Enhancement of alcohol aversion by the nicotinic acetylcholine receptor drug sazetidine-A

Abstract

The prevalence of alcohol use disorders (AUDs) has steadily increased in the United States over the last 30 years. Alcohol acts on multiple receptor systems including the nicotinic acetylcholine receptors (nAChRs), which are known to mediate alcohol consumption and reward. We previously reported that the preclinical drug sazetidine-A, a nAChR agonist and desensitizer, reduces alcohol consumption without affecting nicotine consumption in C57BL/6J mice. Here, we found that sazetidine-A enhances the expression of alcohol aversion without affecting the expression or acquisition of conditioned alcohol reward in C57BL/6J mice. Microinjection of sazetidine-A into the ventral midbrain targeting the ventral tegmental area (VTA) reduced binge alcohol consumption, implicating this region in mediating the effects of sazetidine-A. Furthermore, the sazetidine-A-induced reduction in alcohol consumption was mediated by non-α4 containing nAChRs, as sazetidine-A reduced binge alcohol consumption in both α4 knock-out and wild-type mice. Finally, we found that in mice pretreated with sazetidine-A, alcohol induced Fos transcript in Th-, but not Gad2-expressing neurons in the VTA as measured by increased Fos transcript expression. In summary, we find that sazetidine-A enhances the expression of alcohol aversion, which may underlie the reduction in alcohol consumption induced by sazetidine-A. Elucidating the identity of non-α4 nAChRs in alcohol aversion mechanisms will provide a better understanding the complex role of nAChRs in alcohol addiction and potentially reveal novel drug targets to treat AUDs.

1 |. INTRODUCTION

Alcohol use disorders (AUD) have steadily increased in the United States, along with associated increases in high-risk drinking and mortality.¹ An estimated 30 million Americans met criteria for an AUD in 2013, an increase of 12 million cases compared with the previous decade.¹ Currently, there are only three FDA-approved pharmacotherapies for AUD in the United States (naloxone, acamprosate, and disulfiram), all with variable success rates,² which highlights the need to further understand how alcohol interacts with other receptor systems and identify new drug targets.

The neuronal nicotinic acetylcholine receptors (nAChRs) play an important role in mediating the rewarding properties of alcohol. The nAChRs are pentameric cation channels located on presynaptic terminals and cell bodies of neurons, thus modulating cell excitability and neurotransmitter release.³ There are 11 nAChR subunits found in the brain (α2–7, α9–10, and β2–4) that can combine to form multiple nAChR subtypes, each with its own unique receptor properties such as different ligand binding affinities, calcium permeability, and expression patterns.³ Alcohol does not directly agonize nAChRs but modulates nAChR functional activity⁴ and increases the release of acetylcholine (ACh), the endogenous ligand for nAChRs.⁵ Both nAChR antagonists such as mecamylamine (a nonspecific nAChR antagonist) and partial agonists such as varenicline (an α4β2 partial agonist) reduce alcohol consumption in rodents, illustrating the complex involvement of nAChRs in mediating alcohol consumption.^6–8 Varenicline has been tested in human trials for alcohol dependence with mixed results, with some studies reporting reduced alcohol consumption^9,10 but not others.^11,12

The preclinical drug sazetidine-A initially agonizes and then desensitizes α4β2, α3β4*, α6*, and α7 nAChR subtypes (*denotes additional subunits in the nAChR pentamer).^13–16 High-doses of sazetidine-A have been shown to reduce alcohol and nicotine consumption in rats; however, it also reduces saccharin consumption.¹⁷ Interestingly, we previously reported an alcohol-specific effect of sazetidine-A in mice: We found that a single, low dose of sazetidine-A reduced 24-h continuous and binge drinking-in-the-dark (DID) alcohol consumption in both male and female C57BL/6J mice, whereas sazetidine-A had no effect on the consumption of water, saccharin, or nicotine.¹⁸ Here, we investigated the mechanism by which sazetidine-A reduces alcohol consumption. We report that sazetidine-A enhanced the expression of alcohol aversion without affecting the expression or acquisition of alcohol reward. Sazetidine-A did not require the α4 nAChR subunit to reduce alcohol consumption, implicating non-α4* nAChRs in the mechanism of action of sazetidine-A. Finally, we found that pretreatment with sazetidine-A followed by alcohol injection induced Fos expression in Th-but not Gad2-expressing neurons in the ventral tegmental area (VTA). These findings expand our understanding of the role of the nAChRs in both alcohol-related behaviors and in mechanisms underlying aversion.

2 |. MATERIALS AND METHODS

2.1 |. Animals and drugs

Heterozygous α4 nAChR subunit knock-out (KO) breeder pairs were provided by Dr. Jerry Stitzel at the University of Colorado, Boulder, and α4 nAChR knock-out and their wild-type (WT) littermates were bred on site. We used female α4 WT and KO littermates for Experiment 1 (binge DID). We used 8-week old adult male C57BL/6J mice purchased from Jackson Laboratory (Jackson Laboratory, Sacramento, CA) for Experiments 2–9. The male C57BL/6J mice acclimated to our facility for a minimum of 6 days before behavioral experiments. All mice were group housed in standard cages under a 12-h light/dark cycle until the start of behavioral experiments, when they were individually housed. All animal procedures were performed in accordance with the Institutional Animal Care and Use Committee at the University of Minnesota and conformed to NIH guidelines (National Research Council Committee for the Update of the Guide for the Care and Use of Laboratory Animals, 2010).

