Dimers Mediate Synergy of Dopamine D2 and Adenosine A2 Receptor-Stimulated PKA Signaling and Regulate Ethanol Consumption

Transcription

1 Cell, Vol. 109, , June 14, 2002, Copyright 2002 by Cell Press Dimers Mediate Synergy of Dopamine D2 and Adenosine A2 Receptor-Stimulated PKA Signaling and Regulate Ethanol Consumption Lina Yao, 1,2 Maria Pia Arolfo, 1,2 Douglas P. Dohrman, 1,2 Zhan Jiang, 1 Peidong Fan, 1 Sara Fuchs, 6,8 Patricia H. Janak, 1,2,3 Adrienne S. Gordon, 1,2,3,4 and Ivan Diamond 1,2,3,4,5,7 1 Ernest Gallo Clinic and Research Center 2 Department of Neurology 3 Neuroscience Graduate Program and The Wheeler Center for the Neurobiology of Addiction 4 Department of Cellular and Molecular Pharmacology 5 Department of Pediatrics University of California, San Francisco San Francisco, California Department of Immunology Weizmann Institute Rehovot Israel Summary Dopamine release is activated by ethanol and ad- dicting drugs, but molecular mechanisms linking do- paminergic signaling to neuronal responses and drink- ing behavior are poorly understood. We report that dopamine-d2 receptors induce PKA C translocation and increase CRE-regulated gene expression. Ethanol also activates PKA signaling. Subthreshold concentra- tions of the D2 agonist NPA and ethanol, without effect alone, together cause synergistic PKA translocation and CRE-mediated gene transcription. D2 or adenosine A2 receptor blockade, pertussis toxin, Rp-cAMPS, or overexpression of dominant-negative peptides that se- quester dimers prevent synergy. Importantly, over- expression of a inhibitor peptide in the nucleus accumbens strikingly reduces sustained alcohol con- sumption. We propose that synergy of D2 and A2 confers ethanol hypersensitivity and that dimers are required for voluntary drinking. Introduction administration (Hodge et al., 1997). Furthermore, ethanol preference, sensitivity, and ethanol-conditioned place preference are reduced in mice lacking D2 (Phillips et al., 1998; Cunningham et al., 2000). This may be related to an overall reduction in their motivated response (Risinger et al., 2000). In contrast, overexpression of D2 is reported to reduce ethanol preference and intake (Thanos et al., 2001). D2 activation inhibits adenylyl cyclase activity, reducing camp levels in tissues, cell lines, and brain (Obadiah et al., 1999). D2 activates the trimeric GTP binding proteins Gi and/or Go, consisting of i or o subunits and dimers; i subunits directly inhibit adenylyl cyclase activity. However, dimers have diverse biological effects, which are probably responsible for other intracellular events observed on D2 activation (Lledo et al., 1992; Waszczak et al., 1998; Vallar and Meldolesi, 1989; Piomelli et al., 1991; Gordon et al., 2001). Despite the importance of dopamine in responding to ethanol, the molecular mechanism linking D2 signaling to neuronal activation and drinking behavior is not clear. NG is a clonal neuroblastoma-glioma hybrid cell line that expresses many neuronal properties. We have used these cells as a model system to investigate ethanol-induced changes in camp production and camp-dependent protein kinase (PKA) regulation of in- tracellular signaling. We have demonstrated that short- term exposure to ethanol increases extracellular adenosine (Nagy et al., 1990), which activates adenosine A2 receptors (A2) and increases camp levels (Gordon et al., 1986; Sapru et al., 1994). Receptors that increase camp activate the trimeric GTP binding protein Gs to release s and dimers; s directly activates adenylyl cyclase. camp then binds to the two regulatory subunits in the inactive PKA tetrameric holoenzyme to release two active catalytic (C) subunits. In turn, C subunits phosphorylate diverse proteins to regulate their activity. C subunits can also translocate to the nucleus and regu- late gene expression. One substrate of PKA is the camp element regulatory binding protein (CREB), which binds to the camp regulatory element (CRE) in the regulatory region of certain genes. Phosphorylation of CREB leads to CRE-mediated gene expression. Using NG cells, we have shown that ethanol induces translocation of the catalytic subunit of PKA (C ) to the nucleus (Dohr- Dopaminergic signaling in the nucleus accumbens (NAc) is activated by all addicting drugs (Robbins and Everitt, 1999; Wise and Bozarth, 1987) and appears to be an man et al., 2002), stimulates CREB phosphorylation integral component of the neural circuitry of addiction (Constantinescu et al., 1999), and increases CRE-medi- (Hyman and Malenka, 2001). Release of dopamine in ated gene expression (Asher et al., 2002). anticipation of drinking or during exposure to ethanol is To investigate the relationship of D2 to ethanolthought to contribute to incentive processes and ethanol induced changes in camp-dependent signaling, we deconsumption (Imperato and Di Chiara, 1986; Weiss et veloped an NG108-15/D2 cell line stably expressing D2 al., 1993). The importance of NAc dopamine to ethanol receptors (Asai et al., 1998). We first asked whether D2 intake was confirmed by microinjection of dopaminergic can activate the PKA pathway because of evidence that agents into the NAc: dopamine D2 receptor (D2) ago- D2 activation can increase camp levels (Watts and nists increase and antagonists decrease ethanol self- Neve, 1996, 1997). We show here that the D2 agonist NPA increases camp levels, causes translocation of 7 PKA C, and increases CRE-mediated gene expression. Correspondence: diamond@itsa.ucsf.edu 8 This work was carried out while Dr. Fuchs was on sabbatical leave Next, we found that (1) subthreshold concentrations of at the Ernest Gallo Clinic and Research Center, University of California, NPA and ethanol synergistically induce PKA C translo- San Francisco, San Francisco, California cation and increase CRE-mediated gene expression, (2)

2 Cell 734 Figure 1. D2 Activation or Ethanol Causes Translocation of PKA C in NG108-15/D2 Cells (A) NPA- or (B) ethanol-induced PKA C translocation. Cells expressing D2 were incubated with or without 50 nm of the D2 agonist R( ) 2,10,11- trihydroxy-n-propyl-noraporphine hydrobromide (NPA) or 100 mm ethanol for 10 min. Where indicated, cells were preincubated either with the D2 antagonist spiperone (10 M) for 30 min or 50 ng/ml PTX overnight. Data represent at least three experiments. Staining intensity is standardized by the color bar (orange indicates highest concentration). Scale bar equals 10 m; images are 400. (C) Western blots of nuclear (N), membrane (M), and cytosolic (Cy) fractions from treated cells. (D) NPA-induced PKA C translocation away from the Golgi as a function of time. Cells were incubated with 50 nm NPA and scored positive if staining extended more than one nuclear radius away from the Golgi. Data are the mean SEM (average of 15 cells/field, 3 fields/experiment, n 3). (E) PKA C translocation as a function of NPA concentration. Cells were incubated with the indicated concentrations of NPA for 10 min. free subunits released from G proteins mediate