Convergent evidence from alcohol-dependent humans and rats for a hyperdopaminergic state in protracted abstinence

Significance

A major hypothesis in the addiction field suggests there are deficits in dopamine signaling during abstinence. This hypodopaminergic state is considered a driving mechanism for the relapsing course of the disorder. Experimental support for this view comes mostly from human PET studies that found reduced striatal D2-like receptors in alcoholics. Here we report on surprising findings from postmortem brains of deceased alcoholics and alcohol-dependent rats that show no differences in D2-like receptor binding during withdrawal and prolonged abstinence. Instead we observe a dynamic regulation of D1 receptors, dopamine transporter, dopamine release properties, and phenotypic characteristics that all are in line with a hyperdopaminergic state during protracted abstinence. We propose that both hypo- and hyperdopaminergia are states of vulnerability to relapse.

Abstract

A major hypothesis in addiction research is that alcohol induces neuroadaptations in the mesolimbic dopamine (DA) system and that these neuroadaptations represent a key neurochemical event in compulsive drug use and relapse. Whether these neuroadaptations lead to a hypo- or hyperdopaminergic state during abstinence is a long-standing, unresolved debate among addiction researchers. The answer is of critical importance for understanding the neurobiological mechanism of addictive behavior. Here we set out to study systematically the neuroadaptive changes in the DA system during the addiction cycle in alcohol-dependent patients and rats. In postmortem brain samples from human alcoholics we found a strong down-regulation of the D1 receptor- and DA transporter (DAT)-binding sites, but D2-like receptor binding was unaffected. To gain insight into the time course of these neuroadaptations, we compared the human data with that from alcohol-dependent rats at several time points during abstinence. We found a dynamic regulation of D1 and DAT during 3 wk of abstinence. After the third week the rat data mirrored our human data. This time point was characterized by elevated extracellular DA levels, lack of synaptic response to D1 stimulation, and augmented motor activity. Further functional evidence is given by a genetic rat model for hyperdopaminergia that resembles a phenocopy of alcohol-dependent rats during protracted abstinence. In summary, we provide a new dynamic model of abstinence-related changes in the striatal DA system; in this model a hyperdopaminergic state during protracted abstinence is associated with vulnerability for relapse.

About 10% of the total burden of disease in developed countries is caused by alcohol use alone (1). A large proportion of alcohol-related disability results from alcohol addiction. The condition affects more than 12% of the United States population at some point in their lives and is one of the most prevalent psychiatric disorders in Europe (2, 3). The relapsing course of alcoholism is associated with compulsive drinking, loss of control over intake, and emergence of a negative emotional state during abstinence (4). Afflicted individuals go through repeated cycles of alcohol intoxication and withdrawal leading to persistent alterations in brain activity that are hypothesized to drive relapse and compulsive alcohol use even long after detoxification (5).

Seminal studies in experimental animals established that alcohol’s rewarding properties are associated with increased dopamine (DA) in regions such as the nucleus accumbens (Acb) (6), whereas withdrawal after chronic alcohol use decreases DA neurotransmission (7). In humans, the binding of a DA receptor ligand, typically one for the D2-like receptor subgroup, i.e., [¹¹C] raclopride, can be monitored by PET. Displacement of the radioligand provides an indirect measure of DA release and has been used to demonstrate alcohol-evoked DA release in the accumbens of healthy social drinkers (8, 9). On the other hand, a blunted response of the DA system and reduced availability of the D2-like receptor was found in alcoholics (10–17). The collective interpretation of these studies postulates a hypodopaminergic state, characterized by low extracellular DA levels and reduced D2-like receptor availability in areas of the mesolimbic system, which may drive relapse behavior in alcoholism (18, 19). However, this interpretation of in vivo receptor availability seen in PET studies is inherently ambiguous, because decreased signal can be caused either by a reduced number of receptors or by increased ligand concentration. Furthermore, unchanged or even increased striatal D2-like binding in alcohol-abstinent patients has been found also (20–22). Naltrexone, one of the few medications approved by US Food and Drug Administration for treatment of relapse, reduces alcohol-induced accumbal DA release (23), and this effect seems argue against the importance of a hypodopaminergic state for increased propensity for relapse. Thus, clarification of the role of DA during abstinence is highly important for the development of novel therapeutic strategies.

Here, we used brain samples from deceased alcoholics and controls to study both transcriptional and binding levels for DA D1 and D2-like receptors and transporter (DAT). To investigate the time course of neuroadaptations within the DA system during abstinence, we used an established animal model in which dependence is induced by chronic intermittent intoxication with alcohol vapor leading to long-lasting neuronal and behavioral adaptations that persist even in the absence of the drug (24–27). Although this model is based on experimenter-controlled intoxication, as opposed to the largely voluntary drinking seen in humans, it has been valuable in establishing mechanisms underlying the high propensity for relapse in addicted individuals, i.e., a chronic hyperactivity of central stress systems when access to the drug is prevented (25, 26, 28), a phenomenon that Koob and Le Moal (29) term the “dark side” of addiction. In this animal model we performed microdialysis, electrophysiology, and behavioral studies to demonstrate a hyperdopaminergic state—a condition in which extracellular DA is elevated—during protracted abstinence. The hyperdopaminergic state here refers to the basal tonic state of the system and may underlie a diminished response capability for DA release (30).

Results

Postmortem Brain Analysis Suggests a Hyperdopaminergic State in Human Alcoholics.

Ten alcoholic and 10 control subjects were included in the study. All alcoholics had a history of daily alcohol intake of more than 80 g/d, and the control cases had an average daily consumption of less than 20 g. Subjects were free of detectable alcohol levels at their time of death (see highlighted core sample set in Table S1). There was no significant difference in age, postmortem interval, or brain pH between the groups (Welch’s t test for all variables: P > 0.1).

Table S1.

Demographical data and tissue characteristics of deceased human subjects used for DAT and D1 and D2 receptor autoradiography experiments

All subjects were males of European descent. Tissue was collected at the New South Wales Tissue Resource Centre, University Sydney, Sydney. Core samples of alcoholics and controls are highlighted. BAC, blood alcohol concentration; PMI, postmortem interval.

Sections from the ventral striatum (VS, including the Acb) and nucleus caudatus (NC) were used to measure ligand binding for D1 ([³H]-SCH23390) and D2-like ([³H]-raclopride) receptors and for DAT ([³H]-mazindol) by autoradiography. The number of D1-binding sites in both the VS and NC was strongly decreased in alcoholics as compared with controls (VS: 59%, F_1,15 = 31.7, P < 0.001; NC: 61%, F_1,16 = 104.2, P < 0.001) (Fig. 1A). In contrast, no differences were observed for D2-binding sites (VS: F_1,16 = 0.005, P > 0.5; NC: F_1,15 = 1.3, P > 0.5) (Fig. 1B). Furthermore, samples from alcoholics showed a highly significant reduction in DAT-binding sites (VS: 62%, F_1,14 = 139.8, P < 0.001; NC: 56%, F_1,14 = 65.4, P < 0.001) (Fig. 1C).

Fig. 1.

Expression analysis of the dopaminergic system in postmortem brains of heavy alcoholics suggests a hyperdopaminergic state. (A–C) Bar graphs showing quantitative analysis of D1 receptors (A), D2-like receptors (B), and DAT-binding sites (C) in postmortem striatal brain sections of heavy alcoholics (red bar in A, blue bar in B, green bar in C) and controls (white bars). D1 and DAT are strongly decreased in striatal brain regions, but D2 is not altered. Data are expressed as means (expressed in femtomoles per milligram) ± SEM, n = 8 or 9 per group. (D) Schematic illustration showing a coronal section of the human striatal forebrain region including the NC and VS (including the accumbens).

To support the finding of reduced striatal D1 binding further, we studied a separate, larger cohort of subjects including alcoholics either positive or negative for blood alcohol concentration (BAC) (0.183 ± 0.145 g/100 mL or zero in intoxicated and nonintoxicated alcoholics, respectively, n = 10 per group) and controls (n = 30) (Table S1). As in the first cohort, both striatal regions showed reduced D1 binding in alcoholics with no differences between intoxicated and nonintoxicated subjects (NC: F_2,43 =7.6, P = 0.001; VS: F_2,43 =10.4, P = 0.0002) (Fig. S1). Potentially confounding factors such as tissue pH, postmortem interval, age, and smoking status were included as covariates in the analysis but did not cause any significant effects. In contrast to the protein findings, quantitative real-time PCR (qRT-PCR) to assess mRNA levels for DRD1 and DRD2 did not show any differences between the groups (Table S2). SLC6A3 mRNA encoding the DAT was not determined, because transcripts are located mostly in cell bodies of the nigrostriatal and ventral tegmental area (VTA) neurons.

Fig. S1.

D1 receptor-binding sites are strongly decreased in striatal postmortem tissue of human alcoholics. The bar graphs show D1 receptor-binding sites as measured by [³H]-SCH23390 autoradiography of controls (white bar) compared with not-intoxicated (hatched bar) and intoxicated (black bar) alcoholic subjects. No differences have been detected between not-intoxicated and intoxicated alcoholics. Statistics was performed by ANOVA followed by Fisher’s LSD post hoc test (**P < 0.01, ***P < 0.001 vs. controls). Data (in femtomoles per milligram) are expressed as mean ± SEM; n = 9–26 per group.

