Characterization of Nanoscale Organization of F-Actin in Morphologically Distinct Dendritic Spines In Vitro Using Supervised Learning

Information encoded in volumes and areas of dendritic spines is nearly maximal across mammalian brains

Many experiments suggest that long-term information associated with neuronal memory resides collectively in dendritic spines. However, spines can have a limited size due to metabolic and neuroanatomical constraints, which should effectively limit the amount of encoded information in excitatory synapses. This study investigates how much information can be stored in the population of sizes of dendritic spines, and whether it is optimal in any sense. It is shown here, using empirical data for several mammalian brains across different regions and physiological conditions, that dendritic spines nearly maximize entropy contained in their volumes and surface areas for a given mean size in cortical and hippocampal regions. Although both short- and heavy-tailed fitting distributions approach [Formula: see text] of maximal entropy in the majority of cases, the best maximization is obtained primarily for short-tailed gamma distribution. We find that most empirical ratios of standard deviation to mean for spine volumes and areas are in the range [Formula: see text], which is close to the theoretical optimal ratios coming from entropy maximization for gamma and lognormal distributions. On average, the highest entropy is contained in spine length ([Formula: see text] bits per spine), and the lowest in spine volume and area ([Formula: see text] bits), although the latter two are closer to optimality. In contrast, we find that entropy density (entropy per spine size) is always suboptimal. Our results suggest that spine sizes are almost as random as possible given the constraint on their size, and moreover the general principle of entropy maximization is applicable and potentially useful to information and memory storing in the population of cortical and hippocampal excitatory synapses, and to predicting their morphological properties.

Automatic extraction of actin networks in plants

The actin cytoskeleton is essential in eukaryotes, not least in the plant kingdom where it plays key roles in cell expansion, cell division, environmental responses and pathogen defence. Yet, the precise structure-function relationships of properties of the actin network in plants are still to be unravelled, including details of how the network configuration depends upon cell type, tissue type and developmental stage. Part of the problem lies in the difficulty of extracting high-quality, quantitative measures of actin network features from microscopy data. To address this problem, we have developed DRAGoN, a novel image analysis algorithm that can automatically extract the actin network across a range of cell types, providing seventeen different quantitative measures that describe the network at a local level. Using this algorithm, we then studied a number of cases in Arabidopsis thaliana, including several different tissues, a variety of actin-affected mutants, and cells responding to powdery mildew. In many cases we found statistically-significant differences in actin network properties. In addition to these results, our algorithm is designed to be easily adaptable to other tissues, mutants and plants, and so will be a valuable asset for the study and future biological engineering of the actin cytoskeleton in globally-important crops.

Grading the Severity of Damage to the Perijunctional Actomyosin Ring and Zonula Occludens-1 of the Corneal Endothelium by Ensemble Learning Methods

Purpose: In many epithelia, including the corneal endothelium, intracellular/extracellular stresses break down the perijunctional actomyosin ring (PAMR) and zonula occludens-1 (ZO-1) at the apical junctions. This study aims to grade the severity of damage to PAMR and ZO-1 through machine learning. Methods: Immunocytochemical images of PAMR and ZO-1 were drawn from recent studies on the corneal endothelium subjected to hypothermia and oxidative stress. The images were analyzed for their morphological (e.g., Hu moments) and textural features (based on gray-level co-occurrence matrix [GLCM] and Gabor filters). The extracted features were ranked by SHapley analysis and analysis of variance. Then top features were used to grade the severity of damage using a suite of ensemble classifiers, including random forest, bagging classifier (BC), AdaBoost, extreme gradient boosting, and stacking classifier. Results: A partial set of features from GLCM, along with Hu moments and the number of hexagons, enabled the classification of damage to PAMR into Control, Mild, Moderate, and Severe with the area under the receiver operating characteristics curve (AUC) = 0.92 and F1 score = 0.77 with BC. In contrast, a bank of Gabor filters provided a partial set of features that could be combined with Hu moments, branch length, and sharpness for the classification of ZO-1 images into four levels with AUC = 0.95 and F1 score of 0.8 with BC. Conclusions: We have developed a workflow that enables the stratification of damage to PAMR and ZO-1. The approach can be applied to similar data during drug discovery or pathophysiological studies of epithelia.

