erate delta frequency oscillations, and this reduction in delta Endocannabinoid Modulation of Up-States power is correlated with reduced stability of bouts of NREM even when the drive to sleep has been increased. Discussion We found that the EC system modulates up-states in medial PFC PNs, sleep-wake states, and delta frequency oscillations during NREM. Specifically, the exogenous CB1 agonist, WIN, or inhibitors of EC inactivation augmented up-state amplitude and spiking, and these effects were abolished in CB1 KO cultures. The effects of CB1 agonists are consistent with a reduction in GABAergic signaling onto superficial layer PNs. However, NMDA EPSCs were also highly sensitive to WIN demonstrating that CB1’s effect on up-states is not simply a result of reduced GABAergic inputs onto PNs. In contrast to the effect of agonists, blockade of CB1 profoundly reduced up-state duration and increased sIPSC frequency on layer II/III PNs suggesting that tonic activation of the CB1 receptor regulates up-state duration. Lastly, up-state duration and spiking were paradoxically enhanced in CB1 KO cultures purchase ATL-962 indicating that while CB1 is not required for up-states, neuroadaptive changes likely occur in the absence of CB1 signaling resulting in a dysregulation of cortical network function. Sleep studies in CB1 KO mice found that genetic deletion of CB1 results in an increase in wake via a specific reduction of NREM sleep maintenance. CB1 KO mice display a reduced ability to maintain NREM even following TSD which is indicative of significant dysfunction in the homeostatic sleep response. Furthermore, the rebound in NREM delta power after TSD is reduced in these mice suggesting that EC signaling contributes to the regulation of slow-oscillations that predominate during this state. Endocannabinoid Modulation of Up-States CB1 modulation of up-states The present findings are the first to indicate that EC tone exists in the cortex and that tonic activation of CB1 is important for the regulation of network activity. Up-state activity in prefrontal slice cultures was 
altered by direct activation of CB1 receptors, by modulators of EC inactivation, and by CB1 antagonists. However, slices from KO animals still showed up-state activity demonstrating that CB1 receptors modulate but are not required to generate up-states. WIN’s enhancement of up-state amplitude is consistent with disinhibition of PN’s resulting from a CB1-mediated reduction of GABAergic input. Up-states are characterized by robust and persistent glutamatergic transmission that is countered by GABAAmediated shunting inhibition. Reducing GABAergic input via CB1 activation would decrease this inhibition and further depolarize membrane potential. Despite this idea, in the present study WIN inhibited both NMDA and GABAA PSCs in layers II/ III and V/VI PNs suggesting that these effects might balance one another. Data from previous studies in the PFC and other regions of cortex also show that EC-CB1 signaling can reduce glutamate transmission suggesting that WIN’s effect on up-states is not driven solely by decreased GABAergic signaling. In fact, when IPSC and EPSC effects were compared, WIN produced greater inhibition of EPSCs. One possible explanation for this apparent paradox is that additional CB1-dependent mechanisms are at work in interneuron populations PubMed ID:http://www.ncbi.nlm.nih.gov/pubmed/19643975 resulting in a net loss of GABAergic signaling during up-states. Populations of CB1-positive interneurons in the cortex are electrically coupled, and