Alcohol (Decon Labs, King of Prussia, PA) was mixed with tap water for drinking studies, or diluted in 0.9% saline to a final concentration of 20% v/v for intraperitoneal (i.p.) injections. Sazetidine-A dihydrochloride and mecamylamine hydrochloride were purchased from Tocris Bioscience (Bio-techne, Minneapolis, MN) and made fresh for each experiment by dissolving in saline. Sazetidine-A has been shown to occupy nAChRs in the mouse brain for at least 8 h after injection,¹⁹ and our prior work showed that pretreatment for 1 h with 1 mg/kg sazetidine-A i.p. reduces alcohol consumption in male and female C57BL/6J mice¹⁸; thus, we used the same 1-h time frame and 1 mg/kg dose for the sazetidine-A i.p. injections below. All peripheral injections were administered i.p., and all experiments were performed in separate groups of drug naïve mice.

2.2 |. Experiment 1: Binge alcohol consumption in female α4 wild-type and knock-out mice

Because our previous work showed that sazetidine-A reduces binge consumption in both male and female C57BL/6J mice, and because female mice consume more alcohol than male mice,¹⁸ we used female α4 WT and KO mice on a C57BL/6J background in the binge drinking-in-the-dark (DID) experiment as they were readily available in our facility. For the DID procedure, mice were habituated to three i. p. injections of saline (10 μL/g) 1 week prior to the experiment. Mice were then presented with a bottle of 20% alcohol for 2 h, starting 2 h into the dark cycle for three consecutive days, based on Rhodes et al.²⁰ The water bottle was removed only when the alcohol bottle was present but was freely available at all other times. On day 4, mice were injected with either saline or sazetidine-A (1 mg/kg i.p.) 1 h prior to the presentation of the alcohol bottle for 4 h. Alcohol consumption was calculated by volume, and spillage or evaporation was controlled for by a bottle in an empty control cage.

2.3 |. Experiment 2: VTA microinfusion and binge alcohol consumption in male C57BL/6J mice

Male C57BL/6J mice were anesthetized with ketamine/xylazine (80:10 mg/kg i.p.) and implanted with bilateral 26 gauge stainless steel guide cannula (Plastics One, Roanoke, VA) targeting the VTA (−3.2-mm AP, ±0.5-mm ML, and −4.96 mm ventral to the surface according to the Paxinos and Franklin mouse brain atlas).²¹ The cannulae were fixed with dental adhesive (Geristore, DenMat, Lompoc, CA). Mice then recovered for 3 weeks in their home cage. One week prior to the binge consumption procedure, mice were habituated to handling and received three saline infusions at a volume of 0.5 μL, at a rate of 0.25 μL/min. For the DID procedure, mice were presented with a bottle of 20% alcohol for 2 h, starting 3 h into the dark cycle for four consecutive days, based on Rhodes et al.²⁰ On day 5, mice were infused with either saline (1 μL) or sazetidine-A (1 μg in 1 μL) at a rate of 0.25 μL/min, 20 min prior to the presentation of the alcohol bottle for 4 h. Alcohol consumption was calculated by volume, and spillage or evaporation was controlled for by a bottle in an empty control cage. After the completion of the experiment, 1 μL of ink was bilaterally injected; brains were removed and sectioned to visualize injection sites. Verification of sites was performed in a blinded manner to the behavioral results. One mouse in the sazetidine-A group and one mouse in the saline group had injections outside of the VTA and were excluded from the analysis.

2.4 |. Experiment 3: Effect of sazetidine-A on the expression of alcohol conditioned place preference

All place conditioning experiments were performed in a two-compartment place preference insert inside a 10.8 × 10.8-inch infrared activity field for mice from Med Associates Inc. The place preference insert consisted of two different floor textures (a rod and grid floor) with each floor measuring 5.4 × 10.8 in. The compartments were separated by a divider with a sliding door (Med Associates Inc, St. Albans, VT). All place conditioning procedures were performed in an unbiased manner, with the drug paired chamber counterbalanced across floor textures.

Male C57BL/6J mice were tested in an unbiased alcohol conditioned place preference (CPP) procedure modified from Cunningham et al.²² The procedure consisted of one habituation session, 10 conditioning sessions, one preference test (Test 1), and one test session with sazetidine-A or saline pretreatment (Test 2). For the habituation session, mice were injected with saline and placed in the apparatus with access to both chambers for 30 min, and the baseline time spent in each chamber was recorded. Mice that showed greater than 75% preference for one side of the chamber during baseline testing did not progress to further testing. Mice were then randomly assigned to a saline or alcohol group for conditioning, and the drug-paired floor was also randomly assigned and counterbalanced across groups. The alcohol conditioning group received alcohol (2 g/kg i.p.) and were immediately confined to one chamber for 5 min. On the next day, mice received saline i.p. injection paired with the alternate chamber, and the pairings were alternated each day for 10 total sessions. The saline control group received saline paired with both chambers. For the first preference test (Test 1), mice were injected with saline and had access to both sides for 30 min. The alcohol CPP index was calculated as the time spent in the alcohol-paired chamber during test day minus time spent in that same chamber on habituation day. In the alcohol conditioned group, three out of the 27 mice developed strong aversion to alcohol conditioning (−210 conditioned place aversion [CPA] index or lower) and did not proceed to Test 2, which occurred 24 h after Test 1. To assign the mice for pretreatment with either sazetidine-A (1 mg/kg, i.p.) or saline, the CPP indices of the alcohol-conditioned mice were ranked, and the mice were alternately assigned to either the sazetidine-A or saline group in a manner that ensured that both groups were not statistically different (saline 232.1 ± 38.4, sazetidine-A 270.7 ± 42.2, P = 0.51). Pretreatment with sazetidine-A or saline occurred 1 h prior to the second 30-min preference test (Test 2), which was performed in an identical manner to Test 1. The CPP index for Test day 2 was calculated in the same manner as for Test day 1, which was time spent in the drug-paired chamber minus the time spent in the same chamber on habituation day.

