Chemogenetic inhibition of a monosynaptic projection from the basolateral amygdala to the ventral hippocampus selectively reduces appetitive, but not consummatory, alcohol drinking-related behaviours

Abstract

Alcohol use disorder (AUD) and anxiety/stressor disorders frequently co-occur and this dual diagnosis represents a major health and economic problem worldwide. The basolateral amygdala (BLA) is a key brain region that is known to contribute to the aetiology of both disorders. Although many studies have implicated BLA hyperexcitability in the pathogenesis of AUD and comorbid conditions, relatively little is known about the specific efferent projections from this brain region that contribute to these disorders. Recent optogenetic studies have shown that the BLA sends a strong monosynaptic excitatory projection to the ventral hippocampus (vHC) and that this circuit modulates anxiety- and fear-related behaviours. However, it is not known if this pathway influences alcohol drinking-related behaviours. Here, we employed a rodent operant self-administration regimen that procedurally separates appetitive (e.g. seeking) and consummatory (e.g., drinking) behaviours, chemogenetics and brain region-specific microinjections, to determine if BLA-vHC circuitry influences alcohol and sucrose drinking-related measures. We first confirmed prior optogenetic findings that silencing this circuit reduced anxiety-like behaviours on the elevated plus maze. We then demonstrated that inhibiting the BLA-vHC pathway significantly reduced appetitive drinking-related behaviours for both alcohol and sucrose while having no effect on consummatory measures. Taken together, these findings provide the first indication that the BLA-vHC circuit may regulate appetitive reward seeking directed at alcohol and natural rewards and add to a growing body of evidence suggesting that dysregulation of this pathway may contribute to the pathophysiology of AUD and anxiety/stressor-related disorders.

Keywords: alcohol addiction; anxiety; motivation; seeking; ventral subiculum.

FIGURE 1

Schematic illustration of the time line for the chemogenetic operant self-administration experiments. Using the fading procedure described in Section 2, cohorts of rats were trained over a 3-week period to complete a 30-lever press response requirement to gain access to a sipper tube containing 10% ethanol or 3% sucrose for 20 min drinking sessions that were conducted 5 days/week. During Weeks 3–5, a stable baseline of operant self-administration was obtained and a baseline extinction probe trial (EPT) was conducted. Between Weeks 5 and 10, standard drinking sessions were conducted, and test sessions, in which subjects received intra-vHC VEH or CNO (10, 33 and 100 ng/side) microinjections, were conducted every Wednesday, using a Latin square design. Finally, a similar experimental design was used between Weeks 10 and 12 to test the effect of intra-vHC CNO (33 ng/side) on EPT responding.

FIGURE 2

The BLA makes monosynaptic excitatory projections to the vHC that can be inhibited by CNO in animals expressing ChR2 and Gi-DREADD. (a) Representative images of viral expression at the injection site (top panel) in the BLA and its projections in the vHC (lower panel). Nuclear DAPI staining is shown in blue and viral expression is shown in red. (b) Schematic illustration of viral spread and expression in and around the injection site. (c) Schematic illustration of viral spread throughout the vHC. In (b) and (c), numbers to the right of each brain slice indicate anterior/posterior stereotaxic coordinates relative to bregma. To construct composite maps, expression/projection spread in the target regions (BLA and vHC) from each animal and hemisphere was overlaid on a single hemisphere. Colour gradient (from light to dark red) within injection/projection sites illustrates the consistency of viral expression/projection within the overall area among individual animals and hemispheres. (d) Schematic illustration of cannula placements in the BLA for the Gi-DREADD ethanol cohort (black), the Gi-DREADD sucrose cohort (red), the Gi-DREADD elevated plus maze cohort (blue) and the ‘reporter only’ cohort (green). (e) CNO inhibits optically evoked BLA inputs in the vHC of animals coexpressing ChR2 and Gi-DREADD (baseline: 158.3 ± 48.8 pA, CNO: 32.5 ± 12.7 pA, n = 9, Wilcoxon signed-rank test, ^**p < .01). Representative oEPSCs traces (upper panel) in a vHC neuron during baseline (black traces) and CNO application (purple traces). Thin lines represent individual responses page 35 of 39 European journal of neuroscience while thick lines illustrate average response during baseline and CNO application. Bar graph (lower panel) of oEPSC amplitudes at baseline and during the application of CNO illustrating responses for individual neurons.

FIGURE 3

Chemogenetic inhibition of the BLA-vHC circuit decreases anxiety-like behaviour on the elevated plus maze. (a) In rats expressing Gi-DREADD, intra-vHC infusion of CNO (100 ng/side) increased open arm time and open arm entries with no effect on closed arm entries, a measure of locomotor activity (N = 7; *p < .05, paired t test). (b) In rats only expressing a reporter protein, intra-vHC infusion of CNO had no effect on any measures on the elevated plus maze (N = 5, p > .05, paired t test).

FIGURE 4

Chemogenetic inhibition of the BLA-vHC circuit has no effect on ethanol or sucrose consumption during operant drinking sessions. (a) In Gi-DREADD expressing rats trained to self-administer 10% ethanol (N = 8), intra-vHC injection of CNO (10, 33 and 100 ng/side) had no effect on ethanol intake during regular operant drinking sessions (one-way repeated measures ANOVA, p > .05). (b) Intra-vHC infusion of the reporter protein control group, trained to self-administer 10% ethanol (N = 7), intra-vHC infusion of CNO (100 ng/side) had no effect on operant ethanol intake (paired t test, p > .05). (c) In Gi-DREADD expressing rats trained to self-administer 3% sucrose (N = 7), intra-vHC injection of CNO (10, 33 and 100 ng/side) had no effect on operant sucrose intake (one-way repeated measures ANOVA, p > .05).

FIGURE 5

Chemogenetic inhibition of the BLA-vHC circuit reduces extinction probe trial responding for ethanol and sucrose. (a) Intra-vHC infusion of CNO (33 ng/side) significantly reduced lever responding during extinction probe trials in Gi-DREADD expressing rats trained to self-administer 10% ethanol (N = 8). (b) Intra-vHC infusion of CNO had no effect on lever responding during extinction probe trials in reporter protein control rats trained to self-administer 10% ethanol (N = 7). (c) Intra-vHC infusion of CNO significantly reduced lever responding during extinction probe trials in Gi-DREADD expressing rats trained to self-administer 3% sucrose (N = 7). *p < .05, paired t tests. (d) Time course of lever presses/min in the VEH and CNO-treated ethanol cohort during extinction probe trials. *p < .05. (e) Time course of lever presses/min in the VEH and CNO-treated sucrose cohort during extinction probe trials. *p < .05.

FIGURE 6

Comparisons of the effects of intra-vHC CNO (300 ng/side) on additional consummatory and appetitive measures, assessed during regular drinking sessions, in ethanol (N = 8) and sucrose (N = 7) groups. (a) Chemogenetic inhibition of the BLA-vHC circuit had no effect on any of the consummatory measures (time to first lick, licks/min and lick bouts) in either group. Chemogenetic inhibition of the BLA-vHC circuit significantly reduced two appetitive measures (lever press completion time and lever press bouts with a trend of an inhibition of time to first lever press). No significant effect of reinforcer or interactions between this factor and treatment were observed for any of the consummatory or appetitive measures assessed. ^**p < .01; ^***p < .001, ^#p < .08, two-way mixed-model ANOVAs.