Chronic ethanol exposure in mice evokes pre‐ and postsynaptic deficits in GABAergic transmission in ventral tegmental area GABA neurons

Abstract

Background and purpose

GABAergic neurons in mouse ventral tegmental area (VTA) exhibit elevated activity during withdrawal following chronic ethanol exposure. While increased glutamatergic input and decreased GABA_A receptor sensitivity have been implicated, the impact of inhibitory signaling in VTA GABA neurons has not been fully addressed.

Experimental approach

We used electrophysiological and ultrastructural approaches to assess the impact of chronic intermittent ethanol vapour exposure in mice on GABAergic transmission in VTA GABA neurons during withdrawal. We used CRISPR/Cas9 ablation to mimic a somatodendritic adaptation involving the GABA_B receptor (GABA_BR) in ethanol‐naïve mice to investigate its impact on anxiety‐related behaviour.

Key results

The frequency of spontaneous inhibitory postsynaptic currents was reduced in VTA GABA neurons following chronic ethanol treatment and this was reversed by GABA_BR inhibition, suggesting chronic ethanol strengthens the GABA_BR‐dependent suppression of GABAergic input to VTA GABA neurons. Similarly, paired‐pulse depression of GABA_A receptor‐dependent responses evoked by optogenetic stimulation of nucleus accumbens inputs from ethanol‐treated mice was reversed by GABA_BR inhibition. Somatodendritic currents evoked in VTA GABA neurons by GABA_BR activation were reduced following ethanol exposure, attributable to the suppression of GIRK (K_ir3) channel activity. Mimicking this adaptation enhanced anxiety‐related behaviour in ethanol‐naïve mice.

Conclusions and implications

Chronic ethanol weakens the GABAergic regulation of VTA GABA neurons in mice via pre‐ and postsynaptic mechanisms, likely contributing to their elevated activity during withdrawal and expression of anxiety‐related behaviour. As anxiety can promote relapse during abstinence, interventions targeting VTA GABA neuron excitability could represent new therapeutic strategies for treatment of alcohol use disorder.

Abbreviations

AAV

adeno‐associated virus

Cas9

CRISPR‐associated protein 9

ChR2

channelrhodopsin‐2

CRISPR

clustered regularly interspaced short palindromic repeats

GAD

glutamate decarboxylase

GIRK/K_ir3 channel

G protein‐gated inwardly rectifying K⁺ channel

Kir3.1/2/3

GIRK channel subunit 1, 2 & 3 or K_ir3.1, 3.2 & 3.3

gRNA

guide RNA

NAc

nucleus accumbens

sIPSCs

spontaneous inhibitory postsynaptic currents

VTA

ventral tegmental area

What is already known

• Chronic exposure to ethanol provokes neuroadaptations that contribute to the development of alcohol use disorder.

• GABA neurons in the VTA are hyperexcitable during withdrawal following chronic ethanol exposure in rodents.

What does this study add

• Ethanol exposure in mice provokes multiple adaptations that weaken GABAergic regulation of VTA GABA neurons.

• Mimicking one adaptation in VTA GABA neurons enhances anxiety‐related behaviour in ethanol‐naïve mice.

What is the clinical significance

• Interventions targeting VTA GABA neuron excitability during withdrawal could prevent relapse and support abstinence.

1. INTRODUCTION

Alcohol use disorder (AUD) is a chronic progressive disorder associated with compulsive alcohol consumption, inability to suppress drinking and a negative affective state that emerges with withdrawal and drives relapse (Koob, 2024). Neuroadaptations provoked by chronic alcohol exposure (Roberto & Varodayan, 2017), including those involving γ‐aminobutyric acid (GABA) and its target receptors (Dharavath et al., 2023), are thought to underpin the escalation of drinking and development of compulsive behaviour. Accordingly, understanding key molecular and cellular targets of ethanol and the impact of chronic ethanol exposure and withdrawal on these targets may suggest new opportunities for treatment of alcohol use disorder.

Acute ethanol can potentiate inhibitory currents mediated by ionotropic GABA_A receptors (GABA_AR), an effect proposed to occur through direct binding or facilitation of presynaptic GABA release (Lobo & Harris, 2008; Olsen et al., 2014; Valenzuela & Jotty, 2015). Ethanol can also enhance signalling via the G protein‐coupled GABA_B receptor (GABA_BR) (Federici et al., 2009), likely by binding to and activating its key somatodendritic effector, the G protein‐gated inwardly rectifying K⁺ (GIRK/K_ir3) channel (Aryal et al., 2009; Bodhinathan & Slesinger, 2013; Kobayashi et al., 1999; Lewohl et al., 1999). In addition, ethanol can potentiate GABA release in brain regions associated with reward processing and anxiety (Kelm et al., 2011; Roberto et al., 2003; Theile et al., 2008; Zhu & Lovinger, 2006).

Ethanol‐induced adaptations in GABA_AR expression level, subunit composition and localization have been implicated in ethanol dependence and enhanced voluntary consumption (Olsen & Liang, 2017). The ethanol‐induced potentiation of GABA transmission is limited by presynaptic GABA_BR, whose influence can increase with chronic ethanol exposure (Ariwodola & Weiner, 2004; Silberman et al., 2009). Chronic ethanol exposure also weakens somatodendritic GABA_BR‐GIRK signalling in neurons, including principal neurons of the basal amygdala and orbitofrontal cortex (Marron Fernandez de Velasco et al., 2023; Nimitvilai et al., 2017). The GABA_BR agonist baclofen and positive allosteric modulators are promising therapeutic options for alcohol use disorder (Augier, 2021; de Beaurepaire et al., 2018; Logge et al., 2022), with particular efficacy in cases involving co‐morbid anxiety (Krupitsky et al., 1993; Morley et al., 2014).

Drugs with abuse liability share an ability to increase dopamine in the mesocorticolimbic circuitry (Luscher & Ungless, 2006). While this has prompted significant interest in ventral tegmental area (VTA) dopamine neurons, VTA GABA neurons also play a role in substance use disorders (Bouarab et al., 2019). In rats, ethanol acutely increases VTA dopamine neuron firing rate, while decreasing VTA GABA neuron firing rate (Brodie et al., 1999; Gallegos et al., 1999). With chronic exposure, VTA GABA neurons exhibit elevated activity and are tolerant to the acute inhibitory effect of ethanol (Gallegos et al., 1999). While increased glutamatergic input has been implicated in the increased activity of VTA GABA neurons seen during withdrawal (Williams et al., 2018), diminished GABA_AR‐dependent regulation of VTA GABA neurons and aberrant plasticity of GABAergic inputs may also play a role (Nelson et al., 2018; Nufer et al., 2023).

Here, we examined several facets of GABAergic neurotransmission in VTA GABA neurons from mice during withdrawal following chronic intermittent ethanol (CIE) vapour exposure. We show that chronic ethanol suppresses GABAergic input to VTA GABA neurons, likely via enhanced presynaptic influence of GABA_BR. Chronic ethanol also diminishes somatodendritic GABA_BR‐GIRK signalling in VTA GABA neurons via the GIRK3‐dependent internalization of GIRK channels and GABA_BR. Finally, we found that modelling the latter adaptation is sufficient to enhance anxiety‐related behaviour in ethanol‐naïve mice.

