This raises the question as to how these signaling pathways interact at DMH synapses. If CB1R activation at the presynaptic terminal precludes the effects of NO to enhance GABA release, then the application of a CB1R agonist should block the potentiation of GABA transmission by the NO donor, SNAP. Consistent with this

idea, SNAP failed to increase evoked IPSC amplitude when applied to slices that were continuously perfused with WIN 55,212-2 (104% ± 12.6% of WIN 55,212-2, n = 5, p = 0.646; Figures 5A and 5C). Similarly, it did not affect PPR (baseline: 0.961 ± 0.119; post-drug: 0.883 ± CAL-101 concentration 0.178; p = 0.544) or CV (baseline: 0.502 ± 0.071; post-drug: 0.500 ± 0.045; p = 0.962). Conversely, WIN 55,212-2 still effectively depressed IPSCs that were first potentiated by SNAP (36% ± 12.0% of SNAP,

n = 7; Figures 5B and 5D). This change was accompanied by an increase in PPR (baseline: 0.663 ± 0.109; post-drug: 0.950 ± 0.099; p = 0.048) and CV (baseline: 0.332 ± 0.084; post-drug: 0.593 ± 0.117; p = 0.049), consistent with the effect of WIN 55,212-2 in the absence of SNAP. Interestingly, the onset of the WIN 55,212-2–induced PD0325901 purchase depression was accelerated in the presence of SNAP when compared with WIN 55,212-2 alone, as evidenced by a decrease in the decay constant of the depression after drug application by approximately 80% (from 13.0 ± 2.8 min to 2.5 ± 0.7 min; Figure 5D). These data suggest that activation of CB1Rs attenuates the NO-induced increase in GABA release, whereas NO itself enhances the effects of a CB1R ligand. Next, we conducted experiments to determine the consequences of NO production on eCB-mediated LTDGABA. When NO synthesis was inhibited by L-NAME, HFS (100 Hz for 4 s ×

2, 0.05 Hz interval) failed to elicit LTDGABA (111% ± 11.3% of baseline, n = 8, p = 0.350; Figure 5E), the changes in PPR (baseline: 0.860 ± 0.086; post-HFS: 0.826 ± 0.102; p = 0.369), or the changes in CV (baseline: 0.311 ± 0.028; post-HFS: 0.336 ± first 0.07; p = 0.452). This suggests that NO signaling is required either for eCB production or CB1R signaling. Consistent with the latter idea, direct activation of CB1Rs by WIN 55,212-2 in the presence of L-NAME failed to significantly depress evoked IPSC amplitude (88% ± 10.8% of baseline, n = 6, p = 0.375; Figure 5F), PPR (baseline: 0.903 ± 0.129; post-drug: 0.889 ± 0.092; p = 0.850), or CV (baseline: 0.362 ± 0.067; post-drug: 0.410 ± 0.094; p = 0.168). Overall, these data point to an inherent complexity in the signaling of the retrograde transmitters eCBs and NO in the DMH. Specifically, they argue that eCB signaling prevents NO-mediated potentiation of GABA synapses but that NO signaling is required for eCB-induced depression of GABA signaling. We have demonstrated thus far that CB1R signaling precludes NO-mediated LTPGABA in the DMH. Here, we hypothesized that a physiological state in which CB1R signaling is compromised should favor the induction of LTPGABA.