J.M. Carazo

Disentangling conformational states of macromolecules in 3D-EM through likelihood optimization

Although three-dimensional electron microscopy (3D-EM) permits structural characterization of macromolecular assemblies in distinct functional states, the inability to classify projections from structurally heterogeneous samples has severely limited its application. We present a maximum likelihood–based classification method that does not depend on prior knowledge about the structural variability, and demonstrate its effectiveness for two macromolecular assemblies with different types of conformational variability: the Escherichia coli ribosome and Simian virus 40 (SV40) large T-antigen.

A clustering approach to multireference alignment of single-particle projections in electron microscopy.

Two-dimensional analysis of projections of single-particles acquired by an electron microscope is a useful tool to help identifying the different kinds of projections present in a dataset and their different projection directions. Such analysis is also useful to distinguish between different kinds of particles or different particle conformations. In this paper we introduce a new algorithm for performing two-dimensional multireference alignment and classification that is based on a Hierarchical clustering approach using correntropy (instead of the more traditional correlation) and a modified criterion for the definition of the clusters specially suited for cases in which the Signal-to-Noise Ratio of the differences between classes is low. We show that our algorithm offers an improved sensitivity over current methods in use for distinguishing between different projection orientations and different particle conformations. This algorithm is publicly available through the software package Xmipp.

Insights into the molecular architecture of the 26S proteasome.

Cryo-electron microscopy in conjunction with advanced image analysis was used to analyze the structure of the 26S proteasome and to elucidate its variable features. We have been able to outline the boundaries of the ATPase module in the “base” part of the regulatory complex that can vary in its position and orientation relative to the 20S core particle. This variation is consistent with the “wobbling” model that was previously proposed to explain the role of the regulatory complex in opening the gate in the α-rings of the core particle. In addition, a variable mass near the mouth of the ATPase ring has been identified as Rpn10, a multiubiquitin receptor, by correlating the electron microscopy data with quantitative mass spectrometry.

Pattern recognition and classification of images of biological macromolecules using artificial neural networks

The goal of this work was to analyze an image data set and to detect the structural variability within this set. Two algorithms for pattern recognition based on neural networks are presented, one that performs an unsupervised classification (the self-organizing map) and the other a supervised classification (the learning vector quantization). The approach has a direct impact in current strategies for structural determination from electron microscopic images of biological macromolecules. In this work we performed a classification of both aligned but heterogeneous image data sets as well as basically homogeneous but otherwise rotationally misaligned image populations, in the latter case completely avoiding the typical reference dependency of correlation-based alignment methods. A number of examples on chaperonins are presented. The approach is computationally fast and robust with respect to noise. Programs are available through ftp.

Averaging of Electron Subtomograms and Random Conical Tilt Reconstructions through Likelihood Optimization

The reference-free averaging of three-dimensional electron microscopy (3D-EM) reconstructions with empty regions in Fourier space represents a pressing problem in electron tomography and single-particle analysis. We present a maximum likelihood algorithm for the simultaneous alignment and classification of subtomograms or random conical tilt (RCT) reconstructions, where the Fourier components in the missing data regions are treated as hidden variables. The behavior of this algorithm was explored using tests on simulated data, while application to experimental data was shown to yield unsupervised class averages for subtomograms of groEL/groES complexes and RCT reconstructions of p53. The latter application served to obtain a reliable de novo structure for p53 that may resolve uncertainties about its quaternary structure.

New electron microscopy database and deposition system

To manage, organize and disseminate data on the structure of biological macromolecules solved by 3D electron microscopy, an electron microscopy database has been set up at the European Bioinformatics Institute.

High-performance electron tomography of complex biological specimens.

We have evaluated reconstruction methods using smooth basis functions in the electron tomography of complex biological specimens. In particular, we have investigated series expansion methods, with special emphasis on parallel computation. Among the methods investigated, the component averaging techniques have proven to be most efficient and have generally shown fast convergence rates. The use of smooth basis functions provides the reconstruction algorithms with an implicit regularization mechanism, very appropriate for noisy conditions. Furthermore, we have applied high-performance computing (HPC) techniques to address the computational requirements demanded by the reconstruction of large volumes. One of the standard techniques in parallel computing, domain decomposition, has yielded an effective computational algorithm which hides the latencies due to interprocessor communication. We present comparisons with weighted back-projection (WBP), one of the standard reconstruction methods in the areas of computational demand and reconstruction quality under noisy conditions. These techniques yield better results, according to objective measures of quality, than the weighted backprojection techniques after a very few iterations. As a consequence, the combination of efficient iterative algorithms and HPC techniques has proven to be well suited to the reconstruction of large biological specimens in electron tomography, yielding solutions in reasonable computation times.

Classification of images of biomolecular assemblies: a study of ribosomes and ribosomal subunits of Escherichia coli.

Images of macromolecules obtained in the electron microscope are subjected to correspondence analysis. The structure inherent in the data in the resulting low‐dimensional factor space is characterized by a mixed classification method which combines the dynamic clouds clustering technique with hierarchical ascendant classification (HAC). For our data, the rejection of marginal clusters obtained by dynamic clouds clustering appears as a crucial prerequisite for a stable performance of HAC.

Scipion: A software framework toward integration, reproducibility and validation in 3D electron microscopy

In the past few years, 3D electron microscopy (3DEM) has undergone a revolution in instrumentation and methodology. One of the central players in this wide-reaching change is the continuous development of image processing software. Here we present Scipion, a software framework for integrating several 3DEM software packages through a workflow-based approach. Scipion allows the execution of reusable, standardized, traceable and reproducible image-processing protocols. These protocols incorporate tools from different programs while providing full interoperability among them. Scipion is an open-source project that can be downloaded from http://scipion.cnb.csic.es.

Three-dimensional reconstruction of the connector of bacteriophage φ29 at 1.8 nm resolution

The three-dimensional reconstruction of the connector of bacteriophage phi 29 has been obtained from tilt series of negatively stained tetragonal ordered aggregates under low-dose conditions and up to a resolution of (1/1.8) nm-1. These connectors are built up as dodecamers of only one structural polypeptide (p10). Two connectors form the crystal unit cell, each one facing in the opposite direction with respect to the plane of the crystal and partially overlapping. The main features of the two connectors that build the unit cell were essentially the same, although they were negatively stained in slightly different ways, probably due to their situations with respect to the carbon-coated support grid. The main features of the phi 29 connector structure revealed by this three-dimensional reconstruction are: the existence of two clearly defined domains, one with a diameter of around 14 nm and the other narrower (diameter approximately equal to 7.5 nm); an inner hole running all along the structure (around 7 to 8 nm in height) with a cylindrical profile and an average diameter of 4 nm; a general 6-fold symmetry along the whole structure and a 12-fold one in the wider domain; a clockwise twist of the more contrasted regions of both domains from the narrower towards the wider domain (the direction of DNA encapsidation). These features are compatible with an active role for the connector in the process of DNA packaging.