MarkerDetector: A method for robust fiducial marker detection in electron micrographs using wavelet-based template.

Fiducial marker detection in electron micrographs becomes an important and challenging task with the development of large-field electron microscopy. The fiducial marker detection plays an important role in several steps during the process of electron micrographs, such as the alignment and parameter calibrations. However, limited by the conditions of low signal-to-noise ratio (SNR) in the electron micrographs, the performance of fiducial marker detection is severely affected. In this work, we propose the MarkerDetector, a novel algorithm for detecting fiducial markers in electron micrographs. The proposed MarkerDetector is built upon the following contributions: Firstly, a wavelet-based template generation algorithm is devised in MarkerDetector. By adopting a shape-based criterion, a high-quality template can be obtained. Secondly, a robust marker determination strategy is devised by utilizing statistic-based filtering, which can guarantee the correctness of the detected fiducial markers. The average running time of our algorithm is 1.67 seconds with promising accuracy, indicating its practical feasibility for applications in electron micrographs.

Progressive alterations in polysomal architecture and activation of ribosome stalling relief factors in a mouse model of Huntington's disease.

Given their highly polarized morphology and functional singularity, neurons require precise spatial and temporal control of protein synthesis. Alterations in protein translation have been implicated in the development and progression of a wide range of neurological and neurodegenerative disorders, including Huntington's disease (HD). In this study we examined the architecture of polysomes in their native brain context in striatal tissue from the zQ175 knock-in mouse model of HD. We performed 3D electron tomography of high-pressure frozen and freeze-substituted striatal tissue from HD models and corresponding controls at different ages. Electron tomography results revealed progressive remodelling towards a more compacted polysomal architecture in the mouse model, an effect that coincided with the emergence and progression of HD related symptoms. The aberrant polysomal architecture is compatible with ribosome stalling phenomena. In fact, we also detected in the zQ175 model an increase in the striatal expression of the stalling relief factor EIF5A2 and an increase in the accumulation of eIF5A1, eIF5A2 and hypusinated eIF5A1, the active form of eIF5A1. Polysomal sedimentation gradients showed differences in the relative accumulation of 40S ribosomal subunits and in polysomal distribution in striatal samples of the zQ175 model. These findings indicate that changes in the architecture of the protein synthesis machinery may underlie translational alterations associated with HD, opening new avenues for understanding the progression of the disease.

Architecture of torovirus replicative organelles.

Plus-stranded RNA viruses replicate in the cytosol of infected cells, in membrane-bound replication complexes. We previously identified double membrane vesicles (DMVs) in the cytoplasm of cells infected with Berne virus (BEV), the prototype member of the Torovirus genus (Nidovirales Order). Our previous analysis by transmission electron microscopy suggested that the DMVs form a reticulovesicular network (RVN) analogous those described for the related severe acute respiratory syndrome coronavirus (SARS-CoV-1). Here, we used serial sectioning and electron tomography to characterize the architecture of torovirus replication organelles, and to learn about their biogenesis and dynamics during the infection. The formation of a RVN in BEV infected cells was confirmed, where the outer membranes of the DMVs are interconnected with each other and with the ER. Paired or zippered ER membranes connected with the DMVs were also observed, and likely represent early structures that evolve to give rise to DMVs. Also, paired membranes forming small spherule-like invaginations were observed at late time post-infection. Although resembling in size, the tomographic analysis show that these structures are clearly different from the true spherules described previously for coronaviruses. Hence, BEV shows important similarities, but also some differences, in the architecture of the replication organelles with other nidoviruses.

TomoAlign: A novel approach to correcting sample motion and 3D CTF in CryoET.

TomoAlign is a software package that integrates tools to mitigate two important resolution limiting factors in cryoET, namely the beam-induced sample motion and the contrast transfer function (CTF) of the microscope. The package is especially focused on cryoET of thick specimens where fiducial markers are required for accurate tilt-series alignment and sample motion estimation. TomoAlign models the beam-induced sample motion undergone during the tilt-series acquisition. The motion models are used to produce motion-corrected subtilt-series centered on the particles of interest. In addition, the defocus of each particle at each tilt image is determined and can be corrected, resulting in motion-corrected and CTF-corrected subtilt-series from which the subtomograms can be computed. Alternatively, the CTF information can be passed on so that CTF correction can be carried out entirely within external packages like Relion. TomoAlign serves as a versatile tool that can streamline the cryoET workflow from initial alignment of tilt-series to final subtomogram averaging during in situ structure determination.

