DeepContact: High-throughput quantification of membrane contact sites based on electron microscopy imaging

Mitochondrial Mutation Leads to Cardiomyocyte Hypertrophy by Disruption of Mitochondria-Associated ER Membrane

m.3243A>G is the most common pathogenic mtDNA mutation. High energy-demanding organs, such as heart, are usually involved in mitochondria diseases. However, whether and how m.3243A>G affects cardiomyocytes remain unknown. We have established patient-specific iPSCs carrying m.3243A>G and induced cardiac differentiation. Cardiomyocytes with high m.3243A>G burden exhibited hypertrophic phenotype. This point mutation is localised in MT-TL1 encoding tRNALeu (UUR). m.3243A>G altered tRNALeu (UUR) conformation and decreased its stability. mtDNA is essential for mitochondrial function. Mitochondria dysfunction occurred and tended to become round. Its interaction with ER, mitochondria-associated ER membrane (MAM), was disrupted with decreased contact number and length. MAM is a central hub for calcium trafficking. Disrupted MAM disturbed calcium homeostasis, which may be the direct and leading cause of cardiomyocyte hypertrophy, as MAM enforcement reversed this pathological state. Considering the threshold effect of mitochondrial disease, mito-TALENs were introduced to eliminate mutant mitochondria and release mutation load. Mutation reduction partially reversed the cellular behaviour and made it approach to that of control one. These findings reveal the pathogenesis underlying m.3243A>G from perspective of organelle interaction, rather than organelle. Beyond mitochondria quality control, its proper interaction with other organelles, such as ER, matters for mitochondria disease. This study may provide inspiration for mitochondria disease intervention.

Machine Learning Techniques to Infer Protein Structure and Function from Sequences: A Comprehensive Review

The elucidation of protein structure and function plays a pivotal role in understanding biological processes and facilitating drug discovery. With the exponential growth of protein sequence data, machine learning techniques have emerged as powerful tools for predicting protein characteristics from sequences alone. This review provides a comprehensive overview of the importance and application of machine learning in inferring protein structure and function. We discuss various machine learning approaches, primarily focusing on convolutional neural networks and natural language processing, and their utilization in predicting protein secondary and tertiary structures, residue-residue contacts, protein function, and subcellular localization. Furthermore, we highlight the challenges associated with using machine learning techniques in this context, such as the availability of high-quality training datasets and the interpretability of models. We also delve into the latest progress in the field concerning the advancements made in the development of intricate deep learning architectures. Overall, this review underscores the significance of machine learning in advancing our understanding of protein structure and function, and its potential to revolutionize drug discovery and personalized medicine.

Hexokinase 2 senses fructose in tumor-associated macrophages to promote colorectal cancer growth

Fructose is associated with colorectal cancer tumorigenesis and metastasis through ketohexokinase-mediated metabolism in the colorectal epithelium, yet its role in the tumor immune microenvironment remains largely unknown. Here, we show that a modest amount of fructose, without affecting obesity and associated complications, promotes colorectal cancer tumorigenesis and growth by suppressing the polarization of M1-like macrophages. Fructose inhibits M1-like macrophage polarization independently of fructose-mediated metabolism. Instead, it serves as a signal molecule to promote the interaction between hexokinase 2 and inositol 1,4,5-trisphophate receptor type 3, the predominant Ca2+ channel on the endoplasmic reticulum. The interaction reduces Ca2+ levels in cytosol and mitochondria, thereby suppressing the activation of mitogen-activated protein kinase (MAPK) and signal transducer and activator of transcription 1 (STAT1) as well as NOD-, LRR- and pyrin domain-containing protein 3 (NLRP3) inflammasome activation. Consequently, this impedes M1-like macrophage polarization. Our study highlights the critical role of fructose as a signaling molecule that impairs the polarization of M1-like macrophages for tumor growth.

