Automated Detection and Localization of Synaptic Vesicles in Electron Microscopy Images

CryoVesNet: A dedicated framework for synaptic vesicle segmentation in cryo-electron tomograms

Cryo-electron tomography (cryo-ET) has the potential to reveal cell structure down to atomic resolution. Nevertheless, cellular cryo-ET data is highly complex, requiring image segmentation for visualization and quantification of subcellular structures. Due to noise and anisotropic resolution in cryo-ET data, automatic segmentation based on classical computer vision approaches usually does not perform satisfactorily. Communication between neurons relies on neurotransmitter-filled synaptic vesicle (SV) exocytosis. Cryo-ET study of the spatial organization of SVs and their interconnections allows a better understanding of the mechanisms of exocytosis regulation. Accurate SV segmentation is a prerequisite to obtaining a faithful connectivity representation. Hundreds of SVs are present in a synapse, and their manual segmentation is a bottleneck. We addressed this by designing a workflow consisting of a convolutional network followed by post-processing steps. Alongside, we provide an interactive tool for accurately segmenting spherical vesicles. Our pipeline can in principle segment spherical vesicles in any cell type as well as extracellular and in vitro spherical vesicles.

The synaptic vesicle cluster as a controller of pre- and postsynaptic structure and function

The synaptic vesicle cluster (SVC) is an essential component of chemical synapses, which provides neurotransmitter-loaded vesicles during synaptic activity, at the same time as also controlling the local concentrations of numerous exo- and endocytosis cofactors. In addition, the SVC hosts molecules that participate in other aspects of synaptic function, from cytoskeletal components to adhesion proteins, and affects the location and function of organelles such as mitochondria and the endoplasmic reticulum. We argue here that these features extend the functional involvement of the SVC in synapse formation, signalling and plasticity, as well as synapse stabilization and metabolism. We also propose that changes in the size of the SVC coalesce with changes in the postsynaptic compartment, supporting the interplay between pre- and postsynaptic dynamics. Thereby, the SVC could be seen as an 'all-in-one' regulator of synaptic structure and function, which should be investigated in more detail, to reveal molecular mechanisms that control synaptic function and heterogeneity.

Excitatory synaptic structural abnormalities produced by templated aggregation of α-syn in the basolateral amygdala

Parkinson's disease (PD) and Dementia with Lewy bodies (DLB) are characterized by neuronal α-synuclein (α-syn) inclusions termed Lewy Pathology, which are abundant in the amygdala. The basolateral amygdala (BLA), in particular, receives projections from the thalamus and cortex. These projections play a role in cognition and emotional processing, behaviors which are impaired in α-synucleinopathies. To understand if and how pathologic α-syn impacts the BLA requires animal models of α-syn aggregation. Injection of α-syn pre-formed fibrils (PFFs) into the striatum induces robust α-syn aggregation in excitatory neurons in the BLA that corresponds with reduced contextual fear conditioning. At early time points after aggregate formation, cortico-amygdala excitatory transmission is abolished. The goal of this project was to determine if α-syn inclusions in the BLA induce synaptic degeneration and/or morphological changes. In this study, we used C57BL/6 J mice injected bilaterally with PFFs in the dorsal striatum to induce α-syn aggregate formation in the BLA. A method was developed using immunofluorescence and three-dimensional reconstruction to analyze excitatory cortico-amygdala and thalamo-amygdala presynaptic terminals closely juxtaposed to postsynaptic densities. The abundance and morphology of synapses were analyzed at 6- or 12-weeks post-injection of PFFs. α-Syn aggregate formation in the BLA did not cause a significant loss of synapses, but cortico-amygdala and thalamo-amygdala presynaptic terminals and postsynaptic densities with aggregates of α-syn show increased volumes, similar to previous findings in human DLB cortex, and in non-human primate models of PD. Transmission electron microscopy showed that asymmetric synapses in mice with PFF-induced α-syn aggregates have reduced synaptic vesicle intervesicular distances, similar to a recent study showing phospho-serine-129 α-syn increases synaptic vesicle clustering. Thus, pathologic α-syn causes major alterations to synaptic architecture in the BLA, potentially contributing to behavioral impairment and amygdala dysfunction observed in synucleinopathies.

The Lack of Synapsin Alters Presynaptic Plasticity at Hippocampal Mossy Fibers in Male Mice

Synapsins are highly abundant presynaptic proteins that play a crucial role in neurotransmission and plasticity via the clustering of synaptic vesicles. The synapsin III isoform is usually downregulated after development, but in hippocampal mossy fiber boutons, it persists in adulthood. Mossy fiber boutons express presynaptic forms of short- and long-term plasticity, which are thought to underlie different forms of learning. Previous research on synapsins at this synapse focused on synapsin isoforms I and II. Thus, a complete picture regarding the role of synapsins in mossy fiber plasticity is still missing. Here, we investigated presynaptic plasticity at hippocampal mossy fiber boutons by combining electrophysiological field recordings and transmission electron microscopy in a mouse model lacking all synapsin isoforms. We found decreased short-term plasticity, i.e., decreased facilitation and post-tetanic potentiation, but increased long-term potentiation in male synapsin triple knock-out (KO) mice. At the ultrastructural level, we observed more dispersed vesicles and a higher density of active zones in mossy fiber boutons from KO animals. Our results indicate that all synapsin isoforms are required for fine regulation of short- and long-term presynaptic plasticity at the mossy fiber synapse.

