In Vivo Ethanol Experience Increases D₂ Autoinhibition in the Ventral Tegmental Area

Abstract

Alcoholism is characterized by compulsive alcohol intake after a history of chronic consumption. A reduction in mesolimbic dopaminergic transmission observed during abstinence may contribute to the negative affective state that drives compulsive intake. Although previous in vivo recording studies in rodents have demonstrated profound decreases in the firing activity of ventral tegmental area (VTA) dopamine neurons after withdrawal from long-term ethanol exposure, the cellular mechanisms underlying this reduced activity are not well understood. Somatodendritic dopamine release within the VTA exerts powerful feedback inhibition of dopamine neuron activity via stimulation of D₂ autoreceptors and subsequent activation of G protein-gated inwardly rectifying K⁺ (GIRK) channels. Here, by performing patch-clamp recordings from putative dopamine neurons in the VTA of mouse brain slices, we show that D₂ receptor/GIRK-mediated inhibition becomes more potent and exhibits less desensitization after withdrawal from repeated in vivo ethanol exposure (2 g/kg, i.p., three times daily for 7 days). In contrast, GABA_B receptor/GIRK-mediated inhibition and its desensitization are not affected. Chelating cytosolic Ca²⁺ with BAPTA augments D₂ inhibition and suppresses its desensitization in control mice, while these effects of BAPTA are occluded in ethanol-treated mice. Furthermore, inositol 1,4,5-trisphosphate (IP₃)-induced intracellular Ca²⁺ release and Ca²⁺/calmodulin-dependent protein kinase II are selectively involved in the desensitization of D₂, but not GABA_B, receptor signaling. Consistent with this, activation of metabotropic glutamate receptors that are coupled to IP₃ generation leads to cross-desensitization of D₂/GIRK-mediated responses. We propose that enhancement of D₂ receptor-mediated autoinhibition via attenuation of a Ca²⁺-dependent desensitization mechanism may contribute to the hypodopaminergic state during ethanol withdrawal.

INTRODUCTION

The mesolimbic dopamine system, which originates in the ventral tegmental area (VTA) and projects to the nucleus accumbens (NAc) and other limbic structures, is critically involved in reward processing, behavioral reinforcement, and addictive behaviors. Acute exposure to ethanol, the principal psychoactive constituent of all alcoholic beverages, stimulates mesolimbic dopaminergic transmission (Boileau et al, 2003; Imperato and Di Chiara, 1986; Weiss et al, 1993), primarily via direct excitation of VTA dopamine neuron firing activity (Brodie et al, 1990; Gessa et al, 1985). This effect is believed to mediate the rewarding and reinforcing actions of ethanol. In contrast, profound decreases in NAc dopamine release have been observed in animals withdrawn from long-term ethanol exposure (Rossetti et al, 1992; Weiss et al, 1996) and also in detoxified alcoholics (Martinez et al, 2005; Volkow et al, 2007). In line with these observations, a marked reduction in VTA dopamine neuron activity has been reported in ethanol-withdrawn animals (Diana et al, 1993, 1996). This ‘hypodopaminergic state’, which can be reversed by administration of ethanol itself (Diana et al, 1993; Weiss et al, 1996), may, at least partially, contribute to the emotional/motivational component of ethanol withdrawal symptoms, such as dysphoria and anhedonia (Koob and Volkow, 2010; Melis et al, 2005; Trevisan et al, 1998). However, the cellular mechanisms underlying the hypoactivity of dopamine neurons after ethanol withdrawal remain poorly understood.

It is well known that activation of dopamine D₂ receptors and GABA_B receptors can mediate powerful inhibition of dopamine neurons both in vitro (Beckstead et al, 2004; Labouebe et al, 2007; Lacey et al, 1987) and in vivo (Erhardt et al, 2002; White and Wang, 1984). These receptors are coupled to the activation of G protein-gated inwardly rectifying K⁺ (GIRK) channels via G_i/o G proteins, resulting in membrane hyperpolarization and suppression of neuronal excitability (Luscher and Slesinger, 2010). In vivo recording studies have shown that VTA DA neuron firing activity is tonically inhibited by these receptors (Erhardt et al, 2002; White and Wang, 1984), raising the possibility that enhancement of D₂ and/or GABA_B inhibition might contribute to the tonic hypoactivity of dopamine neurons observed in ethanol-withdrawn animals.

Previous studies have shown that D₂ autoreceptor-mediated inhibition of VTA dopamine neurons is reduced after repeated psychostimulant exposure (Henry et al, 1989; Marinelli et al, 2003; Wolf et al, 1993), whereas the GABA_B receptor-GIRK channel coupling efficiency is increased after repeated exposure to morphine or the GABA_B agonist γ-hydroxybutyrate (Labouebe et al, 2007). In this study, by performing patch-clamp recordings from putative dopamine neurons in the VTA of mouse brain slices, we show that repeated in vivo exposure to ethanol induces sensitization of D₂-mediated inhibition without affecting GABA_B-mediated inhibition. Consistent with this differential modulation, the D₂ receptor signaling is uniquely regulated by a Ca²⁺-dependent desensitization mechanism involving inositol 1,4,5-trisphosphate (IP₃)-induced Ca²⁺ release from intracellular stores and subsequent activation of Ca²⁺/calmodulin-dependent protein kinase II (CaMKII). Furthermore, in vivo ethanol treatment occludes the enhancement of D₂ inhibition and suppression of its desensitization produced by the Ca²⁺ chelator BAPTA observed in control mice, suggesting that the Ca²⁺-dependent desensitization machinery may be suppressed by repeated ethanol exposure.