2. METHODS

2.1. Animals

All animal studies are reported in compliance with the ARRIVE 2.0 guidelines (Percie du Sert et al., 2020) and with the recommendations made by the British Journal of Pharmacology (Lilley et al., 2020). All studies were approved by the Institutional Animal Care and Use Committee at the University of Minnesota. C57BL/6J mice were purchased from The Jackson Laboratory (Bar Harbor, ME, USA). GAD67‐GFP(+) mice, provided by Takeshi Kaneko (Tamamaki et al., 2003), were used in electrophysiological experiments to facilitate the targeted investigation of VTA GABA neurons. Constitutive G irk3 ^−/− (K cnj9 ^−/−) mice were generated previously (Torrecilla et al., 2002) and the complete absence of GIRK3 (K_ir3.3) protein in these mice been established using immunoblotting and immunolabelling approaches (Fernandez‐Alacid et al., 2011; Koyrakh et al., 2005). Girk3 ^−/− mice were crossed with GAD67‐GFP(+) mice for some studies. GADCre (B6J.Cg‐Gad2^tm2(cre)Zjh/MwarJ) mice were obtained from The Jackson Laboratory (JAX #028867) and were crossed with a Cre‐dependent Cas9GFP knock‐in line (B6J.129(B6N)‐Gt (ROSA)26Sor^{tm1(CAG−cas9*,−EGFP)Fezh}/J; JAX #026175) to generate GADCre(+) mice homozygous for the Cas9GFP allele (GADCre(+):Cas9GFP(+) mice); these mice were used for VTA GABA neuron‐specific CRISPR/Cas9 ablation studies. Genotypes were determined using validated PCR‐based genotyping protocols. Male and female mice were used in all studies and group‐housed with no more than 4 males or 5 females in a single microisolator cage. Mice were maintained on a 14:10 h light/dark cycle and were provided ad libitum access to food and water. Mice were randomly assigned to treatment groups and data acquisition or analysis was conducted in blind fashion when feasible.

2.2. Chronic intermittent ethanol exposure

Vapour chambers for chronic intermittent exposure (CIE) studies were constructed as described (Morton et al., 2014). We used a 4‐week chronic intermittent ethanol protocol shown previously to increase handling‐induced convulsions, anxiety‐related behaviour and irritability in ethanol‐treated mice during the first week of withdrawal (Marron Fernandez de Velasco et al., 2023). Each week involved 4 consecutive days of 16‐h overnight vapour sessions (Day 1/1700 h to Day 2/900 h), followed by a 72‐h period of withdrawal. Male and female mice started the chronic intermittent ethanol protocol at 7–9 weeks. To help maintain consistent blood‐ethanol concentrations (BECs) during 16‐h vapour exposure session, mice in the ethanol treatment group were injected with a priming dose containing ethanol (1.5 g·kg⁻¹; 20% vol/vol) and the alcohol dehydrogenase inhibitor pyrazole (68.1 mg·kg⁻¹ intraperitoneal[i.p.]) before being placed in the chamber each day, as described (Borgonetti et al., 2023; Griffin, 2014; Pati et al., 2020; Sidhu et al., 2018). Delivery rate was titrated to yield blood ethanol concentrations of 1.5–2 g·L⁻¹ (150–200 mg·dL⁻¹) as assessed using trunk blood samples from age‐ and strain‐matched sentinels and the EnzyChrom ethanol assay kit (BioAssay Systems, Hayward, CA, USA). Control mice were handled similarly but were administered saline and pyrazole before placement in a chamber where they were exposed to air. As adaptations in GABAergic transmission in VTA GABA neurons reported following chronic ethanol (Nelson et al., 2018) or methamphetamine (Padgett et al., 2012) treatment lasted for at least 7 days, we conducted slice electrophysiological studies 3–7 days after completing the final overnight air‐ or ethanol‐exposure session. No correlations were observed between post‐ chronic intermittent ethanol day and electrophysiological measures.

2.3. Intracranial viral manipulations

Procedures for intracranial viral infusion targeting the VTA (McCall et al., 2019) and nucleus accumbens (NAc; DeBaker et al., 2023) have been described previously. In brief, mice (>50 days) were placed in a stereotaxic frame (David Kopf Instruments, Tujunga, CA, USA) under isoflurane anaesthesia. Body temperature was maintained throughout the surgical procedure using a heated blanket. Mice were administered subcutaneous injections of gentamicin (5 mg/kg) and carprofen (20 mg/kg) prior to surgery. Microinjectors consisting of a 33‐gauge stainless steel tube within a shorter 26‐gauge stainless steel tube were attached to polyethylene‐20 tubing affixed to a 10‐μl Hamilton syringe. Microinjectors were lowered through burr holes in the skull to either the NAc (A/P: +1.6 mm, M/L: ±1.3 mm, D/V: −4.65 mm) or VTA (A/P: −2.75 mm, M/L: ±0.55 mm, D/V: −5.0 mm). AAV vectors (400 nl) were infused at a rate of 100 nl·min⁻¹ and microinjectors were left in place for 5 min to reduce backflow along the injection track upon removal. Upon completion, mice were placed in a clean cage with a heated blanket underneath to allow for recovery within 15–30 minutes. Mice were monitored 3 times daily 24 hr to 72 hr post surgery, during which mice were provided ad libitum access to water containing ibuprofen. Slice electrophysiological and behavioural experiments occurred 4–5 weeks following virus infusion.

2.4. Slice electrophysiology

Slice electrophysiological recordings were performed in acutely isolated slices of the VTA, as described (DeBaker et al., 2021, 2023; McCall et al., 2019). In brief, mice were killed by rapid decapitation following isoflurane anaesthesia (induction chamber with 1 mL isoflurane). Following brain extraction, horizontal slices (225 μM) containing the VTA were prepared and incubated in oxygenated ice‐cold sucrose cutting solution. Slices were hemisected and placed into a chamber containing oxygenated artificial cerebrospinal fluid (aCSF) at 37°C for 10 min, before cooling to room temperature (20–22°C) and acclimation for 1 h prior to recordings. All experiments were conducted in an oxygenated aCSF bath at 32–34°C and only one experiment was conducted per slice. Drugs were applied via an 8‐channel pinch valve perfusion system controlled by a ValveLink 8.3 controller (AutoMate Scientific, Inc; Berkeley, CA, USA). All measured and command potentials accounted for a −10‐mV junction potential calculated using JPCalc software (Molecular Devices, LLC; San Jose, CA). Data were acquired using a MultiClamp 700A amplifier (Molecular Devices), low‐pass filtered at 2 kHz, digitized at 10 kHz and analysed using pCLAMP v.9 software (Molecular Devices). Holding current and series resistance (R_s) were measured before and after each recording. Experiments in which R_s exceeded 20 MΩ were not included in the final analysis.

Cells located medial to the medial terminal nucleus of the optic tract, lateral to midline nuclei (linear nuclei of the raphe or interpeduncular nucleus, when applicable), were targeted for analysis. Consistent with prior reports (Kotecki et al., 2015; Labouebe et al., 2007), GFP‐positive neurons in this region of slices from GAD67‐GFP(+) mice exhibited small (<60 pF) apparent capacitance values and small (<50 pA) I_h currents evoked by a 1‐s voltage ramp from −60 to −120 mV. In studies involving GADCre(+):Cas9GFP(+) mice treated with gRNA‐containing AAV vectors, VTA GABA neurons were identified by Cas9GFP fluorescence; neurons exhibiting GFP and viral‐driven mCherry fluorescence were targeted for analysis. Putative VTA dopamine neurons were identified by lack of Cas9GFP fluorescence; these cells exhibited higher (>60 pF) apparent capacitance, steady baseline firing rate (1–5 Hz) and larger I_h currents (>50 pA). Electrophysiological parameters for VTA GABA neurons in mice exposed to air or ethanol treatments across experimental conditions are summarized in Table 1.

TABLE 1.

Electrophysiological properties of VTA GABA neurons.

