Nucleus Accumbens Shell Orexin-1 Receptors Are Critical Mediators of Binge Intake in Excessive-Drinking Individuals

Abstract

Excessive, binge alcohol drinking is a potent and pernicious obstacle to treating alcohol use disorder (AUD), and heavy-drinking humans are responsible for much of the substantial costs and harms of AUD. Thus, identifying key mechanisms that drive intake in higher-drinking individuals may provide important, translationally useful therapeutic interventions. Orexin-1-receptors (Ox1Rs) promote states of high motivation, and studies with systemic Ox1R inhibition suggest a particular role in individuals with higher intake levels. However, little has been known about circuits where Ox1Rs promote pathological intake, especially excessive alcohol consumption. We previously discovered that binge alcohol drinking requires Ox1Rs in medial nucleus accumbens shell (Shell), using two-bottle-choice Drinking-in-the-Dark (2bc-DID) in adult, male C57BL/6 mice. Here, we show that Shell Ox1Rs promoted intake during intermittent-access alcohol drinking as well as 2bc-DID, and that Shell inhibition with muscimol/baclofen also suppressed 2bc-DID intake. Importantly, with this large data set, we were able to demonstrate that Shell Ox1Rs and overall activity were particularly important for driving alcohol consumption in higher-drinking individuals, with little overall impact in moderate drinkers. Shell inhibition results were compared with control data combined from drug treatments that did not reduce intake, including NMDAR or PKC inhibition in Shell, Ox1R inhibition in accumbens core, and systemic inhibition of dopamine-1 receptors; these were used to understand whether more specific Shell Ox1R contributions in higher drinkers might simply result from intrinsic variability in mouse drinking. Ineffectiveness of Shell inhibition in moderate-drinkers was not due to a floor effect, since systemic baclofen reduced alcohol drinking regardless of basal intake levels, without altering concurrent water intake or saccharin consumption. Finally, alcohol intake in the first exposure predicted consumption levels weeks later, suggesting that intake level may be a stable trait in each individual. Together, our studies indicate that Shell Ox1Rs are critical mediators of binge alcohol intake in higher-drinking individuals, with little net contribution to alcohol drinking in more moderate bingers, and that targeting Ox1Rs may substantially reduce AUD-related harms.

Keywords: SB-334867; alcohol; binge; excessive drinkers; nucleus accumbens; orexin; shell.

FIGURE 1

Ox1R blockade or GABAergic inhibition within medial NAc Shell significantly reduced alcohol drinking in mice. (A) Intra-Shell infusion of the Ox1R inhibitor SB reduced alcohol intake in the 2bc-DID model, replicating previous studies (Lei et al., 2016b). For this and all other bar graphs, open bar is with vehicle injection, shaded bar is with drug injection, tested within-animal. (B) Intra-Shell SB reduced alcohol intake in the 2bc-IA. (C) Global inhibition of Shell with GABA receptor agonists muscimol/baclofen significantly reduced 2bc-DID alcohol intake. (D) Ox1R inhibition in Core did not significantly reduce 2bc-DID alcohol drinking. Shell Ox1R blockage did not alter concurrent water intake in 2bc-DID (E) or 2bc-IA (F); water intake indicated by cross-hatching. (G) Shell GABA-mediated inhibition did not alter concurrent water intake in 2bc-DID. M/B, muscimol/baclofen; SB, SB-334867. ^∗p < 0.05, ^∗∗p < 0.01.