Table S2.

No changes are seen in the expression of DA D1 and D2 receptor mRNA in postmortem striatal tissue from heavy-use alcoholics

Transcript	Region	Controls, ΔCt	Alcoholics, ΔCt	ΔΔCt	F	P
DRD1	VS	4.91 ± 0.17	5.03 ± 0.30	−0.12	(1,18) 0.13	0.73
	NC	4.37 ± 0.10	4.35 ± 0.11	0.02	(1,18) 0.02	0.89
DRD2	VS	0.60 ± 0.30	0.84 ± 0.31	0.24	(1,18) 0.30	0.60
	NC	3.99 ± 0.10	4.23 ± 0.22	−0.24	(1,18) 0.93	0.35

The table shows the results of qRT-PCR for DA DRD1 and DRD2 mRNA; data are expressed as mean ± SEM, n = 9–10 per group. GAPDH Ct values in NC: alcoholics: 22.0 ± 0.3; controls: 21.6 ± 0.1; GAPDH Ct values in VS: alcoholics: 23.6 ± 0.3; controls: 23.2 ± 0.2.

To provide convergent evidence for this surprising finding, we next performed a systematic search and a meta-analysis of existing literature on DA concentrations and its metabolites during abstinence in alcohol-dependent rats and then examined the dopaminergic system at different time points in an established animal model of alcoholism.

Alcohol-Dependent Rats Mirror the Hyperdopaminergic State Observed in Human Alcoholics.

The meta-analysis was based on 16 published studies in rats (a total of 192 rats chronically exposed to ethanol). This analysis revealed an increase in DA release on day 0, followed by a decrease on days 1–3, and an increase again on days 7 and 21 of abstinence (Fig. 2A; for detailed information, see SI Materials and Methods). Accumbal release of DA metabolites [3,4-dihydroxyphenylacetic acid (DOPAC) and homovanillic acid (HVA)] followed a pattern similar to that observed for DA (Fig. S2). In summary, the meta-analysis suggests that dynamic changes in accumbal DA levels occur during abstinence with a hypodopaminergic state during acute withdrawal followed by a hyperdopaminergic state later on.

Fig. 2.

Analyses of the dopaminergic system in alcohol-dependent rats reveals a hyperdopaminergic state in long-term abstinence. (A) The time course of DA release in the Acb was modeled by a meta-analysis of 192 rats derived from 16 animal studies by continuous interpolation of the averages of experimental values with respect to the time of measurement during abstinence. This quantitative analysis suggests a robust pattern of dynamic changes in DA concentrations. During the first 6 d of withdrawal, the DA concentrations decline to 30% of the baseline concentrations (hypodopaminergic state) but then increase again to a hyperdopaminergic state. (B–D) Regulation of D1 receptors (red bars), D2-like receptors (blue bars), and DAT-binding sites (green bars) at different days of abstinence in the AcbS (B), AcbC (C), and CPu (D) of alcohol-dependent rats vs. control rats (set as 0% baseline at each time point). Rats were intermittently exposed to ethanol vapor for 7 wk and were killed immediately after the last exposure cycle (on day 0) or after 1, 3, 7, or 21 d of abstinence. D1 and DAT are dynamically increased and decreased at different times during abstinence, but D2-like binding levels remain unaffected. Statistical analysis was performed by two-way ANOVA followed by Fisher’s post hoc test. Data are expressed as percent of controls ± SEM, n = 4–8 per group. For expression levels in controls at each time point, see Table S3. The shaded areas in the figure indicate a hypo- or a hyperdopaminergic state during abstinence.

Fig. S2.

Meta-analysis predicts a hyperdopaminergic state during protracted alcohol abstinence. DA and its metabolites DOPAC and HVA were investigated during alcohol abstinence by a meta-analysis. During the first 6 d of abstinence DA, HVA, and DOPAC concentrations decline by up to 30% of the baseline condition (hypodopaminergic state). Afterward, concentrations rise above baseline levels (hyperdopaminergic state). The inset shows the dynamic regulation of DA, DOPAC, and HVA during the first 24 h of abstinence.

We then performed a time-course experiment for D1-, D2-like–, and DAT-binding sites during abstinence in alcohol-dependent rats. To be in line with our human studies we focused on the Acb shell (AcbS) and core (AcbC) and the caudate putamen (CPu) (31). Rats exposed to intermittent cycles of alcohol vapor for 7 wk (26, 27) were killed at day 0, 1, 3, 7, or 21 of abstinence as in refs. 24 and 32. DA receptor/transporter-binding sites were quantitatively analyzed (33) and are presented as normalized data respective to control group and time in Fig. 2 B–D. Raw data for controls are summarized in Table S3, and representative images of autoradiographies are shown in Fig. S3.

Table S3.

DAT, D1, and D2-like binding levels in the AcbC, the AcbS, and the CPu at different time points in air-exposed control rats

	D1, fmol/mg	n	D2, fmol/mg	n	DAT, fmol/mg	n
AcbS
0 d	4086.96 ± 78.69	7	328.19 ± 11.30	7	837.00 ± 58.73	5
1 d	3593.05 ± 159.31	6	281.85 ± 24.06	6	755.06 ± 50.06	5
3 d	3533.65 ± 118.09	6	292.87 ± 26.32	4	717.58 ± 59.57	6
7 d	3211.08 ± 129.53	4	267.00 ± 25.44	5	706.73 ± 56.56	7
21 d	3341.14 ± 122.21	6	271.99 ± 10.36	6	786.96 ± 61.20	6
AcbC
0 d	3375.96 ± 63.32	6	342.66 ± 9.57	6	990.85 ± 51.22	6
1 d	2880.18 ± 250.63	7	301.64 ± 6.05	6	1151.70 ± 31.08	7
3 d	3252.50 ± 79.18	6	318.10 ± 10.26	5	1074.52 ± 58.41	6
7 d	2319.81 ± 101.37	8	280.41 ± 21.73	7	1026.81 ± 25.07	7
21 d	2650.33 ± 117.17	7	277.88 ± 14.93	6	1208.06 ± 28.80	7
CPu
0 d	3709.91 ± 58.07	8	627.32 ± 7.98	7	1432.58 ± 53.24	4
1 d	3297.98 ± 45.14	6	610.63 ± 11.03	8	1928.93 ± 73.43	7
3 d	3313.21 ± 69.84	6	614.41 ± 13.30	4	2005.07 ± 34.54	6
7 d	3154.57 ± 69.10	8	627.85 ± 9.01	7	1918.01 ± 32.15	7
21 d	3013.47 ± 74.09	7	573.43 ± 10.56	7	2045.78 ± 32.36	7

Data are expressed as mean values ± SEM; n, number of animals per group.

Fig. S3.

Representative images showing D1 ([³H]SCH23390), D2-like ([³H]raclopride), and DAT ([³H]mazindol) binding (T; total binding) on a coronal striatal rat brain section. Nonspecific (NS) binding was determined on adjacent section by adding flupenthixol (D1), sulpiride (D2-like), and nomifensine (DAT) to radioligand.

AcbS.

D1 and DAT varied as a function of time from exposure (two-way ANOVA, treatment × time; D1 main effect: F_4,54 = 4.6, P < 0.01; DAT main effect: F_4,54 = 4.8, P < 0.01). D1- and DAT-binding sites were strongly regulated at several time points between acute intoxication (day 0) and day 21 of abstinence. On day 0, animals were killed immediately after the last cycle of exposure to ethanol vapor, having positive BACs of 273 ± 52 mg/dL. In comparison to controls, D1 was significantly reduced, by 11%, at this time point but reached control levels 1 d later (day 1). After 3 d of abstinence (day 3), D1 increased (10%, P = 0.07), and this effect reached significance after 7 d. After a further 2 wk of abstinence (day 21), D1 was decreased by 14% (Fig. 2B). The binding sites of DAT at these time points were regulated differentially. On day 0, DAT was increased by 22%; however, this effect failed to reach significance (P = 0.07). On day 1, DAT was significantly decreased by 33% and returned to control levels on day 7. After 21 d of alcohol abstinence, DAT again was significantly reduced by 35% (Fig. 2B). For the AcbC, the D1- and DAT-binding sites followed a pattern similar to that in the AcbS (D1 main effect: F_4,58 = 7.9, P < 0.001; DAT main effect F_4,61 = 6.2, P < 0.001). Post hoc analysis revealed that D1 was decreased on day 0 by 15% and increased on day 7 by 30% relative to baseline. On day 21 D1 again was decreased by 15%. DAT binding was significantly increased by 24% on day 0, returned to control levels on days 1–7, and then tended to decrease (11%, P = 0.05) on day 21 of abstinence (Fig. 2C). In the CPu and in the dorsal striatum D1 and DAT binding paralleled the temporal pattern in the AcbS with the exception that D1 binding was not down-regulated at day 21 (D1 main effect: F_4,58 = 10.8, P < 0.001; DAT main effect: F_4,55 = 25.2, P < 0.001). With chronic alcohol use, D1 binding decreased by 14% on day 0 and increased significantly on day 3 (8%) and on day 7 (11%) of abstinence. DAT binding was strongly increased on day 0 (34%) and decreased on day 1 (9%) and day 21 (13%) (Fig. 2D). D2-like binding sites were not changed at any time point in any region (Fig. 2 B–D).