these NPA-Induced Translocation of PKA C synergistic responses; (3) D2 and A2 activation is required Is Mediated via Increases in camp for synergy; and (4) overexpression in the NAc Ethanol increases camp levels (Nagy et al., 1989; Sapru of a dominant-negative scavenger peptide, which et al., 1994) and causes translocation of PKA C to the sequesters free dimers, strikingly reduces ethanol cytoplasm and nucleus (Figure 1B). Rp-cAMPS, which consumption by rats. prevents binding of camp to the regulatory subunit of PKA, blocks ethanol-induced translocation of PKA C Results (Figure 2A). Rp-cAMPS also blocks NPA-induced translocation (Figure 2A). These data suggested that NPA D2 Activation Induces PKA Translocation might increase camp production. Indeed, Figure 2B NG108-15/D2 cells were incubated for 10 min with the shows that NPA increases camp levels at 10 min followed D2 agonist NPA; anti-pka C antibodies were used to by decreases at 30 min. follow PKA C localization. PKA C is localized primarily to the Golgi in untreated cells (Figures 1A and 1B). D2 NPA Induces CRE-Mediated Gene Expression activation translocates PKA C away from the Golgi to Ethanol induces translocation of PKA C into the nu- the cytoplasm and nucleus (Figure 1A). Ethanol mimics cleus, leading to CREB phosphorylation (Constantinescu NPA-induced translocation (Figure 1B). Western blots et al., 1999) and increased CRE-mediated gene also show that NPA or ethanol each cause PKA C expression (Asher et al., 2002). We asked whether NPA translocation from the membrane fraction to nuclear does the same. Cells were transiently transfected with and cytoplasmic fractions (Figure 1C). Spiperone, a D2 a CRE-luciferase reporter construct to measure gene antagonist, blocks PKA C translocation induced by expression via endogenous CREB. Transfected cells NPA (Figure 1A) but has no effect on ethanol-induced were exposed to NPA and other agents for 10 min and translocation (Figure 1B). NPA-induced translocation assayed for luciferase activity 5 hr later (Figure 2C). NPA occurs within minutes, reaching a maximum at 10 min increases luciferase activity as effectively as forskolin. (Figure 1D); the K M for half-maximal translocation is 1.6 Spiperone blocks NPA stimulation of luciferase but has M (Figure 1E). no effect on ethanol-induced luciferase expression; Rp-

3 Synergy of D2 and Ethanol Requires Dimers 735 Figure 2. NPA- or Ethanol-Induced Translocation of PKA C Requires camp and Induces CRE-Mediated Gene Expression (A) Rp-cAMPS inhibition of NPA- and ethanolinduced PKA C translocation. Cells were pretreated with 20 M Rp-cAMPS for 1.5 hr prior to incubation with NPA or ethanol as in Figure 1. (B) NPA modulation of camp production. Cells were incubated with or without 50 nm NPA for the indicated times. camp was measured by radioimmunoassay (Gordon et al., 1986). Data are the mean SEM of four experiments. Asterisk indicates p 0.05, compared to time zero (one-way ANOVA and Dunett s test). camp levels in the absence of NPA did not change during the experiment. (C) NPA- and ethanol-induced luciferase gene expression. Cells were transiently transfected with a CRE-luciferase construct and preincubated overnight as indicated with PTX, spiperone, or Rp-cAMPS. Data are the mean SEM of at least three experiments. Asterisk indicates p 0.01, compared with control (one-way analysis of variance and Dunnett s test). camps blocks increased gene expression due to NPA or ethanol. These data suggest that D2-mediated in- creases in camp and subsequent PKA C translocation to the nucleus activate CRE-mediated gene expression. Synergy between NPA and Ethanol-Induced Translocation of PKA C and CRE-Mediated Gene Expression Since NPA or ethanol activates the PKA pathway from camp through CRE-mediated gene expression, we searched for synergy between NPA and ethanol. NPA (0.5 nm) or ethanol (25 mm) alone did not cause PKA C translocation (Figure 3A). However, these subthreshold concentrations together produce a striking synergistic increase in PKA C translocation and CRE-mediated luciferase activity (Figures 3A and 3B). Importantly, the synergistic increase in gene expression is as great as that produced by forskolin (Figure 3B). Synergy depends on D2 and camp, because it is blocked by spiperone and Rp-cAMPS (Figures 3A and 3B). Dose responses for ethanol and NPA revealed synergy for each agent with varying concentrations of the other (Figures 3C and 3D). There are no cooperative effects of each agent alone. NPA- and Ethanol-Induced Synergy: PKA C Translocation and CRE-Dependent Gene Expression Is Due to Activation of Adenylyl Cyclase by Dimers Released from Gi/o Gi/o activation is blocked by pertussis toxin (PTX). PTX prevents NPA- and ethanol-induced PKA C translocation and CRE-mediated gene expression (Figures 1A and 2C; Figures 1B and 2C). In addition, PTX inhibits synergy of PKA C translocation (Figure 3A) and CREmediated gene expression (Figure 3B). These results suggest that Gi/o is required for NPA- and ethanolinduced PKA C translocation, gene expression, and synergy. dimers from Gi/o activate adenylyl cyclase (AC) isozymes II and IV (Baker et al., 1999; Federman et al., 1992; Inglese et al., 1994; Tang and Gilman, 1991). Because NG108-15/D2 cells express AC II and IV (Figure 4A), it is possible that dimers are responsible for NPA- and ethanol-induced increases in camp that lead to PKA C translocation, gene expression, and synergy. Free dimers of trimeric G proteins bind to the -adrenergic receptor kinase ( ARK) on activation of -adrenergic receptors (Ruiz-Gomez and Mayor, 1997). Expression of the carboxyl terminus of ARK1 ( ARK1 minigene) binds free dimers and inhibits signaling (Koch et al., 1994). PKA C translocation induced by NPA or ethanol is markedly inhibited by overexpression of the dominant-negative ARK1 minigene (Figure 4B). This scavenger peptide also blocks synergy of PKA C translocation (Figure 4D). The ARK1 minigene could affect the function of other proteins because subunits bind to different amino acid sequences (Hamm, 1998). Therefore, we trans- fected cells with a second dominant-negative scav- enger peptide, QEHA, the specific binding site for on AC II (Chakrabarti et al., 1998; Chen et al., 1995; Weng et al., 1996). QEHA inhibited translocation of PKA C induced by NPA or ethanol (Figure 4C) and synergy of translocation (Figure 4D). QEHA also prevented NPAand ethanol-induced increases in CRE-mediated gene expression (Figure 4F) and synergy (Figure 4G). QEHA did not inhibit forskolin-induced PKA C translocation (Figure 4C), probably because forskolin does not require to activate AC (Yan et al., 1998). Overexpression of the inactive peptide SKEE from an analogous region of ACIII (Chen et al., 1995) did not affect NPA or ethanol- induced PKA C translocation (data not shown). Taken together, these data suggest that release from Gi/o is required for NPA or ethanol-induced translocation of PKA C to the nucleus, subsequent activation of CREmediated gene expression, and synergy.

