Perspective: Emerging strategies for determining atomic-resolution structures of macromolecular complexes within cells

In principle, electron cryo-tomography (cryo-ET) of thin portions of cells provides high-resolution images of the three-dimensional spatial arrangement of all members of the proteome. In practice, however, radiation damage creates a tension between recording images at many different tilt angles, but at correspondingly reduced exposure levels, versus limiting the number of tilt angles in order to improve the signal-to-noise ratio (SNR). Either way, it is challenging to read the available information out at the level of atomic structure. Here, we first review work that explores the optimal strategy for data collection, which currently seems to favor the use of a limited angular range for tilting the sample or even the use of a single image to record the high-resolution information. Looking then to the future, we point to the alternative of so-called “deconvolution microscopy”, which may be applied to tilt-series or optically-sectioned, focal series data. Recording data as a focal series has the advantage that little or no translational alignment of frames might be needed, and a three-dimensional reconstruction might require only 2/3 the number of images as does standard tomography. We also point to the unexploited potential of phase plates to increase the contrast, and thus to reduce the electron exposure levels while retaining the ability align and merge the data. In turn, using much lower exposures per image could have the advantage that high-resolution information is retained throughout the full data-set, whether recorded as a tilt series or a focal series of images.

Comparative image simulations for phase-plate transmission electron microscopy

Numerous physical phase plates (PP) for phase-contrast enhancement in transmission electron microscopy (TEM) have been proposed and studied with the hole-free or Volta PP having a high impact and interest in recent years. This study is concerned with comparative TEM image simulations considering realistic descriptions of various PP approaches and samples from three different fields of application covering a large range of object sizes. The simulated images provide an illustrative characterization of the typical image appearance and common artifacts of the different PPs and the influence of simulation parameters especially important for PP simulations. A quantitative contrast analysis shows the superior phase-shifting properties of the hole-free phase plate for biological applications and the benefits of adjustable phase plates. The application of PPs in high-resolution TEM imaging, especially of weak-phase objects such as (atomically thin) 2D materials, is shown to increase image interpretability. The software with graphical user interface written and used for the presented simulations is available for free usage.

Oxide-free aC/Zr0.65Al0.075Cu0.275/aC phase plates for transmission electron microscopy

Thin-film phase plates (PP) have become a valuable tool for the imaging of organic objects in transmission electron microscopy (TEM). The thin film usually consists of amorphous carbon (aC), which undergoes rapid aging under intense illumination with high-energy electrons. The limited lifetime of aC film PPs calls for alternative PP materials with improved material stability. This work presents thin-film PPs fabricated from the metallic glass alloy Zr_0.65Al_0.075Cu_0.275 (ZAC), which was identified as a promising PP material with beneficial properties, such as a large inelastic mean free path. An adverse effect of the ZAC alloy is the formation of a surface oxide layer in ambient air, which reduces the electrical conductivity and causes electrostatic charging in the electron beam. To avoid surface oxidation, the ZAC alloy is enclosed by thin aC layers. The resulting aC/ZAC/aC layer system is used to fabricate Zernike and Hilbert PPs. Phase-contrast TEM imaging is demonstrated for a sample of carbon nanotubes, which show strong contrast enhancement in PP TEM images.

Carbon contamination in scanning transmission electron microscopy and its impact on phase-plate applications

We analyze electron-beam induced carbon contamination in a transmission electron microscope. The study is performed on thin films potentially suitable as phase plates for phase-contrast transmission electron microscopy. Electron energy-loss spectroscopy and phase-plate imaging is utilized to analyze the contamination. The deposited contamination layer is identified as a graphitic carbon layer which is not prone to electrostatic charging whereas a non-conductive underlying substrate charges. Several methods that inhibit contamination are evaluated and the impact of carbon contamination on phase-plate imaging is discussed. The findings are in general interesting for scanning transmission electron microscopy applications.