Together these data support our notion derived from the meta-analysis that abstinence is characterized by changes in the dynamics of the components of the striatal DA system, with the profile of the intoxicated state being clearly separable from that of acute withdrawal and that of protracted abstinence. The human postmortem findings appear most closely related to those seen with 21-day protracted abstinence in rodents.

Characterization of the Hyperdopaminergic State in Protracted Abstinence.

Extracellular DA levels in the AcbS after 21 d of abstinence were analyzed via in vivo microdialysis. The basal dialysate DA concentrations were significantly elevated in alcohol-dependent rats (F_1,26 = 2.7, P < 0.05) (Fig. 3A). When different doses of ethanol (0, 1, or 2 g/kg, i.p.) were applied, control rats showed an ethanol-induced increase in extracellular DA (on average a 49 ± 33% increase from baseline) after the injection of 2 g/kg ethanol (i.p.), whereas alcohol-dependent rats displayed a blunted response to ethanol treatment (on average a 9 ± 49% increase from baseline) (Fig. 3B). Repeated-measurement ANOVA revealed a significant effect of alcohol injections (F_1,14 = 7.1, P < 0.05), a trend for treatment (alcohol-dependent vs. control, F_1,14 = 3.8, P = 0.07), but no interaction effect (F_1,14 = 0.8, P > 0.5). Further support for increased striatal DA release came from in situ hybridization for TH (tyrosine hydroxylase) mRNA showing an increase in the substantia nigra pars compacta (SNc) by 31% in 3-wk-abstinent rats (F_1,10 = 18.6, P < 0.01) but no changes in the VTA (Fig. S4).

Fig. 3.

The hyperdopaminergic state in 3-wk alcohol-abstinent rats. DA microdialysis displays increased DA levels and a blunted response to ethanol treatment in alcohol-dependent rats. (A) Basal extracellular DA levels within the AcbS are markedly increased in alcohol-dependent rats (n = 15 per group). (B) AcbS DA levels after the application of consecutive doses of ethanol (1 or 2 g/kg, i.p.). Control animals show an increase in extracellular DA levels after ethanol (2 g/kg, i.p.), whereas alcohol-dependent rats show a blunted response to the treatment (n = 8 per group). (C and D) Hyperlocomotion in 3-wk-abstinent rats was detected by records of locomotor activity in the open field (C) and home cage (D). The bar graph in C represents the total track length measured over a 60-min interval in the open-field setting. The graph in D represents the number of body movements per hour during a 72-h period detected using a home cage e-motion system. For the respective time courses, see Fig. S4. (E) Enhanced ethanol cue-induced reinstatement in abstinent rats compared with controls. The graph shows the mean number of responses after the presentation of stimuli previously paired with ethanol. (F) Representative EPSCs recorded at −80 mV in MSNs were evoked by electrical stimulation in the AcbS before (baseline) and during perfusion with 25 mM ethanol or 25 mM ethanol plus 5 μM SKF81297. Current traces represent the average of 10 sweeps. (G) Time courses of the effects shown in F for normalized EPSCs. (H) Summary of the effects on EPSCs (control, n = 12; alcohol-dependent, n = 7). Data are expressed as means ± SEM; *P < 0.05, **P < 0.01, ***P < 0.001 vs. baseline or vs. ethanol; ^#P < 0.05 alcohol-dependent vs. control.

Fig. S4.

Increased locomotor activity and TH expression in 3-wk-abstinent rats. (A) The graph represents 5-min intervals in the open-field setting. *P < 0.05, **P < 0.01. (B) The bar graph shows total track length measured during a 60-min interval in alcohol-dependent and control rats. *P < 0.05. (C) Home cage activity was monitored during a 72-h period using a home cage e-motion system. The hyperactivity in the home cage follows a circadian pattern, i.e., is evident only during the active phase of the rats. (D) The bar graph represents the number of body movements per hour during a 72-h period in alcohol-dependent and control rats. **P < 0.01. The bar graphs in B and C are also shown in Fig. 3. (E) In situ hybridization for TH mRNA revealed increased transcriptional levels in the SNc but no change in the VTA in alcohol-abstinent as compared with control rats. **P < 0.01, alcohol-dependent vs. control rats. There appears to be an increase in habituated activity but not in exploratory activity. The former could reflect hyperactivity of nigrostriatal rather than mesolimbic DA systems; this possibility also is supported by the increase in TH mRNA in the SNc, whereas no difference was observed in the VTA. However, the latter finding is difficult to interpret, because in the VTA, as opposed to the SN, the lack of distinct segregation of cells into TH⁺ and TH⁻ fractions could lead to a dilution of the measured TH signal and could reduce the likelihood of detecting an effect. Also, changes in TH mRNA provide no direct evidence for altered TH synthesis/availability in terminals.

We also assessed basal motor activity following 21 d of abstinence. In the first 20 min in an open-field exposure, i.e., under conditions of novelty, alcohol-dependent and control rats showed no difference in total distance traveled (Fig. S4). However, after habituation, alcohol-dependent rats had significantly enhanced locomotor activity compared with controls (F_1,14 = 5.6, P < 0.05) (Fig. 3C and Fig. S4). Also the alcohol-dependent rats displayed higher activity in the home cage (F_1,28 = 8.3, P < 0.01) (Fig. 3D). This difference resulted exclusively from higher activity during the active phase of the circadian cycle (Fig. S4). Cue-induced reinstatement of alcohol seeking, an established model of relapse (34), was significantly higher in 3-wk-abstinent alcohol-dependent rats than in controls (t₅₄ = 3.8; P < 0.001) (Fig. 3E).

Finally, we validated the down-regulation of D1 in protracted abstinence at the synaptic level by examining glutamatergic inputs to medium spiny neurons (MSNs) of the AcbS in brain slices during electrical stimulation of the AcbS (Fig. 3 F–H). Ethanol perfusion (25 mM; 25 min) increased excitatory postsynaptic currents (EPSCs) similarly in both groups of rats. Subsequent perfusion of the D1 agonist SKF81297 (5 µM; 20 min) in the presence of ethanol further enhanced the EPSCs in control rats but, importantly, not in alcohol-dependent rats (control: F_2,20 = 2115, P < 0.001 vs. baseline and vs. ethanol; alcohol-dependent: F_2,20 = 270, P < 0.001 vs. baseline and P > 0.05 vs. ethanol), as also was apparent from the differences in EPSCs in alcohol-dependent vs. control rats perfused with ethanol plus SKF81297 (P < 0.05).

In summary, these data suggest that 21-d-abstinent alcohol-dependent rats have increased accumbal DA levels but blunted DA responses to alcohol at both presynaptic (DA release) and postsynaptic (D1 agonist) sites. On the behavioral level these plasticity changes are associated with hyperactivity and an increased propensity for relapse in protracted abstinence.

A Transgenic Rat Model for Hyperdopaminergia Is Similar to Rats in Protracted Abstinence.

For further validation of the behavioral consequences of a hyperdopaminergic state, we used a genetic rat model in which a single point mutation in the Slc6a3 gene was introduced. This functional mutation led to an amino acid exchange (N157K) and subsequently to a loss of function of DAT. This loss of function was confirmed by strongly reduced [³H]-mazindol binding to DAT in the Acb in DAT N157K mutants as compared with that in WT rats (t₁₇ = 10.8, P < 0.001) (Fig. 4A). [³H]-SCH23390 binding demonstrated that the mutation also caused a significant reduction in D1 binding (t₁₆ = 16.0, P < 0.001) (Fig. 4B), which was likely caused by compensatory decrease of this receptor resulting from elevated extracellular DA concentration (t₁₀ = 24.1, P < 0.001) (Fig. 4C). The behavioral consequences of this hyperdopaminergic state in DAT N157K mutants are manifested by enhanced locomotor activity in the open-field test (t₃₂ = 8.3, P < 0.001) (Fig. 4D) and enhanced alcohol drinking as compared with WT animals (t₃₂ = 2.5, P < 0.05) (Fig. 4D). This genetic model thus shares several features with alcohol-dependent rats during protracted abstinence.

Fig. 4.

A hyperdopaminergic state in DAT N157K mutant rats is associated with hyperlocomotion and increased alcohol consumption. (A and B) Quantitative analysis of DAT ([³H]-mazindol)-binding sites (A) and D1 ([³H]-SCH23390)-binding sites (B) (expressed in femtomoles per milligram) in the Acb of WT and DAT N157K mutant rats. (C) Basal extracellular DA levels (expressed in femtomoles per microliter) in the Acb of WT and DAT N157K mutant rats. (D) Total track length (in meters) during 60-min testing in the open field in WT and DAT N157K mutant rats. (E) Intake of total ethanol (expressed in grams per kilogram body weight per day) in WT and DAT N157K mutant rats during 6 wk of continuous concurrent access to water and 5%, 10%, and 20% ethanol solutions. Data are expressed as means ± SEM; *P < 0.05, ***P < 0.001 vs. WT.

SI Materials and Methods

Postmortem Expression Studies.

Human brain tissue samples.

