A binary segmentation approach for boxing ribosome particles in cryo EM micrographs

Three-dimensional reconstruction of ribosome particles from electron micrographs requires selection of many single-particle images. Roughly 100,000 particles are required to achieve approximately 10 A resolution. Manual selection of particles, by visual observation of the micrographs on a computer screen, is recognized as a bottleneck in automated single-particle reconstruction. This paper describes an efficient approach for automated boxing of ribosome particles in micrographs. Use of a fast, anisotropic non-linear reaction-diffusion method to pre-process micrographs and rank-leveling to enhance the contrast between particles and the background, followed by binary and morphological segmentation constitute the core of this technique. Modifying the shape of the particles to facilitate segmentation of individual particles within clusters and boxing the isolated particles is successfully attempted. Tests on a limited number of micrographs have shown that over 80% success is achieved in automatic particle picking.     

The reconstruction of structure from electron micrographs of randomly oriented particles

A new method for enhancing and reconstructing the three dimensional structure of randomly oriented particles from their electron micrographs is developed. The method requires as an input many pictures of randomly oriented identical particles. The analysis is based on the calculation and accumulation of the spatial correlation of the densities on the electron micrographs, from which the spherical harmonic coefficients of the structure can be found. The process of enhancement of the spatial correlation and the averaging out of background noise enables reconstructions by use of pictures with low signal-to-noise ratio. The theory is presented and implemented in a computer program package. Simulated electron micrographs of ellipses, rods and a model of hexameric glutamate dehydrogenase are analyzed to demonstrate reconstructions using the computer programs.     

Automatic particle selection from electron micrographs using machine learning techniques

The 3D reconstruction of biological specimens using Electron Microscopy is currently capable of achieving subnanometer resolution. Unfortunately, this goal requires gathering tens of thousands of projection images that are frequently selected manually from micrographs. In this paper we introduce a new automatic particle selection that learns from the user which particles are of interest. The training phase is semi-supervised so that the user can correct the algorithm during picking and specifically identify incorrectly picked particles. By treating such errors specially, the algorithm attempts to minimize the number of false positives. We show that our algorithm is able to produce datasets with fewer wrongly selected particles than previously reported methods. Another advantage is that we avoid the need for an initial reference volume from which to generate picking projections by instead learning which particles to pick from the user. This package has been made publicly available in the open-source package Xmipp.     

The development and application of negative staining to the study of isolated virus particles and their components: A personal account

The subsequent development and application of the negative staining technique to isolated virus particles and their components, following the early studies on the structure of T2 bacteriophage proteins, is described. The use of the method to prepare particles covering a wide range of different types of virus for examination in the electron microscope is reviewed, together with more recent advances in image analysis of electron micrographs obtained from negatively stained virus specimens.     

v-Bodies in mitotic chromatin

Electron micrographs of mitotic chromosomal fibers from Chinese hamster ovary cells resemble very closely published electron micrographs of chromatin from interphase nuclei. The mitotic fibers are composed of particles tandemly aligned along the fiber length, with short connecting strands between the particles. The average diameter of the particles or v-bodies is 89 Å.     

An automatic particle pickup method using a neural network applicable to low-contrast electron micrographs

Three-dimensional reconstruction from electron micrographs requires the selection of many single-particle projection images; more than 10 000 are generally required to obtain 5- to 10-A structural resolution. Consequently, various automatic detection algorithms have been developed and successfully applied to large symmetric protein complexes. This paper presents a new automated particle recognition and pickup procedure based on the three-layer neural network that has a large application range than other automated procedures. Its use for both faint and noisy electron micrographs is demonstrated. The method requires only 200 selected particles as learning data and is able to detect images of proteins as small as 200 kDa.     

Local average intensity-based method for identifying spherical particles in electron micrographs

A method is presented that reliably detects spherical viruses from a wide variety of noisy low-contrast electron micrographs. Such detection is one of the first image analysis steps in the computer-aided reconstruction of three-dimensional density distribution models of viruses. Particle detection is based on the comparison of intensity in a circular area and in the surrounding ring followed by a number of tests to validate the potential particles. The only required input from the user in addition to the micrograph is an approximate radius of the particle. The method has been implemented as program ETHAN that has been tested for several different data sets. ETHAN has also successfully been used to detect DNA-less virus particles for an actual reconstruction.     