Imaging and proteomics toolkits for studying organelle contact sites

Organelle contact sites are regions where two heterologous membranes are juxtaposed by molecular tethering complexes. These contact sites are important in inter-organelle communication and cellular functional integration. However, visualizing these minute foci and identifying contact site proteomes have been challenging. In recent years, fluorescence-based methods have been developed to visualize the dynamic physical interaction of organelles while proximity labeling approaches facilitate the profiling of proteomes at contact sites. In this review, we explain the design principle for these contact site reporters: a dual-organelle interaction mechanism based on how endogenous tethers and/or tethering complexes localize to contact sites. We classify the contact site reporters into three categories: (i) single-protein systems, (ii) two-component systems with activated reporter signal upon organelle proximity, and (iii) reporters for contact site proteomes. We also highlight advanced imaging analysis with high temporal-spatial resolution and the use of machine-learning algorithms for detecting contact sites.

Three-Dimensional Label-Free Observing of the Self-Assembled Nanoparticles inside a Single Cell at Nanoscale Resolution

Understanding the intracellular behavior of nanoparticles (NPs) plays a key role in optimizing the self-assembly performance of nanomedicine. However, conducting the 3D, label-free, quantitative observation of self-assembled NPs within intact single cells remains a substantial challenge in complicated intracellular environments. Here, we propose a deep learning combined synchrotron radiation hard X-ray nanotomography approach to visualize the self-assembled ultrasmall iron oxide (USIO) NPs in a single cell. The method allows us to explore comprehensive information on NPs, such as their distribution, morphology, location, and interaction with cell organelles, and provides quantitative analysis of the heterogeneous size and morphologies of USIO NPs under diverse conditions. This label-free, in situ method provides a tool for precise characterization of intracellular self-assembled NPs to improve the evaluation and design of a bioresponsive nanomedicine.

Stay in touch with the endoplasmic reticulum

The endoplasmic reticulum (ER), which is composed of a continuous network of tubules and sheets, forms the most widely distributed membrane system in eukaryotic cells. As a result, it engages a variety of organelles by establishing membrane contact sites (MCSs). These contacts regulate organelle positioning and remodeling, including fusion and fission, facilitate precise lipid exchange, and couple vital signaling events. Here, we systematically review recent advances and converging themes on ER-involved organellar contact. The molecular basis, cellular influence, and potential physiological functions for ER/nuclear envelope contacts with mitochondria, Golgi, endosomes, lysosomes, lipid droplets, autophagosomes, and plasma membrane are summarized.

Physical Contacts Between Mitochondria and WPBs Participate in WPB Maturation

BACKGROUND

Weibel-Palade bodies (WPBs) are endothelial cell-specific cigar-shaped secretory organelles containing various biologically active molecules. WPBs play crucial roles in thrombosis, hemostasis, angiogenesis, and inflammation. The main content of WPBs is the procoagulant protein vWF (von Willebrand factor). Physical contacts and functional cross talk between mitochondria and other organelles have been demonstrated. Whether an interorganellar connection exists between mitochondria and WPBs is unknown.

METHODS

We observed physical contacts between mitochondria and WPBs in human umbilical vein endothelial cells by electron microscopy and living cell confocal microscopy. We developed an artificial intelligence-assisted method to quantify the duration and length of organelle contact sites in live cells.

RESULTS

We found there existed physical contacts between mitochondria and WPBs. Disruption of mitochondrial function affected the morphology of WPBs. Furthermore, we found that Rab3b, a small GTPase on the WPBs, was enriched at the mitochondrion-WPB contact sites. Rab3b deficiency reduced interaction between the two organelles and impaired the maturation of WPBs and vWF multimer secretion.

CONCLUSIONS

Our results reveal that Rab3b plays a crucial role in mediating the mitochondrion-WPB contacts, and that mitochondrion-WPB coupling is critical for the maturation of WPBs in vascular endothelial cells.

Using DeepContact with Amira graphical user interface

DeepContact is a deep learning software for high-throughput quantification of membrane contact site (MCS) in 2D electron microscopy images. This protocol will guide users through incorporating available DeepContact models in Amira's artificial intelligence module, thereby allowing invoking of DeepContact functions in organelle segmentation and quantifying of MCS with a user-friendly graphical user interface of Amira software. For complete details on the use and execution of this protocol, please refer to Liu et al. (2022).1.

