Automated segmentation of cell organelles in volume electron microscopy using deep learning.

Recent advances in computing power triggered the use of artificial intelligence in image analysis in life sciences. To train these algorithms, a large enough set of certified labeled data is required. The trained neural network is then capable of producing accurate instance segmentation results that will then need to be re-assembled into the original dataset: the entire process requires substantial expertise and time to achieve quantifiable results. To speed-up the process, from cell organelle detection to quantification across electron microscopy modalities, we propose a deep-learning based approach for fast automatic outline segmentation (FAMOUS), that involves organelle detection combined with image morphology, and 3D meshing to automatically segment, visualize and quantify cell organelles within volume electron microscopy datasets. From start to finish, FAMOUS provides full segmentation results within a week on previously unseen datasets. FAMOUS was showcased on a HeLa cell dataset acquired using a focused ion beam scanning electron microscope, and on yeast cells acquired by transmission electron tomography. RESEARCH HIGHLIGHTS: Introducing a rapid, multimodal machine-learning workflow for the automatic segmentation of 3D cell organelles. Successfully applied to a variety of volume electron microscopy datasets and cell lines. Outperforming manual segmentation methods in time and accuracy. Enabling high-throughput quantitative cell biology.

Nuclear envelope budding and its cellular functions.

The nuclear pore complex (NPC) has long been assumed to be the sole route across the nuclear envelope, and under normal homeostatic conditions it is indeed the main mechanism of nucleo-cytoplasmic transport. However, it has also been known that e.g. herpesviruses cross the nuclear envelope utilizing a pathway entitled nuclear egress or envelopment/de-envelopment. Despite this, a thread of observations suggests that mechanisms similar to viral egress may be transiently used also in healthy cells. It has since been proposed that mechanisms like nuclear envelope budding (NEB) can facilitate the transport of RNA granules, aggregated proteins, inner nuclear membrane proteins, and mis-assembled NPCs. Herein, we will summarize the known roles of NEB as a physiological and intrinsic cellular feature and highlight the many unanswered questions surrounding these intriguing nuclear events.

Using reporters of different misfolded proteins reveals differential strategies in processing protein aggregates.

The accumulation of misfolded proteins is a hallmark of aging and many neurodegenerative diseases, making it important to understand how the cellular machinery recognizes and processes such proteins. A key question in this respect is whether misfolded proteins are handled in a similar way regardless of their genetic origin. To approach this question, we compared how three different misfolded proteins, guk1-7, gus1-3, and pro3-1, are handled by the cell. We show that all three are nontoxic, even though highly overexpressed, highlighting their usefulness in analyzing the cellular response to misfolding in the absence of severe stress. We found significant differences between the aggregation and disaggregation behavior of the misfolded proteins. Specifically, gus1-3 formed some aggregates that did not efficiently recruit the protein disaggregase Hsp104 and did not colocalize with the other misfolded reporter proteins. Strikingly, while all three misfolded proteins generally coaggregated and colocalized to specific sites in the cell, disaggregation was notably different; the rate of aggregate clearance of pro3-1 was faster than that of the other misfolded proteins, and its clearance rate was not hindered when pro3-1 colocalized with a slowly resolved misfolded protein. Finally, we observed using super-resolution light microscopy as well as immunogold labeling EM in which both showed an even distribution of the different misfolded proteins within an inclusion, suggesting that misfolding characteristics and remodeling, rather than spatial compartmentalization, allows for differential clearance of these misfolding reporters residing in the same inclusion. Taken together, our results highlight how properties of misfolded proteins can significantly affect processing.

Large organellar changes occur during mild heat shock in yeast.

When the temperature is increased, the heat-shock response is activated to protect the cellular environment. The transcriptomics and proteomics of this process are intensively studied, while information about how the cell responds structurally to heat stress is mostly lacking. Here, Saccharomyces cerevisiae were subjected to a mild continuous heat shock (38°C) and intermittently cryo-immobilised for electron microscopy. Through measuring changes in all distinguishable organelle numbers, sizes and morphologies in over 2100 electron micrographs, a major restructuring of the internal architecture of the cell during the progressive heat shock was revealed. The cell grew larger but most organelles within it expanded even more, shrinking the volume of the cytoplasm. Organelles responded to heat shock at different times, both in terms of size and number, and adaptations of the morphology of some organelles (such as the vacuole) were observed. Multivesicular bodies grew by almost 70%, indicating a previously unknown involvement in the heat-shock response. A previously undescribed electron-translucent structure accumulated close to the plasma membrane. This all-encompassing approach provides a detailed chronological progression of organelle adaptation throughout the cellular heat-stress response.

High-Resolution Episcopic Microscopy (HREM) in Multimodal Imaging Approaches.

High-resolution episcopic microscopy (HREM) is a three-dimensional (3D) episcopic imaging modality based on the acquisition of two-dimensional (2D) images from the cut surface of a block of tissue embedded in resin. Such images, acquired serially through the entire length/depth of the tissue block, are aligned and stacked for 3D reconstruction. HREM has proven to be specifically advantageous when integrated in correlative multimodal imaging (CMI) pipelines. CMI creates a composite and zoomable view of exactly the same specimen and region of interest by (sequentially) correlating two or more modalities. CMI combines complementary modalities to gain holistic structural, functional, and chemical information of the entire sample and place molecular details into their overall spatiotemporal multiscale context. HREM has an advantage over in vivo 3D imaging techniques on account of better histomorphologic resolution while simultaneously providing volume data. HREM also has certain advantages over ex vivo light microscopy modalities. The latter can provide better cellular resolution but usually covers a limited area or volume of tissue, with limited 3D structural context. HREM has predominantly filled a niche in the phenotyping of embryos and characterisation of anatomic developmental abnormalities in various species. Under the umbrella of CMI, when combined with histopathology in a mutually complementary manner, HREM could find wider application in additional nonclinical and translational areas. HREM, being a modified histology technique, could also be incorporated into specialised preclinical pathology workflows. This review will highlight HREM as a versatile imaging platform in CMI approaches and present its benefits and limitations.