinted by permission from Springer Nature: Springer Nature Methods (Nannenga et al., 2014b). have been solved by MicroED (Rodriguez, Eisenberg and suites (EM, tilting stage, carbon grids and liquid ethane vitrifi- Gonen, 2017), with subsequent structures appearing in litera- cation) are sufficient for MicroED. MicroED requires a high ture on a semi-regular basis (Krotee et al.,2017; Gallagher- frame rate camera fast enough to capture individual reflec- Jones et al.,2018; Guenther et al.,2018), many within the sub- tions and prevent spot overlap, and high-speed detectors to angstrom range such as the fused in sarcoma (FUS) protein minimise between-frame readout times, such as direct electron amyloid forming core to 0.73 Å (Luo et al.,2018). Recent detectors (as used previously by the Abrahams group, van reviews highlight MicroED’s history and future improvements Genderen et al., 2016). A previous guide was also outlined (Rodriguez, Eisenberg and Gonen, 2017) and mathematical explaining how to make a device to control continuous crystal and theoretical principles (Clabbers and Abrahams, 2018). tilt (Shi et al., 2016) and which equipment, data acquisition and processing methods are necessary. Briefly, the process involves crystallisation and microcrystal identification, EM/ stage height/tilt calibration and setting up data collection pro- Technical requirements and cesses. Data can now be processed and refined using standard method procedure XRC software, and structural models can be built using COOT, as outlined by Hattne et al., 2015. Thus, aside from Can the typical EM department implement MicroED? A com- an appropriate camera, MicroED is readily available to any prehensive guide with a trouble-shooting section was recently department equipped for single particle analysis, with sup- outlined (Nannenga and Gonen, 2018), explaining that stand- porting protocols and guides to aid researchers. ard equipment used in single particle analysis/typical EM ............................................................................................... .................................................................. 4 Downloaded from https://academic.oup.com/biohorizons/article/doi/10.1093/biohorizons/hzy013/5256436 by DeepDyve user on 16 October 2022 Bioscience Horizons � Volume 11 2018 Review article ............................................................................................... .................................................................. limit for XRC to be 0.5 Å (Jelsch et al., 2000). With more MicroED and XRC development, one might consider it plausible that MicroED To critically assess the efficacy and technical requirements of will surpass XRC. XRC currently achieves higher resolution MicroED, it is useful to compare it with the current leading than MicroED for macromolecules indicating that more methods in XRC. Synchrotron facilities and X-ray free-elec- refinement is required, but understanding how variables affect tron lasers (XFELs) generate continuous or pulsed X-rays use- resolution may significantly close this gap (see ‘Improvements ful for the determination of high-resolution XRC structures. for MicroED’ section). One notable difference between XRC Fortunately, MicroED and XFELs have acquired multiple and MicroED lies in the ‘phase problem’; while XRC rou- lysozyme structures (a molecule often used when developing tinely uses experimental methods to acquire phase informa- and validating crystallographic methods), which offers a use- tion, such methods are challenging in electron ful means of comparison. XFELs have yielded a 0.65 Å lyso- crystallography, requiring a dependence on XRC MR data or zyme structure (Wang et al., 2007) compared to the 1.8 Å ab initio phasing (see ‘Overcoming the phase problem’). A structure from MicroED (de la Cruz et al., 2017), demonstrat- better understanding of the critical MicroED parameters that ing that XRC techniques are currently still superior to affect resolution and development of phasing methods is MicroED for macromolecular crystallography. The difference required to improve resolution. in resolution may be partially due to MicroED’s unique chal- lenges with relation to the manner in which electrons/matter Challenges facing MicroED interactions give rise to IES and MES. Complications arising from IES for XRC/MicroED result from the addition of noise Nannenga et al. (2014b) noted that large crystals may con- (as lower-energy deflected electrons interfere with diffracted tribute more MES events, while separate work predicted that waves) and crystal degradation. It is interesting to note that ≥0.1 μm thickness would give unusable data (Subramanian the ratio of IES/ES is much lower for electron interactions et al., 2015). Surprisingly, proteinase K crystals of 0.1–1 μm than X-ray interactions, implying that with further refine- thickness acquired good resolution when studied by whereby ment, MicroED might have a lower IES-induced noise base- crystal disorder was suggested to explain the disparity line than XRC. Conversely, the fact that electrons interact between the experiment and predictions. Nonetheless, a nega- strongly with matter compared to X-rays creates significantly tive correlation between crystal volume and resolution was more scattering and more MES effects for MicroED, due to observed; thick crystals absorbed electrons more frequently. the weak scattering of X-rays by matter. In conventional This presents a problem; crystals must be large enough for XRC multiple scattering events are rare and presumed negli- observable ES, while minimising IES/MES. IES occurs three gible, with individual diffraction