Rapid astrocyte morphology changes during epileptogenesis in the rodent hippocampus

Abstract

In the past decades astrocytes have been demonstrated to play an important role in physiological brain processes but also neurodegenrative diseases. Their characteristic morphology allows them to closely contact neurons. The close proximity of astrocyte peripheral processes and neuronal compartments enables astrocytes to modulate synaptic transmission and neuronal activity. Changes in this spatial relationship could therefore result in alterations of neuronal activity, which could also be of pathophysiological relevance. Indeed, astrocyte structural and associated functional changes were observed in the late states of epilepsy implying a potential crucial role of astrocytes in epileptogenesis. How early these astrocyte morphology changes occur and to what degree they may contribute to the development of epilepsy is yet unknown. The present study aimed to investigate i) if and on which timescale astrocyte morphology is altered after induction of acute epileptiform activity, ii) which mechanism mediates structural changes of astrocytes and iii) what the functional consequences of astrocyte morphology changes are and if they contribute to the development of epileptiform activity. Combining electrophysiological recordings and two-photon excitation fluorescence microscopy revealed novel results concerning the involvement of astrocyte morphology in the acute development of epileptiform activity. It could be shown that astrocytes respond with a robust structural change to epileptiform activity on a timescale of minutes in vitro and in vivo in different species. This morphology change could be described as shrinkage of peripheral astrocyte processes. Astrocyte processes shrinkage was triggered by the onset of epileptiform activity and did not require ongoing epileptiform activity. By modulating the ROCK pathway it could be demonstrated that these astrocyte morphology changes are ROCK dependent, suggesting a potential involvement of the actin cytoskeleton. As a functional consequence, a decreased astrocyte intra- and intercellular diffusion was observed. Interestingly, the induced astrocyte morphology changes had an overall proepileptic effect. This proepileptic action of rapid astrocyte morphology changes could not be explained by changes in astrocyte K+ buffering. Instead, it was likely mediated by faster accumulation of glutamate, which could result in an amplification of neuronal excitation. These results indicate that astrocyte morphology changes occur on a very rapid time scale after induction of epileptiform activity and are mediated by a fast cytoskeletal restructuring. These astrocyte morphology changes support persistent epileptiform activity. In conclusion, this study provides evidence that astrocyte morphology changes could play a crucial role early during epileptogenesis.