Nanoscale regulation of Ca2+ dependent phase transitions and real-time dynamics of SAP97/hDLG

Synapse associated protein-97/Human Disk Large (SAP97/hDLG) is a conserved, alternatively spliced, modular, scaffolding protein critical in regulating the molecular organization of cell-cell junctions in vertebrates. We confirm that the molecular determinants of first order phase transition of SAP97/hDLG is controlled by morpho-functional changes in its nanoscale organization. Furthermore, the nanoscale molecular signatures of these signalling islands and phase transitions are altered in response to changes in cytosolic Ca²⁺. Additionally, exchange kinetics of alternatively spliced isoforms of the intrinsically disordered region in SAP97/hDLG C-terminus shows differential sensitivities to Ca²⁺ bound Calmodulin, affirming that the molecular signatures of local phase transitions of SAP97/hDLG depends on their nanoscale heterogeneity and compositionality of isoforms.

SAP97/hDLG is a ubiquitous, alternatively spliced, and conserved modular scaffolding protein involved in the organization cell junctions and excitatory synapses. Here, authors confirm that SAP97/hDLG condenses in to nanosized molecular domains in both heterologous cells and hippocampal pyramidal neurons. Authors demonstrate that in vivo and in vitro condensation, molecular signatures of nanoscale condensates and exchange kinetics of SAP97/hDLG is modulated by the local availability of alternatively spliced isoforms. Additionally, SAP97/hDLG isoforms exhibits a differential sensitivity to Ca²⁺ bound Calmodulin, resulting in altered properties of nanocondensates and their real-time regulation

A Ribosomal Perspective on Neuronal Local Protein Synthesis

Continued mRNA translation and protein production are critical for various neuronal functions. In addition to the precise sorting of proteins from cell soma to distant locations, protein synthesis allows a dynamic remodeling of the local proteome in a spatially variable manner. This spatial heterogeneity of protein synthesis is shaped by several factors such as injury, guidance cues, developmental cues, neuromodulators, and synaptic activity. In matured neurons, thousands of synapses are non-uniformly distributed throughout the dendritic arbor. At any given moment, the activity of individual synapses varies over a wide range, giving rise to the variability in protein synthesis. While past studies have primarily focused on the translation factors or the identity of translated mRNAs to explain the source of this variation, the role of ribosomes in this regard continues to remain unclear. Here, we discuss how several stochastic mechanisms modulate ribosomal functions, contributing to the variability in neuronal protein expression. Also, we point out several underexplored factors such as local ion concentration, availability of tRNA or ATP during translation, and molecular composition and organization of a compartment that can influence protein synthesis and its variability in neurons.

A Novel Estimation Method for the Counting of Dendritic Spines

BACKGROUND

Since Cajal's visualisations of the synaptic spine, this feature of the neuron has been of interest to neuroscientists and has been investigated usually in reference to degeneration or proliferation of dendrites and their neurons. Synaptic spine measurement often forms a critical element of any study investigating neuronal morphology. However, the way researchers have counted spines hasn't changed for almost a century. Some of the currently used legacy methods fail to accommodate obscured pisnes or factor-in visibility differences between histological stains.

NEW METHOD

Here we investigate the neuronal dendrite and its synaptic spines, and reveal that using confocal or bright-field technologies may in fact obfuscate spine counts. A mathematical model is developed for the distribution of synaptic spines within the rat, that should, by nature of the formula and the impartiality of probability quotients, be applied to estimate the number of synaptic spines across any length of dendrite that has protrusions within any species.

RESULTS

Using this estimation method, we show that, depending on the method of image capture, there are in fact more spines present than typically counted on lengths of dendrite, something that may have biased morphological studies in the past.

COMPARISON WITH EXISTING METHODS

This new estimation method has been collapsed down into an easy-to-use free website. With input of only four fields, we provide the researcher with a more accurate estimation of the amount of spines on a length of dendrite. This was made possible by fluorescing a Golgi stain and comparing two-photon, bright-field and confocal images.

CONCLUSIONS

An easy web-based resource has been made available to use this new method for spine calculation. Using this method improves the validity of spine measurement and provides a means to review previously published work.

