Ultrastructural differences impact cilia shape and external exposure across cell classes in the visual cortex.

A primary cilium is a membrane-bound extension from the cell surface that contains receptors for perceiving and transmitting signals that modulate cell state and activity. Primary cilia in the brain are less accessible than cilia on cultured cells or epithelial tissues because in the brain they protrude into a deep, dense network of glial and neuronal processes. Here, we investigated cilia frequency, internal structure, shape, and position in large, high-resolution transmission electron microscopy volumes of mouse primary visual cortex. Cilia extended from the cell bodies of nearly all excitatory and inhibitory neurons, astrocytes, and oligodendrocyte precursor cells (OPCs) but were absent from oligodendrocytes and microglia. Ultrastructural comparisons revealed that the base of the cilium and the microtubule organization differed between neurons and glia. Investigating cilia-proximal features revealed that many cilia were directly adjacent to synapses, suggesting that cilia are poised to encounter locally released signaling molecules. Our analysis indicated that synapse proximity is likely due to random encounters in the neuropil, with no evidence that cilia modulate synapse activity as would be expected in tetrapartite synapses. The observed cell class differences in proximity to synapses were largely due to differences in external cilia length. Many key structural features that differed between neuronal and glial cilia influenced both cilium placement and shape and, thus, exposure to processes and synapses outside the cilium. Together, the ultrastructure both within and around neuronal and glial cilia suggest differences in cilia formation and function across cell types in the brain.

Nanometer-scale views of visual cortex reveal anatomical features of primary cilia poised to detect synaptic spillover.

A primary cilium is a thin membrane-bound extension off a cell surface that contains receptors for perceiving and transmitting signals that modulate cell state and activity. While many cell types have a primary cilium, little is known about primary cilia in the brain, where they are less accessible than cilia on cultured cells or epithelial tissues and protrude from cell bodies into a deep, dense network of glial and neuronal processes. Here, we investigated cilia frequency, internal structure, shape, and position in large, high-resolution transmission electron microscopy volumes of mouse primary visual cortex. Cilia extended from the cell bodies of nearly all excitatory and inhibitory neurons, astrocytes, and oligodendrocyte precursor cells (OPCs), but were absent from oligodendrocytes and microglia. Structural comparisons revealed that the membrane structure at the base of the cilium and the microtubule organization differed between neurons and glia. OPC cilia were distinct in that they were the shortest and contained pervasive internal vesicles only occasionally observed in neuron and astrocyte cilia. Investigating cilia-proximal features revealed that many cilia were directly adjacent to synapses, suggesting cilia are well poised to encounter locally released signaling molecules. Cilia proximity to synapses was random, not enriched, in the synapse-rich neuropil. The internal anatomy, including microtubule changes and centriole location, defined key structural features including cilium placement and shape. Together, the anatomical insights both within and around neuron and glia cilia provide new insights into cilia formation and function across cell types in the brain.

Multi-layered maps of neuropil with segmentation-guided contrastive learning.

Maps of the nervous system that identify individual cells along with their type, subcellular components and connectivity have the potential to elucidate fundamental organizational principles of neural circuits. Nanometer-resolution imaging of brain tissue provides the necessary raw data, but inferring cellular and subcellular annotation layers is challenging. We present segmentation-guided contrastive learning of representations (SegCLR), a self-supervised machine learning technique that produces representations of cells directly from 3D imagery and segmentations. When applied to volumes of human and mouse cortex, SegCLR enables accurate classification of cellular subcompartments and achieves performance equivalent to a supervised approach while requiring 400-fold fewer labeled examples. SegCLR also enables inference of cell types from fragments as small as 10 µm, which enhances the utility of volumes in which many neurites are truncated at boundaries. Finally, SegCLR enables exploration of layer 5 pyramidal cell subtypes and automated large-scale analysis of synaptic partners in mouse visual cortex.

CAVE: Connectome Annotation Versioning Engine.