4 Cell 736 Figure 3. Subthreshold Concentrations of NPA and Ethanol Synergistically Induce PKA C Translocation and CRE-Mediated Gene Expression (A) Synergy of PKA C translocation. Cells were incubated for 10 min with 0.5 nm NPA, 25 mm ethanol, or 0.5 nm NPA plus 25 mm ethanol, and localization of PKA C was determined. Where indicated, the cells were preincubated with PTX, spiperone, or RpcAMPS. (B) Synergy of CRE-mediated luciferase activity. Cells were treated as in Figure 2C, followed by addition of 0.5 nm NPA or 25 mm ethanol alone or in combination. Asterisk indicates p 0.01 compared with control (oneway analysis of variance and Dunnett s test). (C) Synergy of CRE-dependent gene expression at a subthreshold concentration of ethanol as a function of NPA concentration. Cells were treated as in (B), with various concentrations of NPA plus 25 mm ethanol. (D) Synergy of CRE-dependent gene expression at a subthreshold concentration of NPA as a function of ethanol concentration. Cells were treated with various concentrations of ethanol plus 0.5 nm NPA. NPA- and Ethanol-Induced Synergy: Role of Adenosine Receptors We have shown that ethanol increases camp levels via A2 activation (Sapru et al., 1994). BW A1434U, an adeno- sine receptor antagonist, blocks ethanol-induced PKA C translocation (Figure 4E) and CRE-mediated gene expression (Figure 4F). In contrast, BW A1434U does not block NPA-induced translocation (Figure 4E) and gene expression (Figure 4F). However, BW A1434U does block synergy (Figures 4E and 4G). BW A1434U inhibition is most likely due to inhibition of A2 receptors, since we do not find evidence for A1 receptors in our NG cell lines (see Discussion). NPA and Ethanol Cause PKA C Translocation and Exhibit Synergy in Primary Hippocampal Neurons Primary hippocampal neurons express dopamine and adenosine receptors coupled to second messengers (Goldsmith and Joyce, 1994; O Kane and Stone, 1998). We used these cells to extend our findings to nontransformed neurons. Figure 5A shows that NPA or ethanol induces PKA C translocation to the cytoplasm, neurites, and nucleus. Spiperone blocks NPA-induced translocation but has no effect on ethanol-treated cells. BW A1434U blocks ethanol-induced PKA C translocation (Figure 5A). Unlike NG108-15/D2 cells, hippocampal neurons express A1 in addition to A2 receptors (O Kane and Stone, 1998). Specific adenosine receptor blockade of either A1 with DPCPX or A2 with DMPX prevents ethanol-induced PKA C translocation (Figure 5A); these antagonists do not affect NPA-induced translocation. Rp-cAMPS, PTX, and overexpression of QEHA prevent ethanol- or NPA-induced PKA C translocation (Figures 5B and 5C). BW A1434U, DPCPX, DMPX, spiperone, QEHA, PTX, or Rp-cAMPS alone do not alter transloca- tion (data not shown). Presynaptic neurotransmitter release was not required for PKA C translocation be- cause preincubation for 2 hr with 1 M TTX, which inhibits release, was without effect (data not shown). Synergy of PKA C translocation occurs in primary hippocampal neurons at subthreshold doses of NPA (0.5 nm) and ethanol (25 mm) added together (Figure 5D). Spiperone, PTX, BW A1434U, and QEHA prevent synergy (Figure 5D). These data suggest that nontrans- formed neurons also exhibit D2- and ethanol-induced changes in PKA C translocation and that synergy depends on release and ethanol-induced accumulation of extracellular adenosine. Overexpression of ARK1, a Dominant-Negative Scavenger Peptide, Decreases Ethanol Consumption NAc neurons express D2 and A2 in the same cells (Fink et al., 1992). Our results suggest that there is synergy

5 Synergy of D2 and Ethanol Requires Dimers 737 Figure 4. Dimers and A2 Are Required for PKA C Translocation, CRE-Mediated Gene Expression, and Synergy (A) Western blots of AC II and IV. Data represent at least three experiments. Whole NG108-15/D2 cell lysates (40 g of protein) were probed with anti-ac II or anti-ac IV antibody before (lane 2) and after (lane 1) preabsorption with peptide antigen. (B) PKA C translocation and overexpression of the carboxyl terminus of ARK1 (Minigene). Cells were incubated with or without NPA or ethanol after transfection with Ad5 ARK1 or the adenovirus Ad5 vector control, and intracellular localization was determined with PKA C antibody. Transfection efficiency was greater than 95% as measured by transfection with Ad5GFP. Data are representative of six experiments with similar results. (C) PKA C translocation and overexpression of the AC II QEHA peptide. Cells were transfected with Ad5 QEHA and incubated with or without ethanol, NPA, or forskolin and analyzed as in (B). Data are representative of four experiments. (D) Synergy of PKA C translocation and overexpression of dominant-negative scavenger peptides. Cells were transfected with Ad5 ARK1, Ad5QEHA, or Ad5 vector control, incubated in the absence or presence of NPA (0.5 nm) and/or ethanol (25 mm), and probed with PKA C antibody. Data are representative of five experiments. (E) NPA- and ethanol-induced PKA C translocation and synergy. Cells were preincubated with or without BW A1434U and further incubated with or without NPA and/or ethanol. Data are representative of three experiments. (F) NPA- and ethanol-induced CRE-mediated gene expression and overexpression of the QEHA peptide. Cells transfected with CRE-luciferase were preincubated with or without Ad5 QEHA or BW A1434U, followed by addition of 50 nm NPA or 100 mm ethanol for 10 min. (G) Synergy of CRE-mediated gene expression and overexpression of QEHA. Cells were treated as in (C), followed by addition of 0.5 nm NPA or 25 mm ethanol alone or in combination. [baseline, day 7 and 14] interaction: F[2,32] 4.23, p 0.024; Ad5 ARK1 day 7 versus baseline, p 0.001; Ad5 ARK1 day 7 versus day 14, p 0.01; Ad5lacZ day 7 versus Ad5 ARK1 day 7, p 0.011). Water intake was unaffected by Ad5 ARK1 (Figure 6B) (ANOVA, treatment by day interaction: F[2,32] 1.9, p 0.16). Ethanol intake (g/kg) decreased significantly on days 3 9 (p 0.05) in rats receiving Ad5 ARK1, but not Ad5LacZ (Fig- ures 6C and 6D). Ethanol