Cohort	Treatment	N	n	Capacitance (pF)	Rm (MΩ)	Ih (pA)	% active	Rate (Hz)
Group 1	Air	20	38	40 ± 2	487 ± 43	13 ± 2	84	5.1 ± 0.7
	Ethanol	21	40	42 ± 2	494 ± 45	21 ± 2	80	5.0 ± 1.0
	Test statistic			t76 = 0.60	U = 747	U = 551	Fisher	U = 455
	P value			0.55	0.90	0.054	0.77	0.45
Group 2	Air	8	26	38 ± 2	558 ± 63	12 ± 3	89	6.4 ± 1.2
	Ethanol	8	26	40 ± 2	504 ± 44	17 ± 3	65	4.1 ± 1.1
	Test statistic			t50 = 0.68	U = 319	U = 231.5	Fisher	U = 147.5
	P value			0.50	0.74	0.051	0.098	0.19
Group 3	Air	10	16	41 ± 2	692 ± 135	9 ± 2	81	6.7 ± 1.5
	Ethanol	11	14	40 ± 3	534 ± 96	12 ± 3	57	6.9 ± 1.0
	Test statistic			U = 108	U = 100	t28 = 0.93	Fisher	U = 42.5
	P value			0.89	0.64	0.36	0.24	0.51
Group 4	Air	4	12	42 ± 4	655 ± 189	34 ± 5	100	14.9 ± 3.2
	Ethanol	4	11	36 ± 5	367 ± 87	27 ± 6	73	12.3 ± 3.2
	Test statistic			U = 47	U = 58	U = 43.5	Fisher	U = 40
	P value			0.26	0.65	0.17	0.093	0.57
Group 5	Air	5	12	38 ± 10	468 ± 133	19 ± 14	83	5.7 ± 5.2
	Ethanol	5	12	41 ± 8	752 ± 590	11 ± 8	75	5.8 ± 4.0
	Test statistic			t22 = 0.65	U = 56	U = 43.5	Fisher	U = 45
	P value			0.52	0.37	0.103	>0.9999	>0.9999

Note: Electrophysiological parameters for VTA GABA neurons extracted from experiments conducted using a 140‐mM KCl pipette solution and an aCSF bath in the absence (Group 1) or presence of 2‐μM CGP54626 (Group 2) or a K‐methanesulfonate pipette solution in an aCSF bath in the absence (Group 3) or presence of 0.3‐mM external Ba²⁺ (Group 4) and in constitutive Girk3 ^−/− mice (Group 5). Apparent capacitance (pF), input resistance (R_M) and I_h current did not differ between VTA GABA neurons from ethanol‐ or air‐treated mice. Similarly, ethanol treatment did not impact firing rate in VTA GABA neurons exhibiting spontaneous activity or the fraction of cells exhibiting spontaneous activity.

Abbreviations: N, number of mice/group; n, number of independent experimental units (recordings)/group; VTA, ventral tegmental area.

Spontaneous inhibitory postsynaptic currents (sIPSCs), paired pulse ratio (PPR) and whole‐cell currents evoked by muscimol were measured at a holding potential of −70 mV in an aCSF bath and with the following pipette solution (in mM):‐ 140 KCl, 2 MgCl₂, 1.1 EGTA, 5 HEPES, 2 Na₂‐ATP, 0.3 Na‐GTP and 5 phosphocreatine (pH = 7.4). sIPSCs and PPR were measured in the presence of 2‐mM kynurenic acid to block ionotropic glutamate receptor currents. Experiments examining the impact of GABA_BR inhibition on sIPSCs and paired pulse ratio involved slices prepared from separate cohorts of mice, using a bath solution containing CGP 54626A (CGP54626; 2 μM). For paired pulse ratio experiments, optical stimulation was delivered through a 4× objective above the slice and consisted of 2 pulses of 470‐nm wavelength light (2–5 mW; 2.5‐ms pulse width) at 20 Hz, allowing an inter‐pulse interval of 50 ms. This stimulation protocol was applied in triplicate for each experiment, with a 1‐ to 2‐min rest period between stimulations. Picrotoxin (100 μM) was used to test the GABA_AR dependence of measured currents.

Whole‐cell somatodendritic currents mediated by activation of GABA_AR (muscimol, 30 μM) or GABA_BR (baclofen, 200 μM) were measured in slices prepared from separate cohorts of mice. Muscimol was applied via fast perfusion after holding current (V_hold = −70 mV) stabilized. Once the peak muscimol‐induced current response was achieved (<3 s), muscimol was washed out with aCSF until the holding current returned to baseline. After 2–3 min, muscimol was again applied to the cell, now in the presence of picrotoxin (100 μM), to test the GABA_AR dependence of observed currents. Baclofen‐evoked current measurements were conducted as described (Kotecki et al., 2015) but with a 140‐mM K‐methanesulfonate pipette solution. In these experiments, spontaneous firing rate was assessed in current‐clamp mode during the 20‐s interval immediately following whole‐cell formation. Cells were then switched to voltage clamp (V_hold = −60 mV) and after holding current stabilized, baclofen was applied via fast perfusion. Once the peak baclofen‐induced current was achieved (<1 min), CGP54626 (2 μM) was applied to test the GABA_BR dependence of the observed response.

2.5. Immunoelectron microscopy

The immuno‐related procedures used comply with the recommendations made by the British Journal of Pharmacology (Alexander et al., 2018). Eight ethanol‐ and air‐treated male and female C57BL/6J mice (2/sex/treatment) were killed with a lethal dose of ketamine (100 mg·kg⁻¹)/xylazine (10 mg·kg⁻¹) and subjected to transcardial perfusion 3 days after the final chronic intermittent ethanol session, and the subcellular distribution of GIRK2 and GABA_B1R was assessed using double pre‐embedding immunoelectron microscopy, with validated approaches and reagents, as described (Koyrakh et al., 2005; Marron Fernandez de Velasco et al., 2023; Padgett et al., 2012). In brief, free‐floating sections containing the VTA were incubated for 48 h in a mixture of two antibodies (anti‐GAD65/67 and either anti‐GIRK2 or anti‐GABA_B1R), at a final protein concentration of 1–2 mg·ml⁻¹ each. GAD65/67 expression was visualized by the immune‐peroxidase reaction, whereas GIRK2 or GABA_B1R labelling was visualized using the silver‐intensified immunogold reaction, as described (Padgett et al., 2012). Ultrastructural analyses were performed in a JEOL‐1400Flash electron microscope equipped with a digital high‐sensitivity sCMOS camera (Jeol Ltd., Tokyo, Japan). Digitized electron images were modified for brightness and contrast by using Adobe Photoshop CS5 (Mountain View, CA). Total number and density of GABA_B1R and GIRK2 immuno‐particles in the plasma membrane and intracellular compartments were quantified and analysed as described (Marron Fernandez de Velasco et al., 2023). Eighteen dendrites per animal were analysed (72 total dendrites analysed per group) to determine the average number of immune‐particles and average density of immune‐particles (# of gold particles/length of dendrite in μm) per dendrite. The Immuno‐related procedures used comply with the recommendations made by the British Journal of Pharmacology (Alexander et al., 2018).

2.6. In situ hybridization

Multi‐channel fluorescent in situ hybridization (RNAscope) was performed according to manufacturer recommendations (Advanced Cell Diagnostics/ACD, Newark, CA, USA) (Wang et al., 2012), using sections of the VTA (16 μm) from 8‐ to 10‐week ethanol‐naïve C57BL/6J mice (4 mice in total, 2 per sex), as described (DeBaker et al., 2023). In brief, mice were killed by rapid decapitation following isoflurane anesthesia (induction chamber with 1 mL isoflurane); brains were removed, snap frozen in isopentane and sectioned by cryostat. Sections were counterstained with DAPI and cover‐slipped with Prolong Gold Antifade (Thermofisher Scientific; Waltham, MA, USA). Probes for detection of specific RNAs were purchased from Advanced Cell Diagnostics, Inc (Newark, CA): GIRK1 (NM_008426.2; target region 658‐1679), GIRK2 (NM_001025584.2; target region 282‐1456), GIRK3 (NM_008429.2; target region 84‐1276) and the GABA neuron marker glutamate decarboxylase 1 (GAD1/GAD67) (NM_008077.4; target region 62‐3113). Fluorescent opal dyes (Opal 520, 620 and 690) for detection of specific probes were purchased from Akoya Biosciences (Marlborough, MA, USA). Fluorescent images of the VTA were acquired using a BZ‐X810 epifluorescent microscope with 40× objective (Keyence, Itasca, IL, USA). Three sections per mouse were incubated with each GIRK subunit probe set and 1–2 images per section were used to determine the fraction of GABA neurons (defined by DAPI and GAD67 labelling) expressing a particular GIRK subunit by manual counting using ImageJ (Schneider et al., 2012). The percentage of VTA GABA neurons expressing each GIRK subunit in each mouse was calculated by dividing the number of identified GIRK‐positive GABA neurons by the total number of GABA neurons evaluated across all images. Subsequently, the percentage of VTA GABA neurons expressing each GIRK subunit was averaged across mice.

