We have developed a statistical approach enabling us to separate large and small events using a blind procedure. Application of the NMDAR antagonists significantly reduced the probability of observing large Ca2+ events, providing the first indication that presynaptic NMDARs contributed to the Ca2+ signal measured in the bouton. Interestingly, inhibition of NR2B containing NMDARs

does not produce a result significantly different from that observed in D-AP5. Because the NMDAR subunit composition is reported to change during development, with the number of NR2A-containing NMDARs thought to increase and perhaps partly replace NR2B-containing receptors within the synapse (Flint et al., 1997, Monyer et al., 1994 and Stocca and Vicini, 1998), our data suggest that we are examining terminals still at an early stage in development. There is literature describing the distribution of NMDAR subunits in the brain, including the hippocampus. learn more NMDAR subunits have been identified at both the pre- and postsynaptic locus. Of relevance here is that although NR1 subunits are reported to localize at CA1 dendrites (Petralia et al., 1994) and in dendritic spines (Petralia

and Wenthold, 1999, Selleckchem Palbociclib Racca et al., 2000 and Takumi et al., 1999), they have not been reported in boutons. In contrast, NR2B subunits have been shown to localize in presynaptic terminals of primate hippocampal CA3-CA1 synapses (Janssen et al., 2005), and NR2B and NR2D Endonuclease subunits have been found in presynaptic terminals of rat CA3-CA1 synapses (Charton et al., 1999 and Thompson et al., 2002) and within the dentate molecular layer (Jourdain et al., 2007). Here we take the immunoEM literature a step further by showing that the NR1 subunit is also present. NMDAR activation is classically dependent on both the presence of glutamate and the depolarization-induced relief of the Mg2+ block, so we manipulated each of these

factors independently in our study. The results inform us that the receptor behaves in a classical way, but they additionally reveal that large transients arise from transmitter release; thus, the variance in Ca2+ transient amplitude is a direct consequence of the stochastic nature of transmitter release and, as such, can be used as a proxy for pr. Although our pharmacological manipulations are consistent with presynaptic NMDARs having an autoreceptor role, we were mindful that the arrival of glutamate at the receptor must coincide with depolarization of the membrane; in essence, this means that the events we observe must be initiated within the duration of a single AP. To assess this, we measured the time required for the NMDAR current to reach its peak following rapid release of glutamate at a bouton. We observe rapid activation kinetics, within the range in which presynaptic NMDARs could function as autoreceptors.