AUTHOR=Aboufares El Alaoui Amina , Jackson Molly , Fabri Mara , de Vivo Luisa , Bellesi Michele TITLE=Characterization of Subcellular Organelles in Cortical Perisynaptic Astrocytes JOURNAL=Frontiers in Cellular Neuroscience VOLUME=14 YEAR=2021 URL=https://www.frontiersin.org/journals/cellular-neuroscience/articles/10.3389/fncel.2020.573944 DOI=10.3389/fncel.2020.573944 ISSN=1662-5102 ABSTRACT=

Perisynaptic astrocytic processes (PAPs) carry out several different functions, from metabolite clearing to control of neuronal excitability and synaptic plasticity. All these functions are likely orchestrated by complex cellular machinery that resides within the PAPs and relies on a fine interplay between multiple subcellular components. However, traditional transmission electron microscopy (EM) studies have found that PAPs are remarkably poor of intracellular organelles, failing to explain how such a variety of PAP functions are achieved in the absence of a proportional complex network of intracellular structures. Here, we use serial block-face scanning EM to reconstruct and describe in three dimensions PAPs and their intracellular organelles in two different mouse cortical regions. We described five distinct organelles, which included empty and full endosomes, phagosomes, mitochondria, and endoplasmic reticulum (ER) cisternae, distributed within three PAPs categories (branches, branchlets, and leaflets). The majority of PAPs belonged to the leaflets category (~60%), with branchlets representing a minority (~37%). Branches were rarely in contact with synapses (<3%). Branches had a higher density of mitochondria and ER cisternae than branchlets and leaflets. Also, branches and branchlets displayed organelles more frequently than leaflets. Endosomes and phagosomes, which accounted for more than 60% of all the organelles detected, were often associated with the same PAP. Likewise, mitochondria and ER cisternae, representing ~40% of all organelles were usually associated. No differences were noted between the organelle distribution of the somatosensory and the anterior cingulate cortex. Finally, the organelle distribution in PAPs did not largely depend on the presence of a spine apparatus or a pre-synaptic mitochondrion in the synapse that PAPs were enwrapping, with some exceptions regarding the presence of phagosomes and ER cisternae, which were slightly more represented around synapses lacking a spine apparatus and a presynaptic mitochondrion, respectively. Thus, PAPs contain several subcellular organelles that could underlie the diverse astrocytic functions carried out at central synapses.