tates in sham treated control tissue = 11.06, p,0.001) and in many cases, reduced up-states to a brief EPSP. In pair-wise comparisons with control treated cultures up-state duration was similarly reduced by the inverse agonists AM281 = 5.51, p,0.001) and AM251 = 3.83, p = 0.004) and the neutral antagonist NESS0327 = 2.91; p = 0.042; ). The peak amplitude of up-states was also significantly reduced by the application of CB1 antagonists = 3.90, p = 0.020); however, pair-wise comparisons with the sham group revealed this effect only reached significance for AM251-treated cultures = 2.93, p = 0.042). Pair-wise comparisons 6 Endocannabinoid Modulation of Up-States 7 Endocannabinoid Modulation of Up-States graphs bars represent mean 6 SEM. For all symbols conferring statistical significance: single symbol p,0.05, double symbol p,0.01, triple 
symbol p,0.001. doi:10.1371/journal.pone.0088672.g004 between drug treatment groups found no significant differences 0.79, p.0.999). Because AM281 produced the largest magnitude inhibition of up-state parameters, this drug was used to confirm that these effects were CB1-dependent. Cultures prepared from CB1 KO mice were exposed to sham treatment or AM281 while up-states were evoked, and these data were then compared to the AM281 and sham treated groups of wt cultures from the previous experiment. To directly compare the effects of AM281 on up-state parameters, data were normalized to pre-drug baseline values for each genotype. AM281 produced only a slight reduction in up-state amplitude in the last 10 min of recording, and a two-way comparison found no interaction = 2.43, p = 0.13) or main effects of genotype = 0.011, p = 0.915) or treatment = 3.46, p = 0.073). However, a comparison of up-state duration from the last 10 min of recordings found a significant interaction between genotype and treatment = 7.16, p = 0.0115) and a main effect of treatment = 10.70, p = 0.0025). As before, the duration of up-states in wt cultures treated with AM281 was significantly reduced compared to sham treated controls = 4.27, PubMed ID:http://www.ncbi.nlm.nih.gov/pubmed/19647483 p,0.001), but there was no difference in the duration of up-states from sham or AM281 treated KO cultures = 0.41, p.0.999). Up-states rely on a balance between synaptic activity at GABAergic and glutamatergic synapses and ongoing EC activation of CB1 decreases GABAergic signaling at layer II/III PNs as well as glutamate synapses throughout cortical layers. If EC tone critically maintains the appropriate ratio between glutamatergic and GABAergic signaling during up-states, then the diminished duration of up-states following application of CB1 antagonists might reflect enhanced GABAergic neurotransmission. To test this hypothesis, baseline recordings of sIPSCs from layer II/III PNs in wt cultures were compared to recordings from cultures incubated in AM281 for at least 1 hr. The inter-event interval of GABAA sIPSCs was significantly reduced in PBTZ 169 biological activity AM281treated cultures = 2.67, p = 0.014), but there was no effect on sIPSC amplitude = 0.79, p = 0.439) supporting the idea that blocking the activation of CB1 enhances GABAergic input onto layer II/III PNs. To identify whether AEA or 2-AG mediates the tonic regulation of up-states, the DAGL inhibitor, THL, was applied to cultures while evoking up-states. Compared to recordings from sham treated cultures, THL application failed to alter either 8 Endocannabinoid Modulation of Up-States 9 Endocannabinoid Modulation of Up-States up-state duration = 1.23 p = 0.238) or
