Katja Graumann

U-Net is a convolutional neural network model developed in 2015 and has proven to be one of the most inspiring deep-learning models for image segmentation. Numerous U-Net-based applications have since emerged, constituting a heterogeneous set of tools that illustrate the current reproducibility crisis in the deep-learning field and that remain slow to spread in application fields such as 3D bioimaging. Here we propose a solution named Biom3d, a modular deep learning framework facilitating the integration and development of novel models, metrics, or training schemes for 3D image segmentation. The development philosophy of Biom3d provides an improved code sustainability and reproducibility in line with the FAIR principles and is available as a graphical user interface and an open-source deep-learning framework to target a large community of users, from end users to deep learning developers.

A distinguishing feature of eukaryotes is the presence of a nuclear envelope (NE) and endomembrane system. The NE is a double-membrane system that surrounds chromatin and is continuous with the endoplasmic reticulum (ER). This interface is crucial in various processes such as calcium signaling and ER-associated degradation. The outer nuclear membrane and ER share a multitude of proteins although some are only functional in one domain, whereas the inner nuclear membrane has its own unique proteome. Until recently, it was not possible to distinguish between the inner and outer nuclear membranes as well as perinuclear ER using light microscopy – only electron microscopy was suitable for this. Now, however, using super-resolution live cell imaging, this can be achieved while still observing protein and membrane dynamics in real time. The protocols described here will allow researchers to determine subcellular localization of potential NE/ER proteins in live plant cells, helping to gain new insights into protein functionality. Key words Nuclear envelope Endoplasmic reticulum Confocal microscopy Proximity distance Fluorescence intensity ratio

Mid-SUN proteins are a neglected family of conserved type III membrane proteins of ancient origin with representatives in plants, animals, and fungi. Previous higher plant studies have associated them with functions at the nuclear envelope and the endoplasmic reticulum (ER). In this study, high-resolution confocal light microscopy is used to explore the localisation of SUN3 and SUN4 in the perinuclear region, to explore topology, and to study the role of mid-SUNs on endoplasmic reticulum morphology. The role of SUN3 in the ER is reinforced by the identification of a protein interaction between SUN3 and the ER membrane-bound transcription factor maMYB. The results highlight the importance of mid-SUNs as functional components of the ER and outer nuclear membrane.

The plant nucleus and the actin cytoskeleton are intimately connected. The actin cytoskeleton is pivotal for nuclear positioning, shape, and dynamics. These properties of the nucleus are important for its functions during normal development and in response to external cues such as biotic and abiotic stresses. Moreover, we know that there is a direct physical connection between the actin cytoskeleton and the nucleus which spans the double-membraned nuclear envelope into the nuclear lamina, and this connection is called the linker of nucleoskeleton and cytoskeleton (LINC) complex. Recently a role for actin in regulating inter-nuclear organization via the control of nuclear invaginations has emerged. Therefore, a detailed understanding of nuclear shape, organization, and dynamics and the techniques used to measure and quantify these metrics will allow us to determine and further understand the contribution made by actin to these parameters. The protocols described here will allow researchers to determine the circularity index of a nucleus, quantify nuclear deformations, and determine dynamics of nuclei within plant cells. Key words Actin Nucleus Klarsicht ANC-1 Syne homology (KASH) Linker of nucleoskeleton and cytoskeleton (LINC) Confocal microscope Circularity index Kymographs

Editorial: Understanding the key border: Structure, function, and dynamics of the plant nuclear envelope

For the past century, the nucleus has been the focus of extensive investigations in cell biology. However, many questions remain about how its shape and size are regulated during development, in different tissues, or during disease and aging. To track these changes, microscopy has long been the tool of choice. Image analysis has revolutionized this field of research by providing computational tools that can be used to translate qualitative images into quantitative parameters. Many tools have been designed to delimit objects in 2D and, eventually, in 3D in order to define their shapes, their number or their position in nuclear space. Today, the field is driven by deep-learning methods, most of which take advantage of convolutional neural networks. These techniques are remarkably adapted to biomedical images when trained using large datasets and powerful computer graphics cards. To promote these innovative and promising methods to cell biologists, this Review summarizes the main concepts and terminologies of deep learning. Special emphasis is placed on the availability of these methods. We highlight why the quality and characteristics of training image datasets are important and where to find them, as well as how to create, store and share image datasets. Finally, we describe deep-learning methods well-suited for 3D analysis of nuclei and classify them according to their level of usability for biologists. Out of more than 150 published methods, we identify fewer than 12 that biologists can use, and we explain why this is the case. Based on this experience, we propose best practices to share deep-learning methods with biologists.