preference was reduced similarly following Ad5 ARK1 injection (data not shown). In separate subjects, Ad5LacZ was expressed initially in glia on day 2, appearing in neurons on day 4 with maxi- mal expression in neurons and neuronal processes on between D2 agonists and ethanol/a2 for stimulation of PKA C translocation and CRE-mediated gene expression. Since dopamine is released in NAc in response to ethanol, NAc neurons would be expected to release dimers synergistically during alcohol consumption. We studied the potential role of in ethanol drinking in a continuous voluntary two-bottle choice paradigm in rats. Ethanol consumption was examined for 2 weeks following bilateral microinjection of Ad5 ARK1 or control Ad5lacZ viral vectors into the medial NAc. Maximal reduction in ethanol consumption relative to controls occurred at day 7 with recovery by day 14 (Figure 6A) (ANOVA, treatment [Ad5lacZ versus Ad5 ARK] by day

6 Cell 738 Discussion Figure 5. NPA and Ethanol Cause Translocation of PKA C in Primary Hippocampal Neurons (A) PKA C translocation and A1, A2, and D2 blockade. Primary hippocampal neurons were incubated in the absence or presence of 50 nm NPA or 75 mm ethanol for 10 min. Where indicated, the cells were preincubated for 30 min with or without spiperone, BW A1434U, the A1 antagonist DPCPX (100 nm), or the A2 antagonist DMPX (10 M). C is indicated by green staining and the neuronspecific marker (NeuN) by red staining. Data are representative of three experiments. Scale bar equals 10 M; image is 400. (B) Rp-cAMPS, PTX, QEHA inhibition of C translocation. Neurons were pretreated for 1.5 hr with or without Rp-cAMPS, or overnight with or without PTX or Ad5QEHA, and then incubated with or without NPA or ethanol as in (A). Data represent at least three experiments. (C) The percentage of primary hippocampal neurons with Golgi, cytoplasmic, and nuclear staining of PKA C. Data are mean SEM (average of 15 cells/field, 3 fields/experiment, n 3). Asterisk indicates p 0.01, compared with control (one-way analysis of variance and Dunnett s test). (D) Synergy of C translocation. Neurons were preincubated with inhibitors as in (A) and (B) and treated with NPA and/or ethanol. Data are representative of three experiments. day 7 (data not shown). Viral expression in neurons and glia decreased by day 10 and was virtually absent by day 14. Rats recovered their preinjection levels of ethanol intake 14 days after injection. These data agree with the time course for results with other adenoviral vectors (Thanos et al., 2001). We show that a D2 agonist activates PKA signaling from increases in camp to CRE-mediated gene expression in a neural cell line. Our results suggest that the mechanism involves D2 coupling to Gi/o, release of dimers, activation of AC II and/or IV, translocation of PKA C to the nucleus, and subsequent PKA-dependent increases in gene expression. Ethanol also activates PKA C translocation and gene expression (Asher et al., 2002). We find remarkable synergy between D2 and ethanol-induced activation. Subthreshold concentrations of NPA or ethanol, which have no effect alone, when added together induce maximal translocation of PKA C and activation of CRE-mediated gene expression. Release of dimers is required for synergy. Synergy appears to be due to stimulation of AC II and/or IV concomitant with ethanol/a2 activation of AC via G s. The functional significance of PKA C translocation induced by subthreshold levels of NPA and ethanol is suggested directly by synergistic increases in CREmediated gene expression and indirectly by decreases in ethanol consumption caused by expression of a dominant-negative scavenger peptide in the NAc. There is extensive evidence that dopaminergic mechanisms contribute to ethanol consumption (Phillips et al., 1998; Cunningham et al., 2000; Risinger et al., 2000). In addition, PKA is also implicated in ethanol drinking. Mice heterozygous for G s with reduced AC activity and mice expressing a dominant-negative inhibitor of a PKA regulatory subunit with reduced PKA activity (Abel et al., 1997) both exhibit increased sensitivity to ethanol sedation and reduced ethanol preference and consumption (Wand et al., 2001). Increased sensitivity to ethanol intoxication has also been reported in Drosophila mutants that lack the gene cheapdate, an allele of amnesiac (Moore et al., 1998). This can be reversed by agents that increase camp/pka signaling. In contrast, mice lacking the RII regulatory subunit of PKA, with reduced camp-dependent PKA activity, exhibit resistance to eth- anol sedation and increased ethanol preference and consumption (Thiele et al., 2000). Similarly, Drosophila that are also deficient in the same Type II camp-dependent protein kinase show decreased sensitivity to etha- nol (Park et al., 2000). It appears, therefore, that the relationship between PKA activity and drinking behavior is not well understood. Here we report that overexpres- sion of dominant-negative scavenger peptides block D2- and ethanol-induced PKA C translocation, gene expression, and synergy, probably by preventing ac- tivation of AC. We also report that expression of a inhibitor peptide in rat medial NAc reduces voluntary ethanol intake, while water consumption is not decreased. Our data provide evidence to suggest that dimers in a specific brain reward location are required to sustain voluntary ethanol consumption. Moreover, this finding depends on direct intervention not affected by developmental compensatory changes, which limit interpretation of data derived from genetically engi- neered mice and Drosophila. Figure 7 illustrates a proposed interrelationship be- tween D2 (red arrows) and ethanol (blue arrows) activa- tion of PKA signaling. D2 and A2 receptors activate AC via and G s, respectively. Ethanol also promotes the