SUBJECTS AND METHODS

Subjects

Male C57BL/6J mice (3–4 weeks old; Jackson Laboratory) were housed under a 12-h light–dark cycle (lights on at 0700 hours). Food and water were available ad libitum. All animal procedures were approved by the University of Texas Institutional Animal Care and Use Committee.

In Vivo Ethanol Treatment

Mice received three times daily i.p. injections of saline or ethanol (2 g/kg, 20% v/v in saline) for 7 days. It should be noted that previous studies reporting reduced dopamine neuron firing in vivo after ethanol withdrawal used comparable ethanol administration protocol (2–5 g/kg, intragastric, four times daily for 6 days) (Diana et al, 1993, 1995). Midbrain slices were prepared 1 day after the final injection. This ethanol treatment protocol produced an increase in the amount of ethanol consumption measured using a 24-h continuous access two-bottle choice paradigm (Supplementary Figure 1).

Slices and Solutions

Mice were killed by cervical dislocation under halothane or isoflurane anesthesia, and horizontal midbrain slices (200–210 μm) were prepared using a vibratome (VT1000S; Leica Microsystems). Slices were routinely cut in ice-cold physiological saline containing (in mM): 126 NaCl, 2.5 KCl, 1.2 NaH₂PO₄, 1.2 MgCl₂, 2.4 CaCl₂, 11 glucose, 21.4 NaHCO₃, saturated with 95% O₂, and 5% CO₂ (pH 7.4, ∼295–300 mOsm/kg). MK-801 was added to this solution to prevent NMDA receptor-mediated excitotoxicity. For experiments performing loose-patch recordings of action potential firing, slices were cut in high-sucrose cutting solution containing (in mM): 205 sucrose, 2.5 KCl, 1.25 NaH₂PO₄, 7.5 MgCl₂, 0.5 CaCl₂, 10 glucose, 25 NaHCO₃, saturated with 95% O₂, and 5% CO₂ (∼305 mOsm/kg). Slices were incubated at 35 °C for >1 h in the physiological saline described above. Recordings were made at 34–35 °C in the same physiological saline perfused at 2–3 ml/min. Unless noted otherwise, the pipette solution contained (in mM): 115 K-gluconate or K-methylsulfate, 20 KCl, 1.5 MgCl₂, 10 HEPES, 0.025 EGTA, 2 Mg-ATP, 0.2 Na₂-GTP, and 10 Na₂-phosphocreatine (pH 7.25, ∼280 mOsm/kg).

Electrophysiological Recordings

Cells were visualized using a 40 × objective on an upright microscope (BX51WI; Olympus) with infrared/oblique illumination optics. Putative dopamine neurons were identified by low-frequency (<4 Hz) pacemaker firing with broad action potentials (>1.2 ms) in cell-attached configuration before break-in and the presence of large whole-cell I_h currents (>200 pA in response to a 1.5-s voltage step from −62 to −112 mV) after break-in. Recordings were restricted to the lateral part of the VTA within ∼150 μm from the medial border of the medial terminal nucleus of the accessory optic tract (MT) in horizontal midbrain slices, where dopamine neurons projecting to the core and lateral shell regions of the NAc are present (Ikemoto, 2007). It should be noted that reduced dopamine release after ethanol withdrawal has been detected in both the core and shell of the NAc (Weiss et al, 1996). Although electrophysiological identification of dopamine neurons has been challenged (Margolis et al, 2006), the criteria described above have been used in recent studies performing recordings from the lateral part of the VTA in horizontal slices (Ahn et al, 2010; Beckstead et al, 2009; Ford et al, 2006; Riegel and Williams, 2008; Wanat et al, 2008; Zhang et al, 2010).

Whole-cell voltage-clamp recordings were routinely made at a holding potential of −62 mV, corrected for a liquid junction potential of −7 mV. Pipettes had an open tip resistance of 1.7–2.3 MΩ when filled with the pipette solution. Series resistance after break-in was typically ∼5–10 MΩ and was periodically monitored during the recording. Recordings were discarded if the series resistance increased beyond 20 MΩ. A Multiclamp 700A or 700B amplifier (Molecular Devices) was used to record the data, which were filtered at 1 kHz, digitized at 2 kHz, and collected using AxoGraph X.

Loose-patch recordings (∼10–20 MΩ seal) of action potential firing were made using glass pipettes filled with 150 mM NaCl (2.0–2.5 MΩ tip resistance). Firing data were filtered at 5 kHz and digitized at 10 kHz. In these recordings, whole-cell break-in was not made afterward to measure I_h currents, and thus low-frequency pacemaker firing and action potential width (>1.2 ms) were the criteria to identify putative dopamine neurons, in accordance with (Ford et al, 2006).

