Challenges in making ideal cryo-EM samples.

Recognizing that interaction with the air-water interface (AWI) is a major challenge for cryo-EM, we first review current approaches designed to avoid it. Of these, immobilizing particles on affinity grids is arguably the most promising. In addition, we review efforts to gain more reliable control of the sample thicknesses, not the least important reason being to prevent immobilized particles from coming in contact with the AWI of the remaining buffer. It is emphasized that avoiding such a contact is as important for cryo-ET as for single-particle cryo-EM. Finally, looking to the future, it is proposed that immobilized samples might be used to perform time-resolved biochemical experiments directly on EM grids rather than just in test tubes or cuvettes.

Streptavidin-Affinity Grid Fabrication for Cryo-Electron Microscopy Sample Preparation.

Streptavidin affinity grids provide strategies to overcome many commonly encountered cryo-electron microscopy (cryo-EM) sample preparation challenges, including sample denaturation and preferential orientations that can occur due to the air-water interface. Streptavidin affinity grids, however, are currently utilized by few cryo-EM labs because they are not commercially available and require a careful fabrication process. Two-dimensional streptavidin crystals are grown onto a biotinylated lipid monolayer that is applied directly to standard holey-carbon cryo-EM grids. The high-affinity interaction between streptavidin and biotin allows for the subsequent binding of biotinylated samples that are protected from the air-water interface during cryo-EM sample preparation. Additionally, these grids provide a strategy for concentrating samples available in limited quantities and purifying protein complexes of interest directly on the grids. Here, a step-by-step, optimized protocol is provided for the robust fabrication of streptavidin affinity grids for use in cryo-EM and negative-stain experiments. Additionally, a trouble-shooting guide is included for commonly experienced challenges to make the use of streptavidin affinity grids more accessible to the larger cryo-EM community.

Perspective: Biochemical and Physical Constraints Associated With Preparing Thin Specimens for Single-Particle Cryo-EM.

While many aspects of single-particle electron cryo-microscopy (cryo-EM) of biological macromolecules have reached a sophisticated level of development, this is not yet the case when it comes to preparing thin samples on specimen grids. As a result, there currently is considerable interest in achieving better control of both the sample thickness and the amount of area that is useful, but this is only one aspect in which improvement is needed. This Perspective addresses the further need to prevent the macromolecular particles from making contact with the air-water interface, something that can result in preferential orientation and even structural disruption of macromolecular particles. This unwanted contact can occur either as the result of free diffusion of particles during the interval between application, thinning and vitrification of the remaining buffer, or-when particles have been immobilized-by the film of buffer becoming too thin prior to vitrification. An opportunity now exists to apply theoretical and practical insights from the fields of thin-film physical chemistry and interfacial science, in an effort to bring cryo-EM sample preparation to a level of sophistication that is comparable to that of current data collection and analysis.

Simple assay for adsorption of proteins to the air-water interface.

A rapid assay is described, based upon the Marangoni effect, which detects the formation of a denatured-protein film at the air-water interface (AWI) of aqueous samples. This assay requires no more than a 20 µL aliquot of sample, at a protein concentration of no more than1 mg/ml, and it can be performed with any buffer that is used to prepare grids for electron cryo-microscopy (cryo-EM). In addition, this assay provides an easy way to estimate the rate at which a given protein forms such a film at the AWI. Use of this assay is suggested as a way to pre-screen the effect of various additives and chemical modifications that one might use to optimize the preparation of grids, although the final proof of optimization still requires further screening of grids in the electron microscope. In those cases when the assay establishes that a given protein does form a sacrificial, denatured-protein monolayer, it is suggested that subsequent optimization strategies might focus on discovering how to improve the adsorption of native proteins onto that monolayer, rather than to prevent its formation. A second alternative might be to bind such proteins to the surface of rationally designed affinity grids, in order to prevent their diffusion to, and unwanted interaction with, the AWI.

Defocus-dependent Thon-ring fading.