spots derived mostly from times more than ES to add noise, presenting an apparent chal- single ES events (aka ‘kinematical scattering’) rather than lenge (Henderson, 1995). Phasing methods present another more complicated multiple scattering events (aka ‘dynamical challenge; acquiring phases is challenging for X-ray (Taylor, scattering’). While dynamic scattering in rotational electron 2003), electron (Dorset, 1997) and neutron crystallography crystallography can be modelled (Oleynikov and Hovmöller, (Hauptman and Langs, 2003). XRC overcomes this using 2007), the complexity often obfuscates correct data interpret- Isomorphous Replacement (IR: heavy metal soaking of crys- ation, and should be reduced as much as possible. While tals to off-set phases and infer the original phases) and anom- MicroED might present with unique challenges, there are alous scattering (AS: whereby X-ray wavelength is altered to potentially some advantages that MicroED has over XRC. deposit energy into heavy metals which off-set phases to infer MicroED undoubtedly offers a cheaper and more accessible original phases) (Hendrickson and Ogata, 1997), but alternative to high-quality X-ray sources; many research insti- MicroED has not implemented IR/AS so far, an area of tutes have an EM suite on site with short wait times, whereas research that requires further investigation. synchrotron/XFEL facilities tend to only be found in regional hubs that typically require long wait times (Shi et al., 2016). MicroED also requires a single or a few crystals compared to Improvements for MicroED XFELs that require many, often taking time to produce. Crystallisation of macromolecules is often the critical time- Rotation scope consuming bottleneck in XRC studies. When crystallisation MicroED typically uses a ±70° crystal tilt, which becomes fails to produce large crystals (for synchrotron sources), such problematic depending on crystal symmetry; most crystals ‘failed’ conditions often produce MicroED-suitable micro- present with space groups (notations denoting the repeating crystals (Stevenson et al., 2014, 2016). While previously symmetry of proteins within unit cells) of P2 2 2 and P2 1 1 1 1 regarded as a by-product of improper crystallisation, (Wukovitz and Yeates, 1995), giving unique diffraction spots MicroED (and XFEL analysis) allows microcrystals to be use- over 180°.A ±70° tilt limits the accessible data to 140° from ful for structural studies, allowing crystallographers to make a single crystal (commonly known as the ‘missing wedge’ of use of more crystallisation conditions. MicroED methodology EM, Bartesaghi et al., 2008). A larger tilt might improve sam- currently uses wavelengths of ~0.025 Å compared to ~0.55 Å pling, and efforts to create innovative solutions (such as those in the highest resolution XRC structure, which is Cambrin attempted by Barnard et al., 1992) could be explored in the (Schmidt et al., 2011), with some considering the wavelength future. ............................................................................................... .................................................................. 5 Downloaded from https://academic.oup.com/biohorizons/article/doi/10.1093/biohorizons/hzy013/5256436 by DeepDyve user on 16 October 2022 Review article Bioscience Horizons � Volume 11 2018 ............................................................................................... .................................................................. Optimal crystal sizes Reducing electron dosage – electron wavelength Crystals require sufficient volume to provide signal, while excessively large crystals generate significant MES/IES, The incident beam might be adjusted to reduce electron dos- prompting to suggest an upper limit of 500 μm for optimal age; electrons with shorter wavelengths degrade crystals diffraction. Microcrystals can be acquired from larger slower and improve longevity (Glaeser, 1971). While (>500 μm) crystals (de la Cruz et al., 2017), but understand- MicroED uses an electron wavelength of 0.025 Å, no work ing the volume/resolution relationship is necessary. assessing wavelength effect on ES/IES/MES has been pub- Interestingly, the highest resolution data currently derives lished. Electron exposure was assessed in proteinase K and from small unit cell crystals with complex symmetry (see GSNQNNF crystals; high-resolution spots degraded at the Table 1), suggesting that perhaps the protein atomicity of a 0.025 Å wavelength (likely due to their typical low signal) crystal (i.e. protein molecules per Å ) or high solvent content with high-resolution data (<2.0 Å) being severely degraded at − −2 (e.g. when protein atomicity is low) contributes to volume/ ≥3e Å (Hattne et al., 2018). At low doses electrons were resolution relationships. It is already known that solvents gen- absorbed and shielded atoms from proper ES, e.g. at ≥0.9 e −2 − −2 erate noise (Bragg and Perutz, 1952)(Fraser, MacRae and Å disulphide bridges deteriorated and broke at 5.78 e Å , Suzuki, 1978), and are typically disordered and not uniformly while glutamate/aspartate residues lost carboxyl density at − −2 − −2 oriented in crystals (Weichenberger et al., 2015), which could ≥2e Å , which was absent at 5 e Å , meaning electron increase noise when protein atomicity is low (since solvents dosage contributes to structural quality and is an important occupy more volume). When reviewing MicroED structures element to control. Higher frequency electrons interact with (Table 1), an observable trend between protein atomicity and matter less frequently (Birkhoff, 1958); further research could resolution appears (Fig. 