7 Synergy of D2 and Ethanol Requires Dimers 739 Figure 6. Expression of ARK1 Inhibitor Peptide in the NAc Reduces Alcohol Consumption Rats were injected with Ad5 ARK1 (n 9) or Ad5LacZ (n 9) viral vector into NAc. (A) Mean SEM ethanol intake (g/kg) is presented for the preinjection baseline period (average of 7 days) and 7 and 14 days after injection. Asterisk indicates significantly different from Ad5LacZ control and Ad5 ARK1 day 14, p Hatchmark indicates significantly different from Ad5 ARK1, baseline, p (B) Mean SEM water intake (ml). (C and D) Ethanol intake as a function of time following Ad5 ARK1 (C) or Ad5LacZ (D) injection. Dashed line indicates baseline. Asterisk indicates p 0.01 versus baseline; hatchmark indicates p 0.05 versus baseline. release of dimers from Gi/o while also activating the A2 Gs-coupled pathway. The resulting increase in camp levels leads to release of the C subunit(s) from regulatory subunit(s) of the PKA holoenzyme and translocation of C to the nucleus where it phosphorylates CREB to initiate gene transcription from CRE. Because D2 and ethanol separately activate the same PKA pathway, we asked whether NPA and ethanol might act synergistically at subthreshold concentrations of each agent. Indeed, we find that NPA-induced PKA C translocation and CRE-mediated gene transcription is synergistic with that produced by ethanol. D2-dependent release of from Gi/o appears to be required for synergy. This is suggested by our observations that the D2 antagonist spiperone or the Gi/o inhibitor PTX each blocks synergy of translocation and gene expression. Moreover, overexpression of inhibitor peptides that prevent activation of AC II and IV also blocks synergy. We previously reported that ethanol stimulates camp production via Gs-coupled A2 (Gordon et al. 1986; Sapru et al., 1994). However, we show here that the Gi/o inhibitor PTX blocks ethanol-induced PKA translocation and CRE-mediated gene expression. An unrecognized Gicoupled receptor could produce a PTX-sensitive increase in. This would require ethanol-induced neurotransmitter release. However, our data suggest that another transmitter receptor is not activated in NG108-15/D2 cells because adenosine receptor blockade completely blocks ethanol-induced increases in camp. Moreover, ethanol-induced PKA translocation in primary hippocampal neurons is not blocked by TTX, which inhibits neurotransmitter release. And, consistent with our findings of PTX sensitivity, ethanol can activate PTX- sensitive G proteins directly, bypassing receptor activation (Klinker et al., 1996). In any event, activation of another Gi-coupled receptor would not alter our conclusion that mediates synergy with ethanol and is required for voluntary alcohol consumption. Ethanol has been reported to stimulate AC VII directly, leading to increased PKA activity in HEK 293 cells (Yoshimura and Tabakoff, 1999). However, this cannot ex- plain our findings in neural NG108-15/D2 cells because ethanol stimulation of AC is blocked by A2 antagonists and inhibition. A2 is also required for synergy of NPA and ethanol. We have no evidence for A1 in NG cell lines either by binding of radiolabeled A1 antagonists or changes in AC activity by A1 agonists (A.S.G. and I.D., unpublished observations). Primary hippocampal neurons, however, express functional A1 (Swanson et al., 1995); ethanol-induced increases in extracellular adenosine in these cells probably activate A1 and A2 receptors. Thus, A1 or A2 antagonists block translocation. A1 probably increases camp via, since A1 cou- ples to Gi/o (Wise et al., 1999) just like D2. Taken together, our data suggest that adenosine receptors are critical for synergy induced by subthreshold concentrations of dopamine agonists and ethanol. This is consistent with reports of adenosine mediating ethanol-

8 Cell 740 not reduced by inhibition, strengthening the interpretation that decreased consumption is specific to ethanol. Although other brain regions were not tested, our results strongly suggest that dimers are required for voluntary alcohol consumption. The exact mechanism whereby inhibition reduces alcohol consumption is not known. However, we find in cell culture that both D2 (via dimers from Gi/o) and ethanol (via A2 and Gs) trigger the PKA-signaling pathway individually and synergistically. Because D2- or ethanol-induced PKA C translocation or synergy requires free subunits, we propose that reduced alcohol consumption produced by inhibition in vivo may reflect a disruption of D2/A2 synergy and/or D2- or ethanol-induced PKA C signaling. The model in Figure 7, based on results in NG108-15/ D2 cells, suggests that neurons expressing both D2 and A2 are uniquely sensitive to low or subthreshold concentrations of ethanol in the presence of endogenous dopamine. Synergy between dopamine and ethanol requires activation of D2 and release of together with adenosine activation of A2. It is tempting to speculate, therefore, that synergy may account for the central role of Figure 7. Synergy of PKA Signaling NAc in regulating ethanol intake. Thus, it is possible that Schematic representation of D2 and ethanol-induced PKA C transdopamine receptors might prevent, attenuate, or re- drugs that target and/or synergy of adenosine and location and CRE-mediated gene expression. A central role for Gi/o subunits is proposed. D2 signaling is indicated by red arrows; verse excessive drinking. Studies are underway to test ethanol signaling is indicated by blue arrows. Ethanol inhibits adeno- this possibility. sine uptake, leading to an increase in extracellular adenosine and activation of A2. This increases camp and activates PKA. D2 activa- Experimental Procedures tion releases dimers, which stimulate AC II and/or IV. Ethanol also induces the release of dimers directly, bypassing D2. The Reagents were from Sigma (St. Louis, MO) except where indicated. final common pathway includes increased camp levels and release Ham s F-12 medium was from GIBCO (Grand Island, NY), NPA and of C from PKA. PKA C translocates to the nucleus, phosphorylates spiperone were from Sigma/RBI (Natick, MA), Rp-cAMPS was from CREB, and increases CRE-mediated gene expression. BioLog (La Jolla, CA), p-cre-luciferase was from Stratagene (La Jolla, CA), and pcmv- -galactosidase was from QIAGEN (Hilden, induced