Deep Learning-Based Segmentation of Cryo-Electron Tomograms.

Cryo-electron tomography (cryo-ET) allows researchers to image cells in their native, hydrated state at the highest resolution currently possible. The technique has several limitations, however, that make analyzing the data it generates time-intensive and difficult. Hand segmenting a single tomogram can take from hours to days, but a microscope can easily generate 50 or more tomograms a day. Current deep learning segmentation programs for cryo-ET do exist, but are limited to segmenting one structure at a time. Here, multi-slice U-Net convolutional neural networks are trained and applied to automatically segment multiple structures simultaneously within cryo-tomograms. With proper preprocessing, these networks can be robustly inferred to many tomograms without the need for training individual networks for each tomogram. This workflow dramatically improves the speed with which cryo-electron tomograms can be analyzed by cutting segmentation time down to under 30 min in most cases. Further, segmentations can be used to improve the accuracy of filament tracing within a cellular context and to rapidly extract coordinates for subtomogram averaging.

Cofilactin filaments regulate filopodial structure and dynamics in neuronal growth cones.

Cofilin is best known for its ability to sever actin filaments and facilitate cytoskeletal recycling inside of cells, but at higher concentrations in vitro, cofilin stabilizes a more flexible, hyper-twisted state of actin known as "cofilactin". While this filament state is well studied, a structural role for cofilactin in dynamic cellular processes has not been observed. With a combination of cryo-electron tomography and fluorescence imaging in neuronal growth cones, we observe that filopodial actin filaments switch between a fascin-linked and a cofilin-decorated state, and that cofilactin is associated with a variety of dynamic events within filopodia. The switch to cofilactin filaments occurs in a graded fashion and correlates with a decline in fascin cross-linking within the filopodia, which is associated with curvature in the bundle. Our tomographic data reveal that the hyper-twisting of actin from cofilin binding leads to a rearrangement of filament packing, which largely excludes fascin from the base of filopodia. Our results provide mechanistic insight into the fundamentals of cytoskeletal remodeling inside of confined cellular spaces, and how the interplay between fascin and cofilin regulates the dynamics of searching filopodia.

Challenges and triumphs in cryo-electron tomography.

Cryo-electron tomography has stepped fully into the spotlight. Enthusiasm is high. Fortunately for us, this is an exciting time to be a cryotomographer, but there is still a way to go before declaring victory. Despite its potential, cryo-electron tomography possesses many inherent challenges. How do we image through thick cell samples, and possibly even tissue? How do we identify a protein of interest amidst the noisy, crowded environment of the cytoplasm? How do we target specific moments of a dynamic cellular process for tomographic imaging? In this review, we cover the history of cryo-electron tomography and how it came to be, roughly speaking, as well as the many approaches that have been developed to overcome its intrinsic limitations.

Cryo-Electron Tomography and Automatic Segmentation of Cultured Hippocampal Neurons.

Cryo-electron tomography is fast becoming a preferred method for studying intracellular environments at the molecular scale. Increases in data collection throughput means that large numbers of tomograms can be generated at rates too fast for humans to easily explore quantitatively. Currently, there is a large effort to make data collection and segmentation tools more automated. Here, we describe a workflow for preparing cultured neurons on electron microscopy grids, batch tomographic data collection, reconstruction and automatic segmentation using freely and commercially available software.