A connectomic study of a petascale fragment of human cerebral cortex

Abstract

We acquired a rapidly preserved human surgical sample from the temporal lobe of the cerebral cortex. We stained a 1 mm3 volume with heavy metals, embedded it in resin, cut more than 5000 slices at ~30 nm and imaged these sections using a high-speed multibeam scanning electron microscope. We used computational methods to render the three-dimensional structure of 50,000 cells, hundreds of millions of neurites and 130 million synaptic connections. The 1.3 petabyte electron microscopy volume, the segmented cells, cell parts, blood vessels, myelin, inhibitory and excitatory synapses, and 100 manually proofread cells are available to peruse online. Despite the incompleteness of the automated segmentation caused by split and merge errors, many interesting features were evident. Glia outnumbered neurons 2:1 and oligodendrocytes were the most common cell type in the volume. The E:I balance of neurons was 69:31%, as was the ratio of excitatory versus inhibitory synapses in the volume. The E:I ratio of synapses was significantly higher on pyramidal neurons than inhibitory interneurons. We found that deep layer excitatory cell types can be classified into subsets based on structural and connectivity differences, that chandelier interneurons not only innervate excitatory neuron initial segments as previously described, but also each others initial segments, and that among the thousands of weak connections established on each neuron, there exist rarer highly powerful axonal inputs that establish multi-synaptic contacts (up to ~20 synapses) with target neurons. Our analysis indicates that these strong inputs are specific, and allow small numbers of axons to have an outsized role in the activity of some of their postsynaptic partners.

The Impact of Neuron Morphology on Cortical Network Architecture

“…Our findings will advance the analysis and interpretation of future connectomics data -i.e., as soon as dense electron microscopy reconstructions become available for large cortical volumes. To illustrate how such datasets could be analyzed in principle, we apply our statistical approach to a densely reconstructed volume that comprises one cubic millimeter of human cortex (23). We analyze this petascale dataset in the preliminary form in which it was reported, which however reflects the current state-of-the-art for dense reconstructions of cortical tissue.…”

“…We show that the neuropil structure predicts cell type-specific connectivity patterns, occurrences of clusters of synapses, as well as of nonrandom network motifs that are in remarkable agreement with the available empirical data. Finally, we apply our statistical approach to a dense reconstruction of a petascale volume of human cortex (23) to illustrate how the impact of neuron morphology on network architecture can be tested beyond the structural model that is used to derive it.…”

Large-scale dendritic spine extraction and analysis through petascale computing

“…We chose the flood filling network (FFN) (Januszewski et al, 2018) segmentation code which has been implemented in multiple, large volume connectomic datasets derived from different species and different connectomics technologies (i.e. ATUM and SEM vs manual section pickup and TEM) (Shapson-Coe et al, 2021;Li et al, 2020) demonstrating that FFN is applicable to a wide range of connectomic studies and thus ideally suited to serve a broad range of samples.…”

“…The copyright holder for this preprint this version posted July 30, 2021. ; https://doi.org/10.1101/2021.07.29.454371 doi: bioRxiv preprint New segmentation algorithms for connectomics are advancing rapidly; however, they often come with increasing demand for more powerful computers, requiring increased processing times and making them less accessible to the general community. Furthermore, as imaging rates increase, producing larger datasets (Yin et al, 2020;Shapson-Coe et al, 2021), this too will put increasing demands for computational resources that can keep pace with these advancements. While advancements in image compression will likely help mitigate the storage load for certain analyses (Minnen et al, 2021) , we envision a continuing need for large computing resources as the number and size of datasets grow.…”

Cell type specific labeling and partial connectomes of dopaminergic circuits reveal non-synaptic communication and large-scale axonal remodeling after exposure to cocaine

“…Connectomics, the large scale and comprehensive 3D reconstructions of neurons and their connections, has emerged as a useful tool for understanding neural circuitry, revealing insights about how neurons connect that could not have been achieved any other way (Bae et al, 2018; Bates et al, 2020; Briggman et al, 2011; Helmstaedter et al, 2013; Karimi et al, 2020; Kasthuri et al ., 2015a; Morgan and Lichtman, 2020). However, most current connectomic efforts in mammals center on cortex and focus almost exclusively on glutamatergic and GABAergic circuits (Gour et al, 2020; Karimi et al ., 2020; Kasthuri et al ., 2015a; Motta et al, 2019; Shapson-Coe et al, 2021). A potential solution would be to label DA axons in large volume EM datasets and recent advances in protein engineering have created such genetic labels (Clarke and Royle, 2018; Shu et al, 2011).…”