responses in vivo. A1 agonists and antagonists Germany). increase and decrease ethanol-induced motor incoordination, respectively (Dar, 2001). In addition, A2-mediated Cell Culture NG108-15/D2 cells stably expressing the rat D2L receptor (Asai et increases in camp appear to account for ethanol-induced al., 1998) were grown on single-well slides in defined media for 3 release of endorphins from hypothalamic neurons days (Dohrman et al., 1996). Media were replaced daily 48 hr after (Boyadjieva and Sarkar, 1999). plating. On day 4, cells were treated as described in the Figure camp and PKA have been implicated in sensitivity to Legends. ethanol and alcohol consumption (Moore et al., 1998; Park et al., 2000; Thiele et al., 2000; Pandey et al., 2001; Rat Hippocampal Primary Culture Wand et al., 2001). Our findings suggest that synergy Cells were prepared from hippocampi of newborn (postnatal day 0) Sprague Dawley rat pups and plated at cells/slide between D2 and ethanol/a2 may be a pathophysiologic (Lissin et al., 1999). One-half of the medium was changed 1 day mechanism that contributes to these responses by after plating and weekly thereafter. Cells were treated on day 10 as changing PKA C localization and increasing camp- in the Figure Legends. dependent gene expression. Indeed, Thibault et al. (2000) have demonstrated that a subfamily of ethanol- Cell Culture Immunocytochemistry regulated genes is also regulated by camp. NG108-15/D2 cells were fixed with cold methanol (Dohrman et al., D2 and A2 receptors are densely expressed in the 1996). Primary hippocampal neurons were fixed in 4% formaldehyde striatum, substantial nigra, and olfactory bulb (Svenphosphate-buffered saline (PBS) and preincubated with blocking containing 120 mm sucrose for 10 min. Fixed cells were rinsed with ningsson et al., 1999; Levey et al., 1993). The striatum buffer followed by incubation with primary antibodies specific for includes the NAc, which regulates ethanol intake. Coex- PKA C (a gift from Dr. Susan Taylor, University of California, San pression of D2 and A2 mrna in the same neurons has Diego, CA), AC type II and type IV (Santa Cruz Biotechnology, Inc., been found in the striatum, but not in other D2 receptorcon International, Inc., Temecula, CA) (Dohrman et al., 1996). Cells Santa Cruz, CA), or neuron-specific nuclear protein (NeuN) (Chemi- expressing sites (Fink et al., 1992). Our results suggest that such neurons will be hypersensitive to ethanol. Our were washed, incubated with FITC- or TR-conjugated anti-rabbit or anti-mouse secondary antibody (Cappel, Aurora, OH) (diluted data also suggest that dimers released in NAc neu- 1:1000), rinsed, and coverslipped. No staining occurred when prirons appear to play a role in sustaining drinking behav- mary antibodies were preincubated with excess peptide antigen or ior. Thus, inhibition of function in medial NAc de- in the absence of FITC- or TR-conjugated anti-rabbit or anti-mouse creased voluntary intake of ethanol. Water intake was secondary antibody. Antibody specificity was confirmed by Western

9 Synergy of D2 and Ethanol Requires Dimers 741 blots. There were no bands after the primary antibodies were preabsorbed with antigen (data not shown). Confocal Microscopy Images were obtained as a single plane near the center of the cell with Bio-Rad 1024 scanning laser confocal microscope (Gordon et al., 2001) and processed using NIH Image and Adobe Photoshop software. Data were quantitated by two observers blind to experimental conditions. Cell Fractionation NG /D2 cells in 100 mm dishes ( cells/dish) were washed with cold PBS and lysed on ice in 0.5 ml lysis buffer containing 50 mm Tris-HCL (ph 7.4), 2.5 mm MgCL 2, 1 mm EDTA, 1 mm DTT, 10% glycerol, and protease inhibitors (0.1 mm phenylmethyl sulfonyl fluoride, 20 g/ml soybean trypsin inhibitor, 25 g/ml aprotinin, 25 g/ml leupeptin, and 1 mm sodium orthovanadate). Cells were homogenized by ten passes through a 26-gauge needle and centrifuged at 3000 rpm for 5 min at 4 C. The nuclear pellet was washed in lysis buffer. The supernatant was centrifuged for 20 min at 150,000 g to separate the membrane pellet from the cytosol. Western Blotting Proteins subjected to SDS/PAGE were transferred to polyvinylidene difluoride membranes and probed with antibodies against PKA C (1:1000) or AC II or IV (Santa Cruz). Secondary antibody was horse- radish peroxidase-linked goat anti-rabbit (1:1000) (NEN BioLabs, Beverly, MA). Proteins were detected using LumiGLO chemilumines- cence substrate (NEN BioLabs) and Kodak Biomax film. Construction and Production of Recombinant Adenoviral Vectors and Transfection of Cells Recombinant Ad5QEHA, Ad5SKEE, Ad5 ARK1, Ad5LacZ, and Ad5GFP vectors under the control of a CMV promoter were created by replacing the E1A gene of the adenovirus with QEHA peptide encoding residues 956 to 982 of AC2, a control peptide (SKEE) corresponding to the cognate region of AC3, the carboxyl terminus (Gly 495 to Leu 689 ) of bovine ARK1, LacZ, or the green fluorescence protein (GFP) gene, respectively. To make the recombinant adenoviruses, cosmids (Stratagene, La Jolla, CA) containing recombinant adenovirus DNA were transfected into 50% 80% confluent HEK 293 cells. Two weeks later, cells were harvested by brisk pipetting, resuspended in PBS, and lysed by three cycles of rapid freezing in ethanol/dry ice and thawing at 37 C. The lysate was then centrifuged at 600 g to remove debris. For large-scale production, HEK293 cells were reinfected with the virus stock and harvested 3 days later. Viruses were purified by CsCl banding and dialyzed three times against PBS. To determine viral titers, HEK293 cells were seeded in 6-well plates and infected with dilutions of the virus stocks 24 hr later. Two hours after infection, medium was removed and cells overlaid with medium containing 2% fetal bovine serum and 1% agarose. Two weeks later, the number of virus plaques was recorded. To transfect NG108-15/D2 cells and primary neurons from hippocampus, culture media were replaced with media containing Ad5QEHA, Ad5SKEE, Ad5 