Montage electron tomography of vitrified specimens

Abstract

Cryo-electron tomography provides detailed views of macromolecules in situ. However, imaging a large field of view to provide more cellular context requires reducing magnification during data collection, which in turn restricts the resolution. To circumvent this trade-off between field of view and resolution, we have developed a montage data collection scheme that uniformly distributes the dose throughout the specimen. In this approach, sets of slightly overlapping circular tiles are collected at high magnification and stitched to form a composite projection image at each tilt angle. These montage tilt-series are then reconstructed into massive tomograms with a small pixel size but a large field of view. For proof-of-principle, we applied this method to the thin edge of HeLa cells. Thon rings to better than 10 Å were detected in the montaged tilt-series, and diverse cellular features were observed in the resulting tomograms. These results indicate that the additional dose required by this technique is not prohibitive to performing structural analysis to intermediate resolution across a large field of view. We anticipate that montage tomography will prove particularly useful for lamellae, increase the likelihood of imaging rare cellular events, and facilitate visual proteomics.

Keywords: Cellular biology; Cryo-electron microscopy; Image processing; Montage tomography; Ultrastructural analysis.

Fig. 1.. Optimization of a montage tiling strategy.

(A) At each tilt angle, a hexagonally-packed set of circular tiles is imaged. Applying a global (B) rotation or (C) translation to the tiles between tilt angles changes which regions of the specimen lie in an overlap region to more uniformly spread the dose. To systematically introduce translational offsets between tilt angles, the global displacements applied to the tiles between tilt angles followed one of three spiral patterns: (D) an Archimedean spiral, (E) a sunflower spiral, or (F) a “snowflake” spiral. The positions of the central tile are indicated by black dots and spiral outwards during the course of the tilt-series. For the snowflake spiral, the pattern is repeated starting from the center if the outermost position is reached before the final tilt angle. One of the tunable parameters for all three spiral patterns is the maximum translation of the central tile (t_max); a second parameter for the Archimedean and snowflake patterns is the number of revolutions (n_rev or n_steps). (G) The dose distributions received by voxels of a discretized specimen during a simulated tilt-series are compared for four tiling strategies. The spatial distribution of the dose is mapped on the specimen (upper). Each tiling strategy was scored by the variance (σ²) of the dose distribution, which is noted at the upper right of each histogram (lower). For the “no offsets” strategy, the hexagonally-packed tiles were rotated by 10° relative to the plane of the detector but no offsets were applied to tile positions between tilt angles. For the “rotations only” strategy, a 20° clockwise rotation was applied to all tiles between tilt angles. For the “translations only” strategy, tile positions were translated along an Archimedean spiral with three revolutions and a maximum translation of 80% of the beam radius for each tile. The best-ranked strategy (right) applies both of these translational and rotational offsets to efficiently distribute the dose.

Fig. 2.. CTF estimation reveals a stable defocus throughout the tilt-series and Thon rings to better than 10 Å in the untilted stitched images.

(A) Defocus values were predicted based on the tiles’ estimated heights in the microscope and are shown as a function of tilt angle (left). By contrast, the per-tile defocus values estimated by CTFFIND4 showed a relatively stable defocus gradient, as plotted for two representative tilt-series (middle and right). This was accomplished by performing an autofocusing step prior to collecting each tile to compensate for the predicted defocus gradient across each tilt angle. (B) The 2d experimental spectrum (upper) and rotationally-averaged 1d CTF fits (lower) are shown for representative tiles (left) or CTF-uncorrected stitched projection images (right) at the indicated tilt angle.

Fig. 3.. Continuity of cellular features in the overlap regions indicates successful stitching.

The montaged projection image at 0° from a representative tilt-series is shown in the upper left inset. The region boxed in red is visualized at higher detail in the main image, with the boundaries of the circular tiles drawn in black. The clear and continuous membranous features visible in the overlap regions between adjacent tiles suggests both successful stitching and minimal radiation damage despite the extra dose. The diameter of each circular tile is 1.08 μm.

Fig. 4.. Diverse cellular features are observed in an example cryotomogram reconstructed from montage tilt-series.

A slice is shown from a representative montage tomogram that spans a 3.3 μm² field of view.

Fig. 5.. Representative montage cryotomogram from the thin edge of a HeLa cell.

A tomographic slice spanning a 3.3 μm² field of view is visualized, with cellular features of interest annotated.

Fig. 6.. Insets from representative montage cryotomograms.

Close-up views from the cryotomograms presented in Figs. 4 and 5 visualize the mitochondria (upper) and microtubules (lower) in greater detail.