In eukaryotes, the nuclear envelope (NE) encloses chromatin and separates it from the rest of the cell. The Linker of Nucleoskeleton and Cytoskeleton (LINC) complex physically bridges across the NE, linking nuclear and cytoplasmic components. In plants, these LINC complexes are beginning to be ascribed roles in cellular and nuclear functions, including chromatin organization, regulation of nuclei shape and movement, and cell division. Homologs of core LINC components, KASH and SUN proteins, have previously been identified in maize. Here, we characterized the presumed LINC-associated maize nucleoskeletal proteins NCH1 and NCH2, homologous to members of the plant NMCP/CRWN family, and MKAKU41, homologous to AtKAKU4. All three proteins localized to the nuclear periphery when transiently and heterologously expressed as fluorescent protein fusions in Nicotiana benthamiana. Overexpression of MKAKU41 caused dramatic changes in the organization of the nuclear periphery, including nuclear invaginations that stained positive for non-nucleoplasmic markers of the inner and outer NE membranes, and the ER. The severity of these invaginations was altered by changes in LINC connections and the actin cytoskeleton. In maize, MKAKU41 appeared to share genetic functions with other LINC components, including control of nuclei shape, stomatal complex development, and pollen viability. Overall, our data show that NCH1, NCH2, and MKAKU41 have characteristic properties of LINC-associated plant nucleoskeletal proteins, including interactions with NE components suggestive of functions at the nuclear periphery that impact the overall nuclear architecture.

The eukaryotic cell nucleus is a central organelle whose architecture determines genome function at multiple levels. Deciphering nuclear organizing principles influencing cellular responses and identity is a timely challenge. Despite many similarities between plant and animal nuclei, plant nuclei present intriguing specificities. Complementary to molecular and biochemical approaches, 3D microscopy is indispensable for resolving nuclear architecture. However, novel solutions are required for capturing cell-specific, sub-nuclear and dynamic processes. We provide a pointer for utilising high-to-super-resolution microscopy and image processing to probe plant nuclear architecture in 3D at the best possible spatial and temporal resolution and at quantitative and cell-specific levels. High-end imaging and image-processing solutions allow the community now to transcend conventional practices and benefit from continuously improving approaches. These promise to deliver a comprehensive, 3D view of plant nuclear architecture and to capture spatial dynamics of the nuclear compartment in relation to cellular states and responses. Abbreviations: 3D and 4D: Three and Four dimensional; AI: Artificial Intelligence; ant: antipodal nuclei (ant); CLSM: Confocal Laser Scanning Microscopy; CTs: Chromosome Territories; DL: Deep Learning; DLIm: Dynamic Live Imaging; ecn: egg nucleus; FACS: Fluorescence-Activated Cell Sorting; FISH: Fluorescent In Situ Hybridization; FP: Fluorescent Proteins (GFP, RFP, CFP, YFP, mCherry); FRAP: Fluorescence Recovery After Photobleaching; GPU: Graphics Processing Unit; KEEs: KNOT Engaged Elements; INTACT: Isolation of Nuclei TAgged in specific Cell Types; LADs: Lamin-Associated Domains; ML: Machine Learning; NA: Numerical Aperture; NADs: Nucleolar Associated Domains; PALM: Photo-Activated Localization Microscopy; Pixel: Picture element; pn: polar nuclei; PSF: Point Spread Function; RHF: Relative Heterochromatin Fraction; SIM: Structured Illumination Microscopy; SLIm: Static Live Imaging; SMC: Spore Mother Cell; SNR: Signal to Noise Ratio; SRM: Super-Resolution Microscopy; STED: STimulated Emission Depletion; STORM: STochastic Optical Reconstruction Microscopy; syn: synergid nuclei; TADs: Topologically Associating Domains; Voxel: Volumetric pixel

This special issue comprises papers based in a meeting of the Nuclear Dynamics Special Interest Group of the Society for Experimental Biology. The session was entitled ‘Functional organisation of t...