Advances in Electron Microscopy, image segmentation and computational infrastructure have given rise to large-scale and richly annotated connectomic datasets which are increasingly shared across communities. To enable collaboration, users need to be able to concurrently create new annotations and correct errors in the automated segmentation by proofreading. In large datasets, every proofreading edit relabels cell identities of millions of voxels and thousands of annotations like synapses. For analysis, users require immediate and reproducible access to this constantly changing and expanding data landscape. Here, we present the Connectome Annotation Versioning Engine (CAVE), a computational infrastructure for immediate and reproducible connectome analysis in up-to petascale datasets (~1mm3) while proofreading and annotating is ongoing. For segmentation, CAVE provides a distributed proofreading infrastructure for continuous versioning of large reconstructions. Annotations in CAVE are defined by locations such that they can be quickly assigned to the underlying segment which enables fast analysis queries of CAVE's data for arbitrary time points. CAVE supports schematized, extensible annotations, so that researchers can readily design novel annotation types. CAVE is already used for many connectomics datasets, including the largest datasets available to date.

Emerging roles of oligodendrocyte precursor cells in neural circuit development and remodeling.

Oligodendrocyte precursor cells (OPCs) are non-neuronal brain cells that give rise to oligodendrocytes, glia that myelinate the axons of neurons in the brain. Classically known for their contributions to myelination via oligodendrogenesis, OPCs are increasingly appreciated to play diverse roles in the nervous system, ranging from blood vessel formation to antigen presentation. Here, we review emerging literature suggesting that OPCs may be essential for the establishment and remodeling of neural circuits in the developing and adult brain via mechanisms that are distinct from the production of oligodendrocytes. We discuss the specialized features of OPCs that position these cells to integrate activity-dependent and molecular cues to shape brain wiring. Finally, we place OPCs within the context of a growing field focused on understanding the importance of communication between neurons and glia in the contexts of both health and disease.

Oligodendrocyte precursor cells ingest axons in the mouse neocortex.

Neurons in the developing brain undergo extensive structural refinement as nascent circuits adopt their mature form. This physical transformation of neurons is facilitated by the engulfment and degradation of axonal branches and synapses by surrounding glial cells, including microglia and astrocytes. However, the small size of phagocytic organelles and the complex, highly ramified morphology of glia have made it difficult to define the contribution of these and other glial cell types to this crucial process. Here, we used large-scale, serial section transmission electron microscopy (TEM) with computational volume segmentation to reconstruct the complete 3D morphologies of distinct glial types in the mouse visual cortex, providing unprecedented resolution of their morphology and composition. Unexpectedly, we discovered that the fine processes of oligodendrocyte precursor cells (OPCs), a population of abundant, highly dynamic glial progenitors, frequently surrounded small branches of axons. Numerous phagosomes and phagolysosomes (PLs) containing fragments of axons and vesicular structures were present inside their processes, suggesting that OPCs engage in axon pruning. Single-nucleus RNA sequencing from the developing mouse cortex revealed that OPCs express key phagocytic genes at this stage, as well as neuronal transcripts, consistent with active axon engulfment. Although microglia are thought to be responsible for the majority of synaptic pruning and structural refinement, PLs were ten times more abundant in OPCs than in microglia at this stage, and these structures were markedly less abundant in newly generated oligodendrocytes, suggesting that OPCs contribute substantially to the refinement of neuronal circuits during cortical development.

Reconstruction of neocortex: Organelles, compartments, cells, circuits, and activity.

We assembled a semi-automated reconstruction of L2/3 mouse primary visual cortex from ∼250 × 140 × 90 µm3 of electron microscopic images, including pyramidal and non-pyramidal neurons, astrocytes, microglia, oligodendrocytes and precursors, pericytes, vasculature, nuclei, mitochondria, and synapses. Visual responses of a subset of pyramidal cells are included. The data are publicly available, along with tools for programmatic and three-dimensional interactive access. Brief vignettes illustrate the breadth of potential applications relating structure to function in cortical circuits and neuronal cell biology. Mitochondria and synapse organization are characterized as a function of path length from the soma. Pyramidal connectivity motif frequencies are predicted accurately using a configuration model of random graphs. Pyramidal cells receiving more connections from nearby cells exhibit stronger and more reliable visual responses. Sample code shows data access and analysis.

A petascale automated imaging pipeline for mapping neuronal circuits with high-throughput transmission electron microscopy.

Electron microscopy (EM) is widely used for studying cellular structure and network connectivity in the brain. We have built a parallel imaging pipeline using transmission electron microscopes that scales this technology, implements 24/7 continuous autonomous imaging, and enables the acquisition of petascale datasets. The suitability of this architecture for large-scale imaging was demonstrated by acquiring a volume of more than 1 mm3 of mouse neocortex, spanning four different visual areas at synaptic resolution, in less than 6 months. Over 26,500 ultrathin tissue sections from the same block were imaged, yielding a dataset of more than 2 petabytes. The combined burst acquisition rate of the pipeline is 3 Gpixel per sec and the net rate is 600 Mpixel per sec with six microscopes running in parallel. This work demonstrates the feasibility of acquiring EM datasets at the scale of cortical microcircuits in multiple brain regions and species.