FIGURE 2

Shell Ox1Rs and activity were important for driving alcohol intake predominantly in higher-drinking individuals. Data combined from Shell inhibition groups (A) and separately combined from control (no-change) groups (where drug treatment had no impact on alcohol drinking) (B) examining how treatment-related drinking change relates to basal alcohol intake levels across individuals (basal consumption determined from vehicle-injected test sessions). The slope of the Shell Inhibition group (red line in A,B) was significantly greater than the slope in Control group (black line in A,B). (C,D) We next calculated the drug-related drinking change relative to basal intake levels in each animal: log[100^∗(intake during drug treatment)/(intake during vehicle)]. Subjects with higher basal drinking showed a significantly greater impact of Shell inhibition, relative to individuals with more moderate basal consumption. Since the change in intake with treatment is on a log scale, yellow–brown dashed lines are included to indicate no change in drinking (0% drop in intake with treatment) compared with 50% drop in intake, and 100% drop in intake is also indicated. (E–J) A median split was used to divide individuals into moderate and higher basal drinkers (Moorman and Aston-Jones, 2009; Alcaraz-Iborra et al., 2017; Moorman et al., 2017). (E,F) Alcohol intake pre- and post-treatment in moderate versus higher basal drinkers, where Shell inhibition strongly reduced alcohol intake levels only in higher drinkers. (G) Shell inhibition produced a significantly greater decrease in alcohol drinking in higher-drinkers relative to moderate-drinkers. (H) In the control group, there was no difference in treatment-related change between higher and moderate basal drinkers. (I,J) Histograms of the number of mice showing different levels of change in drinking with treatment, binned as described in Section “Materials and Methods.” (I) In higher-drinkers, control mice (black) showed a normal distribution with a strong peak at log value of 2, indicating 100% of basal intake (no change). In contrast, higher-drinking mice with Shell inhibition (red) showed a clear shift to the left, indicating inhibition of alcohol consumption. (J) In moderate-drinkers, there was little difference in the distribution between Shell inhibition mice (red) and controls (black). ^∗p < 0.05, ^∗∗∗p < 0.001.

FIGURE 3

Shell regulation of alcohol drinking only in higher-drinking individuals was not due to a floor effect or other confounds. (A) 5 mg/kg baclofen (i.p.) significantly reduced 2bc-DID intake. (B) Baclofen reduction in alcohol drinking occurred in both moderate and higher basal drinkers. Dashed yellow–brown line indicates no effect of treatment on drinking level (0% drop in drinking). Baclofen inhibition of alcohol intake was not due to more general motivational or motor changes, since concurrent water intake (C) and saccharin intake tested in separate mice (D, diagonal lines) were not reduce by 5 mg/kg baclofen. 1 mg/kg baclofen slightly but significantly reduced alcohol intake (E) without altering concurrent water consumption (F). (G,H) While concurrent water intake did not change during Shell Ox1R inhibition, higher basal alcohol intake was not accompanied by a larger decrease in water consumption for 2bc-DID intake (F, data from Figure 1E and Lei et al., 2016b) or 2bc-IA intake (G). In fact, for 2bc-DID, higher basal intake was slightly but significantly correlated with greater water drinking. Together, these are further evidence that the selective effect of Shell inhibition in higher basal drinkers is not inexorably due to mathematical or other differences in higher or moderate intake individuals. BAC, baclofen. ^∗p < 0.05, ^∗∗p < 0.01.

FIGURE 4

Control groups where 2bc-DID alcohol drinking was not altered by drug treatment. (A) Shell PKC inhibition did not alter 2bc-DID intake. (B) Shell NMDAR inhibition did not alter alcohol intake (combined 2bc-DID and 2bc-IA). (C) Systemic inhibition of D1Rs did not reduce 2bc-DID alcohol intake. (D) Inhibition of NMDARs in vmPFC did not reduce 2bc-DID alcohol consumption. OrexinA infusion in the Shell reduced alcohol intake (E) (combined 2bc-DID and 2bc-IA), which was prevented by co-infusion of the Ox2R blocker TCS (F). 10 pmol/side orexinA also tended to decrease drinking (vehicle: 2.11 ± 0.30 g/kg; orexinA: 1.48 ± 0.12 g/kg; p = 0.0577, n = 4). ^∗p < 0.05.

FIGURE 5

Alcohol drinking on the first day of intake predicted consumption 3 weeks later. Alcohol drinking level on the first day of intake (a 24 h session) significantly predicted the average 2 h intake during week 1 (A), week 2 (B), and week 3 (C) of 2bc-DID. This suggests that the basal alcohol consumption level may reflect a more stable trait within each individual, although the variability in mice indicates that some caution is warranted in this interpretation. ^∗∗∗∗p < 0.0001.

FIGURE 6

Histology of cannulae placements. Placements for 2bc-DID Shell SB (A, circles), 2bc-IA Shell SB (B, circles), Shell muscimol/baclofen (C, triangles), Shell orexin (C, circles), Core SB (D, triangles), Shell AP5 (D, circles), Shell 1 μg-side SB (D, X), Shell PKC block (E, circles), orexin+TCS (Ox2R blocker) (E, triangles), vmPFC AP5 (F, circles). All placements are for Shell, except for Core and vmPFC as specifically indicated.