Drosophila melanogaster Astrocytes Respond to and Modulate Synaptic Transmission: A Correlative Anatomical and Electrophysiological Study

Abstract

Astrocytes are the most abundant non-neuronal cells in vertebrate brains. Although Drosophila melanogaster has fewer astrocytic cells relative to neuronal and other glial cell populations, they, like vertebrate astrocytes, are located in synaptic regions, organized into exclusive, minimally-overlapping domains, and play developmental roles in synaptogenesis. But, do Drosophila astrocytes have parallel roles in the regulation of synaptic signaling? Preliminary electron microscopic (EM) data indicates that astrocytic processes are located at a greater distance, on average, from Drosophila synapses than they are from vertebrate synapses, thus raising questions about their capacity to alter synaptic signals. Do astrocytic cells and processes occupy stereotyped synaptic regions across repeating segmental structures and across individuals? In the studies presented here, we have addressed these questions directly in the ventral nerve cord (VNC) of the third-instar larva. We collected the first whole-cell patch-clamp recordings from Drosophila astrocytes. These indicate that intrinsic membrane properties, such as low membrane resistance, high capacitance, a hyperpolarized resting potential relative to neurons, a passive current-voltage relationship, coupling to other astrocytic cells, and an absence of voltage-gated currents, are shared between astrocytes of highly divergent species. Next, we optogenetically activated of a group of glutamatergic pre-motor neurons and showed that astrocytes respond with a glutamate transporter current that is mediated by Eaat1, and that acute, pharmacological and chronic, genetic blockades of this transporter have subsequent effects on the decay of post-synaptic motor neuron currents. Then, we used three-dimensional EM to locate the pre-motor glutamatergic neurons that were activated in the physiological study and measured the distance from each presynaptic site to the nearest astrocytic process. We found that these distances vary 100-fold even along a single neurite and that these structures are rarely in direct contact, but that no synapse is positioned greater than one micron from an astrocytic process. Thus, it is in this anatomical configuration that the regulation of post-synaptic currents by Eaat1 occurs. Finally, we generated a library of single, fluorescently-labeled astrocytes that were co-labeled with fiduciary landmarks, and used this library to compare the placement of astrocyte cell bodies and arbors across VNC segments and individuals. We found substantial variation in the gross shape, size, and territory covered by astrocytes, and conclude that their neuropil domains are not reliably stereotyped. Given the consistent placement of neuronal connectome elements, this indicates that signals of a specific synapse are not regulated by a designated astrocyte. Together, these findings reveal new functional parallels between Drosophila and vertebrate astrocytes. These findings argue for the relevance and applicability of mechanistic discovery in Drosophila astrocytes, and set the stage for further inquiry into the genetic determinants of astrocyte morphology and physiology.