Giant mitochondria in human liver disease

This thematic review aims to provide an overview of the current state of knowledge about the occurrence of giant mitochondria or megamitochondria in liver parenchymal cells. Their presence and accumulation are considered to be a major pathological hallmark of the health and fate of liver parenchymal cells that leads to overall tissue deterioration and eventually results in organ failure. The first description on giant mitochondria dates back to the 1960s, coinciding with the availability of the first generation of electron microscopes in clinical diagnostic laboratories. Detailed accounts on their ultrastructure have mostly been described in patients suffering from alcoholic liver disease, chronic hepatitis, hepatocellular carcinoma and non-alcoholic fatty liver disease. Interestingly, from this extensive literature survey, it became apparent that giant mitochondria or megamitochondria present themselves with or without highly organised crystal-like intramitochondrial inclusions. The origin, formation and potential role of giant mitochondria remain to-date largely unanswered. Likewise, the biochemical composition of the well-organised crystal-like inclusions and their possible impact on mitochondrial function is unclear. Herein, concepts about the possible mechanism of their formation and three-dimensional architecture will be approached. We will furthermore discuss their importance in diagnostics, including future research outlooks and potential therapeutic interventions to cure liver disease where giant mitochondria are implemented.

Subcellular visualization: Organelle-specific targeted drug delivery and discovery

Organelles perform critical biological functions due to their distinct molecular composition and internal environment. Disorders in organelles or their interacting networks have been linked to the incidence of numerous diseases, and the research of pharmacological actions at the organelle level has sparked pharmacists' interest. Currently, cell imaging has evolved into a critical tool for drug delivery, drug discovery, and pharmacological research. The introduction of advanced imaging techniques in recent years has provided researchers with richer biological information for viewing and studying the ultrastructure of organelles, protein interactions, and gene transcription activities, leading to the design and delivery of precision-targeted drugs. Therefore, this reviews the research on organelles-targeted drugs based upon imaging technologies and development of fluorescent molecules for medicinal purposes. We also give a thorough analysis of a number of subcellular-level elements of drug development, including subcellular research instruments and methods, organelle biological event investigation, subcellular target and drug identification, and design of subcellular delivery systems. This review will make it possible to promote drug research from the individual/cellular level to the subcellular level, as well as give a new focus based on newly found organelle activities.

Endoplasmic reticulum membrane contact sites: cross-talk between membrane-bound organelles in plant cells

Eukaryotic cells contain organelles surrounded by monolayer or bilayer membranes. Organelles take part in highly dynamic and organized interactions at membrane contact sites, which play vital roles during development and response to stress. The endoplasmic reticulum extends throughout the cell and acts as an architectural scaffold to maintain the spatial distribution of other membrane-bound organelles. In this review, we highlight the structural organization, dynamics, and physiological functions of membrane contact sites between the endoplasmic reticulum and various membrane-bound organelles, especially recent advances in plants. We briefly introduce how the combined use of dynamic and static imaging techniques can enable monitoring of the cross-talk between organelles via membrane contact sites. Finally, we discuss future directions for research fields related to membrane contact.

Instance segmentation of mitochondria in electron microscopy images with a generalist deep learning model trained on a diverse dataset

Mitochondria are extremely pleomorphic organelles. Automatically annotating each one accurately and precisely in any 2D or volume electron microscopy (EM) image is an unsolved computational challenge. Current deep learning-based approaches train models on images that provide limited cellular contexts, precluding generality. To address this, we amassed a highly heterogeneous ∼1.5 × 10⁶ image 2D unlabeled cellular EM dataset and segmented ∼135,000 mitochondrial instances therein. MitoNet, a model trained on these resources, performs well on challenging benchmarks and on previously unseen volume EM datasets containing tens of thousands of mitochondria. We release a Python package and napari plugin, empanada, to rapidly run inference, visualize, and proofread instance segmentations. A record of this paper's transparent peer review process is included in the supplemental information.