2.5 |. Experiment 4: Effect of sazetidine-A on the acquisition of alcohol CPP

For this experiment, we used a separate group of drug naïve mice. To test the effect of sazetidine-A on the acquisition of alcohol CPP, mice underwent a similar procedure described in Experiment 3 for CPP expression, except that mice were pretreated with sazetidine-A (1 mg/kg, i.p.) or saline, 1 h before every alcohol-paired conditioning session. For the saline-paired sessions, mice received a pretreatment saline injection. The alcohol CPP index was calculated as the time spent in the alcohol-paired chamber during test day minus time spent in that same chamber on habituation day.

2.6 |. Experiment 5: Effect of sazetidine-A on the expression of alcohol CPA

For this experiment, we used a separate group of drug naïve mice. We used a modified alcohol CPA procedure adapted from Cunningham, et al.²³ where mice are injected with alcohol after removal from the chamber. The procedure consisted of one habituation session, 10 conditioning sessions, one preference test (Test 1), and one test session with sazetidine-A pretreatment (Test 2). For the habituation session, mice were injected with saline and placed in the apparatus with access to both chambers for 30 min, and the baseline time spent in each chamber was recorded. Baseline habituation was performed as in Experiment 3. The alcohol conditioning group were placed in the chamber for 5 min and injected with alcohol (2 g/kg i.p.) immediately after removal from the chamber, prior to being placed back into their home cage. The next day, mice received saline i.p. injection after removal from the alternate chamber, and the pairings were alternated each day for 10 total sessions. The saline control group received saline paired with both chambers. For the first aversion test (Test 1), all mice had access to both sides for 30 min. The alcohol CPA index was calculated as the time spent in the alcohol-paired chamber during test day minus time spent in that same chamber on habituation day. Only the alcohol-conditioned mice that had a negative CPA index (22 out of 30 mice) proceeded to Test 2. Randomization to the saline or sazetidine-A pretreatment group occurred in a manner that ensured that both groups were not statistically different (saline −385.2 ± 49.6, sazetidine-A −367.2 ± 45.7, P = 0.80). Pretreatment with sazetidine-A or saline occurred 1 h prior to the second 30 min preference test (Test 2), which was performed in an identical manner to Test 1. The CPA index for Test day 2 was calculated in the same manner as for Test day 1, which was time spent in the drug-paired chamber minus the time spent in the same chamber on habituation day.

2.7 |. Experiment 6: Effect of mecamylamine on the expression of alcohol CPA

For this experiment, we used a separate group of drug naïve mice. We used mecamylamine, a nonspecific nAChR antagonist, to determine whether nAChR antagonism was important for alcohol CPA, in a separate group of drug-naïve mice. All mice were conditioned for alcohol aversion with 2 g/kg alcohol as described above for Experiment 5. We did not include a control saline conditioned group as we had demonstrated that our procedure resulted in significant alcohol CPA in Experiment 5. All mice expressed a negative CPA index on Test day 1; thus, on the next day, all mice were randomized into three groups and pretreated with saline or mecamylamine (2 or 3 mg/kg i.p.) 20 min prior to the second aversion test (Test 2). The alcohol CPA index was calculated as the time spent in the alcohol-paired chamber during test day minus time spent in that same chamber on habituation day.

2.8 |. Experiment 7: Effect of sazetidine-A on alcohol-induced loss-of-righting reflex

For this experiment, we used a separate group of drug naïve mice. Male C57BL/6J mice were pretreated for 1 h with sazetidine-A (1 mg/kg, i.p.) or saline. Mice were then injected with 4 g/kg alcohol i. p. and placed on their backs, alone, in clean cages. Righting reflex was considered lost when the mouse failed to right itself for at least 30 s after administration of the alcohol injection. The mouse was considered to have recovered the righting reflex when it could right itself three times within 30 s.

2.9 |. Experiment 8: Place conditioning with sazetidine-A and sazetidine-A induced locomotor activity

2.9.1 |. Place conditioning

We determined whether our 1-h sazetidine-A pretreatment time frame had any effect on place conditioning. For this experiment, we used a separate group of drug-naive male C57BL/6 mice. Mice were tested in a place conditioning procedure consisting of one habituation session, 10 daily conditioning sessions, and one test. For the habituation session, mice were injected with saline and had access to both chambers for 30 min. For the conditioning sessions, mice were treated with sazetidine-A (1 mg/kg i.p.) or saline 1 h prior to the 30-min conditioning session. On the next day, mice received the opposite injection paired with the alternate chamber, and the pairings were alternated each day. Control mice received saline paired with both sides. For the test session, mice were injected with saline and had access to both chambers for 30 min. The conditioning index was calculated similarly to the alcohol CPP index.

2.9.2 |. Locomotor activity

For this experiment, we used a separate group of drug-naïve male C57BL/6J mice. Mice were pretreated with sazetidine (1 mg/kg, i.p.) or saline for 1 h and placed back in their home cage. The mice were then placed in a locomotor chamber (Med Associates, St. Albans, VT) for 1 h. Total distance traveled, measured via infrared beam breaks, was measured in 5-mintime blocks and compared across the two treatment groups.

