Hypodopaminergic state of the nigrostriatal pathway drives compulsive alcohol use

Abstract

The neurobiological mechanisms underlying compulsive alcohol use, a cardinal feature of alcohol use disorder, remain elusive. The key modulator of motivational processes, dopamine (DA), is suspected to play an important role in this pathology, but its exact role remains to be determined. Here, we found that rats expressing compulsive-like alcohol use, operationalized as punishment-resistant self-administration, showed a decrease in DA levels restricted to the dorsolateral territories of the striatum, the main output structure of the nigrostriatal DA pathway. We then causally demonstrated that chemogenetic-induced selective hypodopaminergia of this pathway resulted in compulsive-like alcohol self-administration in otherwise resilient rats, accompanied by the emergence of alcohol withdrawal-like motivational impairments (i.e., impaired motivation for a natural reinforcer). Finally, the use of the monoamine stabilizer OSU6162, previously reported to correct hypodopaminergic states, transiently decreased compulsive-like alcohol self-administration in vulnerable rats. These results suggest a potential critical role of tonic nigrostriatal hypodopaminergic states in alcohol addiction and provide new insights into our understanding of the neurobiological mechanisms underlying compulsive alcohol use.

Fig. 1. Compulsive alcohol use is specifically associated with a decrease in DA levels in the aDLS.

A Experiment timeline. B Alcohol intake during intermittent-access 20%-ethanol two-bottle-choice (IA 20%-EtOH 2BC). RM two-way ANOVA showed a significant effect of session [F_{(6, 109)} = 7.82, P < 0.001, partial η² = 0.29] but neither effect of group, nor session x group interaction [F_s < 0.47, P > 0.5, partial η² < 0.02]. C Number of active lever presses in 30-min self-administration sessions (SA) of 20% EtOH (FR3), during baseline (BL), “no-shock” and “shock” sessions (see supplementary material for details) in footshock-sensitive (FS, orange, n = 12) and footshock-resistant (FR, purple, n = 9) rats. RM two-way ANOVA showed a significant group x session interaction [F_{(14, 266)} = 2.86, P < 0.001, partial η² = 0.13]. D Mean active lever presses during the last three “no-shock” sessions and the last three “shock” sessions normalized to baseline. RM two-way ANOVA showed a significant shock condition x group interaction [F_{(1, 19)} = 20.89, P < 0.001, partial η² = 0.52]. E NAc, DMS and aDLS DA levels for FR, FS and Water rats (white, n = 8). Mixed-effects analysis found a group x striatal sub-region interaction [F_{(4, 77)} = 2.96, P < 0.05]. F Linear regression between resistance to footshock and DA level in NAc, DMS or aDLS. A significant correlation was found in the aDLS [F_(1,19) = 16.18, P < 0.001], but not in the NAc or DMS [F_s < 1.14, P > 0.3]. Data are expressed as mean ± SEM. Bonferroni correction post-hoc analysis: **, P < 0.01; ***, P < 0.001. BL baseline, DMS dorsomedial striatum, aDLS anterior dorsolateral striatum, FR fixed ratio, NAc nucleus accumbens.

Fig. 2. Chemogenetic inhibition of the SNc DA neurons induces selective nigrostriatal hypodopaminergia.

A–C hM4Di-mCherry and mCherry expression in SNc DA neurons. A TH::Cre rats received bilateral injection of Cre-dependent hM4Di-mCherry or Cre-dependent mCherry virus in the SNc. B Representative images of TH immunostaining and hM4Di-mCherry expression. Scale bar: 250 μm. C Quantification of transgene expression in distal (dSNc), medial SNc (mSNc) and VTA. Three-way ANOVA showed a significant effect of the area [F_(2,129) = 158, P < 0.001, η² = 0.71], but no effect of transgene, treatment or any interaction between these factors [F_s < 1.04, P > 0.36, partial η² < 0.02]. D–G Extracellular DA concentrations in aDLS and NAc throughout eight 45 min-fractions collected by microdialysis. Data are normalized to baseline. D Design of the microdialysis experiment. E Time course of extracellular DA concentrations in the aDLS of hM4Di rats treated with C21 (orange, n = 7) or NaCl (red, n = 7) and mCherry rats treated with C21 (gray, n = 7) or NaCl (white, n = 10). F Time course of extracellular DA concentrations in the NAc of hM4Di rats treated with C21 (orange, n = 10) or NaCl (red, n = 9) and mCherry rats treated with C21 (gray, n = 4) or NaCl (white, n = 6). In the fraction collected between 90 and 135 minutes after injection (dotted squares), two-way ANOVA found a treatment x transgene interaction in the aDLS [F_(1,27) = 8.63, P < 0.01, partial η² = 0.24], but not in the NAc [F_(1,25) = 0.03, P > 0.5, partial η² = 0.001]. Data are expressed as mean ± SEM. Bonferroni correction post-hoc analysis: **, P < 0.01. C21: compound 21, SNc substantia nigra pars compacta, TH tyrosine hydroxylase, VTA ventral tegmental area.

Fig. 3. Chemogenetically-induced nigrostriatal hypodopaminergia induces compulsive alcohol use.