Imaging complement by phase-plate cryo-electron tomography from initiation to pore formation

Phase plates in cryo-electron tomography (cryoET) improve contrast, increasing the ability to discern separate molecules and molecular complexes in dense biomolecular environments. Here, we applied this new technology to the activation of the human complement system. Binding of C1 to antigen-antibody complexes initiates a cascade of proteolytic events that deposits molecules onto adjacent surfaces and terminates with the formation of membrane-attack-complex (MAC) pores in the targeted membranes. We imaged steps in this process using a Volta phase plate mounted on a Titan Krios equipped with a Falcon-II direct electron detector. The data show patches of single-layer antibodies on the surface and C1 bound to antibody platforms, with ca. ∼4% of instances where C1r and C1s proteases have dissociated from C1, and potentially instances of C1 transiently interacting with its substrate C4 or product C4b. Next, extensive deposition of C4b and C3b molecules is apparent, although individual molecules cannot always be properly distinguished with the current methods. Observations of MAC pores include formation of both single and composite pores, and instances of potential soluble-MAC dissociation upon failure of membrane insertion. Overall, application of the Volta phase plate cryoET markedly improved the contrast in the tomograms, which allowed for individual components to be more readily interpreted. However, variability in the phase shift induced by the phase-plate during the course of an experiment, together with incomplete sampling during tomogram acquisition, limited the interpretability of the resulting tomograms. Our studies exemplify the potential in studying molecular processes with complex spatial topologies by phase-plate cryoET.

Imaging by Zernike phase plates in the TEM

The images produced from simple phase objects, lenses and Zernike phase plates when all have rotational symmetry can be calculated by 1D Fourier-Bessel transforms. For a simple disc object producing a uniform phase shift over its diameter, the resulting image can be defined for any size of object phase change. The monotonic range of intensity variation with object phase is found to depend strongly on the phase change introduced by the phase plate; this property of the system is not well predicted by the weak phase approximation. The effect of spreading the phase transition at the plate over a range of radius is beneficial if the plate phase change is sufficiently small. Weak-phase calculations for a phase distribution more typical of a spherical object are also shown.

CTFFIND4: Fast and accurate defocus estimation from electron micrographs

CTFFIND is a widely-used program for the estimation of objective lens defocus parameters from transmission electron micrographs. Defocus parameters are estimated by fitting a model of the microscope's contrast transfer function (CTF) to an image's amplitude spectrum. Here we describe modifications to the algorithm which make it significantly faster and more suitable for use with images collected using modern technologies such as dose fractionation and phase plates. We show that this new version preserves the accuracy of the original algorithm while allowing for higher throughput. We also describe a measure of the quality of the fit as a function of spatial frequency and suggest this can be used to define the highest resolution at which CTF oscillations were successfully modeled.

Towards an optimum design for electrostatic phase plates

Charging of physical phase plates is a problem that has prevented their routine use in transmission electron microscopy of weak-phase objects. In theory, electrostatic phase plates are superior to thin-film phase plates since they do not attenuate the scattered electron beam and allow freely adjustable phase shifts. Electrostatic phase plates consist of multiple layers of conductive and insulating materials, and are thus more prone to charging than thin-film phase plates, which typically consist of only one single layer of amorphous material. We have addressed the origins of charging of Boersch phase plates and show how it can be reduced. In particular, we have performed simulations and experiments to analyze the influence of the insulating Si3N4 layers and surface charges on electrostatic charging. To optimize the performance of electrostatic phase plates, it would be desirable to fabricate electrostatic phase plates, which (i) impart a homogeneous phase shift to the unscattered electrons, (ii) have a low cut-on frequency, (iii) expose as little material to the intense unscattered beam as possible, and (iv) can be additionally polished by a focused ion-beam instrument to eliminate carbon contamination accumulated during exposure to the unscattered electron beam (Walter et al., 2012, Ultramicroscopy, 116, 62-72). We propose a new type of electrostatic phase plate that meets the above requirements and would be superior to a Boersch phase plate. It consists of three free-standing coaxial rods converging in the center of an aperture (3-fold coaxial phase plate). Simulations and preliminary experiments with modified Boersch phase plates indicate that the fabrication of a 3-fold coaxial phase plate is feasible.

Multi-pore carbon phase plate for phase-contrast transmission electron microscopy

A new fabrication method of carbon based phase plates for phase-contrast transmission electron microscopy is presented. This method utilizes colloidal masks to produce pores as well as disks on thin carbon membranes for phase modulation. Since no serial process is involved, carbon phase plate membranes containing hundreds of pores can be mass-produced on a large scale, which allows "disposal" of contaminated or degraded phase modulating objects after use. Due to the spherical shape of the mask colloid particles, the produced pores are perfectly circular. The pore size and distribution can be easily tuned by the mask colloid size and deposition condition. By using the stencil method, disk type phase plates can also be fabricated on a pore type phase plate. Both pore and disk type phase plates were tested by measuring amorphous samples and confirmed to convert the sinus phase contrast transfer function to the cosine shape.