2.10 |. Experiment 9: Fluorescent in situ hybridization

For this experiment, we used a separate group of drug naïve male C57BL/6J mice. We aimed to characterize the impact of sazetidine-A or mecamylamine pretreatment on alcohol-induced neuronal activation within neuronal subpopulations of the VTA. Control groups consisted of mice pretreated with saline, placed back in their home cage for 1 h, then injected with saline (saline-saline group) or 2 g/kg alcohol (saline-alcohol group). To test the impact of sazetidine-A on neuronal activation, another group of animals were pretreated with saline and injected with 1 mg/kg sazetidine-A after 1 h (saline-SAZ group), and a fourth group was pretreated with 1 mg/kg sazetidine-A for 1 h and then injected with 2 g/kg alcohol (SAZ-alcohol group). The effect of mecamylamine was assessed by pretreating a separate group of mice with saline for 20 min followed by injection of 3 mg/kg mecamylamine (saline-MEC group), and a final group was pretreated with 3 mg/kg mecamylamine for 20 min followed by 2-g/kg alcohol (MEC-alcohol group). Drug doses and pretreatment times were chosen to mirror the parameters of the alcohol CPA experiments. Ninety minutes after the second injection, mice were sacrificed and brains were removed, snap frozen in isopentane, and sectioned on a cryostat (HM 525 NX, ThermoFischer) into 16-μM sections. Sections were chosen based on stereotaxic landmarks (acceptance criteria: −2.8 and −3.2 relative to bregma based on the Paxinos and Franklin mouse brain atlas).²¹ Separate sections were collected and processed for colocalization of Fos + Th and Fos + Gad2. The Th and Gad sections were collected consecutively to provide anatomical matching. Sections were adhered to Superfrost® Plus slides, kept at −20°C for 60 min to dry and stored at −80°C until use. Sections were fixed with 4% PFA for 1 h and processed for RNAScope (Advanced Cell Diagnostics) multichannel fluorescent in situ hybridization (FISH) according to manufacturer instructions for fluorescent multiplex assays. Sections were counterstained with DAPI for 20 s at room temperature, coverslipped with Prolong Gold Antifade (ThermoFisher Scientific), and stored at 4°C. Probes for detection of specific targets (Fos, Th, and Gad2) were purchased from Advanced Cell Diagnostics (ACD; http://acdbio.com/).

2.11 |. FISH image acquisition and analysis

Sections containing the VTA were imaged on a Keyence BZ-X700 epifluorescent microscope at 20× magnification. Images for each channel were obtained in z-stack and stitched using Keyence analysis software, and all images were acquired and processed in the same manner. Multichannel images were opened and all channels (including DAPI) were overlaid. A fixed width area was imposed around the VTA based on stereotaxic landmarks to establish regional boundaries. Cells within the boundaries of the VTA were considered positive for each transcript when at least five puncta were observed within each cell, and cell counts were tracked by a blinded researcher using the Cell Counter plugin in ImageJ v1.52e.²⁴ All counts were made bilaterally, with three to four mice analyzed for each treatment group, with four slices (2 Th + Fos and 2 Gad2 + Fos) per mouse.

2.12 |. Statistical analysis

All analyses were calculated using Prism 8.0 (GraphPad, La Jolla, CA). Data were tested for normality and variance, and outliers were detected using the Grubb’s test. Welch’s corrections were used if variances were unequal. Comparison of data across time used two-way repeated measures ANOVA followed by Tukey’s multiple comparisons tests (e.g., α4 nAChR DID and locomotor activity). Comparisons between groups with a single dependent variable used Student’s t-tests or one-way ANOVAs followed by multiple comparisons tests (e.g., FISH RNAScope).

For the conditioning data, comparison across treatment groups was analyzed using Student’s t-tests or one-way ANOVA followed by multiple comparisons test (Dunnett’s multiple comparison tests or unpaired t-tests with Welch’s corrections). To assess the development of conditioning within each group, we used one-sample t-tests compared with a conditioning index of zero (no reward or aversion).

3 |. RESULTS

3.1 |. Experiment 1: Sazetidine-A reduced binge alcohol consumption in female α4 knock-out and wild-type mice

We previously showed that systemic sazetidine-A reduced binge DID alcohol consumption in both male and female C57BL/6J mice.¹⁸ Because α4β2 nAChRs are a major subtype activated and desensitized by sazetidine-A, we sought to determine whether the α4 nAChR subunit was important for the actions of sazetidine-A. We tested the effect of systemic sazetidine-A in female mice that lack the α4 nAChR subunit in the DID procedure. We found that sazetidine-A (1 mg/kg i.p.) reduced alcohol consumption in both α4 WT and KO female mice, as there was a main effect of drug treatment with no effect of genotype or an interaction between drug treatment and genotype (Figure 1, two-way ANOVA F_interaction(1,27) = 0.281, P = 0.60; F_treatment(1,27) = 13.08, P = 0.001; F_genotype(1,27) = 0.004, P = 0.95). These data suggest that sazetidine-A does not require the α4 nAChR subunit to produce a reduction in alcohol consumption.

FIGURE 1.

Experiment 1: Sazetidine-A reduces binge alcohol consumption in α4 knock-out and wild-type mice. (A) Sazetidine-A (1 mg/kg i.p.) reduced binge alcohol consumption in both female α4 knock-out and wild-type litter mates. n = 7–8 per group. Two-way ANOVA for drug treatment: F_treatment(1,27) = 13.08, **P = 0.001

3.2 |. Experiment 2: Microinjection of sazetidine-A into the ventral midbrain reduced alcohol consumption

As our prior work used systemic injections of sazetidine-A,¹⁸ we tested whether administration of sazetidine-A into the ventral midbrain targeting the VTA would reduce acute binge DID alcohol consumption. Male C57BL/6J mice were bilaterally implanted with cannula targeting the VTA prior to starting the binge DID procedure. Mice microinjected with 1-μg/μL sazetidine-A consumed significantly less alcohol during the 4-h binge session compared with mice infused with saline (Figure 2, unpaired t-test, t = 2.965, df = 9, P = 0.02), implicating the VTA in sazetidine-A’s mechanism of action.