PolishEM: image enhancement in FIB-SEM.

SUMMARY: We have developed a software tool to improve the image quality in focused ion beam-scanning electron microscopy (FIB-SEM) stacks: PolishEM. Based on a Gaussian blur model, it automatically estimates and compensates for the blur affecting each individual image. It also includes correction for artifacts commonly arising in FIB-SEM (e.g. curtaining). PolishEM has been optimized for an efficient processing of huge FIB-SEM stacks on standard computers. AVAILABILITY AND IMPLEMENTATION: PolishEM has been developed in C. GPL source code and binaries for Linux, OSX and Windows are available at http://www.cnb.csic.es/%7ejjfernandez/polishem. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Consideration of sample motion in cryo-tomography based on alignment residual interpolation.

Recently, it has been shown that the resolution in cryo-tomography could be improved by considering the sample motion in tilt-series alignment and reconstruction, where a set of quadratic polynomials were used to model this motion. One requirement of this polynomial method is the optimization of a large number of parameters, which may limit its practical applicability. In this work, we propose an alternative method for modeling the sample motion. Starting from the standard fiducial-based tilt-series alignment, the method uses the alignment residual as local estimates of the sample motion at the 3D fiducial positions. Then, a scattered data interpolation technique characterized by its smoothness and a closed-form solution is applied to model the sample motion. The motion model is then integrated in the tomographic reconstruction. The new method improves the tomogram quality similar to the polynomial one, with the important advantage that the determination of the motion model is greatly simplified, thereby overcoming one of the major limitations of the polynomial model. Therefore, the new method is expected to make the beam-induced motion correction methodology more accessible to the cryoET community.

3D electron tomography of brain tissue unveils distinct Golgi structures that sequester cytoplasmic contents in neurons.

Macroautophagy is morphologically characterized by autophagosome formation. Autophagosomes are double-membraned vesicles that sequester cytoplasmic components for further degradation in the lysosome. Basal autophagy is paramount for intracellular quality control in post-mitotic cells but, surprisingly, the number of autophagosomes in post-mitotic neurons is very low, suggesting that alternative degradative structures could exist in neurons. To explore this possibility, we have examined neuronal subcellular architecture by performing three-dimensional (3D) electron tomography analysis of mouse brain tissue that had been preserved through high-pressure freezing. Here, we report that sequestration of neuronal cytoplasmic contents occurs at the Golgi complex in distinct and dynamic structures that coexist with autophagosomes in the brain. These structures are composed of several concentric double-membraned layers that appear to be formed simultaneously by the direct bending and sealing of discrete Golgi stacks. These structures are labelled for proteolytic enzymes, and lysosomes and late endosomes are found in contact with them, leading to the possibility that the sequestered material could be degraded inside them. Our findings highlight the key role that 3D electron tomography, together with tissue rapid-freezing techniques, will have in gaining new knowledge about subcellular architecture.

Cryo-tomography tilt-series alignment with consideration of the beam-induced sample motion.

Recent evidence suggests that the beam-induced motion of the sample during tilt-series acquisition is a major resolution-limiting factor in electron cryo-tomography (cryoET). It causes suboptimal tilt-series alignment and thus deterioration of the reconstruction quality. Here we present a novel approach to tilt-series alignment and tomographic reconstruction that considers the beam-induced sample motion through the tilt-series. It extends the standard fiducial-based alignment approach in cryoET by introducing quadratic polynomials to model the sample motion. The model can be used during reconstruction to yield a motion-compensated tomogram. We evaluated our method on various datasets with different sample sizes. The results demonstrate that our method could be a useful tool to improve the quality of tomograms and the resolution in cryoET.

