By deleting GluN2A or GluN2B during early postnatal development, a period of rapid synaptogenesis, we found that

both subunits negatively regulate synaptic AMPAR expression, but by distinct means. We show that, similar to GluN1 deletion (Adesnik et al., 2008), deletion of GluN2B increases the number of functional synapses, suggesting a basal role for GluN2B-containing NMDARs in Imatinib in vitro maintaining silent synapses in early development. Conversely, deletion of GluN2A increases synaptic strength without affecting the number of unitary connections. These results suggest that when significant bursts of activity drive the synaptic insertion of AMPARs and the recruitment of GluN2A-containing receptors, GluN2A functions to dampen further synapse potentiation. The hippocampal

CA3-to-CA1 synapse is a model excitatory synapse that has been used to delineate the mechanisms of synaptic plasticity. Using conditional KO alleles for GluN2A (Grin2afl/fl; see Figure S1A available online) and GluN2B (Grin2bfl/fl) ( Akashi et al., 2009), we eliminated the target gene in a small subset of hippocampal neurons by transcranial stereotactic injection of P0-P1 mice with a recombinant adeno-associated Dinaciclib concentration virus expressing a Cre-GFP fusion protein (rAAV1-Cre-GFP) ( Kaspar et al., 2002). Figure 1A shows a typical acute slice made from a P18 mouse after P0 injection demonstrating sparse infection of CA1 pyramidal neurons. It has long been suspected Ribonucleotide reductase from in situ hybridization, single-cell reverse-transcriptase polymerase chain reaction, and pharmacologic studies that hippocampal CA1 pyramidal neurons express primarily GluN2A and

GluN2B subunits (Garaschuk et al., 1996, Watanabe et al., 1992 and Zhong et al., 1995). By cross-breeding the Grin2afl/fl (ΔGluN2A) and Grin2bfl/fl (ΔGluN2B) mice, we generated Grin2afl/flGrin2bfl/fl (ΔGluN2AΔGluN2B) mice and simultaneous whole-cell recording from a Cre-expressing cell (green trace in inset), and a control cell in the presence of NBQX revealed a complete loss of NMDAR-EPSCs ( Figure 1B, inset). We followed the time course of subunit depletion by measuring the ratio of NMDAR-EPSCs from Cre-expressing cells to control cells after P0 injection, and demonstrated a gradual decrease in NMDAR-EPSCs and complete loss consistently by P15 ( Figure 1B), similar to the rate of loss of NMDAR-EPSCs in Grin1fl/fl mice (ΔGluN1). These data indicate that, in addition to obligatory GluN1 subunits, synaptic NMDA receptors in CA1 pyramidal neurons contain only GluN2A and GluN2B. Since the NMDA-EPSCs were entirely gone by P15 in the double conditional KO mice, we performed all subsequent analyses of ΔGluN2A and ΔGluN2B mice after P17 unless indicated.