FIGURE 2.

Experiment 2: Microinfusion of sazetidine-A into the ventral midbrain targeting the VTA reduces binge alcohol consumption. (A) Coronal mouse brain atlas diagrams from mouse brain atlas showing the confirmed injection sites in male C57BL/6J mice infused with sazetidine-A (1 μg/μL, red) or saline (1 μL, black). (B) Average alcohol consumption during the 4 h binge session in mice pretreated with saline or sazetidine-A 20 min prior to alcohol access. *P = 0.02, n = 5–6 per group, mean ± SEM

3.3 |. Experiments 3–5: Sazetidine-A enhanced the expression of alcohol aversion without affecting alcohol reward

3.3.1 |. Experiment 3: CPP expression

A reduction in alcohol consumption may be due to a decrease in alcohol reward and/or an increase in alcohol aversion. We first tested the effect of sazetidine-A on the expression of alcohol reward using alcohol CPP. Male C57BL/6J mice were conditioned with 2 g/kg i.p. alcohol injections, which produced significant preference for the alcohol-paired chamber compared with saline conditioning on Test day 1 (Figure 3A, Student’s t-test between saline and alcohol groups: t = 2.675, df = 39, P = 0.01; one-sample t-test compared with a CPP index of zero: saline group t = 1.384, df = 10, P = 0.20; alcohol group t = 3.123, df = 29, P = 0.004). On the next day, the alcohol-conditioned mice were randomized into two groups that were pretreated with sazetidine-A (1 mg/kg i.p.) or saline 1 h prior to a second alcohol preference test (Test day 2). We found no significant difference in the alcohol CPP indices between treatment groups, indicating that sazetidine-A does not impact the expression of alcohol CPP (Figure 3B, t = 0.191, df = 25, P = 0.85).

FIGURE 3.

Experiments 3–6: Sazetidine-A enhanced expression of alcohol aversion without affecting the expression or acquisition of alcohol reward. (A) Experiment 3: Place preference conditioning with 2-g/kg alcohol resulted in significant preference for the alcohol-paired chamber (Test 1), whereas conditioning with saline produced no significant preference. **P < 0.05 using a one-sample t-test compared with a CPP index of zero, n = 30 alcohol, n = 11 saline. (B) Pretreatment with sazetidine-A (1 mg/kg i.p.) 1 h prior to the second test day (Test 2) did not significantly affect the alcohol CPP index versus saline pretreated mice, n = 13–14 per group. (C) Experiment 4: There was no difference in the CPP index between mice pretreated with sazetidine-A (1 mg/kg i.p.) or saline prior to each alcohol conditioning session, n = 10 per group. One-sample t-test compared with a CPP index of zero: saline pretreated group: t = 3.106, df = 9, P = 0.01; sazetidine-A pretreated group: t = 2.250, df = 9, P = 0.05. (D) Experiment 5: Aversion conditioning with 2-g/kg alcohol resulted a negative alcohol CPA index and aversion to the alcohol-paired chamber (Test 1), whereas conditioning with saline produced no significant aversion. ***P = 0.0001 using one-sample t-test compared with a CPA index of zero, n = 30 alcohol, n = 18 saline. (E) Pretreatment with sazetidine-A (1 mg/kg i.p.) 1 h prior to the second test day (Test 2) increased the magnitude of the negative alcohol CPA index compared with saline pretreatment. *P = 0.03 compared with the saline group, n = 9–11 per group. (F) Experiment 6: Aversion conditioning with 2-g/kg alcohol produced a significantly negative alcohol CPA index. ****P < 0.0001 using one-sample t-test compared with a CPA index of zero, n = 35. (G) Pretreatment with mecamylamine at 3 mg/kg i.p. Twenty minutes prior to the second test day (Test 2) increased the magnitude of the negative alcohol CPA index. *P < 0.05 compared with saline group by Dunnett’s post hoc test, n = 10–15 per group. All conditioning indices are in minutes, all mice used were male C57BL/6J, all data shown as mean ± SEM

3.3.2 |. Experiment 4: CPP acquisition

We then tested whether sazetidine-A impaired the acquisition of alcohol CPP during the conditioning sessions. We used a separate group of drug naïve male, C57BL/6J mice. Sazetidine-A (1 mg/kg) or saline was injected 1 h prior to each alcohol conditioning session and the mice were tested in a drug-free state. We found no significant difference in the alcohol CPP index on test day between the sazetidine-A and saline pretreated groups, indicating that sazetidine-A does not impact the acquisition of alcohol CPP (Figure 3C, t = 0.262, df = 18, P = 0.79; one-sample t-test compared with a CPP index of zero: saline pretreated group: t = 3.106, df = 9, P = 0.01; sazetidine-A pretreated group: t = 2.250, df = 9, P = 0.05).

3.3.3 |. Experiment 5: CPA expression

As sazetidine-A had no effect on the expression or acquisition of alcohol CPP, we then tested whether sazetidine-A affected the expression of alcohol aversion using alcohol CPA. Mice were conditioned with 2-g/kg alcohol injections, which produced significant aversion for the alcohol-paired chamber compared with saline conditioning on Test day 1 (Figure 3D, Student’s t-tests between saline and alcohol-treated groups t = 4.204, df = 45, P = 0.0001; one-sample t-test compared with a CPA index of zero: saline-treated group t = 1.18, df = 17, P = 0.25; alcohol-treated group t = 4.47, df = 28, P = 0.001). The alcohol-conditioned mice that developed aversion were randomized into two groups that were pretreated with sazetidine-A (1 mg/kg) or saline prior to a second aversion test on Test day 2. We found that pretreatment with sazetidine-A enhanced the expression of alcohol CPA compared with saline (Figure 3E, t = 2.284, df = 18, P = 0.03).