The Structural Architecture of an Infectious Mammalian Prion Using Electron Cryomicroscopy.

The structure of the infectious prion protein (PrPSc), which is responsible for Creutzfeldt-Jakob disease in humans and bovine spongiform encephalopathy, has escaped all attempts at elucidation due to its insolubility and propensity to aggregate. PrPSc replicates by converting the non-infectious, cellular prion protein (PrPC) into the misfolded, infectious conformer through an unknown mechanism. PrPSc and its N-terminally truncated variant, PrP 27-30, aggregate into amorphous aggregates, 2D crystals, and amyloid fibrils. The structure of these infectious conformers is essential to understanding prion replication and the development of structure-based therapeutic interventions. Here we used the repetitive organization inherent to GPI-anchorless PrP 27-30 amyloid fibrils to analyze their structure via electron cryomicroscopy. Fourier-transform analyses of averaged fibril segments indicate a repeating unit of 19.1 Å. 3D reconstructions of these fibrils revealed two distinct protofilaments, and, together with a molecular volume of 18,990 Å3, predicted the height of each PrP 27-30 molecule as ~17.7 Å. Together, the data indicate a four-rung ß-solenoid structure as a key feature for the architecture of infectious mammalian prions. Furthermore, they allow to formulate a molecular mechanism for the replication of prions. Knowledge of the prion structure will provide important insights into the self-propagation mechanisms of protein misfolding.

Removing Contamination-Induced Reconstruction Artifacts from Cryo-electron Tomograms.

Imaging of fully hydrated, vitrified biological samples by electron tomography yields structural information about cellular protein complexes in situ. Here we present a computational procedure that removes artifacts of three-dimensional reconstruction caused by contamination present in samples during imaging by electron microscopy. Applying the procedure to phantom data and electron tomograms of cellular samples significantly improved the resolution and the interpretability of tomograms. Artifacts caused by surface contamination associated with thinning by focused ion beam, as well as those arising from gold fiducial markers and from common, lower contrast contamination, could be removed. Our procedure is widely applicable and is especially suited for applications that strive to reach a higher resolution and involve the use of recently developed, state-of-the-art instrumentation.

Site-Specific Cryo-focused Ion Beam Sample Preparation Guided by 3D Correlative Microscopy.

The development of cryo-focused ion beam (cryo-FIB) for the thinning of frozen-hydrated biological specimens enabled cryo-electron tomography (cryo-ET) analysis in unperturbed cells and tissues. However, the volume represented within a typical FIB lamella constitutes a small fraction of the biological specimen. Retaining low-abundance and dynamic subcellular structures or macromolecular assemblies within such limited volumes requires precise targeting of the FIB milling process. In this study, we present the development of a cryo-stage allowing for spinning-disk confocal light microscopy at cryogenic temperatures and describe the incorporation of the new hardware into existing workflows for cellular sample preparation by cryo-FIB. Introduction of fiducial markers and subsequent computation of three-dimensional coordinate transformations provide correlation between light microscopy and scanning electron microscopy/FIB. The correlative approach is employed to guide the FIB milling process of vitrified cellular samples and to capture specific structures, namely fluorescently labeled lipid droplets, in lamellas that are 300 nm thick. The correlation procedure is then applied to localize the fluorescently labeled structures in the transmission electron microscopy image of the lamella. This approach can be employed to navigate the acquisition of cryo-ET data within FIB-lamellas at specific locations, unambiguously identified by fluorescence microscopy.

Three-dimensional structure of basal body triplet revealed by electron cryo-tomography.

Basal bodies and centrioles play central roles in microtubule (MT)-organizing centres within many eukaryotes. They share a barrel-shaped cylindrical structure composed of nine MT triplet blades. Here, we report the structure of the basal body triplet at 33 Šresolution obtained by electron cryo-tomography and 3D subtomogram averaging. By fitting the atomic structure of tubulin into the EM density, we built a pseudo-atomic model of the tubulin protofilaments at the core of the triplet. The 3D density map reveals additional densities that represent non-tubulin proteins attached to the triplet, including a large inner circular structure in the basal body lumen, which functions as a scaffold to stabilize the entire basal body barrel. We found clear longitudinal structural variations along the basal body, suggesting a sequential and coordinated assembly mechanism. We propose a model in which δ-tubulin and other components participate in the assembly of the basal body.