A two step approach for semi-automated particle selection from low contrast cryo-electron micrographs

Over recent years advances in cryo-electron microscopy for the study of macromolecular structure have resulted in resolutions in the range 10-15 A becoming routine. With this drive for increased resolution comes the need to collect larger datasets, commonly >10,000 particle images. Manual selection of particles from micrographs is often difficult and with such large numbers of particles now involved it is also laborious and a common bottleneck. Automated methods do exist but are normally restricted to specific samples or data, i.e., spherical particles, no aggregation, high contrast, and low noise. A two step approach has been developed that remains general and can be applied to low contrast, high noise micrographs of small molecules. Specifically, application of the approach is presented using micrographs of Escherichia coli RNA polymerase, which due to low contrast and the relatively small size of the molecule prove difficult to pick manually. To test the automated approach, independent reconstructions of RNA polymerase were carried out using manual and automatically picked data. The two reconstructions are shown to be comparable and the reconstruction from the automatically picked dataset is at a higher resolution, due to an increase in the number of particles picked.     

Highly purified mitochondria from rat brain prepared by phase partition

Mitochondria and synaptosomes from adult rat forebrain can easily be separated by counter-current distribution in an aqueous two phase system composed of Dextran T500 and poly(ethylene glycol) 4000. Both particles may also be separated by a batch procedure in which the same phase system is used. Electron micrographs and enzymatic activities show a high purity of the mitochondria obtained from the dextran-rich lower phase. Electron micrographs and enzymatic activities also show that intact synaptosomes can be obtained from the poly (ethylene glycol)-rich upper phase.The mitochondria purified by this method show good ADP/O ratios, respiratory control ratios, and state 3 rates. Synaptosomes showed a state 2-state 3 transition with no recuperation to state 4.     

TYSON: robust searching, sorting, and selecting of single particles in electron micrographs

We here present TYSON, a new program for automatic and semi-automatic particle selection from electron micrographs. TYSON employs a three-step strategy of searching, sorting and selecting single particles. In the first step, TYSON finds the positions of potential particles by one of three different methods: local averaging, template matching or local variance. The practical merits and drawbacks of these methods are discussed. In the second step, these potential particles are automatically sorted according to their probability of being true positives. Many criteria are provided for this sort. In the final -interactive- step, whole categories of poorly fitting false positives can be removed with a single mouse-click. We present results obtained using cryo-EM micrographs of both spherical virus particles and asymmetric particles. The procedures are fast and use of TYSON allowed, for example, some 20,000 particles to be selected in a single working day.     

Structure of mitochondrial F1-ATPase studied by electron microscopy and image processing

The structure of soluble F₁-ATPase (EC 3.6.1.3) has been investigated by computer analysis of individual molecular images extracted from electron micrographs of negatively stained particles. A total of 1241 images was interactively selected from several digitized micrographs and these images were subsequently aligned relative to different reference images. They were then submitted to a multivariate statistical classification procedure. We have focussed our attention on the main ‘hexagonal’ view which represents some 40% of our population of images. In this view, six masses are located on the outer region of the projection which are associated with the alpha and the beta subunits of the protein. A seventh mass is located close to the centre of the hexagon, but slightly off its exact midpoint. It has the shape of the letter V and its two legs point to two of the outer protein masses, or one alpha-beta subunit pair. The corner of the V has a density as high as those of the large subunits. Possible subunit arrangements and their consequences for the mechanism of ATP synthesis are discussed.     

Spatial disorders and computational cures

Image averaging provides a powerful method for enhancing the yield of interpretable information from electron micrographs of biological macromolecules. However, as originally conceived, the full benefit of averaging is achieved only with perfectly ordered two-dimensional crystals. More recent developments, reviewed here, allow one to rectify disordered lattices, straighten randomly bent filaments, and combine multiple images of free-standing particles, thus extending the advantages of image averaging to virtually every class of macromolecular specimen.     

A rapid method for staining large chylomicrons

In this report, we present a rapid method for producing high-quality micrographs suitable for determining the size distributions of particles in concentrated samples of postprandial chylomicrons and chylomicron remnants. The procedure consists of mixing particles with osmium tetroxide in water to stabilize the lipids of the particles. These fixed and positively stained particles are then negatively stained with phosphotungstate in the presence of dilute sucrose. This dual staining procedure prevents the fusion and clustering of chylomicrons during processing for electron microscopy and is effective with particles of different lipid compositions. In addition, this procedure is simple and rapid, adding only one mixing step and 5 min to the preparation time required for conventional negative stains.     

