CRCNS US-German Research Proposal: Quantitative and computational dissection of
glutamatergic crosstalk at tripartite synapses
(1) Christine R Rose, Heinrich Heine University, Düsseldorf, Germany
(2) Christian Henneberger, University of Bonn, Germany
(3) Ghanim Ullah, University of South Florida, Tampa, FL, USA
Project Description
1 Introduction and Background
Transmission at chemical synapses is the central mechanism by which information is
transferred between neurons. Synaptic connections such as glutamatergic excitatory
synapses are often perceived and modeled as point-to-point connections. However, there is
substantial evidence that crosstalk between various glutamatergic synapses can occur when
the presynaptically released glutamate is sensed not only by its direct postsynaptic partner
but also by nearby synapses of the same and other neurons [4]. Notably, this phenomenon
termed “glutamate spillover” not only defines the input-specificity of a given synaptic
connection and its crosstalk to neighboring synapses, but is also involved in and controlled by
activity-dependent plasticity [1, 7, 8].
How easily glutamate escapes from its release site and how far it spreads into the tissue
depends on the morphological and molecular properties of the extracellular space (ECS) as
well as on the efficacy of glutamate clearance, which primarily depends on astrocytic uptake
[11, 12]. We and others have shown that the efficacy of perisynaptic glutamate uptake by
astrocytes displays a remarkable heterogeneity between brain regions and, importantly, can
vary drastically from one synapse to the next within a brain region [3, 7, 8]. This is in part
because the morphological coverage of synapses by perisynaptic astrocyte processes (PAPs)
can differ strongly between individual synapses [14]. Moreover, the Henneberger lab has
recently shown that higher synaptic coverage by PAPs correlates with a higher local efficacy
of glutamate uptake [3]. We have also demonstrated that in addition to being heterogeneous,
astrocytic glutamate uptake and PAPs morphology both are controlled by neuronal plasticity
[1]. Moreover, glutamate uptake is governed by the transporters’ stoichiometry, importing one
glutamate molecule into the astrocyte by using the energy gained from co-transporting three
Na+ and one proton down the electrochemical gradients, whilst also exporting one K+ [12].
While the inwardly-directed Na+ gradient is the main driving force for glutamate uptake, recent
work by Rose lab and others have shown that glutamatergic activity causes local or global Na+
transients in astrocytes ([Na+]A) [15]. In the mouse hippocampus, astrocytic Na+ signals in fact
arise predominately due to the activity of glutamate transporters themselves, degrading the
Na+ gradient and thereby transiently weakening uptake capacity in a negative feedback-loop
[15-17]. In the neocortex, glutamatergic synaptic activity in addition results in prominent Na+
influx through NMDA receptors, boosting astrocyte Na+ gradients [18].
Thus, it is increasingly appreciated that astrocytic glutamate uptake is neither static nor
uniform. First, it is functionally dependent on the gradients of the transported ions which
dynamically change with synaptic transmission [12]. Second, it is plastic because structural
remodeling of PAPs on time scales of minutes profoundly alters perisynaptic glutamate spread
[1]. Therefore, the emerging hypothesis is that the degree of glutamate spillover and,
therefore, synaptic crosstalk in most brain regions are dynamically regulated and controlled at
the level of the astrocytes. Furthermore, since a single astrocyte can contact thousands of
synapses of various neurons, it has the potential to locally control the crosstalk of many
synapses. In such a scenario, an astrocyte, or a subcellular domain of it, can coordinate
crosstalk between many glutamatergic synapses on different neurons. Thereby, astrocytes
and their PAPs set the spatial fidelity of glutamatergic synaptic transmission and as a
consequence profoundly control neuronal signal exchange.
So far, these important hypotheses remain largely untested. We will fill this gap by
combining quantitative fluorescence imaging, astrocytic manipulations, and predictive
computer modelling. This will be accomplished by investigating perisynaptic astrocytic Na+
gradients, the main driving force of glutamate uptake, and local mechanisms controlling them
1
