Alcohol Self-Administration: Further Examination of the Role of Dopamine Receptors in the Nucleus Accumbens

METHODS

RESULTS

Data were used only from animals who received bilateral infusions in the core region of the N Acc. Histological analysis showed that infusions occurred in areas corresponding to atlas coordinates ranging in the rostral-caudal plane between +1.2 mm and +1.7 mm anterior to bregma. The number of animals used in each condition are indicated in the figure captions. Ethanol consumption during the 30-min session before the start of microinjections averaged 0.58 g/kg (±0.12 g/kg), which are similar to those previously reported for this initiation procedure.⁴⁶

DISCUSSION

In the present study, N Acc infusions of the D₂ agonist quinpirole resulted in a biphasic effect on ethanol self-administration. The low dose of quinpirole (1.0 μg) increased self-administration as a function of increased response rate. Higher doses of quinpirole (4.0 and 10.0 μg) decreased self-administration by a reduction in the time course of ethanol-reinforced responding with no effect on response rate. No dose of quinpirole affected the onset latency of ethanol-reinforced responding. Thus, stimulation of D₂/D₃ receptors in the N Acc influenced the maintenance (response rate) and termination (total responding and time course) of ethanol self-administration, but had no effect on the onset (latency) of responding. These data are not in complete agreement with our previous studies that showed increased ethanol self-administration and delayed termination of responding after quinpirole infusions in the N Acc.^40,41 The reason for these differences are unclear, but in one of the previous studies,⁴⁰ session length was 1 hr rather than 30 min, which may account for some of the differences observed.

One interpretation of the changes in responding observed after quinpirole administration is that D₂ activation in the N Acc has a direct effect on the process of ethanol reinforcement. In vitro slice preparation studies have shown that quinpirole reduces electrically evoked dopamine release⁴⁹ Ethanol self-administration increases DA release in the N Acc,¹¹ which is believed to result from activation of DA cell bodies in the VTA.^7,8 Thus, there is support for the hypothesis that quinpirole altered ethanol-maintained responding by attenuating the ethanol-evoked DA release in the N Acc. Using this hypothesis, the low dose of quinpirole could have increased total responding by only partially attenuating ethanol-stimulated DA release. The higher doses of quinpirole may have attenuated ethanol’s stimulated release of DA to such an extent that responding terminated early because of a lack of appropriate feedback related to consumption. Using this hypothesis, it could be predicted that, for some doses of quinpirole, the reduction of DA release occurring during the ethanol drinking would be such as to alter consumption rate while prolonging responding, explaining the data observed in our previous studies. This interpretation of data suggests that D₂ DA activity within the N Acc is involved in the regulation of maintenance and termination of ethanol self-administration, and the extent of changes in intake resulting from D_2/3 agonist action may be dependent on the extent and timing of altered DA activity during an ethanol drinking episode.

Another explanation of the quinpirole data is that changes in ethanol-reinforced responding after quinpirole infusions resulted from nonspecific effects on locomotor behavior.⁴⁰ data is that changes in ethanol-reinforced Quinpirole (0.3 to 3.0 μg) administration in the N Acc has been shown to increase locomotor activity,³ which could account for the increases in total responding and response rate observed in the present data after the microinjection of the low dose of quinpirole (1.0 μg). However, a higher dose of quinpirole (4.0 μg), similar to one known to also increase locomotor activity,³ decreased ethanol-reinforced responding in the present study without changing response rate. This suggests that nonspecific locomotor-activating effects of quinpirole in the N Acc most likely did not produce the observed decrease in ethanol-maintained responding.

Another hypothesis that might account for the biphasic effect of quinpirole on ethanol-reinforced responding is a possible dose-related differential action at the various DA receptor subtypes due to their affinity for the D₂/D₃ agonist.¹³ For example, at the low dose of quinpirole, only the high-affinity type D₂ autoreceptors might have been involved,¹³ whereas at the higher doses, the lower affinity pre- and postsynaptic D₂ and D₃ receptors might have been activated as well.¹³ As pointed out by Neisewander et al.,²⁸ it is critical to quantify receptors occupied to make statements regarding receptor subtype sensitivity. It is of interest to note, however, that they²⁸ found differential receptor occupation within the dose range of SCH 23390 used in this study.