All human brain tissue samples were obtained from the New South Wales Tissue Resource Centre at the University of Sydney, Sydney (sydney.edu.au/medicine/pathology/trc/index.php). In the first experiment, tissue from 20 age-matched male subjects of European descent consisting of 10 chronic and heavy alcoholics and 10 control subjects was used (see core sample highlighted in Table S1). Subjects’ placement in the alcoholic or control group was confirmed postmortem using the criteria in the Diagnostic Instrument for Brain Studies–Revised, which are consistent with those in the Diagnostic and Statistical Manual for Mental Disorders. All alcoholics had consumed 50 g to more than 80 g of ethanol per day; the control cases had an average daily ethanol consumption of less than 20 g. To reduce the number of confounding factors, the groups did not include any subjects for which the postmortem interval was longer than 36 h or blood alcohol or significant amounts of psychiatric medication (e.g., opioids, benzodiazepines; concentration <0.1 mg/L) were detected during autopsy. For each subject we analyzed tissue samples from the VS including NA and the NC (see the regions identified in Fig. 1D).

In the second sample set, NC and VS brain tissue from 50 additional males of European descent (20 alcoholics and 30 controls; unshaded rows in Table S1) was used for D1 autoradiography (Fig. S1). Ten of the 20 alcoholics (the “intoxicated” group) had positive BACs (0.183 ± 0.145 g/100 mL) at the time of death; the other 10 (the “not-intoxicated” group) had no detectable BAC. No subjects with toxicology results positive for illicit drugs (e.g., morphine, codeine, or tetrahydrocannabinol) or with suicide as cause of death were included in the analysis.

Human postmortem experiments were approved by IRB to Institute for Psychopharmacology, Central Institute for Mental Health, Mannheim (study no. 2009-238-MA).

Receptor/transporter autoradiography on postmortem sections.

Cryosections (10 µm) were taken from dissected human tissue samples of the NC and VS and were used for autoradiography experiments according to the procedures in ref. 33.

For D1-[³H]-SCH23390 autoradiography, sections were incubated twice for 15 min and once for 10 min at room temperature in preincubation buffer [50 mM Tris⋅HCl (pH 7.4), 5 mM MgCl₂, 1 mM EDTA]. Sections then were transferred into humidified chambers, and 800 µL of reaction mix was applied to each slide, followed by incubation for 2 h at 30 °C. The reaction mix contained 3 nM [³H]-SCH23390 [specific activity (Sp. Act.) 60 Ci/mmol] (PerkinElmer) and 10 µM mianserin hydrochloride (Tocris Biosciences) in 50 mM Tris⋅HCl buffer (pH 7.4), 5 mM MgCl2, 100 mM NaCl, 1 mM EDTA, 1 mM DTT, and 0.1% BSA. Nonspecific binding was determined by the addition of a D1 receptor antagonist (10 µM flupenthixol dihydrochloride). Incubation was stopped by washing the slides twice with ice-cold Tris⋅HCl buffer (50 mM) followed by dipping in ice-cold deionized water. Sections were dried under a stream of cold air.

The buffers used for D1 autoradiography were also used for D2-[³H]-raclopride autoradiography. Sections were incubated twice for 15 min in preincubation buffer at room temperature, followed by 2-h incubation at 30 °C with 800 µL of reaction mix containing 5 nM [³H]-raclopride (Sp. Act. 80 Ci/mmol) (PerkinElmer) for both human and rat sections. Nonspecific binding was determined in presence of 30 µM (S)-(-)-sulpiride (Tocris Biosciences). After incubation, slides were washed twice in ice-cold Tris⋅HCl and deionized water and were air-dried.

For DAT-[³H]-mazindol autoradiography, the procedures were similar to those described in ref. 33, with some adaptations. Sections were incubated in ice-cold washing buffer [50 mM Tris⋅HCl (pH 7.9), 300 mM NaCl, and 5 mM KCl] for 5 min. Then the slides were transferred into humidified chambers and were incubated with 800 µL of reaction mix for 40 min at 4 °C. The reaction mix contained 2 nM [³H]-mazindol (Sp. Act. 17.8 Ci/mmol; PerkinElmer) and 0.3 µM desipramine hydrochloride (Tocris Biosciences) in 50 mM Tris⋅HCl (pH 7.9), 300 mM NaCl, and 5 mM KCl. Nonspecific binding was determined by adding 100 µM nomifensine maleate salt (Sigma-Aldrich). Sections then were washed twice in ice-cold washing buffer followed by 30 s in ice-cold deionized water and were dried in a cold air stream.

After drying, FUJI imaging plates (Storage Phosphor Screen BAS-IP TR2025 E Screen; GE Healthcare Life Sciences) were exposed to sections for 6–7 d and afterwards were scanned in a phosphorimager (Fuji PhosphorImager Typhoon FLA 700, GE Healthcare Life Sciences). The MCID program (Interfocus Imaging Ltd.) was used for densitometry analysis. The [³H]-quantitation standard curve (Amersham, GE Healthcare Life Sciences) was used to interpolate the measured optical densities (photostimulable luminescence per square millimeter) of the tissue-equivalent DA receptors and DA transporter densities from sections into nanocuries per milligram. Binding (expressed as femtomoles per milligram) was calculated based on the specific activity of the radioligand and the saturation binding equation B = B_max × [R]/(K_d +[R]), solving for B_max, where B_max is the maximal bound receptor/transporter and K_d is the receptor affinity in nanomoles. Data were expressed as femtomoles per milligram of protein (mean ± SEM). B_max and K_d values of each radioligand are summarized in Table S4.

Table S4.

K_d and B_max values of radioligands used for the DA transporter [³H]mazindol and D1 ([³H]SCH23390) and D2 ([³H]raclopride) receptor binding on human and rat brain sections

	Kd value in rat brain, nM (reference)	Bmax value in rat brain, fmol/mg (reference)	Kd value in human brain, nM (reference)	Bmax value in human brain, fmol/mg (reference)
[3H]raclopride	2.08 (61)	20.0 (61)	1.25 (62)	∼9.5 (62)
[3H]SCH23390	0.7 (63)	347 (63)	1.37 (62)	∼13 (62)
[3H]mazindol	18.2 (64)	0.0073 (64)	18.5 (65)	1.6 (65)

Quantitative RT-PCR from human tissue.

RNA extraction and analysis were performed as described in ref. 51. RNA from brain tissue was isolated using TRIzol (Invitrogen) according to the manufacturer’s protocol. RNA samples underwent a cleanup step using the RNeasy Mini Kit (Qiagen) and then were treated with RQ1 RNase-free DNase (Promega), following the manufacturer’s instructions, to eliminate DNA contamination. All RNA samples had 260/280 ratios between 1.8 and 2.1. RNA samples then were analyzed with an Agilent 2100 Bioanalyzer and the RNA integrity number. One hundred nanograms of total RNA was reverse transcribed to cDNA using the High Capacity RNA-to-cDNA Master Mix (Applied Biosystems, ABI), following the manufacturer’s protocol. Samples were assayed in triplicate in a total reaction volume of 20 μL using Power SYBR Green PCR Master Mix (ABI) on an ABI 7900 HT RT-PCR System (40 cycles of 95 °C for 15 s and 60 °C for 1 min). A melting profile was recorded at the end of each PCR to check for aberrant fragment amplifications. Primers for each target were designed toward the 3′ end of the coding sequence by considering exon–exon junctions, when possible, based on the National Center for Biotechnology Information (NCBI) reference sequence database. Amplicons were designed with a length of 95–110 bp and melting temperatures >75 °C so that amplicons could be distinguished from primer–dimer formations in the melting analysis. Primers for each target transcript were designed based on the NCBI reference sequence database:

• DRD1 (NM_000794.3): forward 5′-ACGACCCCAAGGCAAGGCGT-3′, reverse 5′-TCGGGGCTGTTGCTTTTCTGGT-3′;

• DRD2 (NM_016574.3): forward 5′-CAGACGCCGCAAGCGAGTCA-3′, reverse 5′-TCCTCTCGGGTGGGCTGGTG-3′;

• GAPDH (NM_002046.4): forward 5′-CATGAGAAGTATGACAACAGCCT-3′, reverse 5′-AGTCCTTCCACGATACCAAAGT-3′;

• AluSX: forward 5′-TGGTGAAACCCCGTCTCTACTAA-3′, reverse 5′-CCTCAGCCTCCCGAGTAGCT-3′.

ABI’s SDS 2.2.2 software was used to analyze the SYBR Green fluorescence intensity and to calculate the theoretical cycle number when a defined fluorescence threshold was passed (the Ct value). Relative quantification was done according to the ΔCt method with GAPDH used as internal normalizer. Statistical testing was done by t test on the ΔCt values. The software and ΔCt method were used to determine statistical significance. Melting curves for all primers used in this study exhibited single fluorescence change peaks at the appropriate melting temperatures, indicating the absence of primer–dimer formation. In addition to GAPDH we used AluSX as a second endogenous control. Results were similar for both reference genes.

Experiments in Alcohol-Dependent Rats.

Animals.