Pressure Ejection of Dopamine

Patch pipettes (∼2–3 μm tip diameter), filled with dopamine (100 μM) and ascorbate (1.3 mM), were placed ∼100 μm from recorded cells. Pressure of 20 p.s.i. was applied to rapidly eject dopamine. The duration of pressure application (10–150 ms) was adjusted to obtain dopamine-evoked outward currents of ∼50 pA in each cell.

Drugs

Quinpirole, (RS)-baclofen, CGP54626, U73122, 2-aminoethoxydiphenylborane (2-APB), cyclopiazonic acid (CPA), and (S)-3,5-dihydroxyphenylglycine (DHPG) were obtained from Tocris Bioscience. Heparin and KN-62 were purchased from Calbiochem. All other chemicals used in electrophysiology experiments were from Sigma-RBI.

Data Analyses

In analyzing quinpirole/baclofen-induced outward currents, holding currents during two ∼5-min periods, one immediately before quinpirole/baclofen application and one after reversal of quinpirole/baclofen-induced currents, were fitted to a single straight line, and this line was subtracted from recorded currents to correct for gradual shift in holding current levels frequently observed during whole-cell recordings. Quinpirole/baclofen-induced currents thus obtained were normalized by the membrane capacitance in each cell to estimate current density (expressed in pA/pF). The membrane capacitance was estimated from the fast component of double-exponential fit to the decay of capacitative transients (filtered at 10 kHz and digitized at 20 kHz) evoked by 10-mV hyperpolarizing voltage steps. In vivo saline/ethanol treatments did not affect the membrane capacitance thus estimated in cells reported in this study (naïve: 57.5±1.2 pF, n=63; saline-treated: 57.3±1.1 pF, n=50; ethanol-treated: 59.0±1.4 pF, n=39; F_2, 149=0.52, p=0.59, one-way ANOVA). The difference between the initial peak current amplitude and the current amplitude at the end of continuous agonist application was divided by the peak current amplitude to obtain the magnitude of desensitization (expressed in %).

Data are expressed as mean±SEM. Statistical significance was determined by Student's t-test or ANOVA followed by Bonferroni post hoc test. The difference was considered significant at p<0.05.

RESULTS

D₂ Receptor-Mediated Inhibition is Enhanced after Repeated In Vivo Ethanol Exposure

To test if in vivo ethanol exposure alters D₂ autoreceptor-mediated inhibition, we performed whole-cell voltage-clamp recordings from naïve C57BL/6J mice and from mice that received injections of saline or ethanol (2 g/kg, i.p.) three times daily for 7 days. Recordings were made in midbrain slices prepared 1 day after the final injection. Putative dopamine neurons were identified electrophysiologically (see Subjects and Methods section). Bath application of the D₂ agonist quinpirole (300 nM) produced outward currents that reached peak amplitude in 1–2 min. These currents gradually desensitized during 10 min of continuous quinpirole application and were rapidly reversed by the D₂ antagonist sulpiride (1–2 μM) (Figure 1a). Application of sulpiride by itself elicited no measurable currents (three cells each from saline- and ethanol-treated mice), suggesting the absence of effective dopamine tone in brain slice preparations used in this study. Quinpirole-induced currents exhibited larger peak amplitude and smaller desensitization in ethanol-treated mice compared with naïve or saline-treated mice (peak amplitude: F_2, 34=6.23, p<0.01; desensitization: F_2, 22=15.7, p<0.0001; one-way ANOVA) (Figures 1a–c). Furthermore, concentration-response curves for peak currents produced by quinpirole (3 nM–30 μM) revealed an approximately twofold increase in quinpirole sensitivity, with no change in the maximal response, in ethanol-treated mice compared with saline-treated mice (Figure 1d). Therefore, D₂ receptor-GIRK channel coupling is more potent and also becomes resistant to desensitization after repeated ethanol exposure.

Figure 1

D₂ receptor-mediated outward currents are increased and exhibit less desensitization after in vivo ethanol exposure. (a) Examples of quinpirole-induced outward currents (V_h=−62 mV) in naïve, saline-treated, and ethanol-treated mice. Quinpirole (300 nM) was bath applied for 10 min. Subsequent application of sulpiride (1 μM) reversed quinpirole-induced currents. (b) Summary bar graph showing the peak amplitude of quinpirole-induced outward currents in naïve (n=12 from 8 mice), saline-treated (n=15 from 12 mice), and ethanol-treated mice (n=10 from 9 mice). (c) Summary bar graph plotting the amount of desensitization of quinpirole-induced currents in naïve (n=8 from 5 mice), saline-treated (n=10 from 8 mice), and ethanol-treated mice (n=7 from 7 mice). (d) Concentration-response curves for quinpirole-induced outward currents in saline- and ethanol-treated mice. Each point represents an average of data from 4 to 15 cells. Averaged data points are fitted to a logistic equation. The estimated EC₅₀ values are 100 and 43 nM for saline- and ethanol-treated mice, while the estimated maximal responses are 1.73 and 1.76 pA/pF for saline- and ethanol-treated mice. ^*p<0.05, ^**p<0.01, ^***p<0.001, Bonferroni post hoc test. Error bars indicate SEM.

