, 1997). Similar to humans with AS, mice lacking maternal Ube3a (Ube3am−/p+) have

abnormal EEG activity and are susceptible to cortical seizures, suggesting that loss of Ube3a might disrupt the excitatory/inhibitory balance in the neocortex ( Jiang et al., 1998). Loss of maternally inherited Ube3a results in decreased excitatory synaptic drive onto pyramidal neurons in layer 2/3 (L2/3) of neocortex, as evidenced by a loss of dendritic spines ( Yashiro et al., 2009). Decreased Ube3a-mediated PI3K inhibitor proteasomal degradation of Arc and Ephexin5 proteins may lead to excitatory synaptic defects ( Greer et al., 2010 and Margolis et al., 2010). These observations suggest a mechanism for how the loss of Ube3a may cause fewer and/or weaker excitatory synapses. While these deficits may be relevant to cognitive phenotypes in Ube3am−/p+ mice, they would not on their own predict hyperexcitability and increased seizure susceptibility. We hypothesized that Ube3a loss results in more severe inhibitory deficits, with the net outcome favoring cortical

hyperexcitability. Here, we use the visual cortex as a model to study the role of Ube3a in the establishment and function of inhibitory circuits. We show that Ube3am−/p+ mice have an abnormal accumulation of clathrin-coated vesicles at inhibitory axon terminals, indicating a defect in vesicle cycling. Consistent with this observation, inhibitory synaptic transmission MAPK inhibitor onto L2/3 pyramidal neurons recovers slower following vesicle depletion in Ube3am−/p+ mice, compared to wild-types. Recovery following high-frequency stimulation of excitatory synapses onto L2/3 pyramidal neurons, however, is

normal. This discrepancy among synapse types may further contribute to excitatory/inhibitory also imbalance during high levels of activity. Finally, we show that synaptic inputs onto inhibitory neurons in Ube3am−/p+ mice are largely normal. We conclude that neuron type-specific synaptic deficits are likely to underlie neocortical excitatory/inhibitory imbalance in AS. An excitatory/inhibitory imbalance in AS could arise from reduced numbers of inhibitory interneurons, abnormal inhibitory connectivity, and/or decreased inhibitory neurotransmission. To test the first possibility, we performed immunohistochemistry for three markers—parvalbumin, calretinin, and somatostatin—which together label 96% of the total GABAergic interneurons in L2/3 of mouse primary visual cortex (V1) (Gonchar et al., 2007). We compared Ube3am−/p+ mice and their wild-type (WT) 129Sv/Ev strain littermate controls at postnatal day 80 (P80), an age where AS model mice exhibit abnormal EEG patterns and are susceptible to seizures ( Jiang et al., 1998). We found no differences in the density of inhibitory interneurons expressing these markers ( Figures 1A and 1B), implying that the relative number of inhibitory interneurons is normal in L2/3 of V1.