Introduction
Astrocytes are ubiquitous glial cells of the cortex which display a complex ramified anatomy. Astrocyte processes envelop synapses and dendrites, mediating diverse neuromodulatory pathways. While it is speculated that the domain of astrocyte-mediated neuromodulation is influenced by their anatomy, it is currently unknown what the stereotypical shape of an astrocyte is, beyond the simple observation of their branching. Nor do we know whether astrocyte anatomy abides by universal rules across brain regions and species.
Methods
Employing machine learning, graph theory, and topological analysis, we developed a comprehensive library of morphometric measures that extract quantitative anatomical features of astrocytes and neurons resolved under electron microscopy (Schubert et al., 2022). Based on trends in anatomical features specific to astrocyte morphology, we developed a generative algorithm that synthesizes astrocyte branching structure, inspired by the Minimum Spanning Tree (Cuntz et al., 2011).
Results
Using our library of morphometric features, we can quantitatively differentiate astrocytes both from neurons and from astrocytes of different species. Astrocytes are spatially complex cells with branches packed into a small space. They have large primary branches that define the shape of their territory and fine, diffusive branchlets that fill up the space within their territory. Based on these observations, we created a generative algorithm that replicates astrocyte branched anatomy, including the distinction between primary and terminal processes across different regions and animals.
Discussion
We present the first systematic characterization of cortical astrocyte anatomy from high-resolution EM datasets in different species. Our analysis, in particular, allowed us to identify salient astrocyte features that, in turn, can be used to constrain the minimal spanning tree as a general recipe to grow astrocytes in silico.
References
Cuntz, H., Forstner, F., Borst, A., & Häusser, M. (2010). One rule to grow them all: a general theory of neuronal branching and its practical application. PLoS computational biology, 6(8), e1000877.
Schubert, P. J., Dorkenwald, S., Januszewski, M., Klimesch, J., Svara, F., Mancu, A., ... & Kornfeld, J. (2022). SyConn2: dense synaptic connectivity inference for volume electron microscopy. Nature Methods, 19(11), 1367-1370.
The MICrONS Consortium (2025). Functional connectomics spanning multiple areas of mouse visual cortex. Nature, 640(8058), 435–447.
Acknowledgement
We acknowledge the support of the Natural Sciences and Engineering Research Council of Canada (NSERC) [Grant ID: RGPIN 2024 04333, 589115 2024]