Publisher Correction: Flexible Learning-Free Segmentation and Reconstruction of Neural Volumes.

A correction to this article has been published and is linked from the HTML and PDF versions of this paper. The error has been fixed in the paper.

Flexible Learning-Free Segmentation and Reconstruction of Neural Volumes.

Imaging is a dominant strategy for data collection in neuroscience, yielding stacks of images that often scale to gigabytes of data for a single experiment. Machine learning algorithms from computer vision can serve as a pair of virtual eyes that tirelessly processes these images, automatically detecting and identifying microstructures. Unlike learning methods, our Flexible Learning-free Reconstruction of Imaged Neural volumes (FLoRIN) pipeline exploits structure-specific contextual clues and requires no training. This approach generalizes across different modalities, including serially-sectioned scanning electron microscopy (sSEM) of genetically labeled and contrast enhanced processes, spectral confocal reflectance (SCoRe) microscopy, and high-energy synchrotron X-ray microtomography (µCT) of large tissue volumes. We deploy the FLoRIN pipeline on newly published and novel mouse datasets, demonstrating the high biological fidelity of the pipeline's reconstructions. FLoRIN reconstructions are of sufficient quality for preliminary biological study, for example examining the distribution and morphology of cells or extracting single axons from functional data. Compared to existing supervised learning methods, FLoRIN is one to two orders of magnitude faster and produces high-quality reconstructions that are tolerant to noise and artifacts, as is shown qualitatively and quantitatively.

Fast assembling of neuron fragments in serial 3D sections.

Reconstructing neurons from 3D image-stacks of serial sections of thick brain tissue is very time-consuming and often becomes a bottleneck in high-throughput brain mapping projects. We developed NeuronStitcher, a software suite for stitching non-overlapping neuron fragments reconstructed in serial 3D image sections. With its efficient algorithm and user-friendly interface, NeuronStitcher has been used successfully to reconstruct very large and complex human and mouse neurons.

Sparse reconstruction of brain circuits: or, how to survive without a microscopic connectome.

Inside one voxel of a cubic millimeter of neocortex, fifty to hundred thousand neurons use 4 km of axonal cable to form three to fifteen hundred million synapses with each other. While in the human, such voxel is a small fragment of a cortical area, in the mouse an entire cortical area, like the primary auditory cortex, can be contained in a voxel of this size. This raises the fundamental question of what happens inside such a voxel? Are the circuits contained in this voxel, and their operations, different in every area, or are there general principles that are conserved across cortical areas and species? Such questions go to the heart of understanding how the neocortex wires itself and works. One proposal is to answer these questions by mapping the entire circuit at synaptic resolution to produce a 'connectome' - of the cortical column, or even of the entire brain. However, such a high-resolution connectome is self-evidently unachievable with the tools available and as a strategy it still leaves us short of understanding the 'principles of neural engineering'. We offer an alternative route that uses physiology and computational modeling as a means of generating 'predictive anatomy', where the questions about underlying structure are directed to fundamental principles of organization and operation of the cortical circuits. This approach involves 'sparse' rather than 'dense' reconstructions at light and electron microscope resolution to keep the questions well-matched to current experimental tools. Rather than providing a snap-shot of an entire wiring diagram, our strategy provides for a statistical description of the circuit and integrates theory, function, and structure in a common framework.

Diversity of thalamorecipient spine morphology in cat visual cortex and its implication for synaptic plasticity.

A feature of spine synapses is the existence of a neck connecting the synapse on the spine head to the dendritic shaft. As with a cable, spine neck resistance (R(neck)) increases with increasing neck length and is inversely proportional to the cross-sectional area of the neck. A synaptic current entering a spine with a high R(neck) will lead to greater local depolarization in the spine head than would a similar input applied to a spine with a lower R(neck). This could make spines with high R(neck) more sensitive to plastic changes since voltage sensitive conductances, such as N-methyl-D-aspartic acid (NMDA) channels can be more easily activated. This hypothesis was tested using serial section electron microscopic reconstructions of thalamocortical spine synapses and spine necks located on spiny stellate cells and corticothalamic cells from area 17 of cats. Thalamic axons and corticothalamic neurons were labeled by injections of the tracer biotinylated dextran amine (BDA) in the dorsal lateral geniculate nucleus (dLGN) of anesthetized cats and spiny stellates were filled intracellularly in vivo with horseradish peroxidase. Twenty-eight labeled spines that formed synapses with dLGN boutons were collected from three spiny stellate and four corticothalamic cells and reconstructed in 3D from serial electron micrographs. Spine length, spine diameter, and the area of the postsynaptic density were measured from the 3D reconstructions and R(neck) of the spine was estimated. No correlation was found between the postsynaptic density size and the estimated spine R(neck). This suggests that forms of plasticity that lead to larger synapses are independent of spine neck resistance.