ARK1, or Ad5GFP at a multiplicity of infection of 100 and the cells cultured hr prior to use. CRE-Luciferase Reporter Assay NG108-15/D2 cells were plated at cells/100 mm plate, grown in defined medium for 3 days, transfected with 10 g of CRE-luciferase construct using the LipofectAMINE reagent (GIBCO) as recommended by the manufacturer, and cultured for 24 hr before seeding into 12-well plates. Following a further 2 day culture, cells were treated for 10 min with or without 50 nm NPA or 100 mm ethanol, with or without 0.5 nm NPA or 25 mm ethanol alone or in combination after preincubation with or without 10 M BW A1434U, 10 M spiperone, or 20 M Rp-cAMPS for 30 min, 50 ng/ml PTX or QEHA overnight. Cells were washed with PBS and cultured for a total of 5 hr. Cell extracts were prepared and luciferase activity assayed (Promega, Madison, WI) using a Rosys Anthos Lucy2 microplate lumino- meter (Anthos Labtec Instruments, Austria). Results are expressed as percent increase over control. Data represent three or more separate experiments from triplet wells. Luciferase activities were nor- malized for transfection efficiency determined in cells transfected in parallel with pcmv galactosidase as measured using a kit from Strategene (La Jolla, CA). Two-Bottle Choice Paradigm Voluntary fluid intake and weight were measured daily in singly housed g Male Long Evans rats. Following a 4 day exposure to 10% ethanol as the only liquid source, rats were allowed to choose 10% ethanol in tap water or tap water. Forty-five days later, rats received permanent bilateral cannulae in the NAc (see below). After days to reestablish presurgery baseline drinking, rats received microinjections of either Ad5 ARK1 or Ad5LacZ ( -galactosidase control) and were studied for 14 days. Stereotaxic Surgery, Cannulae Implantation, and Viral Vector Microinjection Rats were anesthetized with isoflurane (Aerrane, Baxter, Deerfield, IL) and implanted with bilateral cannulae (26-gauge; Plastics One, Inc., Roanoke, VA) aimed to the NAc: 1.7 mm anteroposterior (AP), 1.0 mm mediolateral (ML), 6.4 mm dorsoventral (DV) (rela- tive to bregma; Paxinos and Watson, 1998). Twelve to fifteen days later, anesthetized rats received 2 l (10 12 pfu/ml)/side of Ad5 ARK1 or Ad5LacZ viral vector constructs over 10 min through 33-gauge injection cannulae that extended 1 mm below the implanted cannulae (Plastics One, Inc.) via an infusion pump (model 11, Harvard Apparatus, Inc., Holliston, MA). Cannulae were left in place an additional 10 min to ensure diffusion. All procedures performed in this report were approved by our Institutional Animal Care and Use Committee. Histologic Verification of Cannula Placement and Ad5LacZ Expression Rats were deeply anesthetized with isoflurane and transcardially perfused with 4% paraformaldehyde. Cannula placement was evaluated in thionin-stained 50 l coronal brain slices. Only rats with NAc placements were included in the groups for statistical analysis. LacZ expression in a separate group of animals (n 10) was monitored by X-gal histochemistry on postinjection days 2, 4, 7, 10, and 14. Acknowledgments We thank Glaxo/Wellcome for BW A1434U and Dr. Susan Taylor, University of California, San Diego, for antibodies to PKA C. We thank Antonello Bonci, Marc Diamond, William Mailliard, Michael Miles, and Dorit Ron for criticisms and suggestions. We thank Jenni- fer Whistler for instructing and demonstrating cell fractionation pro- cedures, Jeanne Arch for expert manuscript preparation, and Ling Wang for preparing primary hippocampal neurons. Support was from National Institutes of Health grants to A.S.G. and I.D. (AA10030, AA10039), funds provided by the State of California for medical research on ethanol and substance abuse through the University of California, San Francisco (I.D. and P.H.J.), and by the Irwin Green Research Fund in Neurosciences and the Leo and Julia Forchheimer Center for Molecular Genetics at the Weizmann Institute of Science (S.F.). Received: January 4, 2002 Revised: April 25, 2002 References Abel, T., Nguyen, P.V., Barad, M., Deuel, T.A.S., Kandel, E.R., and Bourtchouladze, R. (1997). Genetic demonstration of a role for PKA in the late phase of LTP and in hippocampus-based long-term mem- ory. Cell 88, Asai, K., Ishii, A., Yao, L., Diamond, I., and Gordon, A.S. (1998). Varying effects of ethanol on transfected cell lines. Alcohol. Clin. Exp. Res. 22, Asher, O., Cunningham, T.D., Yao, L., Gordon, A.S., and Diamond, I. (2002). Ethanol stimulates camp responsive element-mediated transcription via CREB and camp-dependent protein kinase. J. Pharmacol. Exp. Ther. 301, Baker, L.P., Nielsen, M.D., Impey, S., Hacker, B.M., Poser, S.W.,

10 Cell 742 Chan, M.Y., and Storm, D.R. (1999). Regulation and immunohisto- Klinker, J.F., Lichtenberg-Kraag, B., Damm, H., and Rommelspacher, chemical localization of betagamma-stimulated adenylyl cyclases H. (1996). Activation of pertussis toxin-sensitive G-proteins in mouse hippocampus. J. Neurosci. 19, in membranes of SH-SY5Y human neuroblastoma cells and bovine Boyadjieva, N., and Sarkar, D.K. (1999). Effects of ethanol on basal transducin by ethanol. Neurosci. Lett. 213, and adenosine-induced increases in -endorphin release and intra- Koch, W.J., Hawes, B.E., Inglese, J., Luttrell, L.M., and Lefkowitz, cellular camp levels in hypothalamic cells. Brain Res. 824, R.J. (1994). Cellular expression of the carboxyl terminus of a G Chakrabarti, S., Rivera, M., Yan, S.-Z., Tang, W.