2.7. Behavioural testing

Subjects underwent 2 days of habituation to procedure rooms; mice spent 1 h in testing rooms on each habituation day and were subjected to experimental handling. Mice were tested in the light/dark box (LDB) test on Day 3 and in the elevated plus maze (EPM) on Day 4, as described (Vo et al., 2021). The light–dark box test was conducted in 2‐compartment chambers housed in sound‐attenuating cubicles (Med Associates Inc, Fairfax, VT, USA). Mice were acclimated to the testing room for 1 h prior to placement in the dark compartment facing the open door to the light compartment. Time spent and movement counts (defined as a change in beam broken) in each side were tracked for 15 min by infrared sensors and extracted using MED‐PC IV 4.2 software (Med Associates Inc). For elevated plus maze testing, mice were acclimated to the testing room for 1 h prior to placement in the centre of the maze facing an open arm. Activity was recorded for 5 min by video camera and time spent in the open arms, closed arms, maze centre and total distance travelled were extracted using ANY‐maze 5.2 software (Stoelting Co., Wood Dale, IL, USA). Following testing, mice were killed by rapid decapitation following isoflurane anaesthesia (induction chamber with 1 mL isoflurane) and the scope and accuracy of viral targeting was assessed in each subject by analysis of viral‐driven mCherry expression in sections of the midbrain. Only mice exhibiting bilateral fluorescence confined primarily to the VTA were included in the final analysis.

2.8. Data and statistical analysis

The data and statistical analysis comply with the recommendations on experimental design and analysis from the British Journal of Pharmacology (Curtis et al., 2022). Data are presented throughout as mean ± SEM. Male and female mice were used in all studies and groups were balanced by sex, and experiments were designed to produce groups of approximately equal size. Comparative statistical analyses were conducted when all groups consisted of at least 5 independent experimental units (n). For slice electrophysiology experiments, (n) was defined as the number of individual experiments (recordings) per group. For immunoelectron microscopy studies, (n) was defined as the number of individual dendrites analysed per group. For behavioural experiments, (n) was defined as the number of mice per group. All analyses were performed using Prism v10 (GraphPad Software, Inc.; Boston, MA, USA). Data points outside ±2 standard deviations from the mean were excluded. The impact of sex on measured parameters was evaluated first using 2‐way analysis of variance (ANOVA). We did not detect a main effect of sex or interaction involving sex in any experiment and no post hoc comparisons were conducted. Therefore, data from male and female mice were pooled for analysis. Squares and triangles are included in most figures to denote individual datapoints from male and female subjects, respectively. Normality and lognormality tests were conducted to inform use of either the unpaired Student's t test or Mann–Whitney test. For all comparisons, differences were considered significant if P < 0.05 and individual values are reported in figure legends and tables.

2.9. Materials

Kynurenic acid, barium chloride and pyrazole were acquired from Sigma‐Aldrich (Saint Louis, MO, USA). Picrotoxin, muscimol, CGP54626 and baclofen were acquired from Tocris (Bristol, UK). Ethanol (200 proof) was purchased from Decon Laboratories, Inc (King of Prussia, PA, USA). For electron microscopy experiments, polyclonal rabbit antibody anti‐GAD65/67 (AB1511; Millipore, Billerica, MA, USA; RRID:AB_90715), monoclonal mouse antibody anti‐GABAB1 (clone N93A/49; Neuromab, Davis, CA, USA; AB_2108180), polyclonal guinea pig antibody anti‐GIRK2 (Rb‐Af830; aa. 390–421 of mouse GIRK2A‐1; Frontier Institute Co. Japan; RRID:AB_2571712), anti‐rabbit IgG conjugated to 1.4‐nm gold particles (1:100; Nanoprobes Inc., Stony Brook, NY, USA) and biotinylated goat anti‐rabbit IgG (Vector Laboratories, Burlingame, CA, USA) were utilized. All adeno‐associated viruses (AAVs) were packaged in AAV8 serotype by the University of Minnesota Viral Vector and Cloning Core (Minneapolis, MN, USA). pAAV‐CaMKIIα‐hChR2(H134R)‐mCherry was a gift from Karl Deisseroth (Addgene plasmid #26975; http://n2t.net/addgene:26975;RRID:Addgene_26975). For CRISPR/Cas9 studies, pAAV‐U6‐guide (g)RNA‐hSyn‐NLSmCherry was generated using the backbone of pAAV‐U6‐gRNA‐hSyn‐Cre‐2A‐EGFP‐KASH (Addgene plasmid #60231), a gift from Dr. Feng Zhang. Guide RNA (gRNA) sequences were as follows:‐ GIRK2, GGAGACGTACCGATACCTGA; LacZ, TGCGAATACGCCCACGCGAT. High titre (>1 × 10¹² genocopies mL⁻¹) AAV8‐U6‐gRNA (GIRK2)‐hSyn‐NLSmCherry and AAV8‐U6‐gRNA (LacZ)‐hSyn‐NLSmCherry vectors stored at −80°C in single‐use aliquots. Details of other materials and suppliers were provided in the specific sections.

2.10. Nomenclature of targets and ligands

Key protein targets and ligands in this article are hyperlinked to corresponding entries in the IUPHAR/BPS Guide to PHARMACOLOGY http://www.guidetopharmacology.org and are permanently archived in the Concise Guide to PHARMACOLOGY 2023/24 (Alexander, Christopoulos, et al., 2023, Alexander, Mathie, et al., 2023).

3. RESULTS

3.1. Impact of chronic ethanol on GABAergic input to VTA GABA neurons

To determine whether chronic ethanol treatment alters GABAergic input to VTA GABA neurons, we began by measuring GABA_AR‐dependent sIPSCs in VTA GABA neurons from GAD67‐GFP(+) mice, 3–7 days after the last session of a 4‐week chronic intermittent ethanol vapour exposure protocol (Figure 1a). This exposure protocol yielded blood ethanol concentrations of 1.5–2 g L⁻¹ (150–200 mg·dl⁻¹) and was shown in a prior study to induce signs of alcohol withdrawal including handling‐induced convulsions and altered performance in multiple mood‐related behavioural tests (Marron Fernandez de Velasco et al., 2023). Chronic ethanol exposure decreased the frequency of sIPSCs in VTA GABA neurons (Figure 1b,c) but had no effect on sIPSC amplitude (Figure 1d,e). Thus, chronic ethanol exposure reduces GABAergic input to VTA GABA neurons.

FIGURE 1.

Chronic ethanol diminishes GABAergic input to ventral tegmental area (VTA) GABA neurons. (a) Spontaneous inhibitory postsynaptic current (sIPSC) traces (V_hold = −70 mV) in VTA GABA neurons from mice treated with air or ethanol (EtOH). The upper grey trace shows that the unitary events observed under these conditions are blocked by picrotoxin (PTX, 100 μM); scale: 25 pA/2 s. (b) sIPSC frequency in VTA GABA neurons from air‐ and ethanol‐treated mice (t₂₅ = 3.157, *P = 0.0041; n = 13–14 recordings/group from 6–8 mice/group). (c) Cumulative distribution function displaying inter‐event intervals of sIPSC events in VTA GABA neurons from air‐ and ethanol‐treated mice. (d) sIPSC amplitude in VTA GABA neurons from air‐ and ethanol‐treated mice (U = 81, P = 0.6500; n = 13–14 recordings/group from 6‐8 mice/group). (e) Cumulative distribution function displaying amplitudes of sIPSC events in VTA GABA neurons from air‐ and ethanol‐treated mice. (f) sIPSC traces (V_hold = −70 mV) in VTA GABA neurons from mice exposed to air or ethanol, measured in the presence of CGP54626 (2 μM); scale: 25 pA/2 s. (g) sIPSC frequency in VTA GABA neurons from air‐ and ethanol‐treated mice measured in the presence of CGP 54626A (U = 89, P = 0.7006; n = 14 recordings/group from 4 mice/group). (h) Cumulative distribution function displaying inter‐event intervals of sIPSC events in VTA GABA neurons from air‐ and ethanol‐treated mice measured in the presence of CGP 54626A (CGP). (i) sIPSC amplitude in VTA GABA neurons from ethanol‐treated mice measured in the presence of CGP 54626A (t₂₆ = 1.372, P = 0.1817; n = 14 recordings/group from 4 mice/group). (j) Cumulative distribution function displaying amplitudes of sIPSC events in VTA GABA neurons from air‐ and ethanol‐treated mice measured in the presence of CGP 54626A.

