Acetylcholine muscarinic M₄ receptors as a therapeutic target for alcohol use disorder: Converging evidence from humans and rodents

Abstract

BACKGROUND:

Alcohol use disorders (AUD) are a major socioeconomic burden on society and current pharmacotherapeutic treatment options are inadequate. Aberrant alcohol use and seeking alters frontostriatal function.

METHODS:

We performed genome-wide RNA sequencing, and subsequent qPCR and receptor binding validation in the caudate-putamen of human AUD samples to identify potential therapeutic targets. We next back-translated our top candidate targets into a rodent model of long-term alcohol consumption to assess concordance of molecular adaptations in the rat striatum. Finally, we adopted rat behavioural models of alcohol intake and seeking to validate a potential therapeutic target.

RESULTS:

We found G protein-coupled receptors (GPCRs) were the top canonical pathway differentially regulated in individuals with AUD. The M₄ muscarinic acetylcholine receptor (mAChR) was down regulated at the gene and protein level in the putamen, but not caudate, of AUD samples. We found concordant downregulation of the M₄ mAChR, specifically on dopamine D₁ receptor-expressing medium spiny neurons in the rat dorsolateral striatum. Systemic administration of the selective M₄ mAChR positive allosteric modulator, VU0467154, reduced home cage and operant alcohol self-administration, motivation to obtain alcohol and cue-induced reinstatement of alcohol seeking in rats. Local microinjections of VU0467154 in the rat dorsolateral striatum reduced alcohol self-administration and cue-induced reinstatement of alcohol seeking.

CONCLUSIONS:

Collectively, these results identify the M₄ mAChR as a potential therapeutic target for the treatment of AUD and D₁ receptor positive medium spiny neurons in the dorsolateral striatum as a key site mediating the actions of M₄ mAChR in relation to alcohol consumption and seeking.

Introduction:

Despite the socioeconomic burden of alcohol use disorders (AUD), therapeutic treatment options are limited(1). Basic neuroscience has identified targets that contribute to alcohol reward, craving and relapse(2-4); however, only an ~8% of drugs that start clinical trial will continue to market(5). Addictions emerge from aberrant learning and motivation towards drugs of abuse, in part via plasticity in the striatum and frontal cortices(6-8). The transition to aberrant seeking is hypothesized to involve a shift in recruitment of ventral striatum (accumbens) in light or social drinkers to dorsal striatum (caudate putamen) in heavy drinkers(9). Within the dorsal striatum, the putamen (dorsolateral striatum, DLS) executes behavioural control after long-term alcohol consumption in rodents(10) and humans(11). Here the cholinergic system regulates motivation for reward(12) and striatal dopamine release(13-16).

Acetylcholine (ACh) has long been implicated in reward and higher order cognitive processes(17), however, focus has historically surrounded nicotinic acetylcholine receptors (nAChRs). Recent development of novel selective drugs to target muscarinic acetylcholine receptors (mAChR) have highlighted their therapeutic potential(18-21). The M₄ muscarinic acetylcholine receptor (mAChR) is a G_i/o coupled G protein-coupled receptor(22, 23), densely expressed in the striatum, on cholinergic interneurons, glutamatergic terminals and medium spiny GABAergic DRD1-expressing projection neurons(24-26). M₄ mAChR KO mice show increased cocaine and alcohol self-administration(27, 28), M₄ selective allosteric modulation reduces cocaine seeking(21) and genetic polymorphisms within the CHRM4 gene are linked to cocaine and heroin addiction(29).

