Inhibition of striatal dopamine release by CB1 receptor activation requires nonsynaptic communication involving GABA, H2O2, and KATP channels

Abstract

The main psychoactive component of marijuana, Delta9-tetrahydrocannabinol (THC), acts in the CNS via type 1 cannabinoid receptors (CB1Rs). The behavioral consequences of THC or synthetic CB1R agonists include suppression of motor activity. One explanation for movement suppression might be inhibition of striatal dopamine (DA) release by CB1Rs, which are densely localized in motor striatum; however, data from previous studies are inconclusive. Here we examined the effect of CB1R activation on locally evoked DA release monitored with carbon-fiber microelectrodes and fast-scan cyclic voltammetry in striatal slices. Consistent with previous reports, DA release evoked by a single stimulus pulse was unaffected by WIN55,212-2, a cannabinoid receptor agonist. However, when DA release was evoked by a train of stimuli, WIN55,212-2 caused a significant decrease in evoked extracellular DA concentration ([DA]o), implicating the involvement of local striatal circuitry, with similar suppression seen in guinea pig, rat, and mouse striatum. Pulse-train evoked [DA]o was not altered by either AM251, an inverse CB1R agonist, or VCHSR1, a neutral antagonist, indicating the absence of DA release regulation by endogenous cannabinoids with the stimulation protocol used. However, both CB1R antagonists prevented and reversed suppression of evoked [DA]o by WIN55,212-2. The effect of WIN55,212-2 was also prevented by picrotoxin, a GABAA receptor antagonist, and by catalase, a metabolizing enzyme for hydrogen peroxide (H2O2). Furthermore, blockade of ATP-sensitive K+ (KATP) channels by tolbutamide or glybenclamide prevented the effect of WIN55,212-2 on DA release. Together, these data indicate that suppression of DA release by CB1R activation within striatum occurs via a novel nonsynaptic mechanism that involves GABA release inhibition, increased generation of the diffusible messenger H2O2, and activation of KATP channels to inhibit DA release. In addition, the findings suggest a possible physiological substrate for the motor effects of cannabinoid agonist administration.

Figure 1. Effect of CB1R activation on DA release in dorsolateral striatum

A) Averaged DA release records after single-pulse stimulation under control conditions and in the presence of WIN55,212-2 (WIN; 5 μM) indicate no effect of WIN on either the amplitude or the time-course of evoked [DA]_o (n = 3). B) Effect of WIN (1-5 μM; see Results) on averaged [DA]_o evoked by pulse-train stimulation (30 pulses, 10 Hz) (n = 13). Inset, representative cyclic voltammograms for peak evoked [DA]_o at a given site under control conditions and in the presence of WIN; DA was identified by characteristic oxidation (ox) and reduction (red) peak potentials (typically +600 and -200 mV vs. Ag/AgCl). C) The effect of WIN (5 μM) persisted when the DA transporter (DAT) was inhibited by GBR-12909 (GBR; 2 μM) (n = 3). D) Activation of CB1Rs by WIN had no effect on single-pulse evoked [DA]_o (p > 0.05 WIN vs. same-site control; n = 3), but suppressed evoked [DA]_o during pulse-train stimulation (***p < 0.001 WIN vs. same-site control; n = 13) in ACSF alone, as well as when the DAT was inhibited by GBR (***p < 0.001 GBR + WIN vs. GBR alone; n = 3).

Figure 2. The effect of WIN55,212-2 on DA release is species independent

A) Activation of CB1Rs by WIN55,212-2 (WIN; 5 μM) caused a similar decrease in pulse-train evoked [DA]_o in the dorsolateral striatum of rats and mice to that seen in guinea pigs striatum (Fig. 1B). B) Comparison of the effect of WIN on average peak evoked [DA]_o in guinea pig (***p < 0.001 WIN vs. same-site control; n = 13), rat (***p < 0.001; n = 4), and mouse (***p < 0.001; n = 8) striatum.

Figure 3. CB1R antagonists prevent and reverse the effect of WIN55,212-2 on DA release, but have no effect alone

A) The effect of WIN55,212-2 (WIN; 1-5 μM) on pulse-train evoked [DA]_o was reversed by AM251 (1 μM) (**p < 0.01 WIN vs. control; ⁺⁺p < 0.01 WIN vs. WIN + AM251; n = 4) and VCHSR1 (2 μM) (**p < 0.01 WIN vs. control; ⁺⁺p < 0.01 WIN vs. WIN + VCHSR1; n = 3). B) Neither AM251 (1-5 μM) or VCHSR1 (2 μM) altered pulse-train evoked [DA]_o, indicating the absence of endocannabinoid regulation of DA release under these conditions. AM251 and VCHSR1 also prevented the effect of WIN (1-5 μM) (*p > 0.05 for all comparisons; n = 3-4 per mean).

Figure 4. CB1R-dependent suppression of DA release is mediated by GABA release inhibition

A) Averaged evoked [DA]_o records showing a significant decrease in pulse-train evoked [DA]_o when GABA_ARs are blocked by pictrotoxin (PTX, 100 μM, n = 5) (left panel). The usual effect of WIN55,212-2 was prevented in the continued presence of PTX (right panel) (n = 5). B) Representative peak evoked [DA]_o versus time profiles showing the effect of WIN55,212-2 (WIN; 5 μM) alone (open circles) and WIN applied in the presence of PTX (100 μM) (filled circles) unmasking an increase in evoked [DA]_o with WIN when GABA_ARs were blocked. Baseline data were normalized to respective control conditions (ACSF or PTX; see Results). C) Comparison of mean peak evoked [DA]_o in the presence of PTX, and PTX +WIN confirmed a slight, but significant increase in evoked [DA]_o with WIN application in the presence PTX (***p < 0.001 control vs. PTX; *p < 0.05 PTX vs. PTX+WIN; n = 5); all means are normalized to control (ACSF alone).

Figure 5. The effect of CB1R activation on evoked [DA]_o is prevented by the H₂O₂-metabolizing enzyme catalase

A) Averaged pulse-train evoked [DA]_o records in the presence of the H₂O₂ metabolizing enzyme, catalase (Cat; 500 U/mL) and in catalase + WIN55,212-2 (WIN; 2 μM).an in heat-inactivated catalase (iCat) and iCat + WIN. B) Catalase prevented the effect of WIN (p > 0.05 Cat vs. Cat + WIN; n = 4), whereas the usual WIN-induced suppression of evoked [DA]_o was seen in iCat (***p < 0.001 vs. same-site control; n = 3).

Figure 6. K_ATP-channel activation is required for CB1R-dependent suppression of DA release

A) Averaged pulse-train evoked [DA]_o records in the presence of a K_ATP channel blocker tolbutamide (Tolb, 200 μM, n = 4), or glybenclamide (Glyb, 10 μM, n = 5) and in Tolb or Glyb + WIN55,212-2 (up to 5 μM); K_ATP channel blockade prevented the effect of WIN on evoked [DA]_o. B) Comparison of the effect of WIN on average peak evoked [DA]_o under control conditions (***p < 0.001; n = 13), in tolbutamide (p > 0.05, n = 4), and in glybenclamide (p > 0.05, n = 5).

Figure 7. Schematic diagram showing the proposed mechanism for the effect of CB1R-dependent suppression of evoked DA release

Glutamate, GABA, and DA synapses converging on a medium spiny neuron (MSN) dendrite. Left panel: AMPA receptor activation by glutamate on a dendritic spine results in increased mitochondrial H₂O₂ production, which diffuses to dopaminergic terminals to open K_ATP channels and thereby decrease DA release. This process is attenuated by activation of GABA_ARs. The amount of H₂O₂ produced is therefore determined by the net effect of glutamatergic excitation and GABAergic inhibition. Right panel: Activation of CB1Rs on GABAergic terminals by the agonist WIN55,212-2 results in the suppression of GABA release, and a consequent facilitation of H₂O₂ production. As a result, more K_ATP channels will be opened on dopaminergic terminals, causing suppression of evoked DA release.