Automated Stitching of Microtubule Centerlines across Serial Electron Tomograms

Abstract

Tracing microtubule centerlines in serial section electron tomography requires microtubules to be stitched across sections, that is lines from different sections need to be aligned, endpoints need to be matched at section boundaries to establish a correspondence between neighboring sections, and corresponding lines need to be connected across multiple sections. We present computational methods for these tasks: 1) An initial alignment is computed using a distance compatibility graph. 2) A fine alignment is then computed with a probabilistic variant of the iterative closest points algorithm, which we extended to handle the orientation of lines by introducing a periodic random variable to the probabilistic formulation. 3) Endpoint correspondence is established by formulating a matching problem in terms of a Markov random field and computing the best matching with belief propagation. Belief propagation is not generally guaranteed to converge to a minimum. We show how convergence can be achieved, nonetheless, with minimal manual input. In addition to stitching microtubule centerlines, the correspondence is also applied to transform and merge the electron tomograms. We applied the proposed methods to samples from the mitotic spindle in C. elegans, the meiotic spindle in X. laevis, and sub-pellicular microtubule arrays in T. brucei. The methods were able to stitch microtubules across section boundaries in good agreement with experts' opinions for the spindle samples. Results, however, were not satisfactory for the microtubule arrays. For certain experiments, such as an analysis of the spindle, the proposed methods can replace manual expert tracing and thus enable the analysis of microtubules over long distances with reasonable manual effort.

Central spindle microtubules are strongly coupled to chromosomes during both anaphase A and anaphase B

“…The IMOD software package (http://bio3d.colourado.edu/imod) was used for the calculation of electron tomograms (Kremer et al, 1996). The Amira software package with an extension to the filament editor was used for the segmentation, automatic tracing, stitching and 3D visualization of microtubules, and for data analysis (Stalling et al, 2005; Weber et al, 2012; Weber et al, 2014).…”

C. eleganschromosomes connect to centrosomes by anchoring into the spindle network

“…We applied the Amira software package with an extension to the filament editor of the Amira visualization and data analysis software for the segmentation and automatic tracing of microtubules 60 . We also used the Amira software to stitch the obtained 3D models in z to create full volumes of the recorded spindles 61 . The automatic segmentation of the spindle microtubules was followed by a visual inspection of the traced microtubules within the tomograms and correction of the individual microtubule tracings.…”

“…In addition, the traced microtubules were manually verified. It is more difficult to estimate the error of the matching algorithm 61 , since it depends on the local density and properties of the microtubules. For this reason, the stitched microtubules were manually verified and corrected for all KMTs.…”

Soluble tubulin is significantly enriched at mitotic centrosomes

“…This correction included manual tracing of undetected microtubules, extending and combining individual traces as well as removing incorrectly traced structures, such as membranes and other cellular components. Using the AMIRA software, corrected 3D models were then stitched in z to obtain complete volumes of the recorded centrosomes (Weber et al, 2014). In total, we recorded eight WT and two nocodazole-treated centrosomes covering an average range of 5 µm in the slice plane.…”