Vortex beam production and contrast enhancement from a magnetic spiral phase plate

Electron vortex beam probes offer the possibility of mapping magnetic moments with atomic resolution. In this work we consider using the stray magnetic field produced from a narrow ferromagnetic rod magnetised along its long axis to produce a vortex beam probe, as an alternative to the currently used holographic apertures or gratings. We show through numerical modelling, electron holography observations and direct imaging of the electron probe, that a long narrow ferromagnetic rod induces a phase shift in the wave-function of passing electrons that approximately describes a helix in the regions near its ends. Directing this rod towards the optical axis of a charged-particle beam probe forming system at a limiting aperture position, with the free-end sufficiently close to the axis, is shown to offer a point spread function composed of vortex modes, with evidence of this appearing in observations of the electron probe formed from inserting a micro-fabricated CoFe rod into the beam path of a 300 keV transmission electron microscope (TEM). If the rod is arranged to contain the magnetic flux of h/e, thus producing a maximum phase shift of 2π, it produces a simple spiral-like phase contrast transfer function for weak phase objects. In this arrangement the ferromagnetic rod can be used as a phase plate, positioned at the objective aperture position of a TEM, yielding enhanced image contrast which is simulated to be intermediate between comparable Zernike and Hilbert phase plates. Though this aspect of the phase plate performance is not demonstrated here, agreement between our observations and models for the probe formed from an example rod containing a magnetic flux of ~2.35h/e, indicate this phase plate arrangement could be a simple means of enhancing contrast and gaining additional information from TEM imaged weak phase samples, while also offering the capability to produce vortex beam probes. However, steps still need to be taken to either remove or improve the support membrane for the rod in our experiments to reduce any effects from charging in the phase plate.

Methods for testing Zernike phase plates and a report on silicon-based phase plates with reduced charging and improved ageing characteristics

Imaging with Zernike phase plates is increasingly being used in cryo-TEM tomography and cryo-EM single-particle applications. However, rapid ageing of the phase plates, together with the cost and effort in producing them, present serious obstacles to widespread adoption. We are experimenting with phase plates based on silicon chips that have thin windows; such phase plates could be mass-produced and made available at moderate cost. The windows are coated with conductive layers to reduce charging, and this considerably extends the useful life of the phase plates compared to traditional pure-carbon phase plates. However, a compromise must be reached between robustness and transmission through the phase-plate film. Details are given on testing phase-plate performance by means of imaging an amorphous thin film and evaluating the power spectra of the images.

Minimizing electrostatic charging of an aperture used to produce in-focus phase contrast in the TEM

Microfabricated devices designed to provide phase contrast in the transmission electron microscope must be free of phase distortions caused by unexpected electrostatic effects. We find that such phase distortions occur even when a device is heated to 300 °C during use in order to avoid the formation of polymerized, carbonaceous contamination. Remaining factors that could cause unwanted phase distortions include patchy variations in the work function of a clean metal surface, radiation-induced formation of a localized oxide layer, and creation of a contact potential between an irradiated area and the surround due to radiation-induced structural changes. We show that coating a microfabricated device with evaporated carbon apparently eliminates the problem of patchy variation in the work function. Furthermore, we show that a carbon-coated titanium device is superior to a carbon-coated gold device, with respect to radiation-induced electrostatic effects. A carbon-coated, hybrid double-sideband/single-sideband aperture is used to record in-focus, cryo-EM images of monolayer crystals of streptavidin. Images showing no systematic phase error due to charging are achievable under conditions of low-dose data collection. The contrast in such in-focus images is sufficient that one can readily see individual streptavidin tetramer molecules. Nevertheless, these carbon-coated devices perform well for only a limited length of time, and the cause of failure is not yet understood.

The use of a central beam stop for contrast enhancement in TEM imaging

Dark field TEM imaging using a stop of the central beam (DF-000) is reported. It is shown that a strong enhancement in the contrast can be obtained for graphene as example of weak phase object and endocytic multivescilar body as example of an unstained biological sample. No charging or significant contamination of the central beam stop is observed. For graphene, a resolution beyond 1Å(-1) was easily obtained. DF-000 imaging can be considered as a good and easy to use alternative of a phase plate.