Digging Deeper: Advancements in Visualization of Inhibitory Synapses in Neurodegenerative Disorders

Recent research has provided strong evidence that neurodegeneration may develop from an imbalance between synaptic structural components in the brain. Lately, inhibitory synapses communicating via the neurotransmitters GABA or glycine have come to the center of attention. Increasing evidence suggests that imbalance in the structural composition of inhibitory synapses affect deeply the ability of neurons to communicate effectively over synaptic connections. Progressive failure of synaptic plasticity and memory are thus hallmarks of neurodegenerative diseases. In order to prove that structural changes at synapses contribute to neurodegeneration, we need to visualize single-molecule interactions at synaptic sites in an exact spatial and time frame. This visualization has been restricted in terms of spatial and temporal resolution. New developments in electron microscopy and super-resolution microscopy have improved spatial and time resolution tremendously, opening up numerous possibilities. Here we critically review current and recently developed methods for high-resolution visualization of inhibitory synapses in the context of neurodegenerative diseases. We present advantages, strengths, weaknesses, and current limitations for selected methods in research, as well as present a future perspective. A range of new options has become available that will soon help understand the involvement of inhibitory synapses in neurodegenerative disorders.

Super-resolution microscopy: a closer look at synaptic dysfunction in Alzheimer disease

The synapse has emerged as a critical neuronal structure in the degenerative process of Alzheimer disease (AD), in which the pathogenic signals of two key players - amyloid-β (Aβ) and tau - converge, thereby causing synaptic dysfunction and cognitive deficits. The synapse presents a dynamic, confined microenvironment in which to explore how key molecules travel, localize, interact and assume different levels of organizational complexity, thereby affecting neuronal function. However, owing to their small size and the diffraction-limited resolution of conventional light microscopic approaches, investigating synaptic structure and dynamics has been challenging. Super-resolution microscopy (SRM) techniques have overcome the resolution barrier and are revolutionizing our quantitative understanding of biological systems in unprecedented spatio-temporal detail. Here we review critical new insights provided by SRM into the molecular architecture and dynamic organization of the synapse and, in particular, the interactions between Aβ and tau in this compartment. We further highlight how SRM can transform our understanding of the molecular pathological mechanisms that underlie AD. The application of SRM for understanding the roles of synapses in AD pathology will provide a stepping stone towards a broader understanding of dysfunction in other subcellular compartments and at cellular and circuit levels in this disease.

Super-resolving Microscopy in Neuroscience

Fluorescence imaging techniques play a pivotal role in our understanding of the nervous system. The emergence of various super-resolution microscopy methods and specialized fluorescent probes enables direct insight into neuronal structure and protein arrangements in cellular subcompartments with so far unmatched resolution. Super-resolving visualization techniques in neurons unveil a novel understanding of cytoskeletal composition, distribution, motility, and signaling of membrane proteins, subsynaptic structure and function, and neuron-glia interaction. Well-defined molecular targets in autoimmune and neurodegenerative disease models provide excellent starting points for in-depth investigation of disease pathophysiology using novel and innovative imaging methodology. Application of super-resolution microscopy in human brain samples and for testing clinical biomarkers is still in its infancy but opens new opportunities for translational research in neurology and neuroscience. In this review, we describe how super-resolving microscopy has improved our understanding of neuronal and brain function and dysfunction in the last two decades.

Cryptic association of B7-2 molecules and its implication for clustering

T-cell co-stimulation through CD28/CTLA4:B7-1/B7-2 axis is one of the extensively studied pathways that resulted in the discovery of several FDA-approved drugs for autoimmunity and cancer. However, many aspects of the signaling mechanism remain elusive, including oligomeric association and clustering of B7-2 on the cell surface. Here, we describe the structure of the IgV domain of B7-2 and its cryptic association into 1D arrays that appear to represent the pre-signaling state of B7-2 on the cell membrane. Super-resolution microscopy experiments on heterologous cells expressing B7-2 and B7-1 suggest, B7-2 form relatively elongated and larger clusters compared to B7-1. The sequence and structural comparison of other B7 family members, B7-1:CTLA4 and B7-2:CTLA-4 complex structures, support our view that the observed B7-2 1D zipper array is physiologically important. This observed 1D zipper-like array also provides an explanation for its clustering, and upright orientation on the cell surface, and avoidance of spurious signaling.

Mechanical Principles Governing the Shapes of Dendritic Spines

Dendritic spines are small, bulbous protrusions along the dendrites of neurons and are sites of excitatory postsynaptic activity. The morphology of spines has been implicated in their function in synaptic plasticity and their shapes have been well-characterized, but the potential mechanics underlying their shape development and maintenance have not yet been fully understood. In this work, we explore the mechanical principles that could underlie specific shapes using a minimal biophysical model of membrane-actin interactions. Using this model, we first identify the possible force regimes that give rise to the classic spine shapes-stubby, filopodia, thin, and mushroom-shaped spines. We also use this model to investigate how the spine neck might be stabilized using periodic rings of actin or associated proteins. Finally, we use this model to predict that the cooperation between force generation and ring structures can regulate the energy landscape of spine shapes across a wide range of tensions. Thus, our study provides insights into how mechanical aspects of actin-mediated force generation and tension can play critical roles in spine shape maintenance.