PowerPoint slide

We next examined the effect of quinpirole on the firing activity of VTA dopamine neurons monitored with a loose-patch configuration. The basal firing frequency was not altered by in vivo ethanol treatment (t₅₄=0.17, p=0.46; unpaired t-test) (Figure 2a). Furthermore, sulpiride (1–2 μM) had no effect on the firing frequency (saline-treated: 1.69±0.43 Hz in control vs 1.65±0.43 Hz in sulpiride, n=5, t₄=1.36, p=0.25; ethanol-treated: 1.75±0.25 Hz in control vs 1.72±0.25 Hz in sulpiride, n=6, t₅=1.04, p=0.35; paired t-test), consistent with the lack of effect of sulpiride in whole-cell voltage-clamp recordings described above. However, bath application of low concentrations of quinpirole (10 and 30 nM) caused significantly larger inhibition of firing in ethanol-treated mice (Figures 2b–d). Here, quinpirole was tested only in cells that exhibited firing frequency in the range of 1.5–3 Hz, because the magnitude of D₂/GIRK-mediated inhibition of firing has been shown to depend on baseline firing frequency (Putzier et al, 2009; Werkman et al, 2001). The firing frequency in the presence of quinpirole was significantly lower in ethanol-treated mice compared with saline-treated controls, although the firing frequency before quinpirole application was comparable in the two groups in these experiments (Figure 2c). A higher concentration of quinpirole (100 nM) invariably abolished dopamine neuron firing in eight cells tested (four cells each from saline- and ethanol-treated mice). These results demonstrate that dopamine neuron firing is more sensitive to D₂ inhibition after repeated ethanol exposure.

Figure 2

Dopamine neuron firing is more sensitive to D₂ inhibition in ethanol-treated mice. (a) Basal firing frequency in individual cells is plotted from saline- and ethanol-treated mice (n=23 from 11 mice in saline group and n=25 from 14 mice in ethanol group). (b) Representative time graphs illustrating quinpirole-induced inhibition of dopamine neuron firing in saline- and ethanol-treated mice. Quinpirole (10 nM) was applied at the time indicated by the horizontal bar. Each point represents average firing frequency during a 60-s period. Example traces (10 s) of dopamine neuron firing in control and in quinpirole are shown in the inset. (c) Summary graphs plotting the firing frequency in control and in quinpirole for individual cells in saline- and ethanol-treated mice. Left: 6 cells from 5 mice in saline group and 4 cells from 4 mice in ethanol group were tested for 10 nM quinpirole (in vivo treatment: F_1, 8=3.30, p=0.11, quinpirole: F_1, 8=74.3, p<0.0001, in vivo treatment × quinpirole: F_1, 8=15.9, p<0.01; mixed two-way ANOVA). Right: 4 cells from 3 mice in saline group and 5 cells from 4 mice in ethanol group were tested for 30 nM quinpirole (in vivo treatment: F_1, 7=4.48, p=0.07, quinpirole: F_1, 7=89.9, p<0.0001, in vivo treatment × quinpirole: F_1, 7=9.13, p<0.05; mixed two-way ANOVA). (d) Summary bar graph demonstrating that quinpirole caused larger firing inhibition in ethanol-treated mice. Data are from the same cells as in c (in vivo treatment: F_1, 15=23.5, p<0.001, quinpirole concentration: F_1, 15=37.3, p<0.0001, in vivo treatment × quinpirole concentration: F_1, 15=1.71, p=0.21; two-way ANOVA). ^*p<0.05, ^**p<0.01, ^***p<0.001, Bonferroni post hoc test. Error bars indicate SEM.

PowerPoint slide

In Vivo Ethanol Exposure Does Not Affect GABA_B Receptor-Mediated Inhibition

D₂ receptors and GABA_B receptors most likely share the same downstream signaling pathway (Beckstead et al, 2004; Labouebe et al, 2007; Luscher and Slesinger, 2010). Thus, we next tested the GABA_B agonist baclofen, which has been shown to produce desensitizing GIRK-mediated currents with an EC₅₀ of ∼10 μM in VTA dopamine neurons while causing non-desensitizing GIRK-mediated currents in VTA GABA neurons (Cruz et al, 2004; Labouebe et al, 2007). Bath application of baclofen (10 μM) induced large outward currents associated with desensitization in both saline- and ethanol-treated mice (Figure 3a). Reponses elicited by baclofen (1 and 10 μM) were comparable in saline- and ethanol-treated mice in terms of both peak outward current amplitude (in vivo treatment: F_1, 33=0.42, p=0.52, baclofen concentration: F_1, 33=126, p<0.0001, in vivo treatment × baclofen concentration: F_1, 33=0.47, p=0.50; two-way ANOVA) and amount of desensitization (in vivo treatment: F_1, 12=0.20, p=0.67, baclofen concentration: F_1, 12=74.0, p<0.0001, in vivo treatment × baclofen concentration: F_1, 12=0.0029, p=0.96; two-way ANOVA) (Figures 3b and c). Thus, GABA_B receptor-mediated inhibition is not affected by ethanol treatment. This finding suggests that the enhancement of D₂ inhibition after in vivo ethanol exposure results from regulation at the level of D₂ receptors and not at downstream signaling components (G_i/o proteins or GIRK channels).