Cortico-amygdala synaptic structural abnormalities produced by templated aggregation of α-synuclein

Parkinson's disease (PD) and Dementia with Lewy bodies (DLB) are characterized by neuronal α-synuclein (α-syn) inclusions termed Lewy Pathology, which are abundant in the amygdala. The basolateral amygdala (BLA), in particular, receives projections from the thalamus and cortex. These projections play a role in cognition and emotional processing, behaviors which are impaired in α-synucleinopathies. To understand if and how pathologic α-syn impacts the BLA requires animal models of α-syn aggregation. Injection of α-synuclein pre-formed fibrils (PFFs) into the striatum induces robust α-synuclein aggregation in excitatory neurons in the BLA that corresponds with reduced contextual fear conditioning. At early time points after aggregate formation, cortico-amygdala excitatory transmission is abolished. The goal of this project was to determine if α-syn inclusions in the BLA induce synaptic degeneration and/or morphological changes. In this study, we used C57BL/6J mice injected bilaterally with PFFs in the dorsal striatum to induce α-syn aggregate formation in the BLA. A method was developed using immunofluorescence and three-dimensional reconstruction to analyze excitatory cortico-amygdala and thalamo-amygdala presynaptic terminals closely juxtaposed to postsynaptic densities. The abundance and morphology of synapses were analyzed at 6- or 12-weeks post-injection of PFFs. α-Syn aggregate formation in the BLA did not cause a significant loss of synapses, but cortico-amygdala and thalamo-amygdala presynaptic terminals and postsynaptic densities with aggregates of α-synuclein show increased volumes, similar to previous findings in human DLB cortex, and in non-human primate models of PD. Transmission electron microscopy showed that PFF-injected mice showed reduced intervesicular distances similar to a recent study showing phospho-serine-129 α-synuclein increases synaptic vesicle clustering. Thus, pathologic α-synuclein causes major alterations to synaptic architecture in the BLA, potentially contributing to behavioral impairment and amygdala dysfunction observed in synucleinopathies.

Dual-channel neural network for instance segmentation of synapse

Detection and segmentation of neural synapses in electron microscopy images are the committed steps for analyzing neural ultrastructure. To date, manual annotation of the structure in synapses has been the primary method, which is time-consuming and restricts the throughput of data acquisition. Recent studies have utilized a series of deformations based on a segmentation model for the detection and segmentation of transmission electron microscope images. However, the analysis of synaptic segmentation and statistics still lacks sufficient automation and high-throughput. Therefore, we developed a dual-channel neural network instance segmentation model with weighted top-down and multi-scale bottom-up schemes, which aid in accurately detecting and segmenting synaptic vesicles and their active zones within presynaptic membranes in complex environments. In addition, we proposed a masked self-supervised pre-training model based on the latest convolutional architecture to improve performance in downstream segmentation tasks. By comparing our model to other state-of-the-art methods, we determined its viability and accuracy. The applicability of our model is thoroughly demonstrated by distinct application scenarios for neurobiological research. These findings indicate that the dual-channel neural network could facilitate the analysis of synaptic structures for the advancement of biomedical research and electron microscope reconstruction techniques.

The proteomic landscape of synaptic diversity across brain regions and cell types

Neurons build synaptic contacts using different protein combinations that define the specificity, function, and plasticity potential of synapses; however, the diversity of synaptic proteomes remains largely unexplored. We prepared synaptosomes from 7 different transgenic mouse lines with fluorescently labeled presynaptic terminals. Combining microdissection of 5 different brain regions with fluorescent-activated synaptosome sorting (FASS), we isolated and analyzed the proteomes of 18 different synapse types. We discovered ∼1,800 unique synapse-type-enriched proteins and allocated thousands of proteins to different types of synapses (https://syndive.org/). We identify shared synaptic protein modules and highlight the proteomic hotspots for synapse specialization. We reveal unique and common features of the striatal dopaminergic proteome and discover the proteome signatures that relate to the functional properties of different interneuron classes. This study provides a molecular systems-biology analysis of synapses and a framework to integrate proteomic information for synapse subtypes of interest with cellular or circuit-level experiments.

Fear memory-associated synaptic and mitochondrial changes revealed by deep learning-based processing of electron microscopy data

Serial section electron microscopy (ssEM) can provide comprehensive 3D ultrastructural information of the brain with exceptional computational cost. Targeted reconstruction of subcellular structures from ssEM datasets is less computationally demanding but still highly informative. We thus developed a region-CNN-based deep learning method to identify, segment, and reconstruct synapses and mitochondria to explore the structural plasticity of synapses and mitochondria in the auditory cortex of mice subjected to fear conditioning. Upon reconstructing over 135,000 mitochondria and 160,000 synapses, we find that fear conditioning significantly increases the number of mitochondria but decreases their size and promotes formation of multi-contact synapses, comprising a single axonal bouton and multiple postsynaptic sites from different dendrites. Modeling indicates that such multi-contact configuration increases the information storage capacity of new synapses by over 50%. With high accuracy and speed in reconstruction, our method yields structural and functional insight into cellular plasticity associated with fear learning.