From this we conclude that these social deficits are not due simply to malnourishment and that the majority small molecule library screening of the behavioral phenotype can be attributed to

loss of GluN2B, although it is possible that premature expression of GluN2A enhances the deficit in social exploration. The primary observation from these studies is that GluN2B-mediated signaling during development cannot be rescued by premature expression of the mature NMDAR subunit GluN2A. This provides strong evidence that these subunits serve distinct functional roles. In terms of synaptic function, we show that GluN2B-containing NMDARs are necessary for appropriate regulation of AMPAR-mediated currents and play a critical role in protein translation-dependent

homeostatic plasticity. Additionally, our data suggest that their unique interaction with CaMKII is responsible for the distinct cellular signaling associated with GluN2B-containing NMDARs and that suppression of protein translation by these receptors involves the mTOR pathway. Interestingly, a recent report has shown that rapid antidepressant effects of NMDAR antagonists involve activation of the mTOR pathway and rapid stabilization of cortical synapses MLN2238 in vivo (Li et al., 2010). The data presented here provide evidence that this may be due to disruption of GluN2B signaling, which acts via mTOR to suppress local protein synthesis in cortical dendrites. GluN2B knockout mice die at P0, and this has largely precluded studies of the role of this subunit during development (Kutsuwada et al., 1996), whereas the absence of any dramatic survival phenotype in GluN2A null mice is consistent with their weak expression during embryogenesis and early postnatal periods (Sakimura et al., 1995). Recent reports have presented data from conditional GluN2B knockout mice; however, these studies have largely focused on the consequences of genetic removal after initial circuit formation (Brigman et al., 2010, von

Engelhardt et al., 2008 and Akashi et al., 2009). In light of this and of evidence that GluN2B may be uniquely required for proper localization of plasticity-related signaling molecules at excitatory synapses (Foster et al., 2010), we focused on examining the specific role of GluN2B during early Amisulpride postnatal development. During this period, homeostatic synaptic plasticity is required to maintain neuronal excitability within a physiologically appropriate range, because both the number of synapses onto individual neurons and sensory-driven activity increases in cortical networks. Homeostatic synaptic plasticity controls synapse excitability in a bidirectional manner, in part by regulating incorporation of synaptic AMPARs (Turrigiano, 2008). Scaling up synaptic AMPAR contribution in response to chronic suppression of neuronal activity has been shown to be dependent upon transcription but independent of NMDAR function (Turrigiano, 2008).