Figure 3

GABA_B receptor-mediated outward currents are not altered by in vivo ethanol exposure. (a) Examples of baclofen-induced outward currents in saline- and ethanol-treated mice. Baclofen (10 μM) was bath applied for 15 min, and then the GABA_B antagonist CGP54626 (200 nM) was applied to reverse baclofen-induced currents. (b) Summary bar graph showing the peak outward current amplitude produced by baclofen (1 and 10 μM) in saline- and ethanol-treated mice (1 μM: n=14 from 11 mice in saline group and n=8 from 6 mice in ethanol group; 10 μM: n=8 from 7 mice in saline group and n=7 from 6 mice in ethanol group). (c) Summary bar graph depicting the amount of desensitization of baclofen-induced currents in saline- and ethanol-treated mice (1 μM: n=3 from 3 mice in saline group and n=3 from 2 mice in ethanol group; 10 μM: n=5 from 4 mice in saline group and n=5 from 5 mice in ethanol group). Error bars indicate SEM.

PowerPoint slide

Regulation of D₂ Receptor-Mediated Inhibition by a Ca²⁺-Dependent Desensitization Mechanism

It has been shown that buffering cytosolic Ca²⁺ with BAPTA suppresses desensitization of D₂ receptor-mediated outward currents in dopamine neurons (Beckstead and Williams, 2007). Therefore, we loaded recorded cells with BAPTA (5 mM) through the whole-cell pipette to test its effect on quinpirole-induced currents in saline- and ethanol-treated mice. Intracellular BAPTA significantly increased the peak amplitude of quinpirole-induced outward currents and reduced the amount of desensitization in saline-treated mice (Figure 4). In contrast, BAPTA failed to significantly affect quinpirole-induced currents in ethanol-treated mice, although the peak current amplitude was slightly increased. Furthermore, Ca²⁺ chelation with BAPTA eliminated the difference in quinpirole responses between saline- and ethanol-treated mice (peak amplitude, BAPTA: F_1, 34=14.3, p<0.001, in vivo treatment: F_1, 34=1.48, p=0.23, BAPTA × in vivo treatment: F_1, 34=1.49, p=0.23; desensitization, BAPTA: F_1, 24=8.38, p<0.01, in vivo treatment: F_1, 24=15.3, p<0.001, BAPTA × in vivo treatment: F_1, 24=6.58, p<0.05; two-way ANOVA). These results suggest that attenuation of a Ca²⁺-dependent desensitization mechanism is responsible for the enhancement of D₂ receptor-mediated inhibition after in vivo ethanol exposure.

Figure 4

Chelating cytosolic Ca²⁺ occludes the enhancement of D₂ receptor-mediated currents caused by ethanol exposure. (a) Examples of quinpirole-induced currents in the presence of BAPTA in saline- and ethanol-treated mice. Quinpirole (300 nM) and sulpiride (1 μM) were applied as indicated by horizontal bars. (b) Summary bar graph showing the effect of BAPTA on the peak amplitude of quinpirole-induced currents in saline group (control: n=15 from 12 mice, BAPTA: n=7 from 4 mice) and in ethanol group (control: n=10 from 9 mice, BAPTA: n=6 from 3 mice). (c) Summary bar graph depicting the effect of BAPTA on desensitization of quinpirole-induced currents in saline group (control: n=10 from 8 mice, BAPTA: n=6 from 4 mice) and in ethanol group (control: n=7 from 7 mice, BAPTA: n=5 from 3 mice). ^**p<0.01, ^***p<0.001, Bonferroni post hoc test. Error bars indicate SEM.

PowerPoint slide

We further explored the intracellular pathways mediating Ca²⁺-dependent desensitization of D₂ receptor signaling in VTA dopamine neurons from naïve mice. D₂ receptor activation may cause release of Ca²⁺ from intracellular stores via the phospholipase C (PLC)/IP₃-mediated pathway (Hernandez-Lopez et al, 2000; Hu et al, 2005; Takeuchi et al, 2002; Vallar et al, 1990). Thus, we examined the effects of pharmacological treatments that interfere with this signaling pathway on outward currents produced by 10-min perfusion of quinpirole (300 nM) (Figure 5). First, the PLC inhibitor U73122 (1 μM, intracellular dialysis through the whole-cell pipette) significantly diminished the desensitization of quinpirole-induced currents (Figures 5a and b). IP₃ receptor antagonists heparin (1 mg/ml, intracellular dialysis) or 2-APB (30 μM, bath application, >10 min pretreatment and present throughout quinpirole application) also significantly attenuated quinpirole-induced desensitization. Furthermore, bath application of CPA (10–20 μM, >10 min pretreatment and present throughout quinpirole application), which depletes intracellular Ca²⁺ stores (Seidler et al, 1989), suppressed quinpirole-induced desensitization. All of these treatments produced ∼50% increases in the average peak amplitude of quinpirole-induced currents compared with the value obtained under control conditions (Figure 5c).