5), however this compares structures build on the work of Hattne et al. (2018) to determine over years from many laboratories, so variability is likely sig- whether shorter wavelengths might reduce interactions and nificant. Interestingly, Sawaya et al. (2016) analysed IES/MES to improve resolution. GNNQQNY in two different symmetries; the higher sym- metry crystal gave better resolution, providing insight into Overcoming the phase problem atomicity/resolution relationships. Further research is essen- tial to characterise the relationship between ES/MES/IES and 2DEC IR techniques have been attempted, but shielding of crystal volume and atomicity. metal nuclei by large electron clouds weakened IR interfer- ence (Ceska and Henderson, 1990), however some argued Reducing electron dosage – crystal rotation that heavy metal phase contributions can theoretically give sufficient information to solve phases (Burmester and speed Schröder, 1997). While MicroED has successfully used direct Electrons deposit less IES per ES than X-rays and less energy per methods to process data (Sawaya et al., 2016; de la Cruz IES, but interact more frequently with matter leading to quicker et al., 2017; Vergara et al., 2017), these computational meth- crystal degradation (Henderson, 1995). Electron dosage is kept ods require resolution to be 1.2 Å or better, (aka ‘Sheldricks below a critical threshold (Shi et al.,2013) and continuous rota- rule’, Sheldrick, 1990; Morris and Bricogne, 2003). tion reduces electron exposure to different crystal locations Consequently, this method is currently limited to small mole- (Nannenga et al., 2014a). Rotation speeds and frame rates are a cules, which tend to give the highest resolution data (Taylor, compromise between adequate sampling of diffraction spots 2010) as seen in the Sawaya et al. (2016) publication. Direct while preventing spot overlap (Hattne et al.,2015). methods have been used on macromolecular data of ~2 Å in XRC by prospective MR using archetypal α-helices/β-sheets A faster rotation with a shorter frame rate might reduce as search models, solving the structure of a previously electron dosage while preventing spot overlap. Rotation unknown 111-residue protein (Rodriguez et al., 2009). While speeds varied over experiments, from 0.1°/s (Vergara et al., useful, this may not be appropriate for macromolecules lack- 2017) to 0.29 /s (de la Cruz et al., 2017), but Table 1 shows ing sufficient α-helix and β-sheet structures. XRC also over- that resolution vs rotation speed (Fig. 6) and frame scope (the comes the phase problem using AS, requiring wave angle covered by a single frame, Fig. 7) appears to show no absorption by heavy atoms (Hendrickson and Ogata, 1997) apparent trend; different rotation speeds and frame scopes an effect that electrons are believed to be capable of achieving did not correlate with resolution; however, the data likely (Burmester and Schröder, 1997). A breakthrough is needed in include significant inter-laboratory and inter-assay variability MicroED analogous to IR/AS to find dependable phasing which may contribute to resolution quality. Indeed, the methods for macromolecular crystals. Gonen group attribute increased resolution over their three lysozyme structures to improved data collection and process- ing (Nannenga and Gonen, 2018). To date, no published Improvements to equipment and data work specifically dedicated to the effect of rotation speeds is processing available. Further research is required to probe rotation speed/frame rate contribution to diffraction patterns using The electron detector used in an EM plays a key part in detec- identical crystals to control for other variables. tion and analysis of diffracted waves. Traditionally, EMs ............................................................................................... .................................................................. 6 Downloaded from https://academic.oup.com/biohorizons/article/doi/10.1093/biohorizons/hzy013/5256436 by DeepDyve user on 16 October 2022 Bioscience Horizons � Volume 11 2018 Review article ............................................................................................... .................................................................. ............................................................................................... .................................................................. Table 1. crystal rotation speed, frame rate and resolution Molecule Crystal rotation Frame rate Rotation per Space Molecules Unit cell Volume per Resolution Reference 3 3 speed (°/s) (s/frame) frame (°) group per unit cell volume (Å ) molecule (Å ) (Å) Proteinase 0.090 4.0 0.360 P4 2 2 8 452 897 56 612 1.60 de la Cruz et al. 3 1 K (2017) 0.089 5.1 0.454 P4 2 2 8 452 267 56 533 1.71 Hattne et al. (2018) 3 1 0.090 4.0 0.360 P43212 8 457 776 57 222 1.75 Hattne et al. (2016) Lysozyme 0.090 4.0 0.360 P4 2 2 8 215 821 26 978 1.80 de la Cruz et al. 3 1 (2017) 0.152 0.5 0.076 P2 2 2 4 228 046 57 012 2.11 Clabbers et al. (2017) 1 1 0.450 2.0 0.900 P2 2 2 4 222 650 55 663 2.20 Xu et al. (2018) 1 1 0.090 4.0 0.360 P4 2 2 8 212 691 26 586 2.50 Nannenga et al. 3 1 (2014b) N/A 10.0 N/A P4 2 2 8 219 373 27 422 2.90 Shi et al. (2013) 3 1 Catalase 0.09 6.0 0.540 P2 2 2 4 2 125 468 531 367 3.20 Nannenga et al. 1 1 1 (2014a) a a a 0.75 2.0 1.500 P2 2 2 4 2466 129 616 532 3.20 Yonekura and Maki- 1 1 1 Yonekura (2016) GNNQQNY 0.30 2.0 0.600 P2 2 2 4 4625 1156 1.05 Sawaya et al. (2016) 1 1 1 0.30 2.0 0.600 P2 2 2726 1363 1.10 Sawaya et al. (2016) The table shows MicroED experiments that used crystals of four proteins (proteinase K, lysozyme, catalase