While previous genome wide profiling has been performed on the prefrontal cortex(30), amygdala(31) and hippocampus(32, 33) of individuals with AUD, alterations in the caudate-putamen remain unknown. Therefore, we performed genome-wide RNA sequencing, RT-qPCR and receptor binding in the caudate-putamen of human AUD individuals to identify potential therapeutic targets. We then probed candidate genes in corresponding rat striatal regions (DMS and DLS) after long-term alcohol consumption and found concordance in the downregulation of the M₄ mAChR. Our data suggest AUD leads to M₄ mAChR downregulation within the DLS, which may facilitate the maintenance of alcohol consumption in humans and rodents after long-term alcohol consumption. We hypothesised that restoring cholinergic balance within the striatum through enhancement of M₄ mAChR signalling would reduce alcohol consumption and seeking following abstinence. Together our data suggest the M₄ mAChR is a potential therapeutic target for the treatment of AUD and dopamine D₁ receptor positive medium spiny neurons (MSNs) in the DLS are a key site where M₄ receptors are dysregulated.

Materials and Methods

Human subjects and sample analysis

Human brain samples were obtained from the New South Wales Tissue Resource Centre at the University of Sydney, Sydney. All procedures were conducted in accordance with the Australian National Health and Medical Research Council’s National Statement on Ethical Conduct in Human Research (2007), the Victorian Human Tissue Act (1982), the National Code of Ethical Autopsy Practice (2002) and the Victorian Government policies and practices in relation to post-mortem tissue. Ethics approval: NSW Health Sydney Local Health District reference number X15-0199 (prev X11-0107) & HREC/11/RPAH/147; Monash University Human Research Ethics Committee CF15/2976-2015001223. Caudate and putamen tissue samples from individuals of European descent were matched as closely as possible according to age, sex and cause of death (n=10/group). To be included as part of the study subjects had to meet the following criteria: age between 18 - 64 years, no head injury at time of death, no recent cerebral stroke, no history of other psychiatric or neurological disorders, no hepatic encephalopathy and post-mortem interval within 72 hours. The AUD cohort, subjects had to satisfy a DSM-IV diagnosis of alcohol use disorder. All subjects were “never in my lifetime” smokers (see Supplementary Figure S1 for subject demographics).

A full description of all materials and procedures used can be found in Supplementary information. RNA libraries were prepared (SureSelect Strand-Specific RNA Library Prep for Illumina Multiplexes Sequencing, Agilent Technologies, CA, USA) and sequenced using the HiSeq 2500 system rapid run mode (Illumina, CA, USA). See Supplementary Tables 4 and 5 for subject metadata, library preparation and sequencing run batching information. Quantitative Real-Time PCR (qRT-PCR) and [³H]-N-methylscopolamine (NMS) binding was used to validate targets using samples from the same AUD and healthy individuals (Supplementary Table 6 for primer sequences).

Animal subjects and analysis

All rodent subjects were adult male inbred Indiana preferring (iP) rats. All animal studies were undertaken in accordance with the Prevention Cruelty to Animals Act (2004) and carried out within the guidelines of the National Health and Medical Research Council (NHMRC) Code of Practice for the Care and Use of Animals for Experimental Purposes in Australia (2013) and approved by the Florey Animal Ethics Committee.

For RT-qPCR and RNAscope analysis rats (N=40) were divided into three groups; alcohol-naïve, alcohol-experienced and alcohol-abstinence. Rats underwent intermittent home-cage two-bottle choice for 57 sessions. Alcohol-experienced rats were sacrificed 24 h after the last ethanol access, while alcohol-abstinence rats were killed after 14 days abstinence. Alcohol-naïve groups were maintained on water. RT-qPCR was conducted in the DMS, DLS, nucleus accumbens core (AcbC) and shell (AcbSh) (n=8/group). See Supplementary Table 6 for primer sequences. The RNAscope Multiplex Fluorescent Reagent Kit (Advanced Cell Diagnostics, CA, USA) was used to detect Chrm4, Chat and Drd1 mRNA in the dorsal striatum (n=5-6/group)(34, 35). Chrm4 puncta were quantified (~Bregma 1.80 to 2.00)⁽³⁶⁾ using image J 3D object counter.