Figure 5

Blocking the PLC-IP₃-CaMKII cascade suppresses desensitization of D₂ receptor-mediated currents. (a) Examples of quinpirole-induced currents in cells loaded with U73122 (1 μM), CaMKII inhibitory peptide (fragment 290–309; 25 μM), or PKC inhibitory peptide (fragment 19–31; 25 μM). Quinpirole (300 nM) and sulpiride (1 μM) were applied as indicated by horizontal bars. (b, c) Summary bar graphs showing the effects of various intracellular Ca²⁺ signaling blockers on desensitization (b) and peak amplitude (c) of quinpirole-induced currents. Each drug was tested in 3–7 cells, while the control data are the same as those from naïve mice shown in Figures 1b and c (desensitization: F_7, 35=9.47, p<0.0001; peak amplitude: F_7, 41=1.61, p=0.16; peak amplitude: F_7, 41=1.61, p=0.16; one-way ANOVA). ^**p<0.01, ^***p<0.001 vs control, Bonferroni post hoc test. Error bars indicate SEM.

PowerPoint slide

D₂ receptor-induced Ca²⁺ rises may lead to activation of CaMKII (Takeuchi et al, 2002) or protein kinase C (PKC) (Gordon et al, 2001). The CaMKII inhibitor KN-62 (10 μM, >1-h preincubation plus intracellular dialysis) and CaMKII inhibitory peptide (fragment 290–309; 25 μM, intracellular dialysis) significantly reduced the amount of quinpirole-induced desensitization, whereas PKC inhibitory peptide (fragment 19–31; 25 μM, intracellular dialysis) had no significant effect (Figures 5a and b). Altogether, these results suggest that PLC/IP₃/CaMKII-mediated pathway is responsible for Ca²⁺-dependent desensitization of D₂ receptor signaling in dopamine neurons.

Cross-Desensitization of D₂ Receptor-Mediated Currents by Activation of mGluRs

Metabotropic glutamate receptors (mGluRs) are coupled to the generation of IP₃ in dopamine neurons (Cui et al, 2007). Thus, we next examined if mGluR activation causes cross-desensitization of D₂ receptor-mediated outward currents. In these experiments, pressure ejection of dopamine (100 μM) was made every 90 s. Transient outward currents thus evoked were abolished by sulpiride (1 μM; n=3) (Figure 6a), confirming that they are mediated by D₂ receptors. Bath application of the mGluR agonist DHPG (1 μM) produced a reversible inhibition of dopamine-evoked currents (29±4% inhibition, n=5) (Figure 6b). To examine the role of IP₃-mediated Ca²⁺ release in the DHPG effect, we depleted intracellular Ca²⁺ stores with CPA (20 μM). Bath application of CPA by itself dramatically augmented dopamine-evoked currents (113±32% increase, n=5) (Figure 6c), suggesting that D₂ receptor signaling is constitutively desensitized by constant Ca²⁺ release from intracellular stores. DHPG failed to affect dopamine-evoked currents when the cells were pretreated with CPA (−4±5% change). Therefore, mGluR activation leads to cross-desensitization of D₂ receptor signaling in a Ca²⁺ store-dependent manner.

Figure 6

mGluR activation cross-desensitizes D₂ receptor-mediated outward currents. (a) Sample traces depicting that dopamine-induced outward currents were abolished by sulpiride (1 μM). Pressure ejection of dopamine (100 μM) was made at the time indicated by the arrow. (b) Sample traces and summary time graph showing inhibition of dopamine-evoked currents by bath application of DHPG (1 μM). (c) Sample traces and summary time graph demonstrating that the DHPG effect on dopamine-evoked currents was suppressed by pretreatment with CPA (20 μM). Note that CPA itself augmented dopamine-induced currents. Error bars indicate SEM.

PowerPoint slide

GABA_B Receptor Signaling is Not Regulated by Ca²⁺-Dependent Desensitization

It has been shown that chelating intracellular Ca²⁺ with BAPTA does not affect desensitization of GABA_B receptor-mediated outward currents in dopamine neurons (Beckstead and Williams, 2007). Consistent with this, blocking IP₃ receptors with heparin or depleting Ca²⁺ stores with CPA had no significant effects on the peak amplitude or the amount of desensitization of baclofen-induced outward currents (peak amplitude: F_2, 8=0.38, p=0.69; desensitization: F_2, 7=0.46, p=0.65; one-way ANOVA) (Figure 7). Thus, the Ca²⁺-dependent mechanism mediating D₂ receptor desensitization does not regulate GABA_B receptor signaling in dopamine neurons.

Figure 7

GABA_B receptor-mediated currents are not regulated by Ca²⁺-dependent desensitization. (a) An example of baclofen-induced current in a cell loaded with heparin (1 mg/ml). Baclofen (10 μM) and CGP54626 (200 nM) were applied as indicated by horizontal bars. (b, c) Summary bar graphs showing that both CPA and heparin failed to affect peak amplitude (b) and desensitization (c) of baclofen-induced currents. Baclofen was applied for 10 min in these experiments. Three to four cells were tested for each condition. Error bars indicate SEM.