Presynaptic GABA release is negatively regulated by GABA_B autoreceptors in some neurons (Kelm et al., 2011) and ethanol enhancement of presynaptic GABA_BR limits the ethanol‐induced potentiation of GABA release in the hippocampus (Ariwodola & Weiner, 2004) and amygdala (Silberman et al., 2009). To test whether the reduced sIPSC frequency provoked by chronic ethanol exposure is linked to altered GABA_BR‐dependent regulation of GABAergic transmission, we measured sIPSCs in the presence of the GABA_B antagonist CGP54626 (Figure 1f). GABA_BR inhibition normalized the ethanol‐induced reduction of sIPSC frequency in VTA GABA neurons (Figure 1g,h). Furthermore, ethanol did not impact sIPSC amplitude in the context of GABA_BR inhibition (Figure 1i,j). These data suggest that chronic ethanol exposure enhances the presynaptic GABA_BR‐dependent inhibition of GABAergic input to VTA GABA neurons.

Medium spiny neurons in the nucleus accumbens (NAc) are a major source of GABAergic input to the VTA (Faget et al., 2016) and stimulation of NAc medium spiny neuron terminals inhibits VTA GABA neurons in a GABA_AR‐dependent manner (Edwards et al., 2017). As such, we used an optogenetic approach to test whether chronic ethanol treatment alters GABAergic input to VTA GABA neurons from the NAc. We expressed channelrhodopsin in NAc medium spiny neurons of GAD67‐GFP(+) mice and optically stimulated NAc terminals adjacent to VTA GABA neurons (Figure 2a). Stimulation of medium spiny neuron terminals reliably evoked picrotoxin‐sensitive IPSCs in VTA GABA neurons from both air‐treated and ethanol‐treated mice (Figure 2b,d). The paired pulse ratio of evoked responses was significantly reduced in VTA GABA neurons from ethanol‐treated mice (Figure 2c) and the ethanol‐induced suppression of paired pulse ratio was normalized by CGP54626 (Figure 2e). These findings suggest that ethanol exposure weakens GABAergic input to VTA GABA neurons from the NAc, due to enhanced influence of GABA_BR.

FIGURE 2.

Chronic ethanol diminishes GABAergic input from the nucleus accumbens (NAc) to ventral tegmental area (VTA) GABA neurons. (a) Experimental strategy to isolate GABAergic input to VTA GABA neurons from the NAc. GAD67‐GFP(+) mice received intra‐NAc infusions of AAV8‐CaMKIIα‐ChR2(H134R)‐mCherry to drive channelrhodopsin expression in NAc medium spiny neurons. The image below shows mCherry fluorescence in the NAc core and shell 5 weeks after infusion; scale: 250 μm. AC, anterior commissure. (b) GABA_AR‐dependent currents (V_hold = −70 mV) in VTA GABA neurons evoked by optical stimulation (blue bars) of NAc terminals, measured after CIE; scale: 200 pA/25 ms. Responses evoked under these conditions were blocked by PTX (100 μM). (c) Paired‐pulse ratio (PPR) of optically evoked GABA_AR‐dependent currents in VTA GABA neurons from air‐ and ethanol‐treated mice (t₂₃ = 4.128, *P = 0.0004; n = 12–13 recordings/group from 4‐5 mice/group). (d) Optically evoked GABA_AR‐dependent currents (V_hold = −70 mV) in VTA GABA neurons, measured in the presence of CGP 54626A (CGP; 2 μM); scale: 100 pA/25 ms. (e) PPR of optically evoked GABA_AR‐dependent currents in VTA GABA neurons from air‐ and ethanol‐treated mice, measured in the presence of CGP 54626A (t₂₂ = 0.6484, P = 0.5234; n = 12 recordings/group from 4 mice/group).

3.2. Impact of chronic ethanol on somatodendritic GABA_A and GABA_BR‐dependent signalling

We next measured somatodendritic currents evoked by saturating concentrations of the GABA_A agonist muscimol (30 μM) (Mortensen et al., 2010) or GABA_B agonist baclofen (200 μM) (Cruz et al., 2004; Labouebe et al., 2007) to test whether ethanol exposure alters maximal postsynaptic GABA_AR‐ and/or GABA_BR‐dependent responses, respectively, in VTA GABA neurons. Consistent with the lack of impact of chronic ethanol on sIPSC amplitude in VTA GABA neurons, chronic intermittent ethanol did not affect whole‐cell currents evoked by muscimol (Figure 3a,b). In contrast, currents evoked by baclofen were smaller in VTA GABA neurons from ethanol‐treated mice (Figure 3c,d). Thus, chronic ethanol exposure weakens the GABA_BR‐dependent inhibition of VTA GABA neurons, without impacting somatodendritic GABA_AR‐dependent signalling.

FIGURE 3.

Chronic ethanol suppresses GABA_BR‐ but not GABA_AR‐dependent somatodendritic currents in ventral tegmental area (VTA) GABA neurons. (a) Inward currents (V_hold = −70 mV) evoked by a saturating concentration of muscimol (30 μM) in VTA GABA neurons, measured after chronic intermittent ethanol; scale: 300 pA/2 s. (b) Peak currents evoked by muscimol (30 μM) in VTA GABA neurons from air‐ and ethanol‐treated mice (t₂₄ = 0.3153, P = 0.7553; n = 12–14 recordings/group from 9 mice/group). (c) Outward currents (V_hold = −60 mV) evoked by a saturating concentration of baclofen (200 μM) in VTA GABA neurons and reversed by CGP 54626A (CGP;2 μM), measured after chronic intermittent ethanol; scale: 5 pA/50 s. (d) Peak currents evoked by baclofen (200 μM) in VTA GABA neurons from air‐ and ethanol‐treated mice (U = 34, *P = 0.007; n = 14–16 recordings/group from 10‐11 mice/group). (e) Outward currents (V_hold = −60 mV) evoked by a saturating concentration of baclofen (200 μM) in VTA GABA neurons, measured after chronic intermittent ethanol and in the presence of 0.3 mM Ba²⁺; scale: 5 pA/25 s. (f) Peak currents evoked by baclofen (200 μM) in VTA GABA neurons from air‐ and ethanol‐treated mice, measured in the presence of 0.3‐mM external Ba²⁺ (t₂₁ = 0.7315, P = 0.4725; n = 11–12 recordings/group from 4 mice/group).

Baclofen‐evoked somatodendritic currents in VTA GABA neurons are a composite response consisting of a Ba²⁺‐sensitive GIRK (K_ir3)‐dependent component and a Ba²⁺‐insensitive GIRK‐independent component (Cruz et al., 2004). To determine whether chronic ethanol exposure impacts one or both components of this composite response, we measured baclofen‐evoked currents in the presence of 0.3‐mM external Ba²⁺. The Ba²⁺‐insensitive (GIRK‐independent) component of the baclofen‐evoked current was not impacted by chronic ethanol treatment (Figure 3e,f), suggesting that chronic ethanol selectively suppresses GABA_BR‐GIRK signalling in VTA GABA neurons.

