Questionable Calcium

See allHide authors and affiliations

Science  05 Mar 2010:
Vol. 327, Issue 5970, pp. 1212-1213
DOI: 10.1126/science.1187420

Astrocytes constitute the major glial cells of the mammalian central nervous system, but they were long regarded as passive elements, providing only structural and nutritional support to neurons. Over the past 25 years, another view has emerged in which astrocytes directly affect neuron function. Just as a neuron releases molecules to signal to another neuron (at a junction called the synapse), astrocytes also transmit molecules to affect neuronal communication, and it is thought that the same release mechanism is used (16). However, conflicting data continue to raise debate about this view, including the contradictory results of mouse studies reported recently (7, 8) and by Agulhon et al. on page 1250 of this issue (9). Can this controversy be resolved?

Thin astrocyte processes wrap tightly around synaptic neurons, and there is good evidence for signaling between these cells at the synaptic junction. Astrocytes respond to the stimulation of nearby neurons with an increase in intracellular calcium (Ca2+) concentration. This increase triggers the release of transmitter molecules such as glutamate, which can potentiate the synaptic activity of neighboring neurons (1, 3), or adensine triphosphate (ATP), which reduces synaptic transmission (4). In addition, restricting the intracellular Ca2+ concentration in astrocytes can prevent the release of transmitters, as shown for d-serine, thus inhibiting modulation of neurotransmission at closely located synapses (5). These findings support a model (see the figure) in which a presynaptic neuron releases transmitters that activate receptors at the postsynaptic neuron and astrocytes. In asytrocytes, this causes the liberation of Ca2+ from the endoplasmic reticulum (ER) and the subsequent release of transmitter molecules that control neurotransmission.

Although this may explain how astrocytes directly affect neuron function, there are concerns as to whether such “gliotransmission” is merely an experimental artifact. For instance, intracellular vesicles harboring transmitter molecules, and the molecular machinery required for their release (in response to intracellular Ca2+), have rarely been detected in astrocytes.

These concerns are now boosted by recent observations using different mouse models (79). One mouse model was genetically engineered to allow for astrocyte-specific increases in intracellular Ca2+. For that purpose, expression of the metabotropic receptor MrgA1, which is normally not expressed in the brain, was induced in astrocytes. The MrgA1 ligand was used to selectively increase the concentration of the intracellular signaling molecule inositol 1,4,5-trisphosphate (IP3). IP3 triggers the liberation of Ca2+ from the ER. The other mouse model was engineered to prevent any IP3-mediated Ca2+ liberation from the ER by deletion of the gene encoding IP3R2, the only astrocyte receptor (expressed on the ER) that responds to IP3. Thus, astrocytes cannot evoke this increase of intracellular Ca2+. Collectively, the studies show that in both paradigms, neurotransmission and synaptic plasticity (long-term potentiation) in hippocampal neurons were unaffected, thus challenging the role of Ca2+ transients (induced by IP3) for gliotransmission.

Calcium trigger.

Glutamate (Glu), ATP, and d-serine (Ser) are gliotransmitters that affect synaptic activity [through ionotropic receptors (postsynaptic neuron) and metabotropic receptors (presynatic neuron and astrocytes)]. Their release is triggered by intracellular Ca2+, but where does the Ca2+ come from?


Differences in experimental setups or method-intrinsic artifacts may be responsible for the divergent results. For example, overexpression of the MrgA1 receptor in mice might not include appropriate receptor localization and embedding in astrocyte signaling pathways. But rather than focus on methodological reasons, it is perhaps more profitable to consider these results in light of several commonly accepted facts: Astrocytes express receptors that respond to neuronal activity; activation of these receptors causes transient increases in intracellular Ca2+; astrocytes release gliotransmitters from intracellular vesicles; these gliotransmitters modulate neuronal activity. However, the data obtained by Agulhon et al. in mice lacking an IP3 receptor suggest that other molecules involved in Ca2+ signaling should be considered, such as voltage-gated Ca2+ channels or transient receptor potential (TRP) channels. TRP channels are Ca2+-permeable, nonselective cation channels that are gated by diverse stimuli such as phospholipids, oxidants, cell volume changes, acidity, or osmolarity (10). In addition, unlike neurons, astrocytes do not necessarily require a process for priming transmitter-containing vesicles for rapid release. The composition and properties of the release machinery also may be different from their neuronal counterparts (11). And the requirements for Ca2+ in releasing gliotransmitters may vary in different subcellular regions of an astrocyte. Perisynaptic astrocyte processes can be thinner than 50 nm. It is conceivable that Ca2+ transients in these regions have been missed by the largely somatic Ca2+ recordings of Agulhon et al. Perhaps TRP or voltage-gated Ca2+ channels, which are both expressed in astrocytes, mediate Ca2+ entry (12).

Technological advancements could help to resolve the mechanism of gliotransmission. More sensitive imaging that operates at submicrometer resolution would allow recording of Ca2+ signals from astrocyte processes just outside the synapse that have thus far been missed. Also, genetic mouse models in which proteins involved in the release of gliotransmitters can be visualized by fluorescence would demonstrate release machinery in astrocytes. Finding out where astrocytic Ca2+ comes from and how it supports gliotransmission is an important step in settling the debate.

References and Notes

  1. I thank N. Brose and S. Wichert for discussions. Supported by the Max Planck Society, the MPI for Experimental Medicine, the European Union (FP7 Neuroglia, EdUGlia), and the Deutsche Forschungsgemeinschaft (CMPB, SPP 1172, TRR 43).
View Abstract

Stay Connected to Science

Navigate This Article