The brightness of modern Schottky field-emission guns can produce electron beams that have very high spatial coherence, especially for the weak-illumination conditions that are used for single-particle electron cryo-microscopy in structural biology. Even so, many users have observed defocus-dependent Thon-ring fading that has led them to restrict their data collection strategy to imaging with relatively small defocus values. In this paper, we reproduce the observation of defocus-dependent Thon-ring fading and produce a quantitative analysis and clear explanation of its causes. We demonstrate that a major cause is the delocalization of high-resolution Fourier components outside the field of view of the camera. We also show that, to correctly characterize the phenomenon, it is important to make a correction for linear magnification anisotropy. Even when the anisotropy is quite small, it is present at all defocus values before circular averaging of the Thon rings, as is also true before merging data from particles in many orientations. Under the conditions used in this paper, which are typical of those used in single-particle electron cryomicroscopy, fading of the Thon rings due to source coherence is negligible. The principal conclusion is that much higher values of defocus can be used to record images than is currently thought to be possible, keeping in mind that the above-mentioned delocalization of Fourier components will ultimately become a limitation. This increased understanding should give electron microscopists the confidence to use higher amounts of defocus to allow, for example, better visibility of their particles and Ewald sphere correction.

Microscale Fluid Behavior during Cryo-EM Sample Blotting.

Blotting has been the standard technique for preparing aqueous samples for single-particle electron cryo-microscopy for over three decades. This technique removes the excess solution from a transmission electron microscope grid by pressing absorbent filter paper against the specimen before vitrification. However, this standard technique produces vitreous ice with inconsistent thickness from specimen to specimen and from region to region within the same specimen, the reasons for which are not understood. Here, high-speed interference contrast microscopy is used to demonstrate that the irregular pattern of fibers in the filter paper imposes tortuous, highly variable boundaries during the removal of excess liquid from a flat, hydrophilic surface. As a result, aqueous films of nonuniform thickness are formed while the filter paper is pressed against the substrate. This pattern of nonuniform liquid thickness changes again after the filter paper is pulled away, but the thickness still does not become completely uniform. We suggest that similar topographical features of the liquid film are produced during the standard technique used to blot EM grids and that these manifest in nonuniform ice after vitrification. These observations suggest that alternative thinning techniques, which do not rely on direct contact between the filter paper and the grid, may result in more repeatable and uniform sample thicknesses.

3.1 Šstructure of yeast RNA polymerase II elongation complex stalled at a cyclobutane pyrimidine dimer lesion solved using streptavidin affinity grids.

Despite significant advances in all aspects of single particle cryo-electron microscopy (cryo-EM), specimen preparation still remains a challenge. During sample preparation, macromolecules interact with the air-water interface, which often leads to detrimental effects such as denaturation or adoption of preferred orientations, ultimately hindering structure determination. Randomly biotinylating the protein of interest (for example, at its primary amines) and then tethering it to a cryo-EM grid coated with two-dimensional crystals of streptavidin (acting as an affinity surface) can prevent the protein from interacting with the air-water interface. Recently, this approach was successfully used to solve a high-resolution structure of a test sample, a bacterial ribosome. However, whether this method can be used for samples where interaction with the air-water interface has been shown to be problematic remains to be determined. Here we report a 3.1 Šstructure of an RNA polymerase II elongation complex stalled at a cyclobutane pyrimidine dimer lesion (Pol II EC(CPD)) solved using streptavidin grids. Our previous attempt to solve this structure using conventional sample preparation methods resulted in a poor quality cryo-EM map due to Pol II EC(CPD)'s adopting a strong preferred orientation. Imaging the same sample on streptavidin grids improved the angular distribution of its view, resulting in a high-resolution structure. This work shows that streptavidin affinity grids can be used to address known challenges posed by the interaction with the air-water interface.

Self-Assembling Micelles Based on an Intrinsically Disordered Protein Domain.

The self-assembly of micellar structures from diblock polymers that contain hydrophilic and hydrophobic domains has been of great interest for the encapsulation of drugs and other hydrophobic molecules. While most commercially used surfactants are derived from hydrocarbon sources, there have been recent efforts to replace these with biodegradable, nontoxic, biologically synthesized alternatives. Previous examples have primarily examined naturally occurring self-assembling proteins, such as silk and elastin-like sequences. Herein, we describe a new series of fusion proteins that have been developed to self-assemble spontaneously into stable micelles that are 27 nm in diameter after enzymatic cleavage of a solubilizing protein tag. The sequences of the proteins are based on a human intrinsically disordered protein, which has been appended with a hydrophobic segment. The micelles were found to form across a broad range of pH, ionic strength, and temperature conditions, with critical micelle concentration (CMC) values in the low micromolar range, 3 orders of magnitude lower than the CMC of commonly used surfactant sodium dodecyl sulfate (SDS). The reported micelles were found to solubilize hydrophobic metal complexes and organic molecules, suggesting their potential suitability for catalysis and drug delivery applications. Furthermore, the inherent flexibility in the design of these protein sequences enables the encoding of additional functionalities for many future applications. Overall, this work represents a new biomolecular alternative to traditional surfactants that are based on nonrenewable and poorly biodegradable hydrocarbon sources.

Opinion: hazards faced by macromolecules when confined to thin aqueous films.

Samples prepared for single-particle electron cryo-microscopy (cryo-EM) necessarily have a very high surface-to-volume ratio during the short period of time between thinning and vitrification. During this time, there is an obvious risk that macromolecules of interest may adsorb to the air-water interface with a preferred orientation, or that they may even become partially or fully unfolded at the interface. In addition, adsorption of macromolecules to an air-water interface may occur even before thinning. This paper addresses the question whether currently used methods of sample preparation might be improved if one could avoid such interfacial interactions. One possible way to do so might be to preemptively form a surfactant monolayer over the air-water interfaces, to serve as a structure-friendly slide and coverslip. An alternative is to immobilize particles of interest by binding them to some type of support film, which-to continue using the analogy-thus serves as a slide. In this case, the goal is not only to prevent the particles of interest from diffusing into contact with the air-water interface but also to increase the number of particles seen in each image. In this direction, it is natural to think of developing various types of affinity grids as structure-friendly alternatives to thin carbon films. Perhaps ironically, if precautions are not taken against adsorption of particles to air-water interfaces, sacrificial monolayers of denatured protein may take the roles of slide, coverslip, or even both.

Monolayer-crystal streptavidin support films provide an internal standard of cryo-EM image quality.

Analysis of images of biotinylated Escherichia coli 70S ribosome particles, bound to streptavidin affinity grids, demonstrates that the image-quality of particles can be predicted by the image-quality of the monolayer crystalline support film. The quality of the Thon rings is also a good predictor of the image-quality of particles, but only when images of the streptavidin crystals extend to relatively high resolution. When the estimated resolution of streptavidin was 5Å or worse, for example, the ribosomal density map obtained from 22,697 particles went to only 9.5Å, while the resolution of the map reached 4.0Å for the same number of particles, when the estimated resolution of streptavidin crystal was 4Å or better. It thus is easy to tell which images in a data set ought to be retained for further work, based on the highest resolution seen for Bragg peaks in the computed Fourier transforms of the streptavidin component. The refined density map obtained from 57,826 particles obtained in this way extended to 3.6Å, a marked improvement over the value of 3.9Å obtained previously from a subset of 52,433 particles obtained from the same initial data set of 101,213 particles after 3-D classification. These results are consistent with the hypothesis that interaction with the air-water interface can damage particles when the sample becomes too thin. Streptavidin monolayer crystals appear to provide a good indication of when that is the case.

Long shelf-life streptavidin support-films suitable for electron microscopy of biological macromolecules.

We describe a rapid and convenient method of growing streptavidin (SA) monolayer crystals directly on holey-carbon EM grids. As expected, these SA monolayer crystals retain their biotin-binding function and crystalline order through a cycle of embedding in trehalose and, later, its removal. This fact allows one to prepare, and store for later use, EM grids on which SA monolayer crystals serve as an affinity substrate for preparing specimens of biological macromolecules. In addition, we report that coating the lipid-tail side of trehalose-embedded monolayer crystals with evaporated carbon appears to improve the consistency with which well-ordered, single crystals are observed to span over entire, 2µm holes of the support films. Randomly biotinylated 70S ribosomes are used as a test specimen to show that these support films can be used to obtain a high-resolution cryo-EM structure.

Factors that Influence the Formation and Stability of Thin, Cryo-EM Specimens.