3.4 |. Experiment 6: The effect of mecamylamine on alcohol CPA

We then tested the effect of mecamylamine on the expression of alcohol aversion to determine whether a nonspecific nAChR antagonist would also enhance the expression of alcohol CPA. We conditioned a separate group of drug naïve mice and showed that alcohol conditioning produced significant aversion on Test day 1 (Figure 3F, one-sample t-test compared with a CPA index of zero: t = 9.655, df = 34, P < 0.0001). On the next day, the alcohol-conditioned mice were randomized into three groups and were pretreated with mecamylamine (2 or 3 mg/kg) or saline 20 min prior to a second aversion test (Test day 2). Mecamylamine enhanced the expression of alcohol aversion at the 3 mg/kg dose but not the 2 mg/kg dose, compared with the saline pretreated group (Figure 3G, one-way ANOVA F_2,32 = 6.38, P = 0.005). These results suggest that antagonism of nAChRs also produces enhanced expression of alcohol aversion.

3.5 |. Experiment 7: Sazetidine-A enhanced the sedative-hypnotic effects of alcohol

To determine whether sazetidine-A affected the sedative-hypnotic effects of high-dose alcohol, we tested whether sazetidine-A altered acute alcohol-induced loss-of-righting-reflex (LORR). Pretreatment with sazetidine-A (1 mg/kg) for 1 h increased alcohol-induced LORR duration produced by 4 g/kg alcohol compared with saline pretreatment (Figure 4A, t = 2.554, df = 7.477, P = 0.04).

FIGURE 4.

Experiments 7–8: Effect of sazetidine-A on alcohol-induced LORR, place conditioning, and locomotor activity. (A) Experiment 7: The average alcohol-induced LORR duration after i.p. injection of 4-g/kg alcohol in mice pretreated with sazetidine-A (1 mg/kg, i.p.) was greater compared with mice pretreated with saline. *P = 0.04 compared with saline, n = 7–8 per group. (B) Experiment 8: Place conditioning with sazetidine-A only using a 1-mg/kg i.p. dose and 1 h timing was not significantly different from saline conditioning, n = 7–8 per group. (C) Average ambulatory distance in mice after injection with sazetidine-A (1 mg/kg, i.p., 1 h) was not significantly different from the saline treated group, n = 8 per group. All data shown as mean ± SEM

3.6 |. Experiment 8: Sazetidine-A treatment alone does not affect place conditioning or locomotor activity

Our experiments used a 1-h pretreatment of 1-mg/kg sazetidine-A, which was based on our prior work.¹⁸ We tested whether this dose and timing of sazetidine-A had any rewarding or aversive effects alone in place conditioning and whether it affected locomotor activity. We found that 1-mg/kg sazetidine-A produced no expression of reward or aversion and was similar to the conditioning produced by saline injections (Figure 4B, t = 0.559, df = 13, P = 0.59). We also found no main effect of sazetidine-A treatment on locomotor activity compared with saline, as measured by beam breaks (Figure 4C, two-way RM ANOVA F_interaction(11,154) = 4.75, P = 0.07; F_treatment(1,154) = 0.89, P = 0.22; F_time(11,154) = 48.80, P < 0.0001).

3.7 |. Experiment 9: Sazetidine-A, but not mecamylamine, induced neuronal activation in Th-expressing neurons in the VTA

As microinjecting sazetidine-A into the ventral midbrain reduced binge DID consumption, we sought to determine the effect of a single systemic injection of sazetidine-A on neuronal activity within the VTA. We additionally sought to compare the neuronal activation induced by sazetidine-A to mecamylamine. FISH (RNA Scope) was used to identify Fos transcript expression in VTA cells that express tyrosine hydroxylase (Th), a common transcript in DA neurons, or glutamate decarboxylase 2 (Gad2), a transcript expressed in GABA neurons (-Figure 5). Pretreatment with sazetidine-A (1 mg/kg) prior to a 2 g/kg alcohol injection produced an increase in Fos expression in Th-, but not Gad2-expressing VTA neurons compared with saline pretreatment (Th neurons: one-way ANOVA F_5,18 = 15.48, P < 0.0001; Tukey’s multiple comparisons test ***P < 0.001 for SAZ/alc compared with all other groups, Figure 6A,B). There was no significant difference in the number of Fos-expressing Th-positive cells between the saline, alcohol, and sazetidine-A only treated groups, suggesting that the increase in Fos expression in Th-expressing neurons is not due to either sazetidine-A or alcohol alone (Tukey’s multiple comparisons tests, P > 0.05 between sal/sal, sal/alc, and sal/SAZ). In addition, we found that pretreatment with mecamylamine (3 mg/kg) prior to a 2-g/kg alcohol injection did not induce Fos expression in Th-expressing VTA neurons (Th neurons: one-way ANOVA F_5,18 = 15.48, P < 0.0001; Tukey’s multiple comparisons test P > 0.05 between MEC/sal and MEC/alc, and compared with the sal/sal, sal/alc, and sal/SAZ groups, Figures 6 and S1). Fos expression in Gad2-expressing neurons did not increase with either sazetidine-A, mecamylamine, or alcohol injections (Gad2 neurons: one-way ANOVA F_5,18 = 0.6097, P = 0.69, Figure 6C, D). These data show that pretreatment with sazetidine-A, but not mecamylamine, followed by alcohol injection selectively increased Fos expression in Th-expressing neurons of the VTA and suggest that the molecular mechanism by which sazetidine-A mediates its actions differs from mecamylamine.