and the hepta-peptide GNNQQNY) under different crystal rotation speeds, frame rates, crystal unit cells and volumes, along with the resultant resolution of the final structures. References to relevant publications are given in the table. The relationship of resolution plotted against protein atomicity (density within the unit cell, defined as the volume per molecule), crystal rotation speed and frame scope (the total angle sampled by a single frame) are shown in Figs 5– 7, respectively. Note: for this catalase experiment, various frame rates and rotation speeds were used from 0.5 to 1.0° over 1–3 s per frame median values of these ranges are displayed in the table. Downloaded from https://academic.oup.com/biohorizons/article/doi/10.1093/biohorizons/hzy013/5256436 by DeepDyve user on 16 October 2022 Review article Bioscience Horizons � Volume 11 2018 ............................................................................................... .................................................................. Resolution plotted against Resolution vs Frame atomicity scope 3.5 3.5 Very high atomicity Proteinase K 2.5 2.5 High Lysozyme atomicity 1.5 Catalase 1.5 Medium 1 atomicity GNNQQNY 0.5 0.5 Very low atomicity 0 200000 400000 600000 800000 01 2 Volume per protein molecule (Å3) Frame scope ( /frame) Figure 5. Resolution vs molecular crystal atomicity from MicroED Figure 7. Resolution vs frame scope from MicroED structures of common structures of common proteins. MicroED structures of proteinase K proteins. The relationship between resolution and frame scope (angle (blue), lysozyme (red), catalase (green) and GNNQQNY hepta-peptide sampled per camera frame) of all Table 1 molecules is shown above, (purple) are plotted to correlate structure resolution to unit cell colour coded for atomicity. Very high atomicity = 1000–1500 A /molecule, atomicity, i.e. volume per protein molecule. Crystals with higher high atomicity = 25 000–30 000 A /molecule, medium atomicity = 3 3 symmetry and tighter packing have less volume per molecule and 55 000–60 000 A /molecule, very low atomicity = 50 0000–650 000 A / higher atomicity. More densely packed cells with higher atomicity give molecule. higher final structure resolution. All experimental information is derived from Table 1. provided that electron dosage was not sufficiently high to dam- age the detector (Clough, Moldovan and Kirkland, 2014). Tim Resolution vs Rotation Gruene’s group is actively engaging in detector development speed (Nederlof et al., 2013; van Genderen et al.,2016; Clabbers 3.5 Very high et al.,2017; Matheson et al., 2017), recently demonstrating a atomicity novel ultrathin hybrid detector that samples smaller pixel areas 2.5 and has shorter dead time between frames, reducing data loss High from pixel-related overlap of intensities (Tinti et al., 2018). The atomicity detector circuitry was specifically designed with beam resistance 1.5 Medium to reduce damage at energies of 100, 200 and 300 keV, while 1 atomicity the Abrahams group applied new DIALS integration software 0.5 Very low (Clabbers et al.,2018). The Gonen group noted that IES elec- atomicity 0 trons may be filtered out of the diffraction data by using energy 0 0.5 1 filters to remove IES (Nannenga et al., 2014b) and improve Frame scope ( /s) SNR in line with other work (Yonekura, Maki-Yonekura and Namba, 2002; Leis et al., 2009), but have seemingly not yet uti- Figure 6. Resolution vs crystal rotation speed from MicroED structures of lised this approach. Future innovation and development of common proteins. The relationship between resolution and crystal equipment and data processing to improve SNR will greatly rotation speed of all Table 1 molecules are shown above, colour coded for 3 improve resolution, and energy filters may be an attractive atomicity. Very high atomicity = 1000–1500 A /molecule, high atomicity = 3 3 option to remove IES from diffracting waves. 25 000–30 000 A /molecule, medium atomicity = 55 000–60 000 A / molecule, Very low atomicity = 500 000–650 000 A /molecule. MicroED in the wider community have used two types of indirect detectors, Charge-Coupled Devices and Complementary Metal-Oxide-Semiconductors The worldwide structural biology community is increasingly (CMOS). Both convert electrons to photons using a scintillator using MicroED in structural studies and are taking an interest which are then detected by the sensor (a comparison is outlined in the technical and theoretical aspects. One group published in Allé et al., 2016). The scintillation inherently introduces scat- Lysozyme structures to 2.2 Å and commented on MicroED’s ter, and bulky detectors can scatter photons laterally into neigh- main challenges, stating that crystal degradation and gonio- bouring pixels, spreading the peak, making indirect detectors metric imperfections in crystal rotation are key challenges to sub-optimal for electron crystallography (Faruqi, 2001). Hybrid overcome (Xu et al., 2018). The first use of MicroED to study detectors detect charge directly from electrons to prevent amyloid fibrils outside of the Gonen group was published scintillator-associated scatter but due to the strong electron/mat- recently, solving the (FUS) amyloid core to 0.73 Å (Luo et al., ter interactions, they are easily damaged by the electron beam. 2018). Other groups are attempting to improve the technique Initial studies suggested that hybrid detectors may nonetheless and contribute to MicroED theory; some are researching and perform better than indirect detectors to generate signal, commenting on the relationship between microcrystal size ............................................................................................... .................................................................. Resolutiopn (Å) Resolution (Å) Resolution (Å) Downloaded from https://academic.oup.com/biohorizons/article/doi/10.1093/biohorizons/hzy013/5256436 by DeepDyve user on 16 October 2022 Bioscience Horizons � Volume 11 2018 Review article ............................................................................................... .................................................................. and diffraction spots (Williams et al., 2017), while others sug- a keen interest in crystallography and wrote this article during gest that maps generated from MicroED might be utilising his M.Res dissertation in Structural Molecular Biology at sub-optimal MR methods as structure factors fail to take Imperial College London, after a BSc in Biomedical Science, negative contributions from IES into account (Wang, 2017). MSc in Biomedical Blood Science at Keele University, and a While the technique is still very young, it is likely that the PGCert in Protein Crystallography at Birkbeck. While work- future will see a growing rate of MicroED structures being ing in clinical laboratories (transfusion and emergency medi- solved and published as the community begins to delve deeper cine in Salisbury and Addenbrookes), Rob undertook the into the intricacies of the method and develop it further. PGCert and decided to pursue a career in basic research. He While the work of the Gonen and Abrahams group have then worked as a Tissue Acquisition Officer for the BCI, undoubtedly contributed significantly to the technique, the before working at UCB Celltech as an Associate Scientist in wider community is now beginning to realise the potential biopharmaceutical method development. In September 2017 that this novel method has within structural biology. he began his MRes, with the PhD beginning in September 2018. Post-PhD Rob would like to contribute to MicroED development within structural biology. Conclusion MicroED is an exciting crystallographic method utilising short Acknowledgements wavelengths to offer better theoretical resolution limits than XRC, while simultaneously sampling more reciprocal space per I would like to thank my supervisor, Thomas Meier, for his illumination. While modern macromolecular EM work typically guidance and feedback. I would like to thank Morgan Beeby relies on single particle analysis without phasing methods or crys- for his invaluable insight concerning electron microscopy and tals (Orlova and Saibil, 2004), MicroED offers significantly high- electron/matter interactions. I would also like to thank Jose er resolution; a characteristic that makes crystallographic Rodriguez for his fruitful correspondence concerning methods so popular (Wilkins, 2013). MicroED offers an afford- MicroED, current research and future directions for the able and accessible XRC alternative by utilising EMs commonly method. used in many structural laboratories with minor adjustments. Continuous rotation MicroED and XRC diffraction software have improved the method, but more work is required to charac- References terise the effect of crystal volume, unit cell packing, electron wave- length, crystal rotation speed, frame rate capture, detector type Abrahams, J. P. (2010) The strong phase object approximation may and the use of energy filters on resolution. While MicroED works allow extending crystallographic phases of dynamical electron dif- very well with small molecule crystals, its use on larger molecules fraction patterns of 3D protein nano-crystals, Zeitschrift fur lacks comparable resolution, likely due to solvent/protein ratios Kristallographie-Crystalline Materials, 225, 67–76. of large unit cells with low symmetry crystals creating noise. For Allé, P., Wenger, E., Dahaoui, S. et al. (2016) Comparison of CCD, CMOS high macromolecular resolution, MicroED depends on MR from and hybrid pixel X-ray detectors: detection principle and data qual- XRC studies. To be a stand-alone technique, MicroED must ity, Physica Scripta, 91, 063001. develop phasing methods for macromolecular datasets. Computational methods have somewhat addressed this challenge Andrews, S., Reichow, S. L. and Gonen, T. (2008) Electron crystallog- at resolutions of 1.2 Å and 2.0 Å, but macromolecules typically raphy of aquaporins, IUBMB Life, 60, 430–436. give lower resolutions than 1.2 Å and even at 2.0 Å may not pre- Barnard, D. P., Turner, J. N., Frank, J. et al. (1992) A 360° single-axis tilt sent sufficient secondary structures to utilise the phasing method, stage for the high-voltage electron microscope, Journal of therefore phasing is particularly troublesome and must be Microscopy, 167, 39–48. addressed by developing innovative phasing methods. As it stands, MicroED is a powerful technique for small molecules, elu- Bartesaghi, A., Sprechmann, P., Liu, J. et al. (2008) Classification and 3D cidating these structures in unprecedented detail and might per- averaging with missing wedge correction in biological electron haps be considered better than XRC in terms of resolution and tomography, Journal of Structural Biology, 162 (3), 436–450. practicality for this class of molecules. Further work and innov- Birkhoff, R. D. (1958) The passage of fast electrons through matter, in ation will be essential for solving the MicroED phase problem, Flügge S. (ed), Corpuscles and Radiation in Matter II, 1st edition, and given that the wider community is beginning to contribute to Springer, Berlin/Heidelberg, pp. 53–138. the technique, we may very well be on the cusp of producing the next series of breakthroughs for this impressive technique. Bragg, W. H. and Bragg, W. L. (1913) The reflection of X-rays by crystals, Proceedings of the Royal Society London, 88, 428–438. Author biography Bragg, W. H. and Perutz, M. F. (1952) The external form of the haemo- globin molecule. I, Acta Crystallographica, 5, 277–283. Rob is now a BBSRC PhD candidate of Structural Biology at the University of Bristol, working with Dr Paul Race studying Burmester, C. and Schröder, R. R. (1997) Solving the phase problem in pro- the structure and function of streptococcal adhesins. Rob has tein electron crystallography: multiple isomorphous replacement and ............................................................................................... .................................................................. 