Search

Ischemia leads to a severe dysregulation of glutamate homeostasis and excitotoxic cell damage in the brain. Shorter episodes of energy depletion, for instance during peri-infarct depolarizations, can also acutely perturb glutamate signaling. It is less clear if such episodes of metabolic failure also have persistent effects on glutamate signaling and how the relevant mechanisms such as glutamate release and uptake are differentially affected. We modeled acute and transient metabolic failure by using a chemical ischemia protocol and analyzed its effect on glutamatergic synaptic transmission and extracellular glutamate signals by electrophysiology and multiphoton imaging, respectively, in the mouse hippocampus. Our experiments uncover a duration-dependent bidirectional dysregulation of glutamate signaling. Whereas short chemical ischemia induces a lasting potentiation of presynaptic glutamate release and synaptic transmission, longer episodes result in a persistent postsynaptic failure of synaptic transmission. We also observed unexpected differences in the vulnerability of the investigated cellular mechanisms. Axonal action potential firing and glutamate uptake were surprisingly resilient compared to postsynaptic cells, which overall were most vulnerable to acute and transient metabolic stress. We conclude that short perturbations of energy supply lead to a lasting potentiation of synaptic glutamate release, which may increase glutamate excitotoxicity well beyond the metabolic incident.

Structural changes of astrocytes and their perisynaptic processes occur in response to various physiological and pathophysiological stimuli. They are thought to profoundly affect synaptic signalling and neuron-astrocyte communication. Understanding the causal relationship between astrocyte morphology changes and their functional consequences requires experimental tools to selectively manipulate astrocyte morphology. Previous studies indicate that RhoA-related signalling can play a major role in controlling astrocyte morphology, but the direct effect of increased RhoA activity has not been documented in vitro and in vivo. Therefore, we established a viral approach to manipulate astrocytic RhoA activity. We tested if and how overexpression of wild-type RhoA, of a constitutively active RhoA mutant (RhoA-CA), and of a dominant-negative RhoA variant changes the morphology of cultured astrocytes. We found that astrocytic expression of RhoA-CA induced robust cytoskeletal changes and a withdrawal of processes in cultured astrocytes. In contrast, overexpression of other RhoA variants led to more variable changes of astrocyte morphology. These induced morphology changes were reproduced in astrocytes of the hippocampus in vivo. Importantly, astrocytic overexpression of RhoA-CA did not alter the branching pattern of larger GFAP-positive processes of astrocytes. This indicates that a prolonged increase of astrocytic RhoA activity leads to a distinct morphological phenotype in vitro and in vivo, which is characterized by an isolated reduction of fine peripheral astrocyte processes in vivo. At the same time, we identified a promising experimental approach for investigating the functional consequences of astrocyte morphology changes.

The concept of the tripartite synapse describes the close interaction of pre- and postsynaptic elements and the surrounding astrocyte processes. For glutamatergic synapses, it is established that the presence of astrocytic processes and their structural arrangements varies considerably between and within brain regions and between synapses of the same neuron. In contrast, less is known about the organization of astrocytic processes at GABAergic synapses although bi-directional signaling is known to exist at these synapses too. Therefore, we established super-resolution expansion microscopy of GABAergic synapses and nearby astrocytic processes in the stratum radiatum of the mouse hippocampal CA1 region. By visualizing the presynaptic vesicular GABA transporter and the postsynaptic clustering protein gephyrin, we documented the subsynaptic heterogeneity of GABAergic synaptic contacts. We then compared the volume distribution of astrocytic processes near GABAergic synapses between individual synapses and with glutamatergic synapses. We made two novel observations. First, astrocytic processes were more abundant at the GABAergic synapses with large postsynaptic gephyrin clusters. Second, astrocytic processes were less abundant in the vicinity of GABAergic synapses compared to glutamatergic, suggesting that the latter may be selectively approached by astrocytes. Because of the GABA transporter distribution, we also speculate that this specific arrangement enables more efficient re-uptake of GABA into presynaptic terminals.