A Experiment timeline. B Quantification of hM4Di-mCherry and mCherry expression in distal (dSNc), medial SNc (mSNc) and VTA. Two-way ANOVA showed significant effect of the area [F_(2,90) = 116.4, P < 0.001, η² = 0.72], but no effect of transgene, treatment or any interaction between these factors [F_s < 1.99, P > 0.14, partial η² < 0.04]. C Alcohol intake during IA2BC 20%-EtOH in hM4Di rats (orange, n = 7) and mCherry rats (black, n = 10). RM two-way ANOVA showed a significant effect of session [F_{(4, 59)} = 10.56], P < 0.001, partial (η² = 0.43), but neither effect of transgene, nor session x transgene interaction [F_s < 0.57, P > 0.5, partial η² < 0.04]. D Number of active lever presses in 30-min self-administration sessions of 20%-EtOH (FR3), during baseline (BL), “Shock/NaCl”, “Wash” and “Shock/C21” sessions. RM two-way ANOVA showed a significant session x transgene interaction [F_(19,266) = 2.18, P < 0.01, partial η² = 0.13]. Mean active lever presses during the last three “Shock/NaCl” and “Shock/C21” sessions normalized to baseline (E) and individual trajectories during these two periods (F). RM two-way ANOVA reported a significant session x transgene interaction [F_(1,14) = 18.41, P < 0.001, partial η² = 0.57]. Correlation between the difference of active lever presses during “Shock/C21” and “Shock/NaCl” sessions and the percent of hM4Di expression within dSNc (G) or mSNc (H). A significant correlation was found for dSNc [F_(1,5) = 7.64, P < 0.05], but not for mSNc [F_(1,5) = 0.18, P > 0.5]. Data are expressed as mean ± SEM. Bonferroni correction post-hoc analysis: **, P < 0.01; ***, P < 0.001.

Fig. 4. Chemogenetically-induced nigrostriatal hypodopaminergia leads to a prolonged negative motivational, but not affective, state.

A Behavioral screening timeline. B Quantification of hM4Di-mCherry and mCherry expression in distal (dSNc), medial SNc (mSNc) and VTA. Three-way ANOVA showed significant effect of the area [F_{(2, 140)} = 346.7, P < 0.01, η² = 0.83], but no effect of transgene, treatment or any interaction between these factors [F_s < 1.31, P > 0.27, partial η² < 0.02]. C Number of 2.5%-sucrose deliveries obtained in 60-min self-administration (SA) sessions under continuous reinforcement (FR1), during baseline (BL), “C21/NaCl” and “Wash” sessions, in hM4Di rats treated with C21 (orange, n = 8) or NaCl (red, n = 9) and mCherry rats treated with C21 (gray, n = 10) or NaCl (white, n = 10). RM three-way ANOVA showed a significant session x transgene x treatment interaction [F_{(11, 363)} = 2.67, P < 0.01, partial η² = 0.07]. D Mean sucrose deliveries obtained during “C21/NaCl” sessions normalized to baseline. Two-way ANOVA showed a significant transgene x treatment interaction [F_{(1, 33)} = 13.38, P < 0.001, partial η² = 0.29]. E—I Distance traveled over the course of a 30-min session in an open area (OA) (E), number of adjusting left and right forepaws in a stepping test (F), sucrose preference measured over a 60-min 2.5%-sucrose two-bottle-choice (2BC) drinking session (G), percentage of time spent in the light area in a light/dark avoidance test (LDA) (H) and time spent immobile in a forced swim test (FST) (I) in hM4Di rats treated with C21 (orange, n = 12) or NaCl (red, n = 11) and mCherry rats treated with C21 (gray, n = 13) or NaCl (white, n = 12). Two- or three-way ANOVA found no interaction implicating the transgene and the treatment [F_s < 1.7, P > 0.2, partial η² < 0.07] in these different tests, except a marginal side x transgene x treatment interaction in stepping test [F_{(1, 44)} = 3.55, P = 0.07, partial η² = 0.07] and a transgene x treatment interaction in LDA [F_{(1, 44)} = 3.68, P = 0.06, partial η² = 0.08], mainly driven by small size effects related to the transgene condition and not to a C21 effect on hM4Di rats. Data are expressed as mean ± SEM. Bonferroni correction post-hoc analysis: ***, P < 0.001.

Fig. 5. OSU6162 compound temporarily decreased compulsive behavior in footshock-resistant rats.

A Experiment timeline. B Alcohol intake during IA 20%-EtOH 2BC in NaCl rats (black, n = 12) and OSU rats (purple, n = 14). RM two-way ANOVA showed a significant effect of session [F_{(4, 115)} = 18.4, P < 0.001, partial η² = 0.43] but no effect of group or session x group interaction [F_s < 0.75, P > 0.5, partial η² < 0.03]. C Number of active lever presses in 30-min self-administration sessions of 20%-EtOH (FR3), during baseline, “Shock”, “Wash” and “Shock/OSU” sessions at either 7.5 mg/kg or 15 mg/kg. Mixed-effects analysis showed a significant session x treatment interaction [F_{(32, 761)} = 1.6, P < 0.05]. D Mean active lever presses during the “No shock” and the first “Shock” session with OSU administration at 7.5 mg/kg (left) or 15 mg/kg (right) normalized to baseline. For both conditions, RM two-way ANOVA showed a significant effect of the treatment [F_s > 7.17; P < 0.05; η² > 0.13] with a marginal treatment x shock interaction [F_s > 3.57; P < 0.07; η² > 0.14] but no effect of the shock [F_s < 0.24; P > 0.5; η² < 0.01]. Data are expressed as mean ± SEM. Bonferroni correction post-hoc analysis: *, P < 0.05; **, P < 0.01; ***, P < 0.001.