, 2007). As in axons, dendritic release of GABA from granule cells requires intracellular Ca2+ (Isaacson, 2001 and Isaacson and Strowbridge, 1998). To test the role of dendritic Ca2+ influx and neurotransmitter release in behavior, Abraham et al. (2010) recently augmented granule cell GABA release by conditionally disrupting the GluR2 AMPA receptor subunit specifically in granule cells. Consistent with the role of GluR2 in conferring Ca2+ impermeability to AMPA receptors (Burnashev et al., 1992, Hollmann et al., 1991 and Verdoorn et al., 1991), removing granule cell

GluR2 resulted in increased Ca2+ influx upon excitation from NVP-AUY922 mitral cells, which in turn triggered more robust GABA release thus increasing mitral cell inhibition. At the behavioral click here level, this manipulation accelerated response latencies in odor discrimination tasks. Conversely, deleting

the obligate NMDA receptor subunit NR1 in granule cells resulted in decreased GABA release and less robust mitral cell inhibition. Behaviorally, this reduction in dendritic GABA release slowed odor discrimination. Together, these data demonstrate the importance of dendritic exocytosis in shaping olfactory sensory processing (Abraham et al., 2010). While dendrodendritic synapses have been characterized anatomically and electrophysiologically,

very little is known about the molecular composition of these synapses. How similar are dendrodendritic synapses to typical axo-dendritic synapses? Immunolabeling EM studies have revealed the presence of canonical glutamatergic postsynaptic scaffolding molecules PSD-93 and PSD-95 at granule/mitral cell dendrodendritic synapses, suggesting DNA ligase that these synapses resemble typical axo-dendritic synapses (Sassoé-Pognetto et al., 2003). Additionally, both AMPA receptors and NMDA receptors are found on granule cells at sites apposed to mitral cell dendritic vesicle release zones (Sassoé-Pognetto et al., 2003). Interestingly, NMDA receptor activation is sufficient to activate dendritic GABA release from granule cells (Chen et al., 2002, Halabisky et al., 2000 and Schoppa et al., 1998). However, Ca2+ influx through NMDA receptors may not be directly coupled to vesicle release. Rather, Ca2+ influx through voltage-gated N- or P/Q-type Ca2+ channels triggered by depolarizing NMDA receptor currents has been shown to mediate vesicle fusion (Isaacson, 2001). Similar to presynaptic axon terminals, Ca2+ influx into granule cells appears to be tightly coupled to vesicle fusion. Introduction of the slow Ca2+ chelator EGTA has no effect on GABA release from granule cells (Isaacson, 2001).