Prefrontal cortex stimulation induces 2-arachidonoyl-glycerol-mediated suppression of excitation in dopamine neurons

Abstract

Endocannabinoids form a novel class of retrograde messengers that modulate short- and long-term synaptic plasticity. Depolarization-induced suppression of excitation (DSE) and inhibition (DSI) are the best characterized transient forms of endocannabinoid-mediated synaptic modulation. Stimulation protocols consisting of long-lasting voltage steps to the postsynaptic cell are routinely used to evoke DSE-DSI. Little is known, however, about more physiological conditions under which these molecules are released in vitro. Moreover, the occurrence in vivo of such forms of endocannabinoid-mediated modulation is still controversial. Here we show that physiologically relevant patterns of synaptic activity induce a transient suppression of excitatory transmission onto dopamine neurons in vitro. Accordingly, in vivo endocannabinoids depress the increase in firing and bursting activity evoked in dopamine neurons by prefrontal cortex stimulation. This phenomenon is selectively mediated by the endocannabinoid 2-arachidonoyl-glycerol (2-AG), which activates presynaptic cannabinoid type 1 receptors. 2-AG synthesis involves activation of metabotropic glutamate receptors and Ca2+ mobilization from intracellular stores. These findings indicate that dopamine neurons release 2-AG to shape afferent activity and ultimately their own firing pattern. This novel endocannabinoid-mediated self-regulatory role of dopamine neurons may bear relevance in the pathogenesis of neuropsychiatric disorders such as schizophrenia and addiction.

Figure 1.

A brief train inhibits EPSCs through activation of CB1 receptors. A, VTA DA cells, stimulated by low-frequency stimulation (0.1Hz), respond to a train (10 stimuli; 5 Hz) at time 0 (arrow) with a transient SE. Time course of train-induced SE (n = 7) is shown. Representative traces for each condition from a single experiment are shown. Calibration: 20 msec, 100 pA. B, Train-induced SE increases paired-pulse ratio (n = 7; p < 0.005). The left graph plots the averaged paired-pulse ratio in a bar graph form, and the right graph plots the paired-pulse ratio for each of the experiments in A before (basal) and 5-15 sec after (SE) the train. C, SE induction depends on number of pulses within the train (n = 7; ^*p < 0.05; ^**p < 0.01) and on stimulation frequency (n = 5; ^*p < 0.05; ^**p < 0.01). D, The magnitude of DSE induced by a depolarizing step (voltage step to +40 mV; 10 sec) is plotted as a function of train-evoked SE (n = 6). Data are fit by linear regression with r² = 0.87 (p < 0.005). E, In the presence of either AM281 (left panel, closed circles; n = 6) or WIN (middle panel, closed circles; n = 5), SE could not be induced. Time course of train-induced SE (control, open circles) is shown. Train-induced suppression of excitation is absent in the CB1^-/- mice (right panel; n = 5). Averaged time courses of SE in the CB1^+/+ (open symbols; n = 5) and CB1^-/- (closed symbols; n = 5) mice are displayed. Representative traces for each condition from a single experiment are shown. The EPSC recorded after the train is superimposed in light gray for comparison. Calibration: 20 msec, 100 pA.

Figure 2.

Blockade of CB1 receptors enhances DA neuron responses to PFC stimulation. A, B, Representative digital storage oscilloscope traces (A) and peristimulus time histograms (B) (1 msec bin) illustrating excitatory responses in VTA DA neurons evoked by the stimulation of the PFC at time 0 (arrows) and the increase in spiking probability after the intravenous administration of SR. C, The probability of stimulus-locked spikes (left), bursts (middle), and the frequency of spikes within bursts (right) is increased in DA neurons after SR administration. Data are expressed as mean ± SEM and normalized to their baseline level (100%). ^*p < 0.05; ^**p < 0.01; ANOVA and Dunnett's test.

Figure 3.

Stimulation of CB1 receptors depresses DA neuron responses to PFC stimulation. A, B, Representative digital storage oscilloscope traces (A) and peristimulus time histograms (B) (1 msec bin) illustrating excitatory responses in VTA DA neurons evoked by the stimulation of the PFC at time 0 (arrows) and displaying the decrease in stimulation-evoked responses induced by the cannabinoid agonist WIN and the reversal by SR. C, WIN decreases spiking probability of DA neurons after PFC stimulation (left), stimulus-locked bursts (middle), and intraburst frequency (right). All effects were fully reversed by SR. Data are expressed as mean ± SEM and normalized to their baseline level (100%).^*p < 0.05; ^**p < 0.01; ANOVA and Dunnett's test.

Figure 4.

Synthesis and release of 2-AG are necessary for train-induced SE. A, Time course of train-induced SE under normal conditions (open circles; n = 7) and in the presence of OMDM-2 (left, closed circles; n = 5; p < 0.01) or THL (right, closed circles; n = 7; p < 0.001). B, Time course of train-induced SE under normal conditions (open circles; n = 7) and in the presence of U73122 (closed circles; n = 7; p < 0.001). C, Magnitude of EPSC amplitude after the train for THL and U73122 plotted as the percentage of baseline before the train (n = 7; p < 0.01).

Figure 5.

Contribution of intracellular calcium, glutamate, and D₂DA receptors to 2-AG-induced effects. A, Magnitude of EPSC amplitude after the train for all conditions [thapsigargin (thapsi); ruthenium red (RR); ryanodine (ryano); n = 7] plotted as the percentage of baseline before the train (dotted line). B, Magnitude of EPSC amplitude after the train for all conditions (AP5, n = 7; MCPG, n = 7; CPCCOEt, n = 6; MPEP, n = 6; eticlopride, n = 7) plotted as the percentage of baseline before the train (dotted line). Representative traces for each condition from a single experiment are shown. The EPSC recorded after the train is superimposed in light gray for comparison. Calibration: 20 msec, 100 pA.