9 Downloaded from https://academic.oup.com/biohorizons/article/doi/10.1093/biohorizons/hzy013/5256436 by DeepDyve user on 16 October 2022 Review article Bioscience Horizons � Volume 11 2018 ............................................................................................... .................................................................. anomalous dispersion as alternatives to imaging, Scanning Microscopy, precession technique and high-resolution electron microscopy, 11, 323–334. Acta Crystallographica A—Foundations and Advances, 59, 117–126. Ceska, T. A. and Henderson, R. (1990) Analysis of high-resolution elec- Georgieva, D., Abrahams, J. P., Zandbergen, H. et al. (2007a) Solving ab- tron diffraction patterns from purple membrane labelled with initio protein and nanostructures with precession electron diffrac- heavy-atoms, Journal of Molecular Biology, 213, 539–560. tion, Microscopy and Microanalysis, 13, 952–953. Cipriani, F., Roewer, M., Landret, C. et al. (2012) CrystalDirect: a new Georgieva, D. G., Kuil, M. E., Oosterkamp, T. H. et al. (2007b) Heterogeneous method for automated crystal harvesting based on laser-induced nucleation of three-dimensional protein nanocrystals, Acta photoablation of thin films, Acta Crystallographica Section D— Crystallographica Section D—Biological Crystallography, 63, 564–570. Biological Crystallography, 68, 1393–1399. Gjonnes, J., Hansen, V., Berg, B. S. et al. (1998) Structure model for the Clabbers, M. T. B. and Abrahams, J. P. (2018) Electron diffraction and phase almfe derived from three-dimensional electron diffraction three-dimensional crystallography for structural biology, intensity data collected by a precession technique. comparison Crystallography Reviews, 24 (3), 176–204. with convergent-beam diffraction, Acta Crystallographica Section A- Foundations and Advances, 54, 306–319. Clabbers, M. T. B., Gruene, T., Parkhurst, J. M. et al. (2018) Electron dif- fraction data processing with DIALS, Acta Crystallographica section Glaeser, R. M. (1971) Limitations to significant information in biological D—Structural Biology, 74, 506–518. electron microscopy as a result of radiation damage, Journal of Ultrastructure Research, 36, 466–482. Clabbers, M. T. B., van Genderen, E., Wan, W. et al. (2017) Protein struc- ture determination by electron diffraction using a single three- Gonen, T., Cheng, Y. F., Sliz, P. et al. (2005) Lipid-protein interactions in dimensional nanocrystal, Acta Crystallographica Section D— double-layered two-dimensional AQP0 crystals, Nature, 438, 633–638. Structural Biology, 73, 738–748. Guenther, E. L., Ge, P., Trinh, H. et al. (2018) Atomic-level evidence for Clough, R. N., Moldovan, G. and Kirkland, A. I. (2014) Direct detectors for packing and positional amyloid polymorphism by segment from electron microscopy, Journal of Physics Conference Series, 522, 012046. TDP-43 RRM2, Nature Structural & Molecular Biology, 25, 311–319. Curry, S. (2015) Structural biology: a century-long journey into an Hattne, J., Reyes, F. E., Nannenga, B. L. et al. (2015) MicroED data collec- unseen world, Interdisciplinary Science Reviews, 40, 308–328. tion and processing, Acta Crystallographica A-Foundation and Advances, 71, 353–360. de la Cruz, M. J., Hattne, J., Shi, D. et al. (2017) Atomic-resolution struc- Hattne, J., Shi, D., de la Cruz, M. J. et al. (2016) Modeling truncated pixel tures from fragmented protein crystals with the cryoEM method values of faint reflections in MicroED images, Journal of Applied MicroED, Nature Methods, 14 (4), 399–402. Crystallography, 49, 1029–1034. Dorset, D. L. (1997) Direct phase determination in protein electron crystal- Hattne, J., Shi, D., Glynn, C. et al. (2018) Analysis of global and site- lography: the pseudo-atom approximation, Proceedings of the National specific radiation damage in cryo-EM, Structure (London, England : Academy of Sciences of the United States of America, 94, 1791–1794. 1993), 26, 759–766. Eckert, M. (2012) Max von Laue and the discovery of X-ray diffraction in Hauptman, H. A. and Langs, D. A. (2003) The phase problem in neutron 1912, Annalen Der Physik, 524 (5), A85. crystallography, Acta Crystallographica Section A, 59, 250–254. Erickson, H. P. (2009) size and shape of protein molecules at the nano- meter level determined by sedimentation, gel filtration, and elec- Henderson, R. (1995) The potential and limitations of neutrons, elec- tron microscopy, Biological Procedures Online, 11, 32–51. trons and X-rays for atomic-resolution microscopy of unstained bio- logical molecules, Quarterly Reviews of Biophysics, 28, 171–193. Ezkurdia, I., Juan, D., Manuel Rodriguez, J. et al. (2014) Multiple evi- dence strands suggest that there may be as few as 19 000 human Henderson, R. and Unwin, P. (1975) 3-Dimensional model of purple protein-coding genes, Human Molecular Genetics, 23, 5866–5878. membrane obtained by electron-microscopy, Nature, 257, 28–32. Faruqi, A. R. (2001) Prospects for hybrid pixel detectors in electron crys- Hendrickson, W. A. and Ogata, C. M. (1997) Phase determination from tallography, Nuclear Instruments and Methods in Physics Research A, multiwavelength anomalous diffraction measurements, Methods in 466, 146–154. Enzymology, 276, 494–523. Fraser,R.D.B., MacRae,T.P.and Suzuki, E.