Male Wistar rats (Charles River Laboratories) with an initial weight of 220–250 g were used for ethanol vapor exposure experiments. The rat mutant DAT target gene Slc6a3_N157K (DAT N157K) was generated at Ingenium Pharmaceuticals. Ten-week-old male F344 rats (generation G0) were injected i.p. with N-ethyl-N-nitrosourea (3 × 65 mg/kg, once weekly). After recovering from transient sterility, the mutagenized males were outcrossed with wild-type F344 females to produce the first generation (G1) of progeny, which were heterozygous for a unique set of point mutations. Heteroduplex analysis by temperature gradient capillary electrophoresis was used to detect the Slc6a3_N157K mutation. Slc6a3_N157K founder rats were backcrossed with F344 female rats for up to 12 generations. Both male and female 6- to 9-mo-old WT and DAT N157K mutant rats were used for experiments. All animals were group-housed in standard rat cages (Type-IV; Ehret) under a 12-h light/dark cycle (lights on at 2:00 PM) with ad libitum access to food and water. All experiments were conducted in accordance with the ethical guidelines for the care and use of laboratory animals and were approved by the local animal care committee (Regierungspräsidium Karlsruhe, Germany, license numbers: 35-9185.81/G-183/09 and 35-9185.81/G-126/13).

Induction of dependence by cyclic intermittent exposure to ethanol vapor.

Rats were exposed to daily intermittent cycles of alcohol vapor intoxication and withdrawal, a paradigm that allows a high degree of control over brain alcohol levels and induces behavioral and molecular changes relevant to the pathophysiology of alcoholism (26, 50).

Rats were weight-matched, assigned to the two experimental groups, and exposed to either ethanol vapor or normal air using a rodent alcohol inhalation system as described previously (24–27, 32, 44, 49–51). Briefly, pumps (Ismatec, Wertheim, Germany) delivered alcohol into electrically heated stainless steel coils (60 °C) connected to an airflow of 18 L/min into glass/steel chambers (1 × 1 × 1 m). Vaporized ethanol was delivered to the individual rat chambers (1 × 1 × 1 m) via the tubes connected to the side arms. Each tube also had its own pressure gauge so the conditions for each chamber could be adjusted evenly. Rats first were allowed to habituate to the chambers for 1 wk. For the next 7 wk rats were exposed to five cycles of 14 h of ethanol vapor per week (0:00 AM–2:00 PM) separated by daily 10-h periods of withdrawal and an additional 58 h of withdrawal at the end of each weekly cycle. Twice per week blood (∼20 µL) was sampled from the lateral tail vein of ethanol-exposed rats for BAC measurements. BACs (range: 150–350 mg⋅dL⁻¹⋅cycle⁻¹) were determined using an AM1 Analox system (Analox Instruments Ltd.). Signs of mild withdrawal, such as tail stiffness and piloerection, were observed during the off intervals by the end of the 7-wk exposure period, but withdrawal intensity never reached seizure levels (27). Following intermittent exposure to ethanol vapor, rats were subjected to various periods of abstinence.

For expression analysis, animals were killed by decapitation immediately at the end of the last alcohol vapor exposure cycle (day 0) or after 1 d (day 1), 3 d (day 3), 7 d (day 7), or 21 d (day 21) of abstinence. Animals were kept under a 12-h light/dark cycle and were killed at the same Zeitgeber time (2 h after light on). Brains were removed quickly, snap-frozen in isopentane at −40 °C, and stored at −80 °C. Coronal brain sections (10–12 µm) at bregma levels +1.70 to +1.20 mm (striatum) and −5.20 to −5.60 mm (midbrain) were prepared using a cryostat (Leica Biosystems) and were thaw-mounted on gelatin-coated and Superfrost Plus (Thermo Fisher Scientific) glass slides for receptor/transporter autoradiography and in situ hybridization experiments, respectively.

Receptor/transporter autoradiography.

D1, D2, and DAT autoradiography experiments were performed as described above for postmortem bindings and as published in ref. 33. However, ligands for D1 and DAT autoradiography were used in different concentrations: for D1, 1 nM [³H]-SCH23390 (+10 mM flupenthixol for nonspecific binding), and for DAT, 4 nM [³H]-mazindol (+100 µM nomifensine for nonspecific binding) were applied. The high quality of D1, D2-like, and DAT binding is illustrated in Fig. S3, and B_max and K_d values of each radioligand are summarized in Table S4.

In situ hybridization.

Rat-specific tyrosine hydroxylase riboprobe was generated from a cDNA fragment [TH, base pairs 7–3,396 on rat cDNA, gene reference sequence: NM_RATDOPER (33)]. Antisense and sense RNA probes were synthesized from 200 ng DNA template, incubated for 90 min at 37 °C with transcription buffer (Ambion Applied Biosystems), 12.5 nmol ATP, CTP, and GTP, 50 pmol UTP, and 125 pmol [α-³⁵S]UTP (1,250 Ci/mmol) (Perkin-Elmer), 1 U RNA polymerase (Roche Molecular Biochemicals), and 1 U RNase inhibitor. DNA was digested by RNase-free DNase at 37 °C for 20 min, followed by purification of the transcripts using spin columns (Illustra MicroSpin S-200 HR columns; GE Healthcare). Counts per minute (cpm) were determined with the Liquid Scintillation Analyzer (1600TR).

In situ hybridization was done as previously described (25, 33, 49, 51). Frozen sections were brought to room temperature, fixed in 4% paraformaldehyde in PBS (pH 7) for 15 min, washed in PBS (pH 7.4) for 10 min, and then washed twice in sterilized water for 5 min. Afterwards, the tissue was deproteinated with 0.1 M HCl for 10 min, washed twice in PBS (pH 7.4) for 5 min, acetylated in 0.1 M triethanolamine (pH 8.0) with 0.25% acetic anhydride for 20 min, again washed twice in PBS (pH 7.4) for 5 min, dehydrated in graded ethanol, and air dried.

Sections were prehybridized with 800 µL prehybridization buffer containing 50% deionized formamide, 50 mM Tris⋅HCl (pH 7.6), 25 mM EDTA (pH 8.0), 20 mM NaCl, 0.25 mg/mL yeast tRNA, and 2.5× Denhardt’s solution (Invitrogen) at 37 °C for 2–4 h in humidified chambers. After prehybridization, 100 µL of hybridization buffer [50% deionized formamide, 20 mM Tris⋅HCl (pH 7.6), 10× Denhardt’s solution, 5 mg/mL yeast tRNA, 1 mg/mL polyadenylic acid, 10 mM EDTA (pH 8.0), 150 mM DTT, 330 mM NaCl, 10% dextran sulfate) containing 1 × 10⁶ cpm of either the labeled antisense RNA or sense RNA was applied, and the sections were covered with siliconized coverslips and incubated at 55 °C overnight in humidified chambers. Coverslips were removed by washing with 1× SSC at 42 °C for 40 min. Sections then were washed twice with 1× SSC for 40 min at 42 °C, once with 0.5× SSC/50% formamide for 1 h at 42 °C, twice with 1× SSC for 30 min, followed by treatment with 1 µg/mL RNaseA in RNase buffer [0.5 M NaCl, 10 mM Tris (pH 8.0), 1 mM EDTA (pH7.5)] for 1 h at 37 °C. After two washing steps in 1× SSC for 30 min at 55 °C followed by a brief washing in sterile water at room temperature, the sections were dehydrated in graded ethanol and air-dried. FUJI imaging plates (Storage Phosphor BAS-IP SR2025 Screen, GE Healthcare Life Sciences) were exposed to sections for 7–10 d and were scanned in a phosphorimager (Fuji PhosphorImager Typhoon FLA 700, GE Healthcare Life Sciences).

Densitometric measurements were taken from the striatum and midbrain, values were converted to nanocuries per gram based on the known radioactivity in the [¹⁴C]-Microscales (Amersham, GE Healthcare Life Sciences), and data were expressed as mean ± SEM.

Microdialysis.

Ethanol nonexposed and exposed Wistar rats and WT and DAT N157K mutant rats were housed in groups of four before surgery and were housed individually after surgery. For surgery, rats were anesthetized with isoflurane (1.5–2%) and were placed in a stereotaxic frame (Kopf Instruments). CMA11 guide cannula (20 gauge, 14 mm; CMA Microdialysis) were implanted unilaterally 2.0 mm above the AcbS (antero-posterior, + 1.6 mm; medio-lateral, ± 0.8 mm; dorso-ventral, 5.6 mm). Coordinates were based on bregma, midline, and dura, respectively. Cannulas were anchored with three stainless steel screws and dental acrylic. Animals were allowed to recover from surgery for 1 wk.

Microdialysis experiments were conducted in conscious, freely moving rats 21 d after the last exposure to ethanol vapor (51). Dialysis probes (CMA11 11/2; CMA Microdialysis) with 2-mm active membranes were introduced into the guide cannula 12 h before the beginning of the microdialysis experiments to minimize damage-induced release of neurotransmitters and metabolites. Each animal participated in only one microdialysis experiment. Samples were collected every 15 min at a flow rate of 1.5 µL/min. On the test day, the system was allowed to equilibrate for 1.5 h before collection started. After six baseline samples were collected, rats were injected i.p. with a saline solution followed by 1.0 g/kg ethanol 30 min later and 2.0 g/kg ethanol 60 min later. Sampling continued throughout the experiment.

Microdialysis probe placements were verified after the microdialysis. The inclusion criterion was that at least 80% of the active membrane was located in the target region. Animals that did not meet the criterion were excluded from the study.