FIGURE 5.

Experiment 9: FISH RNA Scope measurement of Fos, Th, and Gad2 in VTA neurons. Representative images of Fos (white, Cy5) expression in Th-(green, EGFP) or Gad2-positive neurons (red, mCherry) in the VTA. Scale bar = 100 μm. 1,2: magnification of Th and Gad2 in the SAZ/alc group

FIGURE 6.

Experiment 9: Sazetidine-A pretreatment followed by alcohol injection induced Fos expression in Th-expressing VTA neurons. (A,B) The number of Fos-expressing Th-positive neurons in the VTA of mice for all treatment groups. ***P < 0.001 for the group pretreated with sazetidine-A followed by alcohol injection (SAZ/alc group) compared with all other groups. (C,D) The number of Fos-expressing Gad2-positive neurons was similar between all treatment groups. n = 3–5 mice per group, mean ± SEM for (A) and (C). The average number of neurons is reported in parentheses for (B) and (D)

4 |. DISCUSSION

Pharmacological manipulation of nAChRs has been shown to reduce alcohol consumption in animal models. Mecamylamine, varenicline, and sazetidine-A all reduce alcohol consumption in rodents despite their different mechanisms of action and nAChR subtype targets.^6–8,18 Mecamylamine is a nonspecific, noncompetitive nAChR antagonist, whereas varenicline is best known as a partial agonist at α4β2 receptors, but in vitro studies show that it is also a partial agonist at α3* and α6* receptors and a full agonist at α7 receptors.^14,25–28 in vitro studies show that sazetidine-A agonizes and then desensitizes α4β2, α3β4*, α6*, and α7 nAChRs.^13–16 Sazetidine-A and varenicline act on similar nAChR subtypes but with differing potencies.¹³

These nAChR-acting drugs have been used in rodent models to help elucidate the role of nAChRs in alcohol consumption, and the vast majority of studies have been performed in male animals (see Table 1). Both mecamylamine and nicotine reduce binge DID consumption in male mice.⁷ Varenicline has been shown to reduce binge consumption in both male and female C57BL/6J mice.^29,30 Moreover, varenicline requires the α4 nAChR subunit to reduce alcohol consumption, as it does not reduce alcohol consumption in male α4 KO mice.²⁹ In our previous work, we found that sazetidine-A reduced both continuous and binge alcohol consumption in male and female C57BL/6J mice.¹⁸ In Experiment 1, we used female α4 KO and WT mice to determine whether the decrease in binge alcohol consumption by sazetidine-A required the α4 nAChR subunit. In contrast to varenicline, we found that sazetidine-A decreased binge alcohol consumption in both α4 KO and WT female mice, indicating that sazetidine-A does not require the α4 subunit to decrease alcohol consumption. Constitutive deletion of the α4 subunit also produces a decrease in α6 and β3 expression in the α4 KO mice,³³ and data collected using the α4 KO mice require caution in interpretation. Divergent effects of varenicline and sazetidine-A have been demonstrated by other studies. Turner et al³⁴ showed that intrahippocampal injections of sazetidine-A, but not varenicline, reduced nicotine withdrawal signs in male mice. Furthermore, acute varenicline injection, but not acute sazetidine-A injection, produces anxiolytic effects in the marble-burying test and in novelty-induced hypophagia in male mice.³⁵ Sazetidine-A, but not varenicline, produced antidepressant effects on the forced swim test and tail suspension test in male mice.³⁵ Together, these data suggest that despite the similarity in nAChR targets in vitro, sazetidine-A and varenicline likely act on different nAChR subtypes in vivo with different functional effects to produce distinct behavioral outcomes.

TABLE 1.

Comparison of sazetidine-A, varenicline, and mecamylamine effects on alcohol DID consumption and alcohol place conditioning

	Sazetidine-A	Mecamylamine	Varenicline
Binge DID consumption	↓18 Does not require α4 subunit	↓7	↓29,30 Requires α4 subunit
CPP expression	No effect	↓31	No effect32
CPP acquisition	No effect	↓31	Unknown
CPA expression	↑	↑	Unknown

A decrease in alcohol consumption may be due to a decrease in alcohol reward, an increase in alcohol aversion, or a combination of both. We found no effect of sazetidine-A on the expression or acquisition of alcohol CPP in male C57BL/6J mice. In contrast, mecamylamine has been shown to inhibit the acquisition and expression of alcohol CPP in male mice when administered intracerebroventricularly,³¹ which suggests that antagonism of nAChRs reduces the development alcohol reward associations and the expression of such previously learned associations. Interestingly, varenicline has been shown to have no effect on the expression of alcohol CPP in male and female mice,³² similar to our findings with sazetidine-A. Notably, the role of nAChRs in alcohol aversion has not been extensively explored in preclinical models. Here, we found that 1-mg/kg sazetidine-A and 3-mg/kg mecamylamine both enhanced the expression of conditioned alcohol aversion. Importantly, pretreatment with 1-mg/kg sazetidine-A did not impact locomotor activity in our study. Previous studies have additionally shown that mecamylamine doses of 1–4 mg/kg do not affect locomotor activity or induce aberrant behavior in male mice,³⁶ suggesting that the observed enhancement of alcohol aversion is not due to the changes in locomotion. The mechanism underlying the contribution of nAChRs to alcohol aversion is unclear. As sazetidine-A initially activates and then desensitizes nAChRs,^13–16 and we showed that both sazetidine-A and mecamylamine enhance alcohol aversion, we speculate that reduced nAChR function, through desensitization or antagonism, mediates the enhanced expression of alcohol aversion in the CPA assay. Overall, these data demonstrate an important role of nAChRs in conditioned alcohol aversion and suggest that enhancing alcohol aversion may be one mechanism by which nAChR drugs mediate a decrease in alcohol consumption. Therefore, developing additional pharmacological agents that increase the expression of previously conditioned alcohol aversion may be a potentially useful and viable strategy to reduce alcohol consumption.