How thalamus connects to spiny stellate cells in the cat's visual cortex.

In the cat's visual cortex, the responses of simple cells seem to be totally determined by their thalamic input, yet only a few percent of the excitatory synapses in layer 4 arise from the thalamus. To resolve this discrepancy between structure and function, we used correlated light and electron microscopy to search individual spiny stellate cells (simple cells) for possible structural features that would explain the biophysical efficacy of the thalamic input, such as synaptic location on dendrites, size of postsynaptic densities, and postsynaptic targets. We find that thalamic axons form a small number of synapses with the spiny stellates (188 on average), that the median size of the synapses is slightly larger than that of other synapses on the dendrites of spiny stellates, that they are not located particularly proximal to the soma, and that they do not cluster on the dendrites. These findings point to alternative mechanisms, such as synchronous activation of the sparse thalamic synapses to boost the efficacy of the thalamic input. The results also support the idea that the thalamic input does not by itself determine the cortical response of spiny stellate cells, allowing the cortical microcircuit to amplify and modulate its response according to the particular context and computation being performed.

The synaptic organization of the claustral projection to the cat's visual cortex.

The claustrum is a subcortical structure reciprocally connected with most areas of neocortex. This strategic location suggests an integrative role of the claustrum across different sensory modalities. However, our knowledge of the synaptic relationship between the neocortex and the claustrum is basic. In this study, we address this question through a structural investigation of the claustral projection to the ipsilateral primary visual cortex of the cat. Light microscopic reconstructions of axons from the entire thickness of cortex showed a very sparse innervation of the entire cortical depth, with most synaptic boutons in layers 2/3 and 6. Axons bearing numerous boutons terminaux and boutons en passant branched in these laminae. The sparse innervation did not seem to be compensated by particularly large synapses, given that the postsynaptic densities in the superficial layers are of comparable sizes (0.1 µm(2)) to other cortical synapses. All claustral synapses were asymmetric and in most cases targeted spines (87% in layer 4, 94% in layers 2/3 and 97% in layer 6). The pattern of innervation together with the known physiology of this projection suggests that the claustrum has a modulatory effect on visual cortex.

Whose Cortical Column Would that Be?

The cortical column has been an invaluable concept to explain the functional organization of the neocortex. While this idea was born out of experiments that cleverly combined electrophysiological recordings with anatomy, no one has 'seen' the anatomy of a column. All we know is that when we record through the cortex of primates, ungulates, and carnivores in a trajectory perpendicular to its surface there is a remarkable constancy in the receptive field properties of the neurons regarding one set of stimulus features. There is no obvious morphological analog for this functional architecture, in fact much of the anatomical data seems to challenge it. Here we describe historically the origins of the concept of the cortical column and the struggles of the pioneers to define the columnar architecture. We suggest that in the concept of a 'canonical circuit' we may find the means to reconcile the structure of neocortex with its functional architecture. The canonical microcircuit respects the known connectivity of the neocortex, and it is flexible enough to change transiently the architecture of its network in order to perform the required computations.

Selective targeting of the dendrites of corticothalamic cells by thalamic afferents in area 17 of the cat.