HPLC analysis.

The DA content in the dialysate samples was determined by HPLC. Electrochemical detection was acquired with the ALEXIS 100 cooled-micro LC-EC system (Antec Leyden BV) equipped with a microbore VT-03 flow cell. The working potential of the cell was set at 400 mV, and the oven temperature of the DECADE II electrochemical detector (Antec) was set at 35 °C. The mobile phase of pH 6 contained 50 mM phosphoric acid, 400 mg/L octanesulfonic acid sodium salt (OSA), 0.1 mM EDTA, 8 mM KCl, and 15% methanol and was perfused at a flow rate of 200 µL/min. Duplicate 4-µL aliquots of each sample were injected onto a reversed-phase column (C18, ALF-205 column, 50 × 2,1-mm i.d., 3 µm) (Axel Semrau GmbH & Co. KG), and the DA content was determined by the area under the peak, using an external standard curve as a reference. The detection limit for DA was 200 pM with a signal-to-noise ratio of 2.

Locomotor activity.

Locomotor activity was measured in the home cage during a 72-h interval and in an open field for 1 h. For home-cage locomotion, rats were singly housed for 24 h before the experiment. An infrared sensor (Infra E-Motion GmbH) was placed over each cage, and body movements were monitored during a 72-h period starting on day 17 of abstinence.

The open field consisted of four equal arenas (51 ×51 × 50 cm) made of dark PVC. The distance traveled (in centimeters) was recorded for 60 min at a light intensity of 50 lx on day 25 of abstinence. The same system was used for DAT N157K mutant rats. The observation program Viewer2 (Biobserve GmbH) was used for the analysis of locomotor activity.

Cue-induced reinstatement of alcohol-seeking behavior in alcohol-dependent rats.

Operant alcohol self-administration apparatus.

All alcohol-seeking experiments were carried out in operant chambers (MED Associates Inc.) enclosed in ventilated sound-attenuating cubicles. The chambers were equipped with a response lever on each side panel of the chamber. Responses at the appropriate lever activated a syringe pump that delivered an ∼30-µL drop of fluid into a liquid receptacle next to the lever. A light stimulus (house light) was mounted above the right response lever of the self-administration chamber. An IBM-compatible computer controlled the delivery of fluids, presentation of stimuli, and data recording.

Alcohol self-administration training.

All animal training and testing sessions were performed during the dark phase of the light cycle (51, 52). Animals (n = 16) were trained to press the lever reinforced with 10% (vol/vol) ethanol in daily 30-min sessions using a fixed-ratio 1 schedule. During the first 3 d of training, animals were fluid-deprived for 20 h/d to facilitate the acquisition of an operant response to a liquid reinforcer. Responses at the left lever were reinforced by the delivery of 0.2% (wt/vol) saccharin solution. (Throughout the training phase responses at the inactive (right) lever were recorded but not reinforced.) For the next 3 d, animals underwent the same procedure without fluid deprivation. Following the acquisition of a saccharin-reinforced response, rats were trained to self-administer ethanol. Thus, rats had access to 0.2% saccharin with 5% ethanol for 1 d, 5% ethanol for 1 d, 0.2% saccharin with 8% ethanol for 1 d, 8% ethanol for 1 d, 0.2% saccharin with 10% ethanol for 1 d, and 10% ethanol for 1 d.

Conditioning phase.

The purpose of the conditioning phase was to train the animals to associate the availability of ethanol with the presence of specific discriminative stimuli. This phase started after the completion of the saccharin-fading procedure. Discriminative stimuli predicting ethanol (10%) availability were presented during each subsequent daily 30-min session. An orange-flavored extract served as the cue stimulus (S)for ethanol. This olfactory stimulus was generated by depositing six drops of orange extract into the bedding of the operant chamber before each session. In addition, each lever press resulting in ethanol delivery was accompanied by a 5-s presentation of the house light (conditioned light stimulus). The 5-s period served as a “time-out,” during which responses were recorded but not reinforced. At the end of each session the bedding of the chamber was changed, and trays were cleaned thoroughly. The animals received a total of 10 ethanol conditioning sessions. Throughout the conditioning phase, responses at the inactive (right) lever were recorded but not reinforced.

Conditioning and extinction phase in alcohol-dependent rats.

After a 2-wk abstinence phase, all animals were reconditioned to self-administer 10% ethanol in 10 daily conditioning sessions. After completing the reconditioning phase, rats were subjected to daily 30-min extinction sessions for seven consecutive days, which in total were sufficient to reach the extinction criterion of fewer than 10 lever responses per session. Extinction sessions began by extending the levers without presenting olfactory discriminative stimuli. Responses at the previously active lever activated the syringe pump, without resulting in the delivery of ethanol or the presentation of response-contingent cues (stimulus light).

Reinstatement testing.

Reinstatement tests began 3 d after the final extinction session. In these tests, rats were exposed to the same conditions as during the conditioning phase, except that ethanol was not made available. Sessions were initiated by the extension of both the ethanol-associated and inactive levers and the presentation of the discriminative stimulus predicting ethanol (S). Responses at the ethanol-associated lever were followed by the activation of the syringe pump and the presentation of the conditioned stimulus (light). The first two lever presses resulted in the delivery of ethanol, i.e., ∼60 µL of liquid that served as an additional olfactory/gustatory cue.

Long-term voluntary alcohol consumption.

All animals were singly housed. After 2 wk of habituation to the animal room, rats were given ad libitum access to tap water and to 5%, 10%, and 20% ethanol solutions (vol/vol) for six successive weeks. Alcohol drinking solutions were made up from 96% ethanol (Sigma-Aldrich) diluted with tap water to the different concentrations. Spillage and evaporation were minimized by the use of special bottle caps. The positions of bottles were changed weekly to avoid location preferences. Intake of ethanol (measured in grams of pure alcohol per kilogram of body weight per day) was calculated as the daily average across the seven measuring days.

Electrophysiology.

Brain slices.

Coronal slices (300 µm) containing the AcbS were prepared using the HM 650 V microtome (Microm International) from seven 15- to 16-wk-old male Wistar rats (n = 12 MSNs from four controls and n = seven MSNs from three alcohol-dependent rats in protracted abstinence, i.e., 3 wk after completing the chronic intermittent exposure to alcohol protocol as described above). Rats were anesthetized by inhalation of isofluorane and killed. The brains were removed rapidly and placed in dissection buffer at 4 °C containing (in mM): 220 sucrose, 3.5 KCl, 1.25 NaH₂PO₄, 3 MgCl₂, 1 CaCl₂, 25 NaHCO₃, and 10 dextrose. Individual slices were stored gently in artificial cerebrospinal fluid (ACSF) for at least 1.5 h before recording. ACSF was similar to the dissection buffer, except that sucrose was replaced by 124 mM NaCl, MgCl₂ was lowered to 1.5 mM, and CaCl₂ was raised to 2.5 mM. Both dissection buffer and ACSF were saturated with 95% O₂/5% CO₂ (pH 7.4).

Patch-clamp whole-cell recordings.

The slices were transferred to a submerged recording chamber, perfused with ACSF at 2 mL/min, and imaged using a Zeiss Axioskop 2 microscope (Carl Zeiss AG). Whole-cell recordings were performed at 30 °C from MSNs located in the AcbS with the EPC-9 amplifier interfaced to Patchmaster software (HEKA Elektronik). Borosilicate recording pipettes (o.d., 1.5 mm; 2–4 MΩ) were pulled on the Flaming/Brown puller P-97 (Sutter Instruments) and were filled with internal solution containing (in mM): 130 K-Gluconate, 10 KCl, 0.2 EGTA, 10 Hepes, 4 Mg-ATP, 0.5 Na_-GTP, and 10 Na-Phosphocreatine (pH 7.25, 280–290 mOsm). For electrical stimulation, borosilicate glass pipettes filled with ACSF were placed in the AcbS to evoke EPSCs in MSNs around 200 pA at the holding potential (V_h) of −80 mV. GABAergic transmission was antagonized by picrotoxin (1 mM), which was added to the internal solution (56). Electrophysiological data were filtered at 2 kHz and digitized at 10 kHz. Input resistance was monitored via hyperpolarizing pulses (−10 mV, 100 ms). Only cells with holding currents ≤100 pA at V_h = −80 mV and series resistance ≤20 MΩ were studied. Cells were discarded if any of these parameters changed by ≥20% during the course of the experiment.

Statistics.

All expression data are expressed as means ± SEM and were statistically analyzed within a region by one-way ANOVA (treatment effect) or two-way ANOVA (time and treatment effect in a time-course experiment) followed by Fisher’s LSD post hoc test. Data from the microdialysis experiment were analyzed by two-way repeated-measures ANOVA followed by Fisher’s LSD post hoc test. Locomotion tests were analyzed by two-way ANOVA followed by Fisher’s LSD post hoc test when appropriate. Analysis of cue-induced reinstatement in alcohol-dependent rats and comparisons between WT and N157K DAT mutant rats were done by independent two-tailed t tests.

