Alignment algorithms and per-particle CTF correction for single particle cryo-electron tomography

Abstract

Single particle cryo-electron tomography (cryoSPT) extracts features from cryo-electron tomograms, followed by 3D classification, alignment and averaging to generate improved 3D density maps of such features. Robust methods to correct for the contrast transfer function (CTF) of the electron microscope are necessary for cryoSPT to reach its resolution potential. Many factors can make CTF correction for cryoSPT challenging, such as lack of eucentricity of the specimen stage, inherent low dose per image, specimen charging, beam-induced specimen motions, and defocus gradients resulting both from specimen tilting and from unpredictable ice thickness variations. Current CTF correction methods for cryoET make at least one of the following assumptions: that the defocus at the center of the image is the same across the images of a tiltseries, that the particles all lie at the same Z-height in the embedding ice, and/or that the specimen grid and carbon support are flat. These experimental conditions are not always met. We have developed a CTF correction algorithm for cryoSPT without making any of the aforementioned assumptions. We also introduce speed and accuracy improvements and a higher degree of automation to the subtomogram averaging algorithms available in EMAN2. Using motion-corrected images of isolated virus particles as a benchmark specimen, recorded with a DE20 direct detection camera, we show that our CTF correction and subtomogram alignment routines can yield subtomogram averages close to 4/5 Nyquist frequency of the detector under our experimental conditions.

Integrative structure and functional anatomy of a nuclear pore complex

“…A subsequent 2D classification identified near top, tilted and side views; the latter provided an independent estimate of NPC thickness perpendicular to the NE (620–640 Å). A 3D reconstruction with the best sub-tomograms produced a map at ~35 Å resolution (data not shown) with similar features to those obtained with the e2spt package 75,76 , including distinct connections between each spoke and the transporter, further validating the features in our Cryo-ET map (Fig. 3).…”

“…After interactive test runs with etomo 74 , we processed the tilt series in an automated fashion with batchtomo using a 7 × 7 patch tracking to create aligned tilt series and calculated back-projection and SIRT tomograms which were CTF-corrected by phase-flipping each image in the tilt series (Supplementary Power Point Presentation, slides 3–6). The final SIRT tomograms were binned 3× and used for interactive sub-tomogram "particle" picking with e2spt_boxer.py in the EMAN2 single particle tomography package 75,76 with a low pass filter of 100 Å. In total 6,416 fully sampled unfiltered sub-tomograms were extracted from the back-projection tomograms in 300 × 300 × 300 voxel volumes.…”

“…In the alignment and averaging process, new algorithms for high-speed 3D alignment with automatic missing wedge compensation and averaging 76 , were employed throughout, and were critical for processing such large sub-tomograms. An initial reference was prepared by averaging a small subset of sub-tomograms, to produce a low-resolution reference using the C 8 -symmetry of the complex 77 .…”

The advent of structural biology in situ by single particle cryo-electron tomography

“…), using only a few hundred icosahedrally symmetric virus particles (Galaz-Montoya et al 2016a). Most importantly, the defocus gradient due to tilting was fitted by directly measuring the power spectrum in strips of constant defocus, similar to the method proposed by Fernández et al (2006), except that it was done on each individual image ( i.e.…”

The integrative role of cryo electron microscopy in molecular and cellular structural biology

“…The images of a tilt series can then be aligned to reconstruct a tomogram (see for example Briggs, ; Lučič et al., ; Dubrovsky et al., ; Asano et al., ). If common sub‐structures exist these can be extracted as sub‐tomograms, aligned and averaged to form sub‐tomogram averages with an improved signal‐to‐noise ratio (Wan et al., ; Galaz‐Montoya et al., ). Such approaches have been used in the past to analyse, for example, the cellular structure of the nuclear pore complex (NPC), of membrane‐bound ribosomes, of actin filaments in combination with segmentation tools and template‐based identification of the components (Frangakis et al., ; Medalia et al., ; Beck et al., ; Ortiz et al., ).…”