Pyramidal cells of layer 6 in cat visual cortex are the source of the corticothalamic projection, and their recurrent collaterals provide substantially more excitatory synapses in layer 4 than does the thalamic input. They have predominantly simple receptive fields and can be driven monosynaptically by electrically stimulating thalamic relay cells. Layer 6 cells could thus provide a significant disynaptic amplification of the thalamic input to layer 4, particularly since their synapses facilitate, unlike the thalamic afferents whose synapses depress. However, purely geometric considerations of the relation of their dendritic trees to the thalamic input indicate that they should form a far smaller number of synapses with thalamic afferents than do the simple cells of layer 4. We thus analyzed quantitatively the thalamic input to identified corticothalamic cells by labeling the thalamic afferents and corticothalamic cells in vivo. We made a correlated light and electron microscopic study of 73 "contacts" between thalamic afferents and five corticothalamic cells. The electron microscope revealed that only 24 of the contacts identified at light microscope level were indeed synapses and, contrary to geometric predictions, virtually all were located on spines on the basal dendrites. Our quantitative estimates indicate that the corticothalamic cells form even fewer synapses with the thalamic afferents than predicted by geometric considerations and only 1/10 as many as do the layer 4 simple cells. These data strongly suggest it is the collective computation of cortical neurons, not the monosynaptic thalamic input, that determines the output of the corticothalamic cells.

The proportion of synapses formed by the axons of the lateral geniculate nucleus in layer 4 of area 17 of the cat.

The connection between the dorsal lateral geniculate nucleus (dLGN) and area 17 of the cat is a classical model for studying thalamocortical relations. We investigated the proportion of asymmetric synapses in layer 4 of area 17 of cats formed by axons of the dLGN, because this is an important morphological parameter in understanding the impact of dLGN axons on their target neurons. Although the present consensus is that this proportion is small, the exact percentage remains in doubt. Most previous work estimated that the thalamus contributes less than 10% of excitatory synapses in layer 4, but one estimate was as high as 28%. Two issues contribute to these widely different estimates, one being the tracers used, the other being the use of biased stereological approaches. We have addressed both of these issues. Thalamic axons were labeled in vivo by injections of biotinylated dextran amine into the A lamina of the dLGN of anesthetized cats. After processing, the brain was cut serially and prepared for light and electron microscopy. The density of asymmetric synapses in the neuropil and the density of synapses formed by labeled dLGN boutons were measured by using an unbiased sampling method called the physical disector. Our counts indicate that, in the fixed cat brain, there are 5.9 x 10(8) +/- 0.9 x 10(8) asymmetric synapses per cubic millimeter of layer 4 in area 17, and the dLGN input provides only 6% of all asymmetric synapses in layer 4. The vast majority of synapses of layer 4 probably originate from other neurons in area 17.

The W cell pathway to cat primary visual cortex.

The thalamic input to area 17 in the cat can be divided into at least three parallel pathways, the W, X, and Y. Although the latter two are some of the best studied synaptic connections in the brain, the former remains poorly understood both in structure and in function. By combining light and electron microscopy, we have reconstructed in 3-D single W axons and described quantitatively the synapses that they form. We have also made a structural comparison of reconstructed synapses from the three visual pathways. Thalamic axons were labeled in vivo by injections of biotinylated dextran amine into the dLGN. W axons originating from C laminae injections arborized in layers 1, 2/3, and 5. Axons that traversed layer 1 supplied a few descending collaterals to layer 2/3, but the most extensive innervation in layer 2/3 was provided by axons ascending from the white matter. Most W boutons formed a single synapse, dendritic spines being the most common target, with dendritic shafts forming the remaining targets. In layer 1, the area of the postsynaptic density of spine synapses (0.16 microm(2)) was significantly larger than that of layers 2/3 (0.11 microm(2)) and 5 (0.09 microm(2)). Synapses from X and Y axons in layer 4 were similar in size to synapses formed by W boutons in layer 1. In layer 1, the main targets of the W axons are likely the apical dendrites of pyramidal cells, so that both proximal and distal regions of pyramidal cell dendritic trees can be excited by the W pathway.

A systematic random sampling scheme optimized to detect the proportion of rare synapses in the neuropil.

Synapses can only be morphologically identified by electron microscopy and this is often a very labor-intensive and time-consuming task. When quantitative estimates are required for pathways that contribute a small proportion of synapses to the neuropil, the problems of accurate sampling are particularly severe and the total time required may become prohibitive. Here we present a sampling method devised to count the percentage of rarely occurring synapses in the neuropil using a large sample (approximately 1000 sampling sites), with the strong constraint of doing it in reasonable time. The strategy, which uses the unbiased physical disector technique, resembles that used in particle physics to detect rare events. We validated our method in the primary visual cortex of the cat, where we used biotinylated dextran amine to label thalamic afferents and measured the density of their synapses using the physical disector method. Our results show that we could obtain accurate counts of the labeled synapses, even when they represented only 0.2% of all the synapses in the neuropil.