3.3. Mechanism underlying the ethanol‐induced suppression of GABA_BR‐GIRK(K_ir3) signalling

Increased internalization of GABA_BR and/or GIRK channels underlies the suppression of GABA_BR‐GIRK signalling induced by psychostimulants in VTA GABA neurons (Padgett et al., 2012) and by chronic ethanol in principal neurons of the basal amygdala (Marron Fernandez de Velasco et al., 2023). To test whether increased internalization of GABA_BR and/or GIRK channels could explain the smaller baclofen‐evoked currents in VTA GABA neurons from ethanol‐treated mice, we used an established pre‐embedding immunoelectron microscopy approach and knockout‐validated antibodies targeting GABA_B1R and GIRK2 (Figure 4a) (Marron Fernandez de Velasco et al., 2023). We detected a significant reduction in the number and density of GABA_B1R (Figure 4b,c) and GIRK2 (Kir3.2; Figure 4d,e) immune‐particles on the plasma membrane of VTA GABA neuron dendrites following chronic ethanol treatment and a parallel increase in the number and density of receptor and channel in the intracellular compartment. Thus, chronic ethanol exposure provokes internalization of GABA_BR and GIRK channels in VTA GABA neuron dendrites.

FIGURE 4.

Chronic ethanol promotes internalization of GABA_BR and GIRK (K_ir3) channels in ventral tegmental area (VTA) GABA neurons. GABA_B1R (top row) and GIRK2 (bottom row) immunolabelling in the dendritic compartment (Den) of a VTA GABA neuron identified by GAD65/67‐HRP immunoreactivity, taken from air‐ and ethanol‐treated C57BL/6J mice 3 days after completing chronic intermittent ethanol. GABA_B1R and GIRK2 immuno‐particles were most commonly observed at the plasma membrane (arrows) and intracellular sites (hatched arrows), with minimal labelling in axon terminals (at, arrowhead); scale: 500 nm. (b) Average number of GABA_B1R immune‐particles on the plasma membrane (PM) (U = 838, *P < 0.0001) and intracellular compartment (intra) (U = 883, *P < 0.0001) of dendrites (n = 72 dendrites/group from 4 mice/group). (c) Density of GABA_B1R immuno‐particles on the PM (U = 643, *P < 0.0001) and intracellular compartment (intra) (U = 1195, *P < 0.0001) of dendrites (n = 72 dendrites/group from 4 mice/group). (d) Total number of GIRK2 immuno‐particles on the PM (U = 838, *P < 0.0001) and intracellular compartment (intra) (U = 467, *P < 0.0001) of dendrites (n = 72 dendrites/group from 4 mice/group). (e) Density of GIRK2 immuno‐particles on the PM (U = 637, *P < 0.0001) and intracellular compartment (intra) (U = 1269, *P < 0.0001 of dendrites (n = 72 dendrites/group from 4 mice/group).

GIRK (K_ir3) channels in most neurons are heterotetramers containing GIRK1(K_ir3.1) and GIRK2 (K_ir3.2) subunits (Lujan & Aguado, 2015), and ablation of either subunit eliminates the GIRK component of the baclofen‐evoked current in VTA GABA neurons (Cruz et al., 2004; Kotecki et al., 2015). While ablation of GIRK3 (K_ir3.3) did not impact baclofen‐evoked current amplitude in VTA GABA neurons (Kotecki et al., 2015), GIRK3 expression has been detected in these neurons by single‐cell reverse transcription polymerase chain reaction (RT‐PCR) (Cruz et al., 2004). We confirmed this finding in ethanol‐naïve adult C57BL/6J mice using fluorescent multi‐plex in situ hybridization with probes for GIRK1, GIRK2 and GIRK3, along with the GABA neuron marker GAD67 (Figure 5a). Most GAD67‐positive neurons in the VTA express mRNA for GIRK1 (96%), GIRK2 (83%) and GIRK3 (94%).

FIGURE 5.

GIRK3 is required for the ethanol‐induced suppression of GABA_BR‐GIRK signalling in ventral tegmental area (VTA) GABA neurons. (a) RNAscope images for GAD67 (cyan, upper row), as well as GIRK1, GIRK2 and GIRK3 subunits (red, lower row), in the VTA of adult C57BL/6J mice; scale: 25 μm. White circles highlight examples of GAD67‐positive neurons exhibiting expression of a particular GIRK subunit. (b) Outward currents (V_hold = −60 mV) evoked by baclofen (200 μM) and reversal by CGP54626 (2 μM) in VTA GABA neurons from Girk3 ^−/− mice, measured after chronic intermittent ethanol; scale: 10 pA/25 s. (c) Peak currents evoked by baclofen (200 μM) in VTA GABA neurons from air‐ and ethanol‐treated Girk3 ^−/− mice (t₂₂ = 0.2072, P = 0.8378; n = 12 recordings/group from 5 mice/group).

GIRK3 has been implicated in subcellular trafficking of neuronal GIRK channels and is required for some forms of GIRK channel plasticity (Luo et al., 2022). For example, GIRK3 is required for the suppression of GABA_BR‐GIRK signalling in VTA dopamine neurons provoked by neuronal activity (Lalive et al., 2014) or in vivo methamphetamine exposure (Munoz et al., 2016), as well as the ethanol‐induced suppression of GIRK channel activity in basal amygdala principal neurons (Marron Fernandez de Velasco et al., 2023). We crossed GAD67‐GFP(+) mice with Girk3(Kcnj9 ^−/− ) mice to test whether GIRK3 is required for the expression of ethanol‐induced suppression of GABA_BR‐GIRK signalling in VTA GABA neurons. Baclofen‐evoked currents in VTA GABA neurons from GAD67‐GFP(+):Girk3 ^−/− mice were unaffected by ethanol (Figure 5b,c), indicating that GIRK3 is required for ethanol‐induced suppression of GABA_BR‐dependent signalling in VTA GABA neurons.

3.4. Impact of GABA_BR‐GIRK(K_ir3) signalling plasticity on anxiety‐related behaviour

In vivo activity of VTA GABA neurons tracks aversive stimuli and negative states (Cohen et al., 2012; Root et al., 2020; Tan et al., 2012), and manipulation of VTA GABA neuron activity can modulate anxiety‐related behaviours in mice (Chen et al., 2020; Jennings et al., 2013; Yu et al., 2021). To examine the relevance of the ethanol‐induced suppression of GABA_BR‐GIRK signalling in VTA GABA neurons to anxiety‐related behaviour, we used a neuron‐specific CRISPR/Cas9 approach to ablate GIRK channels in VTA GABA neurons in ethanol‐naïve mice. GADCre(+):Cas9GFP(+) mice received intra‐VTA infusions of AAV vectors harbouring guide RNAs for LacZ (control) or GIRK2 (Figure 6a,b). GIRK2 was targeted in these experiments as ablation of this subunit eliminates the GIRK component of the composite GABA_BR‐induced current in VTA GABA neurons (Cruz et al., 2004; Kotecki et al., 2015). Indeed, baclofen‐evoked currents were ~50% smaller in VTA GABA neurons from mice treated with the GIRK2 gRNA (Figure 6c,d), comparable with the suppression observed during withdrawal following chronic ethanol exposure. Importantly, baclofen‐evoked currents in neighbouring VTA dopamine neurons were normal in animals treated with the GIRK2 gRNA (Figure 6e,f), attesting to the neuron selectivity of this intervention.

FIGURE 6.