Detecting circular and rectangular particles based on geometric feature detection in electron micrographs

Accurate and automatic particle detection from cryo-electron microscopy (cryo-EM images) is very important for high-resolution reconstruction of large macromolecular structures. In this paper, we present a method for particle picking based on shape feature detection. Two fundamental concepts of computational geometry, namely, the distance transform and the Voronoi diagram, are used for detection of critical features as well as for accurate location of particles from the images or micrographs. Unlike the conventional template-matching methods, our approach detects the particles based on their boundary features instead of intensities. The geometric features derived from the boundaries provide an efficient way for locating particles quickly and accurately, which avoids a brute-force searching for the best position/orientation. Our approach is fully automatic and has been successfully applied to detect particles with approximately circular or rectangular shapes (e.g., KLH particles). Particle detection can be enhanced by multiple sets of parameters used in edge detection and/or by anisotropic filtering. We also discuss the extension of this approach to other types of particles with certain geometric features.     

Automatic cryo-EM particle selection for membrane proteins in spherical liposomes

Random spherically constrained (RSC) single particle reconstruction is a method to obtain structures of membrane proteins embedded in lipid vesicles (liposomes). As in all single-particle cryo-EM methods, structure determination is greatly aided by reliable detection of protein “particles” in micrographs. After fitting and subtraction of the membrane density from a micrograph, normalized cross-correlation (NCC) and estimates of the particle signal amplitude are used to detect particles, using as references the projections of a 3D model. At each pixel position, the NCC is computed with only those references that are allowed by the geometric constraint of the particle’s embedding in the spherical vesicle membrane. We describe an efficient algorithm for computing this position-dependent correlation, and demonstrate its application to selection of membrane-protein particles, GluA2 glutamate receptors, which present very different views from different projection directions.     

Fourier amplitude decay of electron cryomicroscopic images of single particles and effects on structure determination

Several factors, including spatial and temporal coherence of the electron microscope, specimen movement, recording medium, and scanner optics, contribute to the decay of the measured Fourier amplitude in electron image intensities. We approximate the combination of these factors as a single Gaussian envelope function, the width of which is described by a single experimental B-factor. We present an improved method for estimating this B-factor from individual micrographs by combining the use of X-ray solution scattering and numerical fitting to the average power spectrum of particle images. A statistical estimation from over 200 micrographs of herpes simplex virus type-1 capsids was used to estimate the spread in the experimental B-factor of the data set. The B-factor is experimentally shown to be dependent on the objective lens defocus setting of the microscope. The average B-factor, the X-ray scattering intensity of the specimen, and the number of particles required to determine the structure at a lower resolution can be used to estimate the minimum fold increase in the number of particles that would be required to extend a single particle reconstruction to a specified higher resolution. We conclude that microscope and imaging improvements to reduce the experimental B-factor will be critical for obtaining an atomic resolution structure.     

A quantitative method for assessing co-localization in immunolabeled thin section electron micrographs

A method is introduced for the analysis of nearest neighbor distances between immunogold particles marking proteins on electron micrographs. Deviation from the distribution that is predicted by chance indicates co-localization of the labeled species, and the potential for productive interaction in vivo. Application of this method to the analysis of nearest neighbor distances in experiments with pea leaf thin sections and isozyme-directed antibodies indicates that glyceraldehyde-3-P dehydrogenase is located near P-glycerate kinase and near aldolase in the chloroplast stroma, consistent with the notion that these enzymes are part of a multi-enzyme photosynthetic CO(2)-fixation complex in situ.     

Electron microscopic studies on liver 5-aminolaevulinate synthase

The structure of chick embryo liver 5-aminolaevulinate synthase has been examined by electron microscopic studies using negative staining. From the different projections of the enzyme particles observed in electron micrographs, a model for the enzyme molecule has been proposed. In this model, an enzyme molecule consists of two curved and identical subunits associated in opposite polarities. From the dimensions of an enzyme molecule subunit measured from electron micrographs, the relative molecular mass of each subunit is estimated to be 70 000.     

An approach to automated particle picking from electron micrographs based on reduced representation templates

Reduced representation templates are used in a real-space pattern matching framework to facilitate automatic particle picking from electron micrographs. The procedure consists of five parts. First, reduced templates are constructed either from models or directly from the data. Second, a real-space pattern matching algorithm is applied using the reduced representations as templates. Third, peaks are selected from the resulting score map using peak-shape characteristics. Fourth, the surviving peaks are tested for distance constraints. Fifth, a correlation-based outlier screening is applied. Test applications to a data set of keyhole limpet hemocyanin particles indicate that the method is robust and reliable.