Poor consistency of the ice thickness from one area of a cryo-electron microscope (cryo-EM) specimen grid to another, from one grid to the next, and from one type of specimen to another, motivates a reconsideration of how to best prepare suitably thin specimens. Here we first review the three related topics of wetting, thinning, and stability against dewetting of aqueous films spread over a hydrophilic substrate. We then suggest that the importance of there being a surfactant monolayer at the air-water interface of thin, cryo-EM specimens has been largely underappreciated. In fact, a surfactant layer (of uncontrolled composition and surface pressure) can hardly be avoided during standard cryo-EM specimen preparation. We thus suggest that better control over the composition and properties of the surfactant layer may result in more reliable production of cryo-EM specimens with the desired thickness.

Automated particle correspondence and accurate tilt-axis detection in tilted-image pairs.

Tilted electron microscope images are routinely collected for an ab initio structure reconstruction as a part of the Random Conical Tilt (RCT) or Orthogonal Tilt Reconstruction (OTR) methods, as well as for various applications using the "free-hand" procedure. These procedures all require identification of particle pairs in two corresponding images as well as accurate estimation of the tilt-axis used to rotate the electron microscope (EM) grid. Here we present a computational approach, PCT (particle correspondence from tilted pairs), based on tilt-invariant context and projection matching that addresses both problems. The method benefits from treating the two problems as a single optimization task. It automatically finds corresponding particle pairs and accurately computes tilt-axis direction even in the cases when EM grid is not perfectly planar.

Disulfide linkage and structure of highly stable yeast-derived virus-like particles of murine polyomavirus.

VP1 is the major coat protein of murine polyomavirus and forms virus-like particles (VLPs) in vitro. VLPs consist of 72 pentameric VP1 subunits held together by a terminal clamp structure that is further stabilized by disulfide bonds and chelation of calcium ions. Yeast-derived VLPs (yVLPs) assemble intracellularly in vivo during recombinant protein production. These in vivo assembled yVLPs differ in several properties from VLPs assembled in vitro from bacterially produced pentamers. We found several intermolecular disulfide linkages in yVLPs involving 5 of the 6 cysteines of VP1 (Cys(115)-Cys(20), Cys(12)-Cys(20), Cys(16)-Cys(16), Cys(12)/ Cys(16)-Cys(115), and Cys(274)-Cys(274)), indicating a highly coordinated disulfide network within the in vivo assembled particles involving the N-terminal region of VP1. Cryoelectron microscopy revealed structured termini not resolved in the published crystal structure of the bacterially expressed VLP that appear to clamp the pentameric subunits together. These structural features are probably the reason for the observed higher stability of in vivo assembled yVLPs compared with in vitro assembled bacterially expressed VLPs as monitored by increased thermal stability, higher resistance to trypsin cleavage, and a higher activation enthalpy of the disassembly reaction. This high stability is decreased following disassembly of yVLPs and subsequent in vitro reassembly, suggesting a role for cellular components in optimal assembly.

Electron microscopy of biotinylated protein complexes bound to streptavidin monolayer crystals.

Chemical biotinylation of protein complexes followed by binding to two-dimensional (monolayer) crystals of streptavidin is shown to be an effective way to prepare cryo-EM specimens from samples at low protein concentration. Three different multiprotein complexes are used to demonstrate the generality of this method. In addition, native thermosomes, purified from Sulfolobus solfataricus P2, are used to demonstrate that a uniform distribution of Euler angles is produced, even though this particle is known to adopt a preferred orientation when other methods of cryo-EM specimen preparation are used.

Experimental evaluation of support vector machine-based and correlation-based approaches to automatic particle selection.

The goal of this study is to evaluate the performance of software for automated particle-boxing, and in particular the performance of a new tool (TextonSVM) that recognizes the characteristic texture of particles of interest. As part of a high-throughput protocol, we use human editing that is based solely on class-average images to create final data sets that are enriched in what the investigator considers to be true-positive particles. The Fourier shell correlation (FSC) function is then used to characterize the homogeneity of different single-particle data sets that are derived from the same micrographs by two or more alternative methods. We find that the homogeneity is generally quite similar for class-edited data sets obtained by the texture-based method and by SIGNATURE, a cross-correlation-based method. The precision-recall characteristics of the texture-based method are, on the other hand, significantly better than those of the cross-correlation based method; that is to say, the texture-based approach produces a smaller fraction of false positives in the initial set of candidate particles. The computational efficiency of the two approaches is generally within a factor of two of one another. In situations when it is helpful to use a larger number of templates (exemplars), however, TextonSVM scales in a much more efficient way than do boxing programs that are based on localized cross-correlation.