GIRK channel ablation in ventral tegmental area (VTA) GABA neurons induces anxiety‐related behaviour. (a) Modelling the ethanol‐induced suppression of GABA_BR‐GIRK signalling in VTA GABA neurons of ethanol‐naïve mice. AAV8‐U6‐gRNA (GIRK2)‐hSyn‐mCherry or control (AAV8‐U6‐gRNA (LacZ)‐hSyn‐mCherry) vectors were infused bilaterally into the VTA of GADCre(+):Cas9GFP(+) mice, followed 4–5 weeks later by electrophysiological and behavioural assessments. (b) Outward currents (V_hold = −60 mV) evoked by baclofen (200 μM) in VTA GABA neurons of GADCre(+):Cas9GFP(+) mice treated with LacZ or GIRK2 gRNAs, measured 5 weeks after viral infusion; scale: 10 pA/25 s. (c) Peak currents evoked by baclofen (200 μM) in VTA GABA neurons from LacZ and GIRK2 gRNA‐treated mice (t₉ = 3.971, *P = 0.0033; n = 5–6 recordings/group from 2 mice/group). (d) Outward currents (V_hold = −60 mV) evoked by baclofen (200 μM) in VTA dopamine neurons of GADCre(+):Cas9GFP(+) mice treated with LacZ or GIRK2 gRNAs, measured 5 weeks after viral infusion; scale: 50 pA/25 s. (e) Peak currents evoked by baclofen (200 μM) in VTA dopamine neurons from LacZ and GIRK2 gRNA‐treated mice (t₈ = 0.8567, P = 0.4165; n = 5 recordings/group from 2 mice/group). (f) Percentage of time spent in the light side of the light–dark box test by GADCre(+):Cas9GFP(+) mice treated with LacZ or GIRK2 gRNA (U = 22, P = 0.0197; n = 10–11 mice/group). (g) Movement counts during light–dark box test (U = 35, P = 0.1734; n = 10–11 mice/group). (h) Percentage of time spent in the open arms of the elevated plus maze (t₁₉ = 2.918, *P = 0.0088; n = 10–11 mice/group). (i) Total distance travelled during the elevated plus maze test (t₁₉ = 1.197, P = 0.2460; n = 10–11 mice/group).

A separate cohort of gRNA‐treated mice were evaluated in light–dark box and elevated plus maze tests. GIRK channel ablation in VTA GABA neurons correlated with a significant reduction in the percentage of time spent in the light side of the light–dark box apparatus (Figure 6g) and a significant reduction in the percentage of time spent in the open arms of the elevated plus maze (Figure 6i). No change in locomotor activity was observed in either test (Figure 6h,j). Thus, reduced GIRK channel activity in VTA GABA neurons is sufficient to enhance anxiety‐related behaviour in ethanol‐naïve mice.

4. DISCUSSION

Ventral tegmental area (VTA) GABA neurons exhibit elevated activity during withdrawal from chronic ethanol exposure (Gallegos et al., 1999). Increased glutamatergic input has been implicated in this phenomenon (Williams et al., 2018), along with diminished GABA_AR‐dependent regulation of VTA GABA neurons (Nelson et al., 2018). Here, we identify two GABA_BR‐dependent mechanisms that may contribute to elevated activity of VTA GABA neurons during ethanol withdrawal. Specifically, we show that chronic ethanol suppresses GABAergic input to VTA GABA neurons via an enhanced presynaptic influence of GABA_BR, while also suppressing GABA_BR‐GIRK signalling in VTA GABA neurons.

4.1. Impact of ethanol on GABAergic input to VTA GABA neurons

We began by examining the impact of chronic ethanol exposure on sIPSCs in VTA GABA neurons, as this approach permits insight into changes in action potential‐dependent and independent mechanisms that contribute to synaptic GABA release. Chronic ethanol exposure decreased sIPSC frequency, consistent with decreased GABAergic input to these neurons. The lack of impact of chronic ethanol on sIPSC amplitude and on whole‐cell currents evoked by a saturating muscimol concentration suggest that postsynaptic GABA_AR function was unaffected by ethanol. This was somewhat surprising given that dietary ethanol consumption for 12 weeks in rats correlated with decreased levels of the GABA_AR α₁ subunit, which is enriched in VTA GABA neurons (Tan et al., 2010), in tissue punches containing the VTA (Ortiz et al., 1995). This reduction was not seen with shorter (2 or 4 weeks) ethanol consumption periods (Charlton et al., 1997; Papadeas et al., 2001), however, which better aligns with the duration of our chronic intermittent ethanol protocol.

Interestingly, the spontaneous firing of VTA GABA neurons from GAD67‐GFP(+) mice (on a CD‐1 strain background) was less sensitive to muscimol‐induced inhibition, as assessed 7 days after a 3‐week chronic intermittent ethanol or repeated ethanol injection protocol (Nelson et al., 2018). The impact of chronic ethanol in this case was only detected in the 0.1‐ to 1‐μM muscimol range, with a higher concentration (10 μM) completely suppressing VTA GABA neuron firing in both ethanol‐ and air‐treated mice. Here, we used a high muscimol concentration (30 μM) to ensure maximal activation of somatodendritic GABA_AR in VTA GABA neurons. Muscimol is a pan‐selective GABA_A agonist, but it exhibits a particularly high affinity for δ subunit‐containing extrasynaptic GABA_AR (Benkherouf et al., 2019). As such, a small reduction in the level or function of δ subunit‐containing GABA_AR in VTA GABA neurons could explain decreased sensitivity of muscimol in suppressing VTA GABA neuron firing rate following chronic ethanol exposure (Yang et al., 2023). It is also possible that spontaneous firing of VTA GABA neurons is driven by glutamatergic input (Williams et al., 2018) and that presynaptic GABA_AR on glutamatergic terminals mediates the suppression of firing. In this event, ethanol exposure could suppress this specific presynaptic GABA_AR influence without impacting the muscimol‐induced somatodendritic currents examined in our study.

4.2. Role of GABA_BR in the ethanol‐induced suppression of sIPSC frequency

Our data suggest that GABA_BR tempers GABAergic input to VTA GABA neurons, as has been reported at other synapses (Ariwodola & Weiner, 2004; Silberman et al., 2009). Indeed, sIPSC frequencies measured in the presence of CGP 54626A (GABA_B antagonist) were higher than those measured in the absence of antagonist (5 Hz vs. 2 Hz, respectively, in VTA GABA neurons from air‐treated controls), consistent with substantial tonic GABA_BR‐dependent suppression of GABA input to VTA GABA neurons in acute slices from ethanol‐naïve mice. The ethanol‐induced suppression of GABA_AR‐dependent sIPSCs in VTA GABA neurons was also normalized by GABA_BR inhibition. Collectively, these findings suggest that chronic ethanol provokes a durable inhibition of GABAergic input to VTA GABA neurons that is dependent on the activity of presynaptic GABA_BR. Given that GABA_BR inhibition also normalized the ethanol‐induced suppression of paired pulse ratio of IPSCs evoked by optical stimulation of NAc medium spiny neurons, it is tempting to speculate that presynaptic GABA_BR on GABAergic terminals originating from NAc medium spiny neurons mediates the chronic ethanol‐induced suppression of GABA release. We cannot rule out, however, the potential contribution of somatodendritic GABA_BR and its influence on action potential‐dependent release of GABA. Indeed, pre‐ and/or postsynaptic GABA_BR on GABAergic inputs that remain intact in the acute horizontal slice preparation, including those deriving from the rostromedial tegmental nucleus, interpeduncular nucleus and VTA GABA neurons themselves, could underlie the impact of CGP 54626A on sIPSCs in VTA GABA neurons (Bouarab et al., 2019).

4.3. Suppression of GABA_BR‐GIRK (K_ir3) signalling as a common consequence of drug exposure

Plasticity of neuronal GABA_BR‐GIRK signalling has been reported in response to drugs with abuse potential (e.g. cocaine, methamphetamine, morphine and ethanol) (Luo et al., 2022), electrical stimulation (Huang et al., 2005; Lalive et al., 2014) and aversive stimuli (Lecca et al., 2016). A single injection of cocaine or methamphetamine in mice can suppress GABA_BR‐dependent signalling in VTA GABA neurons (Padgett et al., 2012). Interestingly, acute exposure to several drugs of abuse, including ethanol (Gallegos et al., 1999), cocaine (Bocklisch et al., 2013) and Δ9‐tetrahydrocannabinol (Friend et al., 2017), inhibits VTA GABA neurons in rodents. Thus, suppression of GABA_BR‐GIRK signalling may represent a shared homeostatic plasticity mechanism to compensate for repeated VTA GABA neuron inhibition.