In the second postmortem experiment (Fig. S1 and Table S1) an analysis of covariance was performed to examine the differences in the D1-binding sites in the controls and the not-intoxicated and intoxicated alcoholic groups. The variables “smoker” and “non-smoker” were included in the analysis. Active and ex-smokers were grouped as “smokers,” and non-smokers and subjects with unknown smoker status were grouped as “non-smokers.” Tissue pH, postmortem interval, age, and smoker state were considered as candidate covariates. Nonsignificant variables were removed sequentially by a stepwise backward procedure, and no covariant was identified as significant. This procedure was followed by a Fisher’s post hoc test.

For the electrophysiological experiments, statistical analysis during the perfusion of drugs was performed during the last 10 min of every condition using two-way ANOVA followed by Bonferroni’s post hoc test, and comparisons between control and alcohol-dependent rats were performed using an unpaired Student’s t test. The significance level was P = 0.05. All results are shown as averages ± SEM.

Meta-Analysis.

A literature search was conducted on PubMed (www.ncbi.nlm.nih.gov/pubmed/). No particular journal was preferred. The search included the keywords “alcohol/ethanol” AND “withdrawal/abstinence” AND “dopamine” AND “accumbens” OR “striatum” OR “Ventral Tegmental Area/VTA”. Literature selection criteria further included (i) chronic administration of only alcohol (no further pharmacological agents were administered either acutely or chronically) and (ii) presence of withdrawal symptoms. Articles that did not comply with these criteria were excluded. Of ∼225 publications, 29 publications (based on 352 rodents chronically exposed to ethanol and 96 alcoholic individuals) fulfilled the selection criteria.

The following variables were obtained from the publications and used for further analysis:

• i)Weight, age, gender, and consciousness (if anesthetics were applied: agent and dose)
• ii)Exact method of measurement (in vivo microdialysis, patch-clamp recordings, tissue HPLC, PET, and others)
• iii)The administration paradigm (self-administration, free-choice, i.p injections, and others) and daily doses of alcohol in animals; the history of alcohol dependence and average daily alcohol consumption in humans
• iv)The number of the alcoholic individuals and ethanol-exposed animals used in each experiment
• v)Extracellular and in situ DA, DOPAC, and HVA concentrations, firing frequency and burst rates of dopaminergic neurons, availability of D1 and D2 receptors and DAT.
• vi)Time of measurement after alcohol withdrawal
• vii)Relative change (percentage) of the obtained variable (v) in comparison with the controls

For the meta-analysis of the effects of abstinence from chronic ethanol exposure on mesolimbic dopaminergic systems, we used a fixed-effect model (57–60) with respect to the extracted variables (v) and analyzed the withdrawal interval of [0, 60] days. $\overline{x}=\frac{1}{N}{\displaystyle \sum _{i=1}^R{n}_i{x}_i}$ represents the weighted average effect of the concentrations of DA and its metabolites, respectively, as the weighted sum of the products of the mean effects x_i from each experiment i and the number of animals used in that particular study n_i, whereby $N={\displaystyle {\sum }_{i=1}^k{n}_i}$ denotes the total number of animals considered in the meta-analysis of the k studies. If the amount of extracellular DA was not directly specified by the measurement (e.g., tissue punches), the ratio of DOPAC to DA was calculated as an estimate for active DA concentrations.

Meta-analysis results.

Because human imaging studies (Table S5, refs. 1–6), animal studies of electrophysiological changes (Table S5, refs. 7–12), and D1 and D2 receptors and DAT availability (Table S5, refs. 10, 11, 13, 14) do not provide sufficient data (n = 96 alcoholic patients; n = 122 rodents) for a robust meta-analysis of the dynamics of dopaminergic processes, the investigation focused on alterations in DA concentrations and its metabolites.

Table S5.

Data sources for meta-analysis

Reference number	Year of publication	Authors	Journal	Journal issue: Page
1	1999	Laine TP, Ahonen A, Rasanen P, Tiihonen J	Psychiatry research	90:153–157
2	1996	Volkow ND, Wang GJ, Fowler JS, Logan J, Hitzemann R, Ding YS, Pappas N, Shea C, Piscani K	Alcoholism, clinical and experimental research	20:1594–1598
3	2002	Volkow ND, Wang GJ, Maynard L, Fowler JS, Jayne B, Telang F, Logan J, Ding YS, Gatley SJ, Hitzemann R, Wong C, Pappas N	Psychiatry research	116:163–172
4	2007	Volkow ND, Wang GJ, Telang F, Fowler JS, Logan J, Jayne M, Ma Y, Pradhan K, Wong C	The Journal of neuroscience	27:12700–12706
5	2005	Martinez D, Gil R, Slifstein M, Hwang DR, Huang Y, Perez A, Kegeles L, Talbot P, Evans S, Krystal J, Laruelle M, Abi-Dargham A	Biological psychiatry	58:779–786
6	1994	Hietala J, West C, Syvalahti E, Nagren K, Lehikoinen P, Sonninen P, Ruotsalainen U	Psychopharmacology	116:285–290
7	1993	Diana M, Pistis M, Carboni S, Gessa GL, Rossetti ZL	PNAS	90:7966–7969
8	1996	Diana M, Pistis M, Muntoni A, Gessa G	Neuroscience	71:411–415
9	1993	Shen RY, Chiodo LA	Brain research	622:289–293
10	1998	Bailey CP, Manley SJ, Watson WP, Wonnacott S, Molleman A, Little HJ	Brain research	803:144–152
11	2001	Bailey CP, O'Callaghan MJ, Croft AP, Manley SJ, Little HJ	Neuropharmacology	41:989–999
12	2011	Perra S, Clements MA, Bernier BE, Morikawa H	Neuropsychopharmacology	36:993–1002
13	2001	Cowen MS, Lawrence AJ	Alcoholism, clinical and experimental research	25:1126–1133
14	1988	Lucchi L, Moresco RM, Govoni S, Trabucchi M	Brain research	449:347–351
15	1992	Rossetti ZL, Hmaidan Y, Gessa GL	European journal of pharmacology	221:227–234
16	1991	Rossetti ZL, Melis F, Carboni S, Gessa GL	European journal of pharmacology	200:371–372
17	1992	Rossetti ZL, Longu G, Mercuro G, Hmaidan Y, Gessa GL	Alcohol and alcoholism	27:477–480
18	1999	Rossetti ZL, Isola D, De Vry J, Fadda F	Neuropharmacology	38:1361–1369
19	1996	Weiss F, Parsons LH, Schulteis G, Hyytia P, Lorang MT, Bloom FE, Koob GF	The Journal of neuroscience	16:3474–3485
20	2001	Rothblat DS, Rubin E, Schneider JS	Neuroscience letters	300:63–66
21	2000	Bailey CP, Andrews N, McKnight	Pharmacology,	66:153–161
		AT, Hughes J, Little HJ	biochemistry, and behavior
22	2004	Thielen RJ, Engleman EA, Rodd ZA, Murphy JM, Lumeng L, Li TK, McBride WJ	The Journal of pharmacology and experimental therapeutics	309:216–225
23	1992	Gil E, Colado I, Lopez F, Fernandez-Briera A, Fernandez-Lopez A, Calvo P	Neuropharmacology	31:1151–1156
24	1984	Dar MS, Wooles WR	Alcohol	1:453–458
25	2011	Barak S, Carnicella S, Yowell QV, Ron D	The Journal of neuroscience	31:9885–9894
26	1997	Nestby P, Vanderschuren LJ, De Vries TJ, Hogenboom F, Wardeh G, Mulder AH, Schoffelmeer AN	Psychopharmacology	133:69–76
27	2006	Zhao RJ, Yoon SS, Lee BH, Kwon YK, Kim KJ, Shim I, Choi KH, Kim MR, Golden GT, Yang CH	Neuroscience letters	395:28–32
28	2006	Halladay AK, Wagner GC, Sekowski A, Rothman RB, Baumann MH, Fisher H	Synapse	59:277–289
29	1995	George SR, Fan T, Ng GY, Jung SY, O'Dowd BF, Naranjo CA	The Journal of pharmacology and experimental therapeutic	273:373–379

Sixteen studies (Table S5, refs. 7, 15–29) using a total of 192 ethanol-exposed rodents were weighted for each time step and integrated into the meta-analysis procedures. The time course of DA, DOPAC, and HVA concentrations in the Acb (Fig. 2A and Fig. S2) was obtained by continuous interpolation of the averages of experimental values with respect to the time of measurement after alcohol withdrawal. This quantitative analysis suggests a robust pattern of dynamical changes in DA, DOPAC, and HVA concentrations. Although the paradigm to history (pattern) of ethanol intake did not affect the analysis, the withdrawal period may be considered as a vanishing swing between two states. Within the first 6 d of withdrawal, the dopaminergic system undergoes a substantial decline to up to 30% of the baseline concentrations (hypodopaminergic state). However, around the sixth day of abstinence the system moves toward a nonallostatic hyperdopaminergic state, reaching its peak between the second and third week of abstinence and converging to the baseline concentration after 2 mo.