Tomo3D 2.0--exploitation of advanced vector extensions (AVX) for 3D reconstruction.

Tomo3D is a program for fast tomographic reconstruction on multicore computers. Its high speed stems from code optimization, vectorization with Streaming SIMD Extensions (SSE), multithreading and optimization of disk access. Recently, Advanced Vector eXtensions (AVX) have been introduced in the x86 processor architecture. Compared to SSE, AVX double the number of simultaneous operations, thus pointing to a potential twofold gain in speed. However, in practice, achieving this potential is extremely difficult. Here, we provide a technical description and an assessment of the optimizations included in Tomo3D to take advantage of AVX instructions. Tomo3D 2.0 allows huge reconstructions to be calculated in standard computers in a matter of minutes. Thus, it will be a valuable tool for electron tomography studies with increasing resolution needs.

A marker-free automatic alignment method based on scale-invariant features.

In electron tomography, alignment accuracy is critical for high-resolution reconstruction. However, the automatic alignment of a tilt series without fiducial markers remains a challenge. Here, we propose a new alignment method based on Scale-Invariant Feature Transform (SIFT) for marker-free alignment. The method covers the detection and localization of interest points (features), feature matching, feature tracking and optimization of projection parameters. The proposed method implements a highly reliable matching strategy and tracking model to detect a huge number of feature tracks. Furthermore, an incremental bundle adjustment method is devised to tolerate noise data and ensure the accurate estimation of projection parameters. Our method was evaluated with a number of experimental data, and the results exhibit an improved alignment accuracy comparable with current fiducial marker alignment and subsequent higher resolution of tomography.

Robust membrane detection based on tensor voting for electron tomography.

Electron tomography enables three-dimensional (3D) visualization and analysis of the subcellular architecture at a resolution of a few nanometers. Segmentation of structural components present in 3D images (tomograms) is often necessary for their interpretation. However, it is severely hampered by a number of factors that are inherent to electron tomography (e.g. noise, low contrast, distortion). Thus, there is a need for new and improved computational methods to facilitate this challenging task. In this work, we present a new method for membrane segmentation that is based on anisotropic propagation of the local structural information using the tensor voting algorithm. The local structure at each voxel is then refined according to the information received from other voxels. Because voxels belonging to the same membrane have coherent structural information, the underlying global structure is strengthened. In this way, local information is easily integrated at a global scale to yield segmented structures. This method performs well under low signal-to-noise ratio typically found in tomograms of vitrified samples under cryo-tomography conditions and can bridge gaps present on membranes. The performance of the method is demonstrated by applications to tomograms of different biological samples and by quantitative comparison with standard template matching procedure.

Markerauto2: A fast and robust fully automatic fiducial marker-based tilt series alignment software for electron tomography.

Cryoelectron tomography (cryo-ET) has become an indispensable technology for visualizing in situ biological ultrastructures, where the tilt series alignment is the key step to obtain a high-resolution three-dimensional reconstruction. Specifically, with the advent of high-throughput cryo-ET data collection, there is an increasing demand for high-accuracy and fully automatic tilt series alignment, to enable efficient data processing. Here, we propose Markerauto2, a fast and robust fully automatic software that enables accurate fiducial marker-based tilt series alignment. Markerauto2 implements the following novel pipelines: (1) an accelerated high-precision fiducial marker detection with wavelet multiscale template, (2) an ultra-fast and robust fiducial marker tracking supported by hashed geometric features, (3) a high-angle fiducial marker supplementation strategy to produce more complete tracks, and (4) a precise and robust calibration of projection parameters with group-weighted parameter optimization. Comprehensive experiments conducted on both simulated and real-world datasets demonstrate the robustness, efficiency, and effectiveness of the proposed software.

Disruption of the mitochondrial network in a mouse model of Huntington's disease visualized by in-tissue multiscale 3D electron microscopy.