Intra-accumbens infusions of the D₁ agonist SKF 38393 alone had no effect on any measured parameter of ethanol-reinforced responding at the doses tested. This is in contrast to numerous studies showing that SKF 38393 influenced measures of locomotor, consummatory, and reinforced behaviors. When administered systemically, SKF 38393 reduced alcohol intake in the selected line of high alcohol drinking rats.³⁰ As well, systemic SKF 38393 reduced food intake in rats^50–52 and sucrose (10%) sham feeding in rats.⁵³ Both cocaine and food-reinforced lever pressing in squirrel monkeys was reduced after systemic SKF 38393.⁵⁴ Intra-accumbens infusions of SKF 38393 stimulated motor behavior at doses similar to those tested in this study.^55–57 However, a N Acc-microinjected dose of SKF 38393 that enhanced locomotion (1.0 μg) had no effect on quantity of food consumed or other collateral eating behaviors.⁵⁸ These studies compared with the present study suggest that systemically administered SKF 38393 is most likely acting through DA or other receptor systems [i.e., serotonin (5-HT)] in brain regions other than the N Acc to reduce the intake of alcohol.

N Acc microinjections of the DA D₁ antagonist SCH 23390 had no effect on ethanol-reinforced responding at a low dose (0.05 μg), but resulted in significant decreases in total responding and time course at the two higher doses (0.5 and 2.0 μg). Thus, termination of ethanol-reinforced responding may be partly regulated by D, receptor function in the N Acc. SCH 23390 did not alter response rate or latency, suggesting that D₁ receptor activation in this brain region is not necessary for either the onset or maintenance of responding. However, because SCH 23390 has binding affinity for 5-HT₂ and 5-HT_lc receptors,^59,60 it is possible that the reductions in ethanol-reinforced responding observed in the present experiment were not specific to D₁ receptor blockade.

When administered systemically⁶¹ or in the N Acc,⁶² SCH 23390 decreases locomotor behavior, suggesting that the observed changes in ethanol-reinforced responding may have resulted from nonspecific motor effects. However, reductions in locomotor behavior produced by N Acc infusions of SCH 23390 (0.5 and 2.5 μg) were not accompanied by decreases in heroin self-administration in rats,⁶² indicating a dissociation of the nonspecific locomotor actions and opioid reinforcement function. This data⁶² raises the possibility that the observed effect of SCH 23390 in the present study was also not a result of nonspecific motor actions. Taken together with the finding that SKF 38393 produced no effect on ethanol self-administration, the finding that SCH 23390 was able to decrease responding suggests that the process of ethanol reinforcement may be partly dependent on tonic D₁ receptor function, but is not enhanced by increased D₁ stimulation.

Intra-accumbens injections of the D₂ antagonist raclopride (1.0 μg) decreased ethanol-reinforced lever pressing by reductions in both time course and response rate. This replicates previous findings in which intra-accumbens infusions of raclopride decreased ethanol⁴¹- and sucrose⁴²-reinforced responding similarly when behavioral baselines were matched. In these previous studies, there was a shift to the right in the dose effect curve for sucrose reinforcement, suggesting again that the effects on ethanol responding were not a result of interference with motor performance.