(1978) An improved method for Hughes, M. P., Sawaya, M. R., Goldschmidt, L. et al. (2017) Low- calculating the contribution of solvent to the X-ray diffraction pattern complexity domains adhere by reversible amyloid-like interactions of biological molecules, Journal of Applied Crystallography, 11, 693–694. between kinked β-sheets. bioRxiv (in press) doi: 10:10:/153817 Gallagher-Jones, M., Glynn, C., Boyer, D. R. et al. (2018) Sub-angstrom Jaskolski, M., Dauter, Z. and Wlodawer, A. (2014) A brief history of cryo-EM structure of a prion protofibril reveals a polar clasp, Nature macromolecular crystallography, illustrated by a family tree and its Structural & Molecular Biology, 25 (2), 131–134. Nobel fruits, FEBS, 281, 3985–4009. Gemmi, M., Zou, X., Hovmöller, S. et al. (2003) Structure of Ti2P solved Jelsch, C., Teeter, M. M., Lamzin, V. et al. (2000) Accurate protein crystal- by three-dimensional electron diffraction data collected with the lography at ultra-high resolution: valence electron distribution in ............................................................................................... .................................................................. 10 Downloaded from https://academic.oup.com/biohorizons/article/doi/10.1093/biohorizons/hzy013/5256436 by DeepDyve user on 16 October 2022 Bioscience Horizons � Volume 11 2018 Review article ............................................................................................... .................................................................. crambin, Proceedings of the National Academy of Sciences of the from submicrometre three-dimensional protein crystals, Acta United States of America, 97, 3171–3176. Crystallographica Section D—Biological Crystallography, 69, 1223–1230. Jiang, L., Georgieva, D., Zandbergen, H. W. et al. (2009) Unit-cell determin- Oleynikov, P. and Hovmöller, S. (2007) Precession electron diffraction: ation from randomly oriented electron-diffraction patterns, Acta observed and calculated intensities, Ultramicroscopy, 107, 523–533. Crystallographica Section D—Biological Crystallography, 65, 625–632. Orlova, E. V. and Saibil, H. R. (2004) Structure determination of macro- Koshland, D. E. (1994) The key-lock theory and the induced fit theory, molecular assemblies by single-particle analysis of cryo-electron Angewandte Chemie-International Edition, 33, 2375–2378. micrographs, Current Opinion in Structural Biology, 14, 584–590. Krotee, P., Rodriguez, J. A., Sawaya, M. R. et al. (2017) Atomic structures Rodriguez, J. A., Eisenberg, D. S. and Gonen, T. (2017) Taking the meas- of fibrillar segments of hIAPP suggest tightly mated beta-sheets are ure of MicroED, Current Opinion in Structural Biology, 46, 79–86. important or cytotoxicity, Elife, 6, e19273. Rodriguez, D. D., Grosse, C., Himmel, S. et al. (2009) Crystallographic ab Leis, A., Rockel, B., Andrees, L. et al. (2009) Visualizing cells at the nano- initio protein structure solution below atomic resolution, Nature scale, Trends in Biochemical Sciences, 34, 60–70. Methods, 6, U39. Lindsay, D. G. and Shall, S. (1969) Acetoacetylation of insulin, Rodriguez, J. A., Ivanova, M. I., Sawaya, M. R. et al. (2015) Structure of Biochemical Journal, 115, 587–595. doi:10.1042/bj1150587 the fig- the toxic core of alpha-synuclein from invisible crystals, Nature, 525, ure used in Figure 1 was original published by Portland Press under 486–490. the STM permission guidelines (https://www.stm-assoc.org/ Rolland, T., Tasan, M., Charloteaux, B. et al. (2014) A proteome-scale copyright-legal-affairs/permissions/permissions-guidelines/), and map of the human interactome network, Cell, 159, 1212–1226. request has been granted to use the figure. Sawaya, M. R., Rodriguez, J., Cascio, D. et al. (2016) Ab initio structure Luo, F., Gui, X., Zhou, H. et al. (2018) Atomic structures of FUS LC determination from prion nanocrystals at atomic resolution by domain segments reveal bases for reversible amyloid fibril forma- MicroED, Proceedings of the National Academy of Sciences of the tion, Nature Structural & Molecular Biology, 25, 341–346. United States of America, 113, 11232–11236. Matheson, J., Moldovan, G., Kirkland, A. et al. (2017) Testing and com- Schmidt, A., Teeter, M., Weckert, E. et al. (2011) Crystal structure of small parison of imaging detectors for electrons in the energy range protein crambin at 0.48 Å resolution, Acta Crystallographica 10–20 keV, Journal of Instrumentation, 12, C11016. International Section F- Structural Biology Communications, 67, 424–428. Conference on Position Sensitive Detectors. Sheldrick, G. M. (1990) Phase annealing in Shelx-90—direct methods Morris, R. J. and Bricogne, G. (2003) Sheldrick’s 1.2 angstrom rule and for larger structures, Acta Crystallographica Section A, 46, 467–473. beyond, Acta Crystallographica Section D—Biological Crystallography, 59, 615–617. Shi, D., Nannenga, B. L., de la Cruz, M. J. et al. (2016) The collection of MicroED data for macromolecular crystallography, Nature Protocols,11, 895–904. Nannenga, B. L. and Gonen, T. (2014) Protein structure determination by MicroED, Current Opinion in Structural Biology, 27, 24–31. The fig- Shi, D., Nannenga, B. L., Iadanza, M. G. et al. (2013) Three-dimensional