Bed Nucleus of the Stria Terminalis (BNST) neurons containing the serotonin 5HT2c receptor modulate operant alcohol self-administration behavior in mice

Abstract

The serotonin 5HT_2c receptor has been widely implicated in the pathophysiology of alcohol use disorder (AUD), particularly alcohol seeking and the affective consequences of chronic alcohol consumption. However, little is known about the brain sites in which 5HT_2c exerts its effects on specific alcohol-related behaviors, especially in females. Here, we investigated the effects of site-specific manipulation of the 5HT_2c receptor system in the BNST on operant alcohol self-administration behaviors in adult mice of both sexes, including the acquisition and maintenance of fixed-ratio responding, motivation for alcohol (progressive ratio), and quinine-adulterated responding for alcohol on a fixed-ratio schedule (punished alcohol seeking). Knockdown of 5HT_2c in the BNST did not affect the acquisition or maintenance of operant alcohol self-administration, nor did it affect progressive ratio responding for alcohol. This manipulation had only a subtle effect on responding for quinine alcohol selectively in females. On the other hand, chemogenetic inhibition of BNST 5HT_2c-containing neurons (BNST^5HT2c) increased operant alcohol self-administration behavior in both sexes on day 2, but not day 9, of testing. It also increased operant responding for 1000 μM quinine-adulterated alcohol selectively in males. Importantly, chemogenetic inhibition of BNST^5HT2c did not alter operant sucrose responding or motivation for sucrose in either sex. We then performed cell-type specific anterograde tracing, which revealed that BNST^5HT2c project to similar regions in males and females, many of which have been previously implicated in AUD. We next used chemogenetics and quantification of the immediate early gene cFos to characterize the functional influence of BNST^5HT2c inhibition on vlPAG activity. We show that chemogenetic inhibition of BNST^5HT2c reduces vlPAG cFos in both sexes, but that this reduction is more robust in males. Together these findings suggest that BNST^5HT2c neurons, and to a small extent the BNST 5HT_2c receptor, serve to promote aversive responses to alcohol consumption, potentially through sex-dependent disinhibition of vlPAG neurons.

Fig. 1. Genetic knockdown of the serotonin 5HT_2c receptor in the BNST promotes punished alcohol seeking in females.

A Representative viral infection in BNST. B Active and inactive lever presses in FR4, females (Active: n = 13 GFP, n = 12 Cre; Two-way repeated measures ANOVA). C Average active lever presses in FR4, both sexes (n = 13 GFP females, n = 12 Cre females, n = 17 GFP males, n = 13 Cre males; Two-way ANOVA). D Active lever presses in PR, both sexes (n = 13 GFP females, n = 12 Cre females, n = 17 GFP males, n = 13 Cre males; Two-way ANOVA). E Active and inactive lever presses in FR4, males (Active: n = 17 GFP, n = 13 Cre; Two-way ANOVA). F Average alcohol intake in FR4, both sexes (n = 13 GFP females, n = 12 Cre females, n = 17 GFP males, n = 13 Cre males). G Break point in PR, both sexes (n = 13 GFP females, n = 12 Cre females, n = 17 GFP males, n = 13 Cre males). H Active lever presses for quinine-adulterated alcohol seeking, females (n = 13 GFP, n = 12 Cre; Two-way repeated measures ANOVA). I Alcohol intake for quinine-adulterated alcohol seeking, females (n = 13 GFP, n = 12 Cre; Two-way repeated measures ANOVA). J Active lever presses for quinine-adulterated alcohol seeking, males (n = 17 GFP, n = 13 Cre; Two-way repeated measures ANOVA). K Alcohol intake for quinine-adulterated alcohol seeking, males (n = 17 GFP, n = 13 Cre; Two-way repeated measures ANOVA). # denotes quinine concentration x virus effect, & denotes effect of quinine concentration, @ denotes effect of sex. Figure made with Biorender.com. All data presented as mean ± SEM.

Fig. 2. Chemogenetic inhibition of BNST^5HT2c promotes punished alcohol seeking in males.

A Representative image of viral infection in BNST. B Lever presses in FR4, females (n = 8 mCherry, n = 8 hM4di; Two-way repeated measures ANOVA (d2–d8)). C Active lever presses day 2 FR4 during chemogenetic inhibition, both sexes (n = 8 mCherry females, n = 8 hM4di females, n = 7 mCherry males, n = 7 hM4di males; Two-way ANOVA). D Active lever presses in PR, both sexes (n = 8 mCherry females, n = 8 hM4di females, n = 7 mCherry males, n = 7 hM4di males). E Lever presses in FR4, males (n = 7 mCherry, n = 7 hM4di; Two-way repeated measures ANOVA (d2–d8)). F Alcohol intake on d2 FR4 during chemogenetic inhibition, both sexes (n = 8 mCherry females, n = 8 hM4di females, n = 7 mCherry males, n = 7 hM4di males; Two-way ANOVA). G Break point in PR, both sexes (n = 8 mCherry females, n = 8 hM4di females, n = 7 mCherry males, n = 7 hM4di males; Two-way ANOVA). H Active lever presses for quinine-adulterated alcohol seeking without chemogenetic manipulation, females (n = 8 mCherry, n = 8 hM4di; Two-way repeated measures ANOVA). I Active lever presses for alcohol + 250 uM quinine during chemogenetic inhibition, females (n = 8 mCherry, n = 8 hM4di; Two-way repeated measures ANOVA). J Active lever presses for alcohol + 1000 uM quinine during chemogenetic inhibition, females (n = 8 mCherry, n = 8 hM4di; Two-way repeated measures ANOVA). K Alcohol + 1000 uM quinine intake during chemogenetic inhibition, females (n = 8 mCherry, n = 8 hM4di; Two-way repeated measures ANOVA). L Active lever presses for quinine-adulterated alcohol seeking without chemogenetic inhibition, males (n = 7 mCherry, n = 7 hM4di; Two-way repeated measures ANOVA). M Active lever presses for alcohol + 250 uM quinine during chemogenetic inhibition, males (n = 7 mCherry, n = 7 hM4di; Two-way repeated measures ANOVA). N Active lever presses for alcohol + 1000 uM quinine during chemogenetic inhibition, males (n = 7 mCherry, n = 7 hM4di; Two-way repeated measures ANOVA). O Alcohol + 1000 uM quinine intake during chemogenetic inhibition, males (n = 7 mCherry, n = 7 hM4di; Two-way repeated measures ANOVA). $ denotes effect of virus, & denotes effect of quinine concentration, @ denotes effect of sex, # denotes CNO x virus interaction effect, * denotes post-hoc effects. Figure made with Biorender.com. All data presented as mean ± SEM.