Automated and semi-automated enhancement, segmentation and tracing of cytoskeletal networks in microscopic images: A review

Cytoskeletal filaments are structures of utmost importance to biological cells and organisms due to their versatility and the significant functions they perform. These biopolymers are most often organised into network-like scaffolds with a complex morphology. Understanding the geometrical and topological organisation of these networks provides key insights into their functional roles. However, this non-trivial task requires a combination of high-resolution microscopy and sophisticated image processing/analysis software. The correct analysis of the network structure and connectivity needs precise segmentation of microscopic images. While segmentation of filament-like objects is a well-studied concept in biomedical imaging, where tracing of neurons and blood vessels is routine, there are comparatively fewer studies focusing on the segmentation of cytoskeletal filaments and networks from microscopic images. The developments in the fields of microscopy, computer vision and deep learning, however, began to facilitate the task, as reflected by an increase in the recent literature on the topic. Here, we aim to provide a short summary of the research on the (semi-)automated enhancement, segmentation and tracing methods that are particularly designed and developed for microscopic images of cytoskeletal networks. In addition to providing an overview of the conventional methods, we cover the recently introduced, deep-learning-assisted methods alongside the advantages they offer over classical methods.

Real-time nanoscale organization of amyloid precursor protein

Despite an intuitive understanding of the role of APP in health and disease, there exist few attempts to dissect its molecular localization at excitatory synapses. Though the biochemistry involved in the enzymatic processing of APP is well understood, there is a void in understanding the nonuniformity of the product formation in vivo. Here, we employed multiple paradigms of single molecules and ensemble based nanoscopic imaging to reveal that APP molecules are organized into regulatory nanodomains that are differentially compartmentalized in the functional zones of an excitatory synapse. Furthermore, with the aid of high density single particle tracking, we show that the lateral diffusion of APP in live cells dictates an equilibrium between these nanodomains and their nano-environment, which is affected in a detrimental variant of APP. Additionally, we incorporate this spatio-temporal detail 'in silico' to generate a realistic nanoscale topography of APP in dendrites and synapses. This approach uncovers a nanoscale heterogeneity in the molecular organization of APP, depicting a locus for differential APP processing. This holistic paradigm, to decipher the real-time heterogeneity of the substrate molecules on the nanoscale, could enable us to better evaluate the molecular constraints overcoming the ensemble approaches used traditionally to understand the kinetics of product formation.

Glutaredoxin1 Diminishes Amyloid Beta-Mediated Oxidation of F-Actin and Reverses Cognitive Deficits in an Alzheimer's Disease Mouse Model

Aims: Reactive oxygen species (ROS) generated during Alzheimer's disease (AD) pathogenesis through multiple sources are implicated in synaptic pathology observed in the disease. We have previously shown F-actin disassembly in dendritic spines in early AD (34). The actin cytoskeleton can be oxidatively modified resulting in altered F-actin dynamics. Therefore, we investigated whether disruption of redox signaling could contribute to actin network disassembly and downstream effects in the amyloid precursor protein/presenilin-1 double transgenic (APP/PS1) mouse model of AD. Results: Synaptosomal preparations from 1-month-old APP/PS1 mice showed an increase in ROS levels, coupled with a decrease in the reduced form of F-actin and increase in glutathionylated synaptosomal actin. Furthermore, synaptic glutaredoxin 1 (Grx1) and thioredoxin levels were found to be lowered. Overexpressing Grx1 in the brains of these mice not only reversed F-actin loss seen in APP/PS1 mice but also restored memory recall after contextual fear conditioning. F-actin levels and F-actin nanoarchitecture in spines were also stabilized by Grx1 overexpression in APP/PS1 primary cortical neurons, indicating that glutathionylation of F-actin is a critical event in early pathogenesis of AD, which leads to spine loss. Innovation: Loss of thiol/disulfide oxidoreductases in the synapse along with increase in ROS can render F-actin nanoarchitecture susceptible to oxidative modifications in AD. Conclusions: Our findings provide novel evidence that altered redox signaling in the form of S-glutathionylation and reduced Grx1 levels can lead to synaptic dysfunction during AD pathogenesis by directly disrupting the F-actin nanoarchitecture in spines. Increasing Grx1 levels is a potential target for novel disease-modifying therapies for AD.