Alcohol sedation and alcohol consumption often show a negative correlational relationship, as genetic deletion of postsynaptic density protein 95 or dopamine beta-hydroxylase concurrently increase alcohol sedation and decrease consumption.^37,38 In addition, some genetic hybrid mouse models show the same negative correlation, with increased alcohol sedation and decreased alcohol consumption in B6 X NZB/B1NJ mice compared with B6 X FVB/NJ mice.³⁹ We showed that sazetidine-A produced an increase in alcohol sedation, which may be a contributing factor to sazetidine-A’s ability to reduce alcohol consumption.

We microinjected sazetidine-A injected directly into the ventral midbrain targeting the VTA and found that it reduced binge alcohol consumption in C57BL/6 mice. The VTA is an important structure that is critical for drug consumption, reward, and aversion. The VTA contains 8 of the 11 nAChR subunits found in the brain^40,41 and consists of DA, GABA, and glutamate neurons in distinct circuits that mediate reward and aversion.^42,43 In particular, optogenetic activation of VTA GABA neurons or inhibition of VTA DA neurons during place conditioning has been shown to produce CPA.⁴⁴ The neural circuitry that mediates aversion also includes projections from the lateral habenula (LHb) to the GABAergic rostromedial tegmental nucleus (RMTg), which then provides inhibitory input to the VTA.^45–47 Our microinjection study utilized a 1-μL infusion volume per side, which could potentially result in sazetidine-A diffusion into other proximal brain regions such as the substantia nigra and RMTg. Identifying whether these brain structures are involved in mediating sazetidine-A’s effects will provide further insight into the role of nAChRs in the neurobiology of alcohol aversion.

To determine the effect of sazetidine-A and mecamylamine on VTA neurons, and to begin identifying the neural circuitry that was important for nAChR-mediated alcohol aversion, we examined neuronal activation by measuring Fos induction in the VTA using the same time frame and doses of sazetidine-A and mecamylamine as in our CPA experiments. An examination of transcript expression using FISH RNA Scope showed that pretreatment with 1 mg/kg sazetidine-A followed by 2-g/kg alcohol (SAZ-alc group) increased Fos transcript expression in Th-expressing neurons and had no effect on Fos expression in Gad2-expressing neurons of the VTA. There was no significant change in Fos transcript expression in Th-expressing neurons in the saline, sazetidine-A or alcohol only groups. We also found that pretreatment with 3-mg/kg mecamylamine followed by 2 g/kg alcohol (MEC-alc group) did not induce Fos expression in either Th- or Gad2-expressing neurons in the VTA. Moreover, there was no change in the number of Gad2-expressing neurons that showed Fos expression across all groups. These data suggest that although both sazetidine-A and mecamylamine produce enhanced alcohol CPA, the circuit mechanisms by which both drugs mediate the enhanced alcohol aversion may differ.

Based on our data and the known actions of sazetidine-A and mecamylamine in vitro,^13–16 we speculate that the Fos induction in Th-expressing cells is produced by sazetidine-A agonism of nAChRs. This would be followed by prolonged sazetidine-A induced nAChR desensitization, as sazetidine-A is cleared slowly from the brain and can occupy brain nAChRs for ~8 h.¹⁹ Because nAChR desensitization is thought to decrease neuronal activity,^48,49 sazetidine-A-induced desensitization may lead to decreased activity of putative VTA DA neurons, thus enhancing alcohol aversion and reducing alcohol consumption. In contrast, our data suggest that mecamylamine enhances alcohol CPA via a different molecular mechanism. Because we found no Fos induction in either Th- or Gad2-expressing VTA neurons, we speculate that mecamylamine mediates its effects by directly antagonizing nAChRs, regardless of neuronal cell type.

Because we showed that sazetidine-A did not require the α4 nAChR subunit to reduce alcohol consumption, our data further suggest that non-α4 nAChRs on Th-expressing neurons may be important in sazetidine-A mediated alcohol aversion. Future studies will test this working hypothesis and determine which nAChR subunits are present in Fos expressing Th-positive neurons. Chronic exposure to sazetidine-A does not produce nAChR upregulation in rats or mice,⁵⁰ suggesting that repeated exposure to sazetidine-A would produce cycles of receptor desensitization and resensitization without increasing receptor expression.

5 |. CONCLUSIONS

In summary, we found that sazetidine-A, a nAChR agonist and desensitizer, enhanced the expression of alcohol CPA without affecting the expression or acquisition of alcohol CPP. Mecamylamine, a nAChR antagonist, also enhanced alcohol CPA. We showed that pretreatment with sazetidine-A, followed by alcohol injection, increased Fos expression in Th-, and not Gad2-, expressing VTA neurons, whereas pretreatment with mecamylamine had no effect on Fos expression in either cell type. We speculate that the increase in alcohol aversion contributes to the reduction in alcohol consumption we observed after treatment with sazetidine-A. Further elucidating the role of specific nAChR subunits in alcohol aversion mechanisms may identify novel drug targets for the development of pharmacotherapies to treat alcohol use disorder.