An increasing array of clinical and preclinical data suggest that dysregulation of GIRK channel function contributes to substance use disorders, including alcohol use disorder (Kotajima‐Murakami et al., 2022; Luo et al., 2022). For example, several single nucleotide polymorphisms in the GIRK2/KCNJ6 gene, including both synonymous and non‐synonymous polymorphisms, were identified in a family‐based, genome‐wide association study evaluating determinants of a frontal theta event‐related oscillation phenotype that is associated with reward processing and alcohol use disorder (Clarke et al., 2011; Kamarajan et al., 2017; Kang et al., 2012). Excitatory neurons derived from human induced pluripotent stem cells harbouring alcohol use disorder‐associated polymorphisms in GIRK2/KCNJ6 showed varying levels of excitability correlated inversely with GIRK2 expression (Popova et al., 2023). Interestingly, ethanol exposure (20 mM) induced GIRK2 expression, ameliorating hyperexcitability in neurons harbouring some non‐coding GIRK2 polymorphisms on the excitability of human glutamatergic neurons.

The GIRK3/KCNJ9 gene emerged as a candidate‐of‐interest in a genome‐wide association study involving more than 20,000 participants who completed the alcohol use disorder identification test (Sanchez‐Roige et al., 2019). The mouse GIRK3/Kcnj9 gene was identified in an unbiased quantitative trait locus analysis that mapped genetic determinants of barbiturate withdrawal, with lower levels of GIRK3 expression associated with less severe withdrawal from pentobarbitone and ethanol (Kozell et al., 2009). Indeed, Girk3 ^−/− mice exhibit an almost complete lack of physical and affective ethanol withdrawal symptoms (Marron Fernandez de Velasco et al., 2023). While Girk3 ^−/− mice exhibited normal ethanol metabolism (Herman et al., 2015), they do show several aberrant ethanol‐related behaviours including enhanced ethanol reward (Tipps et al., 2016) and enhanced binge‐like ethanol consumption (Herman et al., 2015). Notably, enhanced binge‐like ethanol consumption in Girk3 ^−/− mice was reversed by viral reconstitution of GIRK3 in the VTA (Herman et al., 2015).

4.4. GIRK3 (K_ir3.3) as a key regulator of sensitivity to ethanol‐induced neuroadaptations

GIRK3 is required for several forms of GIRK channel plasticity, including the ethanol‐induced suppression of GABA_BR‐GIRK signalling in VTA GABA neurons (this study) and principal neurons of the basal amygdala (Marron Fernandez de Velasco et al., 2023). Ectopic expression of GIRK3 can exert a dominant‐negative influence on GIRK channel activity (Ma et al., 2002; McCall et al., 2019), likely via its influence on the subcellular trafficking of GIRK channels. Indeed, GIRK3 possesses a unique lysosomal targeting sequence sufficient to inhibit plasma membrane trafficking of GIRK1/GIRK2 channels (Ma et al., 2002). GIRK3 also interacts with sorting nexin 27 (SNX27) in neurons and this interaction promotes the endosomal movement and reduced surface expression of GIRK channels (Lunn et al., 2007). These reports align with our observations that chronic ethanol exposure enhances internalization of GABA_B1 and GIRK2 in these neurons. At present, it is unclear how chronic ethanol exposure or other stimuli that provoke GIRK3‐dependent adaptations in GIRK channel activity engage this type of trafficking mechanism. As GIRK channels may form macro‐complexes that include GABA_BR (Fajardo‐Serrano et al., 2013), mechanisms that regulate either GIRK channel or GABA_BR trafficking could play a role (Rose & Wickman, 2022).

4.5. Behavioural relevance of the suppression of GABA_BR‐GIRK signalling in VTA GABA neurons

GABA_BR agonists and positive allosteric modulators reduce symptoms of alcohol use disorder in preclinical and clinical settings (Augier, 2021; Holtyn & Weerts, 2022). Alcohol use disorder and anxiety disorders are often highly comorbid, with diagnosis of one disorder significantly increasing risk for the development of the other (Anker & Kushner, 2019) and symptoms of one generally aggravating symptoms of the other (Smith & Randall, 2012). A growing body of evidence supports the contention that manipulating GABA_BR‐dependent signalling can have therapeutic benefits in the context of alcohol use disorder, particularly in cases of co‐morbid anxiety and/or stress (Addolorato et al., 2009; Agabio & Colombo, 2015; Felice et al., 2022; Morley et al., 2014).

Given these insights, we were motivated to assess the relevance of diminished GABA_BR‐GIRK signalling in VTA GABA neurons on measures of anxiety‐related behaviour. Light–dark box and elevated plus maze tests measure hypervigilance, a common anxiety‐related symptom associated with heightened attention towards predicting potential dangers or threats (Richards et al., 2014), which can manifest as avoidance of contexts where threats are thought to be present (Kimble et al., 2014). VTA GABA neurons have been implicated in the expression of anxiety‐related behaviour in mice, including hypervigilance (Chen et al., 2020; Jennings et al., 2013; Tan et al., 2012; Zhou et al., 2019). We reported previously that the 4‐week chronic intermittent ethanol protocol used in this study is sufficient to enhance anxiety‐related behaviour in the light–dark box test (Marron Fernandez de Velasco et al., 2023); this behavioural adaptation required GIRK3 (Kir3.3) but did not involve suppression of GABA_BR‐GIRK signalling in basal amygdala principal neurons. Here, we show that suppressing GIRK channel activity in VTA GABA neurons is sufficient to enhance anxiety‐related behaviour as assessed in light–dark box and elevated plus maze tests. While GIRK2 ablation may also impact signalling involving other inhibitory GPCRs in VTA GABA neurons, these findings suggest the ethanol‐induced suppression of GABA_BR‐GIRK signalling in VTA GABA neurons might contribute to anxiety‐related behaviours that emerge during ethanol withdrawal.

In sum, our data show that chronic ethanol exposure in mice reduces GABAergic input to VTA GABA neurons, while also diminishing their postsynaptic GABA_BR‐dependent responses. These findings add to our understanding of the increased activity of VTA GABA neurons reported during ethanol withdrawal (Gallegos et al., 1999; Nelson et al., 2018; Williams et al., 2018) and suggest that this adaptation may contribute to heightened anxiety associated with alcohol use disorder and could be implicated in relapse (Koob & Volkow, 2010, 2016).

AUTHOR CONTRIBUTIONS

Eric H. Mitten: Conceptualization (equal); data curation (lead); formal analysis (lead); funding acquisition (supporting); investigation (lead); methodology (supporting); writing—original draft (equal). Anna Souders: Data curation (supporting); formal analysis (supporting); investigation (supporting); methodology (supporting). Ezequiel Marron Fernandez de Velasco: Data curation (supporting); investigation (supporting); methodology (supporting); resources (supporting); supervision (supporting). Carolina Aguado: Data curation (supporting); formal analysis (supporting); investigation (supporting); methodology (supporting). Rafael Luján: Formal analysis (supporting); funding acquisition (supporting); investigation (supporting); methodology (supporting). Kevin Wickman: Conceptualization (equal); funding acquisition (lead); resources (lead); supervision (lead); writing—original draft (equal).

CONFLICT OF INTEREST STATEMENT

The authors declare that they have no competing interests.

DECLARATION OF TRANSPARENCY AND SCIENTIFIC RIGOUR

This declaration acknowledges that this paper adheres to the principles for transparent reporting and scientific rigour of preclinical research as stated in the BJP guidelines for Design and Analysis, Immunoblotting and Immunochemistry and Animal Experimentation, and as recommended by funding agencies, publishers and other organizations engaged with supporting research.

Mitten, E. H. , Souders, A. , Marron Fernandez de Velasco, E. , Aguado, C. , Luján, R. , & Wickman, K. (2025). Chronic ethanol exposure in mice evokes pre‐ and postsynaptic deficits in GABAergic transmission in ventral tegmental area GABA neurons. British Journal of Pharmacology, 182(1), 69–86. 10.1111/bph.17335

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available from the corresponding author upon reasonable request.