For systemic operant studies male iP rats were trained to self-administer ethanol (10% (v/v)) or sucrose (0.25–2.5% w/v) under FR3 operant conditions 5 days/week for >60 sessions⁽¹⁹⁾. To test the involvement of the M₄ mAChR, vehicle (10% Tween80 in water; 10 mL/kg, p.o.) or the selective M₄ PAM VU0467154 was administered orally as a microsuspension in vehicle (10 or 30 mg/kg; 10 mL/kg, p.o.) 2 or 2.5 h prior to testing respectively(37). Separate cohorts of rats were tested for operant self-administration of ethanol (10 mg/kg, n=9; 30 mg/kg, n=7) or sucrose (30 mg/kg, n=8) ^{(38, 39)}, progressive ratio (PR3-4, 30 mg/kg, n=10)(40, 41), or extinction followed by cue-induced reinstatement to alcohol seeking (30 mg/kg, n=8)(19). Rats were tested in a randomised counterbalanced manner after baseline responding was re-established (5 days). To assess the role of M₄ mAChRs in voluntary alcohol consumption rats (n=19) were allowed to self-administer alcohol in a continuous access two-bottle choice paradigm for 11 weeks. Alcohol (10% v/v) and water consumption were measured 2, 4, 12 and 24 h from the start of the dark cycle after vehicle or VU0467154 (30mg/kg p.o.) . Two separate cohorts were examined on the rotarod after receiving vehicle or VU0467154 (30 or 100 mg/kg, p.o.) 2.5 or 3 hours prior to testing respectively (see Supplementary Information).

For intra-cranial infusions, vehicle (60% DMSO in saline, 500nL/hemisphere) or VU0467154 (600 pmol or 3 μmol/hemisphere, 500nL/hemisphere) was infused directly into the DLS via implanted cannulae directly prior to behavioural testing⁽¹⁹⁾. Separate cohorts of rats then underwent self-administration (600 pmol, n=7; 3 μmol, n=14) or cue-induced reinstatement of alcohol seeking (3 μmol, n=7) as described above. Site validations were conducted by an experimenter blinded to treatment group and outcomes (Supplementary information).

Results

GPCRs are dysregulated in caudate putamen of individuals with AUD

The development of AUD is hypothesized to involve recruitment of caudate putamen(9). Understanding changes induced following long-term alcohol consumption may identify targets to treat AUD. We first conducted genome-wide RNA sequencing on age and gender matched non-smoking individuals with AUD (n=10) and controls (n=10). Unpaired t-test of patient demographics revealed individuals with AUD had significantly increased alcohol intake per day, standard drinks consumed per week and number of years drinking compared to controls. A significant difference was also observed in post-mortem interval. No difference in age, BMI, ratio of brain volume/weight, brain pH level or RNA quality were observed between AUD and controls (Supplementary Figure 1).

Using the cut off (p < 0.05 and FC > 1.6) we observed 303 differentially expressed genes (DEGs) within the caudate-putamen, of which 171 genes were upregulated and 132 genes were downregulated in AUD compared to controls (Figure 1C & Supplementary Tables 1 and 2). Ingenuity pathway analysis (IPA) demonstrated G protein-coupled receptors (GPCRs) as the top differentially regulated canonical pathway in AUD (overlap 5% 15/272; p = 3.36×10⁻⁶, Figure 1D & E, Supplementary Table 3), including CHRM4 (M₄ mAChR; p = 1.8×10⁻³, log 2 fold change (FC) = −1.96), CNR1 (cannabinoid receptor 1; p = 4.8×10⁻⁴, FC = −2.13), DRD1 (dopamine D1 receptor; p = 8.0×10⁻⁴, FC = −2.13) and GPR88 (G protein-coupled receptor 88; p = 5.7×10⁻³, FC = −2.10).

Figure 1. G protein-coupled receptor expression is dysregulated in individuals with AUD.