Co-administration of SKF 38393 (0.03 μg) blocked the biphasic effects of the two low doses of quinpirole (1.0 and 4.0 μg) on ethanol-reinforced responding. Although the D₁ agonist failed to alter ethanol maintained responding when administered alone, it attenuated the effects of the D₂ agonist. This result would seem to be in contrast with the notion that D₁ receptor stimulation is required for the expression of D₂-mediated effects.²¹ Electrophysiological experiments have shown that a high percentage of N Acc neurons are insensitive to SKF 38393 alone, but demonstrate a potentiated inhibition after quinpirole application when SKF 38393 is present in the medium.^26,63 Synergistic interactions between D₁ and D₂ receptors have been found in behavioral studies of grooming,²⁵ locomotor activity^3,27,64 and stereotypy.⁵⁵ However, behavioral studies involving reinforcement, such as d-amphetamine self-administration⁶⁵ and conditioned reinforcement,⁶⁶ have reported an additive interaction between D₁ and D₂ receptors. Systemic administration of either quinpirole or SKF 38393 decreased food intake in rats, but the combination did not potentiate the decrease produced by either compound alone.⁵² It would appear, therefore, that co-administration of SKF 38393 and quinpirole may have a variety of effects which can be either additive or antagonistic, depending on a variety of factors, including the behavior under study.⁶⁷ With regard to ethanol-reinforced responding, the present data suggest that D₁ and D₂ receptor stimulation in the N Acc is antagonistic when the D₂ agonist effects are expressed. In the dose ranges tested, it seems that, when both agonists bind competitively to their respective receptor subtypes, the efficacy of the D₂ agonist to affect ethanol self-administration is decreased.

Co-administration of an ineffective dose of SCH 23390 (0.05 μg) on ethanol self-administration blocked the low-dose increasing effects of quinpirole, but did not prevent the decrease in total responding observed after the higher dose of quinpirole (10.0 μg). Thus, the D₁ agonist SKF 38393 and the D₁ antagonist SCH 23390 resulted in a similar blockade of the behavioral expression of the low dose D₂ agonist effects. This suggests that some level of tonic DA stimulation of D₁ receptors in the N Acc may be necessary for the low-dose effects of the D₂ agonist on ethanol-reinforced responding to be observed.

The present data show that co-administration of ineffective doses of the D₁ agonist SKF 38393 (0.03 μg) with raclopride (0.1 μg) resulted in an enhancement of the response rate decreasing effects of the D₂ antagonist. This suggests quantitative synergism⁶⁸ in the expression of the D₂ antagonist-induced effects on ethanol-reinforced responding. The D₁ antagonist SCH 23390 did not affect the behavioral effects of raclopride. These data suggest that the expression of D₂ antagonist-induced effects on ethanol-reinforced responding are enhanced by stimulation of D₁ receptors, but are not dependent on D₁ receptor function.

The behavioral effects (i.e., increased time course and decreased response rate) produced by the co-administration of SKF 38393 with the high dose of quinpirole were similar to previous studies that examined the effects of intra-accumbens infusions of the nonspecific mixed agonist d-amphetamine^40,41 This result could be explained based on the hypothesis that relative levels of D₁ and D₂ activity may determine the particular form of behavioral expression.⁶⁹ However, added studies are required to test this hypothesis.

One important finding of the present study is that none of the DA agonists or antagonists, at the doses tested alone or in combination, influenced the onset of ethanol self-administration (i.e., latency of the first response). In the operant behavioral situation, discriminative stimuli set the occasion for (i.e., signal) response sequences that have a higher probability of resulting in the availability of a reinforcer. The role of the N Acc in behavioral control by conditioned stimuli has been demonstrated by several different studies. For instance, N Acc infusions of d-amphetamine enhanced responding that resulted in presentation of a conditioned reinforcer.^70,71 Increases in extracellular DA in the N Acc have been observed preceding the onset of an alcohol self-administration session when the animals are placed into the operant apparatus in which they have previously received ethanol reinforcement.¹¹ Thus, one proposed function of N Acc is to integrate the actions of discriminative stimuli and/or conditioned reinforcers in the organization of motor activity.⁷² However, the present data suggest that DA transmission in the N Acc is not involved in regulating the onset of ethanol self-administration, which would not support a role for DA specifically in the initial events that control an ethanol consumption event. In conclusion, data from the present study demonstrate that D₁ and D₂ receptors in the N Acc are involved in the maintenance and termination, but not onset, of ethanol-reinforced responding in a limited access operant paradigm.