Glutamate can be released from astrocytes by several mechanisms, including exocytosis, volume regulated anion channels, as well as hemichannels (Hamilton and Attwell, 2010). Because this field is still at a relatively early stage, there is still discussion over which mechanism predominates under specific conditions and whether there are region-specific variations in release pathway. In the dentate gyrus, there is considerable evidence for an exocytotic mechanism of release of glutamate from astrocytes. For example, astrocyte-to-synapse modulation is attenuated by intracellular glial dialysis of tetanus toxin, and immunoelectron microscopy has revealed the presence

of vesicular glutamate transporter associated with small electron lucent vesicles in astrocytes associated with synapses in the dentate gyrus (Bezzi et al., 2004 and Jourdain et al., 2007). Because BTK phosphorylation it is not feasible with current technology to image exocytosis in brain slices, they examined exocytosis in cultured astrocytes using total internal reflection microscopy (TIRF) to see if it could provide

a clue to a gliotransmission switching mechanism. In TIRF one images the first ∼100 nm adjacent to the plasma membrane allowing the resolution of vesicles that are docked and ready to be released. Santello et al. (2011) transfected astrocytes with two vesicle-targeted constructs, one is pH sensitive and dequenches upon exocytosis, while the other is pH insensitive and allowed for the examination of the location of vesicles in relation to the plasma membrane. With this strategy, they showed that TNFα regulates the number of vesicles resident at the astrocytic plasma membrane. check details TNFα does not change the total number of vesicles but increases two-fold those that are present at the plasma membrane ready for exocytosis. As a consequence the presence of TNFα regulates the rate of exocytosis. In the absence of TNFα, exocytosis is slow and asynchronous with kinetics determined by the rate of

delivery of distant vesicles. However, the ability of TNFα to increase those vesicles already resident at the plasma membrane permits burst-mode exocytosis (Figure 1). How could this change in exocytosis mode gate gliotransmission in situ? Santello et al. (2011) reasoned that glutamate released by a slow asynchronous mechanism would be scavenged by local glutamate transporters almost which would prevent this gliotransmitter from accessing neuronal NMDA receptors that are required to modulate presynaptic release. Under the influence of TNFα, burst-mode release would provide sufficient temporally coincident glutamate to allow this transmitter to escape reuptake transporters. If this is the case, they predicted, and demonstrated, that pharmacological attenuation of glutamate transporters would permit astrocytic glutamate to access neuronal NMDA receptors and induce the consequent increase in mEPSC frequency under conditions of slow asynchronous release (TNFα−/−).