PowerPoint slide

DISCUSSION

In this study, we found that repeated in vivo exposure to ethanol results in an enhancement of D₂ autoreceptor-mediated inhibition of VTA dopamine neurons. This enhancement results from downregulation of a Ca²⁺-dependent desensitization mechanism comprising IP₃-induced Ca²⁺ release from intracellular stores and activation of CaMKII. In contrast, GABA_B receptor-mediated inhibition is not regulated by this Ca²⁺-dependent desensitization mechanism, and hence is not altered by repeated ethanol exposure. Upregulation of dopamine autoinhibition may have an important role in regulating the mesolimbic dopaminergic output during ethanol withdrawal in vivo.

Ca²⁺-Dependent Desensitization of D₂ Receptor-Mediated Inhibition

A previous study has demonstrated that chelating cytosolic Ca²⁺ with BAPTA in dopamine neurons results in an increase in the amplitude of D₂ receptor-mediated outward currents together with a decrease in the amount of desensitization (Beckstead and Williams, 2007). In that study, desensitization was induced either via sustained stimulation of dendritic dopamine release, which caused long-term depression of D₂ IPSCs, or via prolonged application of dopamine. Our data also show that BAPTA significantly increased the peak amplitude of quinpirole-induced outward currents and suppressed their desensitization in VTA dopamine neurons from saline-treated mice. These observations indicate the role of cytosolic Ca²⁺ in regulating basal, as well as agonist induced, desensitization of D₂ receptor signaling. However, depletion of intracellular Ca²⁺ stores with CPA had no effect on D₂ receptor desensitization in the study by Beckstead et al, whereas CPA largely suppressed desensitization of quinpirole-induced currents in this study. Although the reason for this discrepancy is not clear, it may be accounted for, at least partially, by sampling of different populations of dopamine neurons (mostly from the substantia nigra in the study by Beckstead et al vs from the lateral VTA in our study). Indeed, it has been shown that different populations of dopamine neurons in different areas of the ventral midbrain have distinct functional and chemical properties, such as the expression levels of Ca²⁺-binding proteins (Lammel et al, 2008; Neuhoff et al, 2002).

Our data suggest that D₂ receptor signaling is constitutively desensitized by constant Ca²⁺ release from intracellular stores under basal conditions, because Ca²⁺ store depletion by CPA, similarly to Ca²⁺ chelation by BAPTA, increased the peak amplitude of D₂-mediated outward currents. A number of studies have demonstrated that D₂ receptor activation leads to Ca²⁺ release from intracellular stores via the G_i/o/PLC/IP₃-mediated pathway (Hernandez-Lopez et al, 2000; Hu et al, 2005; Takeuchi et al, 2002; Vallar et al, 1990). In line with this, blockade of PLC or IP₃ receptors significantly attenuated quinpirole-induced desensitization. Thus, D₂ receptor-induced rises in cytosolic Ca²⁺ levels, in addition to basal levels, may mediate the desensitization observed during continuous D₂ receptor activation. Although a rise in Ca²⁺ levels following iontophoretic application of dopamine has not been observed in dopamine neurons (Beckstead and Williams, 2007), small but continuous Ca²⁺ release, which cannot be readily detected with fluorescence Ca²⁺ imaging, may be responsible for agonist-induced desensitization of D₂ receptor signaling. Moreover, prolonged mGluR activation by bath perfusion of the agonist DHPG, which would partially deplete Ca²⁺ stores via continuous IP₃-mediated Ca²⁺ release (Cui et al, 2007), cross-desensitized dopamine-induced currents in a Ca²⁺ store-dependent manner, further supporting the idea that continuous Ca²⁺ release can mediate desensitization of D₂ receptor signaling.

It has been reported recently that CaMKII phosphorylation causes desensitization of D₃ receptors (Liu et al, 2009) and also D₂ receptors (JQ Wang, personal communication), both of which belong to the D₂-like dopamine receptor family coupled to G_i/o G proteins. Although dopamine neurons express both D₂ and D₃ receptors, only D₂ receptors are coupled to GIRK activation in dopamine neurons (Beckstead et al, 2004; Davila et al, 2003). Furthermore, activation of D₂ receptors, as well as D₄ receptors, has been shown to activate CaMKII via G_i/o and intracellular Ca²⁺ release (Gu and Yan, 2004; Takeuchi et al, 2002). Consistent with these previous findings, CaMKII inhibition suppressed quinpirole-induced desensitization in the current study. In contrast, PKC, which can also desensitize D₂ receptors (Morris et al, 2007; Namkung and Sibley, 2004), appears not to have a role in dopamine neurons.

D₂ and GABA_B receptor-mediated outward currents in dopamine neurons are largely eliminated by genetic deletion of GIRK2 but not GIRK3 (Beckstead et al, 2004; Labouebe et al, 2007), indicating that these two receptors most likely share the same signaling targets. Furthermore, GABA_B receptor activation has also been shown to cause IP₃-mediated Ca²⁺ release in certain cells (Michler and Erdo, 1989; Parramon et al, 1995). However, blockade of IP₃-mediated Ca²⁺ release failed to affect desensitization of baclofen-induced currents in this study, demonstrating that Ca²⁺-dependent desensitization is selectively involved in the regulation of D₂, but not GABA_B, receptor signaling in dopamine neurons. It has been shown that GABA_B receptor-GIRK channel coupling efficiency, as well as agonist-induced desensitization of GABA_B receptor-mediated currents, are regulated by regulator of G protein signaling (RGS) proteins (Labouebe et al, 2007; Mutneja et al, 2005). These findings strongly suggest that these two G_i/o-coupled receptors are differentially regulated by distinct desensitization mechanisms in dopamine neurons.