Huntington's disease (HD) is an inherited neurodegenerative disorder caused by an expanded CAG repeat in the coding sequence of huntingtin protein. Initially, it predominantly affects medium-sized spiny neurons (MSSNs) of the corpus striatum. No effective treatment is still available, thus urging the identification of potential therapeutic targets. While evidence of mitochondrial structural alterations in HD exists, previous studies mainly employed 2D approaches and were performed outside the strictly native brain context. In this study, we adopted a novel multiscale approach to conduct a comprehensive 3D in situ structural analysis of mitochondrial disturbances in a mouse model of HD. We investigated MSSNs within brain tissue under optimal structural conditions utilizing state-of-the-art 3D imaging technologies, specifically FIB/SEM for the complete imaging of neuronal somas and Electron Tomography for detailed morphological examination, and image processing-based quantitative analysis. Our findings suggest a disruption of the mitochondrial network towards fragmentation in HD. The network of interlaced, slim and long mitochondria observed in healthy conditions transforms into isolated, swollen and short entities, with internal cristae disorganization, cavities and abnormally large matrix granules.

A ridge-based framework for segmentation of 3D electron microscopy datasets.

Three-dimensional (3D) electron microscopy (EM) has become a major player in structural cell biology as it enables the analysis of subcellular architecture at an unprecedented level of detail. Interpretation of the resulting 3D volumes strongly depends on segmentation, which consists in decomposing the volume into their structural components. The computational approaches proposed so far have not turned out to be of general applicability. Thus, manual segmentation still remains a prevalent method. Here, a new computational framework for segmentation of 3D EM datasets is introduced. It relies on detection and characterization of ridges (i.e. local maxima). The detected ridges are modelled as asymmetric Gaussian functions whose parameters constitute ridge descriptors. This local information is then used to cluster the ridges, which leads to the ultimate segmentation. In this work we focus on membranes and locally planar structures in general. The performance of the framework is illustrated with its application to a number of complex 3D datasets and a quantitative analysis.

Electron cryo-tomography structure of axonemal doublet microtubule from Tetrahymena thermophila.

Doublet microtubules (DMTs) provide a scaffold for axoneme assembly in motile cilia. Aside from α/ß tubulins, the DMT comprises a large number of non-tubulin proteins in the luminal wall of DMTs, collectively named the microtubule inner proteins (MIPs). We used cryoET to study axoneme DMT isolated from Tetrahymena We present the structures of DMT at nanometer and sub-nanometer resolution. The structures confirm that MIP RIB72A/B binds to the luminal wall of DMT by multiple DM10 domains. We found FAP115, an MIP-containing multiple EF-hand domains, located at the interface of four-tubulin dimers in the lumen of A-tubule. It contacts both lateral and longitudinal tubulin interfaces and playing a critical role in DMT stability. We observed substantial structure heterogeneity in DMT in an FAP115 knockout strain, showing extensive structural defects beyond the FAP115-binding site. The defects propagate along the axoneme. Finally, by comparing DMT structures from Tetrahymena and Chlamydomonas, we have identified a number of conserved MIPs as well as MIPs that are unique to each organism. This conservation and diversity of the DMT structures might be linked to their specific functions. Our work provides structural insights essential for understanding the roles of MIPs during motile cilium assembly and function, as well as their relationships to human ciliopathies.

A differential structure approach to membrane segmentation in electron tomography.

Electron tomography allows three-dimensional visualization of cellular landscapes in molecular detail. Segmentation is a paramount stage for the interpretation of the reconstructed tomograms. Although several computational approaches have been proposed, none has prevailed as a generic method and thus segmentation through manual annotation is still a common choice. In this work we introduce a segmentation method targeted at membranes, which define the natural limits of compartments within biological specimens. Our method is based on local differential structure and on a Gaussian-like membrane model. First, it isolates information through scale-space and finds potential membrane-like points at a local scale. Then, the structural information is integrated at a global scale to yield the definite segmentation. We show and validate the performance of the algorithm on a number of tomograms under different experimental conditions.