ure used in Figure 3 was originally published by Elsevier and electron crystallography of protein microcrystals, Elife, 2, e01345. request has been granted to use the figure by the Copyright The figure used in Figure 2 was original published under the Clearance Centre (license 4367131293431). Creative Commons Attribution 3.0 Unported (CC BY 3.0) license (https://creativecommons.org/licenses/by/3.0/), and request has Nannenga, B. L. and Gonen, T. (2018) MicroED: a versatile cryoEM meth- been granted to use the figure. od for structure determination, Emerging Topics in Life Sciences,2, 1–8. Stevenson, H. P., Lin, G., Barnes, C. O. et al. (2016) Transmission electron microscopy for the evaluation and optimization of crystal growth, Nannenga, B. L., Shi, D., Hattne, J. et al. (2014a) Structure of catalase Acta Crystallographica Section D—Structural Biology, 72, 603–615. determined by MicroED, Elife, 3, e03600. Stevenson, H. P., Makhov, A. M., Calero, M. et al. (2014) Use of transmis- Nannenga, B. L., Shi, D., Leslie, A. G. W. et al. (2014b) High-resolution sion electron microscopy to identify nanocrystals of challenging structure determination by continuous-rotation data collection in protein targets, Proceedings of the National Academy of Sciences of MicroED, Nature Methods, 11, 927–930. The figure used in Figure 4 the United States of America, 111, 8470–8475. was original published by Springer Nature and request has been granted to use the figure by the Copyright Clearance Centre Subramanian, G., Basu, S., Liu, H. et al. (2015) Solving protein nanocrys- (license 4367140499674). tals by cryo-EM diffraction: Multiple scattering artifacts, Ultramicroscopy, 148, 87–93. Nederlof, I., Georgieva, D. and Abrahams, J. P. (2011) Electron diffrac- tion of submicron 3D protein crystals, Acta Crystallographica Section Sweet, R. M. (1985) Introduction to Crystallography, Methods in A-Foundation and Advances, 67, C228. Enzymology, 114, 19–46. Nederlof, I., van Genderen, E., Li, Y. et al. (2013) A Medipix quantum Taylor, G. (2003) The phase problem, Acta Crystallographica Section D— area detector allows rotation electron diffraction data collection Biological Crystallography, 59, 1881–1890. ............................................................................................... .................................................................. 11 Downloaded from https://academic.oup.com/biohorizons/article/doi/10.1093/biohorizons/hzy013/5256436 by DeepDyve user on 16 October 2022 Review article Bioscience Horizons � Volume 11 2018 ............................................................................................... .................................................................. Taylor, G. L. (2010) Introduction to phasing, Acta Crystallographica Wilkins, S. W. (2013) Celebrating 100 years of X-ray crystallography, Section D—Biological Crystallography, 66, 325–338. Acta Crystallographica Section A, 69, 1–4. Tinti, G., Frojdh, E., van Genderen, E. et al. (2018) Electron crystallog- Williams, S. R., Dilanian, R. A., Quiney, H. M. et al. (2017) Analysis of dif- raphy with the EIGER detector, IUCRJ, 5, 190–199. fracted intensities from finite protein crystals with incomplete unit cells, Crystals, 7, 220. Unwin, P. and Henderson, R. (1975) Molecular-structure determination by electron-microscopy of unstained crystalline specimens, Journal Wisedchaisri,G., Reichow, S. L. andGonen,T.(2011) Advances in of Molecular Biology, 94, 425–440. structural and functional analysis of membrane proteins by elec- tron crystallography, Structure (London, England : 1993),19, van Genderen, E., Clabbers, M. T. B., Das, P. P. et al. (2016) Ab initio struc- 1381–1393. ture determination of nanocrystals of organic pharmaceutical com- pounds by electron diffraction at room temperature using a Timepix Wukovitz, S. W. and Yeates, T. O. (1995) Why protein crystals favor quantum area direct electron detector, Acta Crystallographica some space-groups over others, Nature Structural Biology,2, A-Foundation and Advances,72, 236–242. 1062–1067. Vergara, S., Lukes, D. A., Martynowycz, M. W. et al. (2017) MicroED struc- Xu, H., Lebrette, H., Yang, T. et al. (2018) A rare lysozyme crystal form ture of Au-146(p-MBA)(57) at subatomic resolution reveals a twinned solved using highly redundant multiple electron diffraction data- FCC cluster , Journal of Physical Chemistry Letters,8,5523–5530. sets from micron-sized crystals, Structure (London, England : 1993), 26, 667–675. Wang, J. (2017) On the appearance of carboxylates in electrostatic potential maps, Protein Science, 26, 396–402. Yonekura, K. and Maki-Yonekura, M. (2016) Refinement of cryo-EM structures using scattering factors of charged atoms, Journal of Wang, J.,Dauter, M.,Alkire, R. et al.(2007) Tricliniclysozymeat0.65 Å reso- Applied Crystallography, 49, 1517–1523. lution, Acta Crystallographica Section D—Biological Crystallography, 63, 1254–1268. Yonekura, K., Maki-Yonekura, S. and Namba, K. (2002) Quantitative comparison of zero-loss and conventional electron diffraction from Weichenberger, C. X., Afonine, P. V., Kantardjieff, K. et al. (2015) The solv- two-dimensional and thin three-dimensional protein crystals, ent component of macromolecular crystals, Acta Crystallographica Biophysical Journal, 82, 2784–2797. Section D—Biological Crystallography, 71, 1023–1038. ............................................................................................... ..................................................................

Journal

BioScience HorizonsOxford University Press

Published: Jan 1, 2018

There are no references for this article.