-J., and Gintzler, protein-coupled receptor kinase attenuates G beta gamma-medi- A.R. (1998). Chronic morphine augments G /Gs stimulation of ated signaling. J. Biol. Chem. 269, adenylyl cyclase: relevance to opioid tolerance. Mol. Pharmacol. Levey, A.J., Hersch, S.M., Rye, D.B., Sunahara, R.K., Niznik, H.B., 54, Kitt, C.A., Price, D.L., Maggio, R., Brann, M.R., and Ciliax, B.J. (1993). Chen, J., DeVivo, M., Dingus, J., Harry, A., Li, J., Sui, J., Carty, D.J., Localization of D1 and D2 dopamine receptors in brain with subtype- Blank, J.L., Exton, J.H., Stoffel, R.H., et al. (1995). A region of adenylyl specific antibodies. Proc. Natl. Acad. Sci. USA 90, cyclase 2 critical for regulation by G protein subunits. Science Lissin, D.V., Carroll, R.C., Nicoll, R.A., Malenka, R.C., and von Zastrow, 268, M. (1999). Rapid, activation-induced redistribution of iono- Constantinescu, A., Diamond, I., and Gordon, A.S. (1999). Ethanolrosci. tropic glutamate receptors in cultured hippocampal neurons. J. Neuinduced translocation of camp-dependent protein kinase to the 19, nucleus. Mechanism and functional consequences. J. Biol. Chem. Lledo, P.M., Homburger, V., Bockaert, J., and Vincent, J.D. (1992). 274, Differential G protein mediated coupling of D2 dopamine receptors Cunningham, C.L., Howard, M.A., Gill, S.J., Rubinstein, M., Low, to K and Ca 2 currents in rat anterior pituitary cells. Neuron 8, M.J., and Grandy, D.K. (2000). Ethanol-conditioned place preference is reduced in dopamine D2 receptor-deficient mice. Pharmac. Bio- Moore, M.S., DeZazzo, J., Luk, A.Y., Tully, T., Singh, C.M., and chem. Behav. 67, Heberlein, U. (1998). Ethanol intoxication in Drosophila: genetic and Dar, M.S. (2001). Modulation of ethanol-induced motor incoordinapathway. pharmacological evidence for regulation by the camp signaling tion by mouse striatal A(1) adenosinergic receptor. Brain Res. Bull. Cell 93, , Nagy, L.E., Diamond, I., Collier, K., Lopez, L., Ullman, B., and Gordon, Dohrman, D.P., Diamond, I., and Gordon, A.S. (1996). Ethanol causes A.S. (1989). Adenosine is required for ethanol-induced heterologous translocation of camp-dependent protein kinase catalytic subunit desensitization. Mol. Pharmacol. 36, to the nucleus. Proc. Natl. Acad. Sci. USA 93, Nagy, L.E., Diamond, I., Casso, D.J., Franklin, C., and Gordon, A.S. Dohrman, D.P., Chen, H.-M., Gordon, A.S., and Diamond, I. (2002). (1990). Ethanol increases extracellular adenosine by inhibiting aden- Ethanol-induced translocation of PKA occurs in two phases: control osine uptake via the nucleoside transporter. J. Biol. Chem. 265, by different molecular mechanisms. Alcohol. Clin. Exp. Res. 6, Obadiah, J., Avidor-Reiss, T., Fishburn, C.S., Carmon, S., Bayewitch, Federman, A.D., Conklin, B.R., Schrader, K.A., Reed, R.R., and M., Vogel, Z., Fuchs, S., and Levavi-Sivan, B. (1999). Adenylyl cy- Bourne, H.R. (1992). Hormonal stimulation of adenylyl cyclase clase interaction with the D2 dopamine receptor family; differential through Gi-protein subunits. Nature 356, coupling to Gi, Gz, and Gs. Cell. Mol. Neurobiol. 19, Fink, J.S., Weaver, D.R., Rivkees, S.A., Peterfreund, R.A., Pollack, O Kane, E.M., and Stone, T.W. (1998). Interaction between adeno- A.E., Adler, E.M., and Reppert, S.M. (1992). Molecular cloning of the sine A1 and A2 receptor-mediated responses in the rat hippocamrat A2 adenosine receptor: selective co-expression with D2 dopamine pus in vitro. Eur. J. Pharmacol. 362, receptors in rat striatum. Brain Res. Mol. Brain Res. 14, Pandey, S.C., Roy, A., and Mittal, N. (2001). Effects of chronic ethanol intake and its withdrawal on the expression and phosphorylation of Goldsmith, S.K., and Joyce, J.N. (1994). Dopamine D2 receptor ex- the CREB gene transcription factor in rat cortex. J. Pharmacol. Exp. pression in hippocampus and parahippocampal cortex of rat, cat, Ther. 296, and human in relation to tyrosine hydroxylase-immunoreactive fi- Park, S.K., Sedore, S.A., Cronmiller, C., and Hirsh, J. (2000). Type bers. Hippocampus 4, II camp-dependent protein kinase-deficient Drosophila are viable Gordon, A.S., Collier, K., and Diamond, I. (1986). Ethanol regulation but show developmental, circadian, and drug response phenotypes. of adenosine receptor-stimulated camp levels in a clonal neural cell J. Biol. Chem. 275, line: an in vitro model of cellular tolerance to ethanol. Proc. Natl. Paxinos, G., and Watson, C. (1998). The Rat Brain in Stereotaxic Acad. Sci. USA 83, Coordinates, 4th Edition (San Diego: Academic Press). Gordon, A.S., Yao, L., Jiang, Z., Fishburn, C.S., Fuchs, S., and Dia- Phillips, T.J., Brown, K.J., Burkhart-Kasch, S., Wenger, C.D., Kelly, mond, I. (2001). Ethanol acts synergistically with a D2 dopamine M.A., Rubinstein, M., Grandy, D.K., and Low, M.J. (1998). Alcohol agonist to cause translocation of protein kinase C. Mol. Pharmacol. preference and sensitivity are markedly reduced in mice lacking 58, dopamine D2 receptors. Nat. Neurosci. 1, Hamm, H.E. (1998). The many faces of G protein signaling. J. Biol. Piomelli, D., Pilon, C., Giros, B., Sokoloff, P., Martres, M.P., and Chem. 273, Schwartz, J.C. (1991). Dopamine activation of the arachidonic acid Hodge, C.W., Samson, H.H., and Chappelle, A.M. (1997). Alcohol cascade as a basis for D1/D2 receptor synergism. Nature 353, self-administration: further examination of the role of dopamine receptors in the nucleus accumbens. Alcohol. Clin. Exp. Res. 21, 1083 Risinger, F.O., Freeman, P.A., Rubinstein, M., Low, M.J., and Grandy, D.K. (2000). Lack of operant ethanol self-administration in dopamine Hyman, S.E., and Malenka, R.C. (2001). Addiction and the brain: the D2 receptor knockout mice. Psychopharmacology (Berl.) 152, neurobiology of compulsion and its persistence. Nat. Rev. Neurosci , Robbins, T.W., and Everitt, B.J. (1999). Drug addiction: bad habits Imperato, A., and Di Chiara, G. (1986). Preferential stimulation of add up. Nature 398, dopamine release in the nucleus accumbens of freely moving rats Ruiz-Gomez, A., and Mayor, F.J. (1997). Beta-adrenergic receptor by ethanol. J. Pharmacol. Exp. Ther. 239, kinase (GRK2) colocalizes with beta-adrenergic receptors during Inglese, J., Luttrell, L.M., Iniguez-Lluhi, J.A., Touhara, K., Koch, W.J., agonist-induced receptor internalization. J. Biol. Chem. 272, 9601 and Lefkowitz, R.J. (1994). Functionally active targeting domain of the beta-adrenergic receptor kinase: an inhibitor of G beta gamma- Sapru, M.K., Diamond, I., and Gordon, A.S. (1994). Adenosine receptors mediated stimulation of type II adenylyl cyclase. Proc. Natl. Acad. mediate cellular adaptation to ethanol in NG cells. J. Sci. USA 91, Pharmacol. Exp. Ther. 271,