Discussion

In this translational study we provide evidence for the development of a hyperdopaminergic state during protracted alcohol abstinence. We found striatal D1 receptor and DAT binding to be strongly decreased in postmortem brain tissue from alcoholics. These results were supported by a dynamic regulation of D1 and DAT at different times during abstinence in an animal model of alcohol dependence, with pronounced reductions of D1 and DAT during protracted abstinence. Functionally, the reduction in D1 receptors resulted in a lack of modulation of glutamatergic transmission upon D1 stimulation. Further, we found elevated basal extracellular DA levels in the AcbS associated with a blunted DA response to alcohol challenges. On the behavioral level we saw hyperactivity and enhanced alcohol-seeking during protracted abstinence. A transgenic rat model for hyperdopaminergia demonstrated a decrease in accumbal DAT and D1, increased extracellular DA levels and locomotor activity, and increased alcohol intake similar to that seen in 3-wk-abstinent rats. Together with a meta-analysis of rodent studies, our data provide convergent evidence that a hyperdopaminergic state occurs during protracted abstinence. According to the literature and the time-course studies we present here, we propose that dynamical changes take place in the mesolimbic DA system during withdrawal and protracted abstinence, resulting in a hypodopaminergic state that characterizes acute withdrawal (19, 35) and a hyperdopaminergic state that characterizes protracted abstinence (Fig. 2). These dynamic alterations of the mesolimbic DA system in the course of the addiction cycle have implications for our understanding of the mechanisms underlying alcoholism, the interpretation of PET results in alcoholic patients, and the development of effective therapeutic strategies.

A major hypothesis suggests deficits in DA signaling, leading to a hypodopaminergic state, as a driving mechanism for the relapsing course of the disorder (18, 19). Experimental support for this view comes from animal work (35) and human PET studies that found reduced availability of striatal D2-like receptors compared with controls (10–16). Because these PET studies do not provide a coherent picture (20–22), we used saturated receptor autoradiography techniques to measure the number of DA receptor and DAT-binding sites in postmortem brain tissues of alcoholics and controls. Surprisingly, we found a highly significant reduction in both D1- and DAT-binding sites in striatal tissue but no change in D2-like binding; these results imply a hyper- rather than a hypodopaminergic state, especially in protracted abstinence. These findings show that PET studies should be interpreted with caution, because the commonly used low-affinity radiotracer can be displaced by competing levels of endogenous DA even though the number of binding sites remains unchanged. In fact, such competition has been observed for [¹¹C]raclopride upon pharmacological manipulation of DA levels (36). Thus, psychostimulant-increased DA levels lead to a reduction in striatal D2-binding potential. In line with our own data from human postmortem brain tissue, a recent study with the high-affinity D2 ligand [¹⁸F]fallypride, which is less sensitive to endogenous DA levels (37), found unaltered striatal binding potential in alcoholics during protracted abstinence as compared with healthy controls (22). In vivo data for D1 are not available for alcoholics. Similar to our data, Tupala et al. (38) found reduced striatal D1 binding in postmortem brain samples from alcoholics. In apparent contrast to our data, however, they found decreased D2 in striatal regions. This difference could be caused by various factors. We noticed that most of the subjects in their sample had high alcohol or medication levels at the time of death. Further, human in vivo studies found increased DA synthesis rates in alcoholics as assessed by the uptake of [¹⁸F]DOPA, an immediate precursor of DA synthesis, as well as reduced DAT availability (39). Reduction of DAT is in line with our observations and those of others (40) in postmortem samples.

Animal studies performed during withdrawal consistently found reward deficits associated with the suppression of accumbal DA release (4, 41, 42). Remarkably, although protracted abstinence is the most relevant clinical condition in alcohol and other substance use disorders (4, 5, 43), it has not been the focus of preclinical DA research. To investigate the neuroadaptations of the DA system in more detail, we first performed a meta-analysis of the existing rodent literature on DA concentrations and its metabolites in the AcbS at different time points during abstinence. We found evidence for alcohol-induced DA elevation on day 0 followed by a decline during acute withdrawal, but around the sixth day of abstinence an increase in DA is found, which is most augmented during protracted abstinence. Although in this quantitative evaluation of the literature the methods used to induce dependence in the rats vary among studies, the general pattern of dynamic changes appears to be robust and seems to follow an oscillatory-like mode over time.

To confirm such oscillatory-like regulation of DA at the membrane level, we analyzed binding sites of D1, D2-like receptors, and DAT in three striatal regions of alcohol-dependent rats at various time points during abstinence according to our previous studies (24, 32). A similar temporal pattern of regulation was found for D1 and DAT (Fig. 2). During protracted abstinence (day 21), both D1- and DAT-binding sites are decreased. D2-like receptors were not changed at any time during abstinence, a finding that is in agreement with our human postmortem brain data. Functional evidence is given by electrophysiological data showing a blunted modulation of glutamatergic transmission upon D1 activation in the presence of ethanol in the MSNs of the AcbS and thus confirming a strong reduction of D1 in these neurons in response to elevated DA during protracted abstinence.

Our microdialysis studies in the AcbS of 21-d-abstinent rats found elevated extracellular DA levels. In addition, we found a blunted accumbal DA response to acute alcohol in abstinent rats, a finding that is in line with human PET data obtained after psychostimulant challenge in alcoholics (15, 16). This lack of responsiveness could be interpreted in two ways, reflecting state-specific response dynamics dependent on low or high extracellular DA levels or a relative DA deficit caused by high chronic demands that have exhausted compensatory mechanisms. The latter interpretation is supported by a recent metabolomics study showing deficits in central energy metabolism in the AcbS of alcohol-abstinent rats (44). In fact, it has long been proposed that blunted DA responses represent an endophenotype for substance use disorders (17, 18), and thus the blunted DA response to alcohol may represent a potential factor for enhanced alcohol-seeking behavior. It is worth noting that these reciprocal effects of increased steady-state activity and blunted challenge-induced responses occur in other biological systems as well. For example, some alcoholic populations show enhanced basal production of cortisol but have a blunted response to acute intervening stress, leading to impaired or inappropriate stress responses (45).

Mechanistically, the reduction we observed in D1 and DAT densities in alcoholics and alcohol-dependent rats after 21 d of abstinence can be explained in several ways. Chronic stimulation of D1 by repeated intoxication may lead to internalization and degradation of D1. Such a mechanism has been demonstrated after repeated administration of DA agonists and produces a lack of sensitivity to subsequent administration of DA agonists on behavioral, biochemical, and electrophysiological levels (46, 47). Also, DAT and D1 expression seem intrinsically related: Postsynaptic D1 is reduced in both DAT N157K mutant rats and DAT-KO mice (48).

On the behavioral level, alcohol-dependent rats in protracted abstinence show enhanced motor activity, enhanced alcohol consumption (24, 25, 27, 49, 50), and augmented reinstatement of cue-induced alcohol seeking, a finding that has been replicated consistently in several studies (51–53). A similar phenotype is observed in DAT N157K mutant rats. These mutant rats exhibit hyperdopaminergia, resembling on a molecular level alcohol-dependent rats in protracted abstinence (i.e., strongly reduced DAT and D1), and also display enhanced motor activity and augmented alcohol consumption. Thus, on both the molecular and the behavioral level, DAT-mutant rats represent a phenocopy of alcohol-dependent rats during protracted abstinence.

Taken together, our studies provide convergent evidence for a hyperdopaminergic state of the reward system during protracted abstinence. This hyperdopaminergic state is associated with increased motor activity and augmented alcohol seeking and use. We suggest that an enhanced risk for relapse exists both during acute withdrawal and long into protracted abstinence, but, according to our data, this vulnerability can be associated with either hypo- or hyperdopaminergia. Although the link between the early withdrawal phenomena and subsequent dysregulations remains unclear, many biological functions depend on homeostatic regulation, so that either a deficit or an excess in regulation results in worsening performance. Such a model was proposed for the role of DA in cognitive functioning in ref. 54. In this sense, an increased risk for relapse in a hypodopaminergic state could be caused by reward deficiency, whereas hyperdopaminergia might cause hyperactivity, which often is associated with poor impulse control.

In conclusion, our study extends the current neurobiological understanding of alcohol dependence, proposing that a hyperdopaminergic state may exist during protracted abstinence, at least in some alcoholics. We show dynamic changes within the mesolimbic DA system during alcohol exposure, withdrawal, and prolonged abstinence. Enhanced dopaminergic activity during alcohol exposure is followed by a hypodopaminergic system that characterizes the first few days of acute withdrawal; subsequently, counter adaptive changes that involve D1-, DAT-, and DA-releasing properties ensue, leading to a hyperdopaminergic state during protracted abstinence. Clinical studies are now warranted to define whether this hyperdopaminergic state is a marker for vulnerability to craving and relapse and whether it provides a window for specific intervention.

Materials and Methods

Human postmortem brain samples of males of European ancestry were used for expression analysis. Dependence was induced in male Wistar rats by cyclic intermittent ethanol vapor exposure, and the animals were used for expression analysis, microdialysis, electrophysiology, and behavioral tests. A meta-analysis was performed for accumbal DA concentration during abstinence. DAT N157K mutant rats and all experimental procedures are described in detail in SI Materials and Methods. Human postmortem brain experiments were approved by the institutional review board (IRB study no. 2009-238-MA licensed to the Institute for Psychopharmacology, Central Institute for Mental Health, Mannheim, Germany), and all animal experiments were approved by the local animal care committee (Regierungspräsidium Karlsruhe, Germany, license numbers: 35-9185.81/G-183/09 and 35-9185.81/G-126/13).