Enhanced D₂ Autoinhibition and Hypodopaminergic State During Ethanol Abstinence

Repeated ethanol exposure resulted in a dramatic reduction in the amount of desensitization observed during quinpirole application. This attenuated desensitization was accompanied by an increase in the potency of quinpirole to activate GIRK channels. The maximal quinpirole-induced currents were comparable in saline- and ethanol-treated mice, suggesting that the expression levels of D₂ receptors and GIRK channels were not altered. It is of note that a similar increase in the potency of baclofen has been observed in VTA dopamine neurons after repeated in vivo exposures to morphine or γ-hydroxybutyrate (Labouebe et al, 2007). This increase in baclofen potency results from downregulation of RGS proteins, which control GABA_B receptor desensitization as described above.

It is well known that repeated exposure to cocaine or amphetamine leads to a decrease in D₂ agonist-induced inhibition of VTA dopamine neuron firing (Henry et al, 1989; Marinelli et al, 2003; Wolf et al, 1993). Furthermore, repeated cocaine administration has been shown to cause upregulation of CaMKII expression in the VTA (Backes and Hemby, 2003; Licata et al, 2004). Therefore, differential regulation of the CaMKII-dependent desensitization mechanism may contribute to the difference in D₂ autoreceptor sensitivity, that is, increase vs decrease, following repeated intermittent administration of alcohol vs psychostimulants.

In vivo recording studies have found profound reductions in VTA dopamine neuron firing activity after withdrawal from repeated ethanol exposure ((Diana et al, 1993, 1995), but also see (Shen, 2003; Shen and Chiodo, 1993)). Our data show that the basal firing frequency of dopamine neurons measured in brain slices was not altered by in vivo ethanol exposure, as has been reported in previous brain slice recording studies (Brodie, 2002; Hopf et al, 2007; Okamoto et al, 2006). However, inhibition of firing produced by low concentrations of quinpirole was significantly augmented in ethanol-treated mice, consistent with the increased potency of quinpirole to activate GIRK channels. Although our data failed to detect functional dopamine tone in VTA slices, as evidenced by the lack of effect of the D₂ antagonist sulpiride, autoinhibitory regulation mediated by dopamine tone has been shown to have an important role in controlling the basal firing of dopamine neurons in vivo (Henry et al, 1989; Pucak and Grace, 1994; White and Wang, 1984). Therefore, increased D₂ autoreceptor-mediated inhibition may well lead to reduced dopamine neuron firing in vivo after withdrawal from repeated ethanol exposure.

In addition to dopamine-mediated autoinhibition, the activity of dopamine neurons is regulated by numerous neurotransmitter inputs that may undergo ethanol-induced adaptive changes (Morikawa and Morrisett, 2010). For example, it has been reported that VTA GABA neuron firing, as well as GABA release onto dopamine neurons, are increased following ethanol withdrawal (Gallegos et al, 1999; Melis et al, 2002), which would result in enhanced GABAergic inhibition of dopamine neurons. The relative contribution of different neurotransmitter inputs to dopamine neuron hypoactivity remains to be determined.

Ample evidence indicates the impact of stress on the dopaminergic system and its role in the development of addictive behaviors (Sinha, 2008; Ungless et al, 2010). Thus, repeated stress associated with daily ethanol intoxication and withdrawal may have played a role in causing the alterations in D₂ signaling in this study and reductions in dopamine neuron firing reported in previous studies (Diana et al, 1993, 1995).

The hypodopaminergic state observed in ethanol-withdrawn animals or in detoxified alcoholics is characterized by hypoactivity of dopamine neurons and the resulting decrease in dopamine release in the NAc and striatum, together with downregulation of postsynaptic D₂ receptors in those dopamine projection areas (Martinez et al, 2005; Melis et al, 2005; Volkow et al, 2007). It has been postulated that this hypodopaminergic state contributes to the emotional/motivational component of ethanol dependence, although activation of other neurotransmitter systems, such as those involving corticotropin-releasing factor or the endogenous κ-opioid dynorphin, has also been implicated (Koob, 2009; Wee and Koob, 2010). In support of the role of the hypodopaminergic state in dependence, ethanol-withdrawn animals have been shown to self-administer ethanol until NAc dopamine levels are restored to the pre-withdrawal baseline levels (Weiss et al, 1996). Furthermore, animals that exhibit high ethanol preference and consumption have low basal dopamine levels in the NAc (George et al, 1995; McBride et al, 1995). Our study suggests that promoting the Ca²⁺-dependent desensitization mechanism that regulates D₂ autoinhibition in the VTA might alleviate the hypodopaminergic state in alcohol-dependent individuals, thus aid in reducing compulsive alcohol drinking.