Visual proteomics.

Visual proteomics attempts to generate molecular atlases by providing the position and angular orientation of protein complexes inside of cells. This is accomplished by template matching (pattern recognition), a cross-correlation-based process that matches the structure of a specific protein complex to the densities of the whole volume or subvolume of a cell, that is typically acquired by cryoelectron tomography. Thereby, a search is performed that scans the entire volume for structural templates contained in a database. In this chapter, we primarily describe the practical experiences gained with visual proteomics during the Leptospira interrogans proteome project [Beck et al. (2009). Visual proteomics of the human pathogen Leptospira interrogans. Nat. Methods 6, 817.]. We give a practical guide how to implement the method and review critical experimental and computational aspects in detail. Based on a survey that has been undertaken for protein complexes from Desulfovibrio vulgaris, we review the difficulty of generating reference structures in detail. Finally, we discuss the high yield targets for technical improvements.

Survey of large protein complexes in D. vulgaris reveals great structural diversity.

An unbiased survey has been made of the stable, most abundant multi-protein complexes in Desulfovibrio vulgaris Hildenborough (DvH) that are larger than Mr approximately 400 k. The quaternary structures for 8 of the 16 complexes purified during this work were determined by single-particle reconstruction of negatively stained specimens, a success rate approximately 10 times greater than that of previous "proteomic" screens. In addition, the subunit compositions and stoichiometries of the remaining complexes were determined by biochemical methods. Our data show that the structures of only two of these large complexes, out of the 13 in this set that have recognizable functions, can be modeled with confidence based on the structures of known homologs. These results indicate that there is significantly greater variability in the way that homologous prokaryotic macromolecular complexes are assembled than has generally been appreciated. As a consequence, we suggest that relying solely on previously determined quaternary structures for homologous proteins may not be sufficient to properly understand their role in another cell of interest.

Water transport in AQP0 aquaporin: molecular dynamics studies.

A double lipid bilayer structure containing opposing tetramers of AQP0 aquaporin, in contact through extracellular face loop regions, was recently modeled using an intermediate-resolution map obtained by electron crystallographic methods. The pores of these water channels were found to be critically narrow in three regions and subsequently interpreted to be those of a closed state of the channel. The subsequent determination of a high-resolution AQP0 tetramer structure by X-ray crystallographic methods yielded a pore model featuring two of the three constrictions as noted in the EM work and water molecules within the channel pore. The extracellular-side constriction region of this AQP0 structure was significantly larger than that of the EM-based model and similar to that of the highly water permeable AQP1. The X-ray-based study of AQP0 however could not ascertain if the water molecules found in the pore were the result of water entering from one or both ends of the channel, nor whether water could freely pass through all constriction points. Additionally, this X-ray-based structure could not provide an answer to the question of whether the double lipid bilayer configuration of AQP0 could functionally maintain a water impermeable state of the channel. To address these questions we conducted molecular dynamics simulations to compare the time-dependent behavior of the AQP0 and AQP1 channels within lipid bilayers. The simulations demonstrate that AQP0, in single or double lipid bilayers, is not closed to water transport and that thermal motions of critical side-chains are sufficient to facilitate the movement of water past any of its constriction regions. These motional requirements do however lead to significant free energy barriers and help explain physiological observations that found water permeability in AQP0 to be substantially lower than in the AQP1 pore.

Crystal structure of human calmodulin-like protein: insights into its functional role.

A calmodulin (CaM)-like protein (hCLP) is expressed in human mammary epithelial cells but appears to be limited to certain epithelial cells such as those found in skin, prostate, breast and cervical tissues. A decrease in the expression of this protein is associated with the occurrence of tumors in breast epithelium. The structure of hCLP determined to 1.5 A resolution by X-ray crystallography shows a distinct 30 degrees displacement along the interconnecting central helix, when compared to the highly conserved structure of vertebrate CaM, resulting in a difference in the relative orientation of its two globular domains. Additionally, the electric surface potential landscape at the target protein binding regions on the two globular domains of hCLP is significantly different from those of CaM, indicating that the respective ranges of hCLP and hCaM target proteins do not fully overlap. Observations that hCLP can competitively inhibit CaM activation of target proteins also imply a role for hCLP in which it may also serve as a modulator of CaM activity in the epithelial cells where hCLP is expressed.