Skip to main content

Plasticity at the tripartite synapse: an in vivo study of astrocyte-synapse interactions in the mammalian cortex

Final Report Summary - PLASTICASTROS (Plasticity at the tripartite synapse: an in vivo study of astrocyte-synapse interactions in the mammalian cortex.)

Proper functioning of the central nervous system (CNS) requires cooperation between its cellular components - neurons and glial cells. For years the “holy grail” of neuroscience has been deciphering how neurons transmit information. Glial cells have received marginal attention, as they were regarded merely as support cells for neurons. However, research over the last 30 years has revealed that glial cells, namely astrocytes, are incredibly influential in overall brain function. They maintain brain homeostasis, regulate local blood flow, participate in the immune response of the brain and influence synaptic transmission. Additionally, astrocytes are involved in a number of devastating disorders, such as neurodegenerative diseases, epilepsy and mental retardation and they mediate some of the effects of acute brain trauma and stroke. Thus, targeting astrocytes represents a genuine alternative strategy to the targeting of neurons as a treatment for various brain pathologies. To make this strategy possible, we need first to understand what astrocytes do.

In particular, astrocytes were shown to be in constant dialogue with synapses and secrete factors that regulate synaptic transmission and synapse number. However, the identity of molecules responsible for the physical interaction of astrocytic processes and synapses, as well as the molecular mechanisms underlying astrocyte-synapse interaction(s) were largely unknown at the start of this project. Additionally, most of the data available on neuro-glial interactions originate from ex vivo studies, which potentially biases the results due to altered astrocyte physiology, hampering the interpretation of data. These limits are due to a long-standing lack of appropriate tools for in vivo studies of astrocyte function.

The main goal of this proposal was to investigate the influence of astrocytes on the structure of neuronal networks in the brain of a living mouse. To reach this ambitious target, we split the work into several specific sections, each representing a separate technical challenge, completion of which would result in significant progress towards understanding astrocyte function.

Biochemical purification of presynaptic terminals results in the co-purification of astrocyte specific proteins, suggesting that a strong interaction between astrocytes and neurons is responsible for maintaining these interactions. To identify putative cell adhesion molecules (CAMs) involved in the interaction between astrocytic processes and synapses we performed an in silico screen of available databases for genes expressed in astrocytes, which encode proteins containing structural features common in CAMs, such as membrane spanning domains and IgG domains. As a result, we obtained a list of more than 100 molecules fulfilling these criteria. To narrow down this list of candidate molecules, we also performed a thorough literature search, as well searching open source repositories of gene expression data for brain. The result of this search is a list of 25 proteins, some of which are common CAMs, but for which the specific function in astrocytes remains unknown.

The era of whole-transcriptome profiling has resulted in numerous lists of genes potentially relevant for any process of interest (e.g. neurodegenerative disease) being produced. We faced this situation after performing our screen for CAMs. With pre-existing tools it would have been virtually impossible to functionally screen even a few candidate genes, due to time and cost constraints. To beat this limit we decided to establish a new method of generating transgenic mice that allows cell type-specific gene manipulation on a wild-type background. To this end we extended the in utero electroporation method to allow astrocyte specific expression of various transgenes in mouse over the long-term. Using a novel three electrode system for current delivery to target desired brain regions we achieved stable, efficient and specific expression of transgene in astrocytes in the brain of the adult mouse. We further explored the potential of the new method for numerous applications, including overexpression of multiple genes, gene silencing and in vivo imaging of different cell types.
In summary, our method represents a flexible, fast and affordable alternative for genetic manipulation of selected pathways in brain cells. The “blocks” building the construct can be easily swapped, so this technique can be employed to express in a constitutive or inducible fashion, genes encoding fluorescent proteins, proteins of interest for over-expression studies (including dominant negatives or constitutively active versions) or shRNA for loss-of-function studies. In theory, this method should also be transferrable to other species, for example rats, which are the animal model of choice for behavioral studies, but for which readily available transgenic lines do not exist.

What is the impact of the candidate CAMs (or any astrocytic protein) on astrocyte-synapse interactions? This question can be answered only when approached taking into account the natural surroundings of the cell, i.e. how the cell sits in the brain of a living animal. Until we started this project, most of the data on astrocyte physiology and its relationship to neuronal network function originated using ex vivo approaches. Therefore, to get proper insights into this process, we decided to monitor astrocytic activity in the awake, behaving mouse in response to sensory stimulation. To this end we chose to work in visual cortex, since this region is optically accessible for in vivo 2-photon imaging and well characterized in terms of the neuronal response to visual stimulation. We specifically expressed the ultrasensitive genetically encoded calcium indicator (GCaMP6) in astrocytes, and monitored calcium dynamics in astrocytes of the visual cortex through a chronic cranial window during visual stimulation of awake mice (in a collaboration with Dr V. Bonin at the Neuro-Electronic Research Flanders (NERF) Institute). By using a specific set of stimuli we were able to reveal, for the first time, spatiotemporal dynamics of these signals and to distinguish the visually-driven component of the calcium signal in individual cells, as well as within the wider astrocyte network. In addition, we could distinguish these responses from calcium transients induced by arousal. To link physiological data with the structure of individual cell processes at high resolution we turned to correlative light and electron microscopy (CLEM, in collaboration with the VIB Imaging Core Facility). After in vivo imaging, brains were processed for serial block face electron microscopy and (2-photon) imaged cells were identified and reconstructed.

These studies provide unique pioneering data on the astrocytic response to sensory stimulation. As our approach relates physiological and structural information, it provides a framework for future work on astrocyte signaling, including in pathology.

After establishing the appropriate toolset we approached our final aim - the functional characterization of identified astrocytic proteins in dendritic spine plasticity in vivo. We selected two CAMs and confirmed the expression of both proteins in astrocytes, as well as their developmental expression profile in the brain. To analyze their participation in signaling complexes we have generated constructs for the overexpression of these selected proteins labeled with genetic tags (to be delivered to astrocytes using the in utero electroporation method developed during this project), allowing subsequent purification and biochemical analysis of the proteins and associated complexes. In future, we plan to eliminate these proteins selectively from astrocytes and to evaluate the impact of this manipulation on astrocyte physiology (calcium imaging), synapse structure (CLEM) and mouse behavior.

In summary, during this Marie Sklodowska-Curie IEF-founded project, we made significant progress in understanding the role of astrocytes in the brain by: -) identification of potential molecules mediating astrocyte-synapse interaction; -) generating a new tool for genetic manipulation of astrocytes in vivo and -) revealing the response of astrocytes to physiological (sensory stimulation) in the intact brain. Thus, we established a complete workflow from molecule identification, through genetic manipulation of proteins of interest to the recording of astrocyte activity following manipulation. These studies provide a baseline for investigating physiological signaling in astrocytes, which can be used to interpret changes in diseases where information flow is affected (such as schizophrenia).

Further information on this and other projects can be found on the host lab website: