Acute cocaine exposure weakens GABA(B) receptor-dependent G-protein-gated inwardly rectifying K+ signaling in dopamine neurons of the ventral tegmental area

Abstract

Enhanced glutamatergic neurotransmission in dopamine (DA) neurons of the ventral tegmental area (VTA), triggered by a single cocaine injection, represents an early adaptation linked to the more enduring effects of abused drugs that characterize addiction. Here, we examined the impact of in vivo cocaine exposure on metabotropic inhibitory signaling involving G-protein-gated inwardly rectifying K(+) (Girk) channels in VTA DA neurons. Somatodendritic Girk currents evoked by the GABA(B) receptor (GABA(B)R) agonist baclofen were diminished in a dose-dependent manner in mice given a single cocaine injection. This adaptation persisted for 3-4 d, was specific for DA neurons of the VTA, and occurred in parallel with an increase in spontaneous glutamatergic neurotransmission. No additional suppression of GABA(B)R-Girk signaling was observed following repeated cocaine administration. While total Girk2 and GABA(B)R1 mRNA and protein levels were unaltered by cocaine exposure in VTA DA neurons, the cocaine-induced decrease in GABA(B)R-Girk signaling correlated with a reduction in Girk2-containing channels at the plasma membrane in VTA DA neurons. Systemic pretreatment with sulpiride, but not SCH23390 (7-chloro-3-methyl-1-phenyl-1,2,4,5-tetrahydro-3-benzazepin-8-ol), prevented the cocaine-induced suppression of GABA(B)R-Girk signaling, implicating D(2/3) DA receptor activation in this adaptation. The acute cocaine-induced weakening of somatodendritic Girk signaling complements the previously demonstrated cocaine-induced strengthening of glutamatergic neurotransmission, likely contributing to enhanced output of VTA DA neurons during the early stages of addiction.

Figure 1.

Cocaine suppresses inhibitory signaling in VTA DA neurons. a, Total distance traveled on test day by C57BL/6J mice (7–25 per group) following administration of saline (sal) or cocaine (3, 15, 30 mg/kg; F_(3,54) = 21.8; p < 0.001); activity was monitored for 1 h after injection. The diagram above the plot depicts the schedule of subject handling, which included 4 acclimation days, a test day, and three distinct time frames (1–2, 3–5, 10–12 d after injection) within which all electrophysiological assessments were made. b, Representative somatodendritic currents (V_hold = −60 mV) evoked by a saturating concentration of baclofen (200 μm) in VTA DA neurons from mice treated on test day with saline or cocaine (15 mg/kg). The lines above the traces denote the duration of drug application. Baclofen-induced currents were reversed by the GABA_BR antagonist CGP54626 (2 μm; CGP). c, Mean baclofen-induced currents in VTA DA neurons from saline- and cocaine-treated mice (n = 5–24 per group). The x-axis labels refer to the type of test day injection. The labels above the bars denote the time frame (in days after injection) of the recordings. Significant effects of cocaine dose (F_(3,57) = 7.2; p < 0.001), in addition to recording time frame (1–2, 3–5, 10–12 d) on baclofen-induced currents were observed among animals treated with 15 mg/kg cocaine (F_(3,59) = 8.0; p < 0.001). d, Mean peak currents (in picoamperes) triggered by increasing concentrations of baclofen (2, 20, 200 μm) applied to the same cell in slices from mice treated with either saline or 30 mg/kg cocaine (1–2 d after injection). Two-way ANOVA revealed significant main effects of drug (F_(1,45) = 15.1; p < 0.001) and baclofen concentration (F_(2,45) = 33.7; p < 0.001), but no drug by baclofen concentration interaction (F_(2,45) = 1.4; p = 0.3), consistent with the lack of impact of cocaine treatment on the sensitivity of GABA_BR–Girk signaling in VTA DA neurons. Maximum current amplitudes were slightly smaller than those measured in cells receiving a single saturating baclofen exposure, likely attributable to desensitization triggered by successive baclofen applications. e, Representative sEPSCs measured in VTA DA neurons 1–2 d after the injection of saline or cocaine. sEPSCs in control and cocaine-treated mice were blocked by kynurenic acid (2 mm). f, Mean sEPSC frequency (F_(4,61) = 15.5; p < 0.001) across the six groups (n = 4–23 per group). Data from saline-treated subjects were pooled as recording time frame did not affect currents, sEPSC amplitudes, or sEPSC frequencies. *p < 0.01, **p < 0.05, and ***p < 0.001, respectively, versus sal. Error bars indicate SEM.

Figure 2.

Repeated cocaine dosing decreases GABA_BR–Girk signaling in VTA DA neurons. a, Total distance traveled by mice (n = 5–17 per group) at three stages of a repeated cocaine dosing regimen, depicted left of the plot. Subjects were randomly assigned to saline (white) or cocaine (15 mg/kg; gray) groups; activity was assessed after injection of saline (s₂ acclimation day) and after the first and fifth cocaine injection days. Two-way ANOVA revealed a drug by day interaction (F_(2,56) = 26.1; p < 0.001). Only within-drug comparisons are displayed on the plot. ***p < 0.001 versus saline (s₂) activity; ^###p < 0.001. b, Representative currents evoked by 200 μm baclofen in VTA DA neurons from mice given repeated saline or cocaine (15 mg/kg) injections; recordings were made 1–2 d following the last injection. c, Mean baclofen-induced peak currents measured 1–2 and 10–12 d following the last injection of saline or cocaine (n = 6–14 per group; F_(2,25) = 5.4; p < 0.05). Data from saline-treated subjects were pooled as recording time frame did not affect baclofen-induced currents. **p < 0.01 versus saline group. Error bars indicate SEM.

Figure 3.

Acute cocaine decreases the density of Girk2 on the plasma membrane of VTA DA neurons. a–f, Representative images of GABA_BR1 and Girk2 immunoreactivity in TH-positive VTA neurons. Immunoreactivity for TH (a, d), GABA_BR1 (b), and Girk2 (e) in VTA-containing sections from a saline-treated mice. c, f, High-magnification views of the regions outlined in b and e, showing immunoreactivity for TH in red and either GABA_BR1 (c) or Girk2 (f) in green. The arrowheads show some examples of neurons exhibiting clear somatic labeling for TH and either GABA_BR1 or Girk2. Abbreviations: MN, Mammillary nuclei; PBP, parabrachial pigmented nucleus; SN, substantia nigra. Immunoreactivity for Girk2 in VTA DA neurons from saline- (g) or cocaine-treated (h) dopaminergic neurons of the VTA, as detected using preembedding dual-labeling immunoelectron microscopy. In control conditions, the peroxidase reaction product (TH immunoreactivity) filled dendritic shafts (Den), whereas immunoparticles (Girk2 immunoreactivity) were located along the extrasynaptic plasma membrane (arrows) and at intracellular sites (crossed arrows). A reduction in Girk2 immunoreactivity along the plasma membrane (arrows) was evident 24 h following an acute injection of 30 mg/kg cocaine. Scale bar, 0.5 μm.

Figure 4.

The cocaine-induced suppression of GABA_BR–Girk signaling in VTA DA neurons is D_2/3R dependent. a, Total distance traveled on test day by mice (6–25 per group) following administration of saline or cocaine (15 mg/kg), 30 min after pretreatment with saline (sal), the D_1/5 DA receptor (D_1/5R) antagonist SCH23390 (SCH) (0.18 mg/kg), or the D_2/3R antagonist sulpiride (sulp) (50 mg/kg). Two-way ANOVA revealed a significant pretreatment by drug (saline, cocaine) interaction (F_(2,61) = 4.7; p < 0.05). Only within-drug comparisons are shown. ^#p < 0.05 and ^###p < 0.001, respectively, versus sal pretreatment group. b, Representative currents evoked by 200 μm baclofen in VTA DA neurons from cocaine-treated mice pretreated with either SCH23390 or sulpiride. c, Mean baclofen-induced somatodendritic currents in neurons from saline- and cocaine-treated mice (n = 7–24 per group). All recordings were made 1–2 d after test day injection. Two-way ANOVA revealed main effects of pretreatment (F_(2,72) = 6.9; p < 0.01) and drug (F_(1,72) = 11.1; p < 0.01), but no interaction (F_(2,72) = 0.8; p = 0.5). As such, only within-drug comparisons are presented. Note that pretreatment condition had no significant effect on baclofen-induced currents among saline-treated mice. **p < 0.01 and ***p < 0.001, respectively. Error bars indicate SEM.