Fig. 3. Chemogenetic inhibition of BNST^5HT2c does not influence sucrose seeking.

A Representative viral infection in the BNST. B Lever presses in FR4 without chemogenetic inhibition, females (n = 8 mCherry, n = 9 hM4di). C Active lever presses on d2 FR4 during chemogenetic inhibition, both sexes, all mice received 3 mg/kg CNO (n = 8 mCherry females, n = 9 hM4di females, n = 8 mCherry males, n = 7 hM4di males; Two-way ANOVA). D Active lever presses in PR, both sexes during chemogenetic inhibition, both sexes, all mice received 3 mg/kg CNO (n = 8 mCherry females, n = 9 hM4di females, n = 8 mCherry males, n = 7 hM4di males; Two-way ANOVA). E Lever presses in FR4 without chemogenetic inhibition, males (n = 8 mCherry, n = 7 hM4di). F Sucrose intake on d2 FR4 during chemogenetic inhibition, all mice received 3 mg/kg CNO, both sexes (n = 8 mCherry females, n = 9 hM4di females, n = 8 mCherry males, n = 7 hM4di males; Two-way ANOVA). G Break point in PR during chemogenetic inhibition, all mice received 3 mg/kg CNO, both sexes (n = 8 mCherry females, n = 9 hM4di females, n = 8 mCherry males, n = 7 hM4di males; Two-way ANOVA). @ denotes effect of sex. Figure made with Biorender.com. All data represented as mean + SEM.

Fig. 4. BNST^5HT2c anterograde tracing in both sexes.

A Representative viral infections in BNST, females (left) and males (right). B BNST^5HT2c post-synaptic terminals in Nucleus Accumbens (Nac). C BNST^5HT2c post-synaptic terminals in Lateral Septum (LS). D BNST^5HT2c post-synaptic terminals in Lateral Hypothalamus (LH). E BNST^5HT2c post-synaptic terminals in Lateral Habenula (LHb). F BNST^5HT2c post-synaptic terminals in Ventromedial Hypothalamus (VMH). G BNST^5HT2c post-synaptic terminals in Central Amygdala (CeA). H BNST^5HT2c post-synaptic terminals in Ventral Tegmental Area (VTA). I BNST^5HT2c post-synaptic terminals in Dorsal Raphe Nucleus (DRN). J BNST^5HT2c post-synaptic terminals in ventro-lateral Periacqueductal Gray (vlPAG); males show qualitatively more dense terminal expression than females.

Fig. 5. Chemogenetic inhibition of BNST^5HT2c reduces BNST and vlPAG activity in a sex-specific manner.

A Quantification of mCherry+ nuclei in DREADD (hM4di) males and females (n = 9 females, n = 8 males; Student’s two-tailed unpaired t-test). B Immunohostochemistry for cFos (green), mCherry (red), and DAPI (blue) in the BNST following chemogenetic inhibition of BNST^5HT2c. C Immunohistochemistry for cFos (green) in the vlPAG following chemogenetic inhibition of BNST^5HT2c. D Quantification of BNST cFos + /5HT2c+ nuclei in both sexes (Two-way ANOVA and Bonferroni post-hoc; n = 7 mCherry males, n = 8 hM4di males, n = 8 mCherry females, n = 9 hM4di females). E Quantification of BNST cFos as % of mcherry cells (Two-way ANOVA and Bonferroni posthoc, n = 8 hM4di males, n = 9 hM4di females). F Quantification of total BNST cFos+ nuclei (Two-way ANOVA and Bonferroni post-hoc; n = 7 mCherry male, n = 8 hM4di male, n = 8 mCherry female n = 9 hM4di female). G Quantification of vlPAG cFos+ neurons in males and females (n = 7 mCherry male, n = 8 hM4di male, n = 8 mCherry female n = 9 hM4di female). H Quantification of vlPAG cFos+ nuclei in males (Student’s two-tailed unpaired t-tests; n = 7 mCherry, n = 8 hm4d). I Quantification of vlPAG cFos+ nuclei in females (Student’s two-tailed unpaired t-tests; n = 8 mCherry, n = 9 hM4d). J vlPAG cFos in males and females normalized to number of BNST mCherry cells in DREADD mice (# vlPAG cFos/# BNST mCherry) (Student’s two-tailed unpaired t-test; n = 8 males, n = 9 females). K vlPAG cFos in males and females normalized to number of BNST mCherry cells in DREADD mice (# vlPAG cFos/# BNST mCherry) (Two-way ANOVA, n = 8 males, n = 9 females). @ denotes effect of sex, $ denotes effect of virus, # denotes sex x virus interaction, * denotes post-hoc effect. Figure made with Biorender.com. All data represented as mean + SEM.