(A) Sagittal and coronal schematics of brain indicating where caudate and putamen were dissected for RNA sequencing. (B) Volcano plot and (C) summary of differentially expressed genes (DEGs) in the caudate and putamen of individuals with alcohol use disorders (AUD) compared to healthy controls showed 171 upregulated (pink) and 132 significantly downregulated (blue) genes. (D) Ingenuity Pathway Analysis of the top 300 differentially expressed genes identified the top 5 most significant canonical pathways, of which GPCRs were number 1 [overlap 5% 15/272; p = 3.36 × 10⁻⁶]. (E) Summary of differential expression of G protein-coupled receptors in individuals with AUD compared to healthy control. PTH1R (p = 3.3 × 10⁻⁴, FC = 1.69), ADGRG1 (p = 4.6 × 10⁻⁵, FC = = 1.70), S1PR3, (p = 2.2 × 10⁻³, FC = 1.85), GPR4 (p = 7.8 × 10⁻⁵, FC = 2.13), GPRC5C (p = 1.4 × 10⁻⁵, FC = 2.14), GRM5 (p = 1.3 × 10⁻², FC = 1.69), GPR158 (p = 1.4 × 10⁻³, FC = 1.76), HTR6 (p = 9.3 × 10⁻³, FC = 1.79), CHRM4 (p = 1.8 × 10⁻³, FC = 1.96), GPR88 (p = 5.7 × 10⁻³, FC = −2.10), GPR6 (p = 1.2 × 10⁻², FC = −2.13), DRD1 (p = 8.0 × 10⁻⁴, FC = −2.13), CNR1 (p = 4.8 × 10⁻⁴, FC = −2.13) and HTR2C (p = 6.0 × 10⁻⁴, FC = −2.54). n=10/treatment per region. Abbreviations; FC, fold change; ADGRG1, adhesion G-protein coupled receptor G1; GPR, G protein-coupled receptor; GRM5, glutamate metabotropic receptor 5; S1PR3, sphingosine 1 phosphate receptor 3; PTH1R, parathyroid hormone 1 receptor; HTR, 5-hydoxytryptamine receptor; CHRM4, M₄ muscarinic acetylcholine receptor; DRD1, dopamine D₁ receptor.

CHRM4 is dysregulated in the putamen of individuals with AUD

Within the caudate-putamen, long-term alcohol consumption is hypothesised to result in a shift from dorsomedial (caudate) to dorsolateral (putamen) engagement(10). Therefore, to determine whether GPCR dysregulation is localised to a specific subdivision we conducted RT-qPCR in the caudate vs. putamen from the same individuals with AUD and controls (n=10/group/region). In line with RNA sequencing, CNR1 (t = 2.25, df = 18, p = 0.037), CHRM4 (t = 2.21, df = 18, p = 0.040), DRD1 (t = 2.12, df = 18, p = 0.048) and GPR88 (t = 2.46, df = 18, p = 0.0246) were downregulated, while DRD2 (t = 0.650, df = 18, p = 0.52) and DRD3 (t = 0.772, df = 18, p = 0.45) mRNA expression were not altered. These differences were restricted to the putamen (Figure 2A), suggesting GPCR dysregulation is subregion specific in AUD. We focused on the M₄ mAChR as it indirectly modulates dopamine release(13) and has been implicated in addictive behaviours(21, 28, 42). Additionally, there has been recent progress in the development of mAChR subtype selective allosteric modulators(37, 43). Therefore, we examined mAChR protein binding in the caudate vs. putamen of post-mortem tissue from the same individuals with AUD and controls using [³H]-NMS (n=10/group/region). AUD samples had decreased mAChR density in the putamen (B_max= 51.5 fmol/mg protein control vs. 36.1 fmol/mg protein in AUD; F_(1,76) = 8.153, p = 0.006, Figure 2B), but not caudate (B_max= 37.9 fmol/mg protein control vs. 42.6 fmol/mg protein AUD; F_(1,76) = 0.776, p = 0.381, Figure 2B). As M₁ and M₄ receptor subtypes are the most abundant in the dorsal striatum(44) we conducted assays in the presence of either the selective M₁ mAChR toxin (MT-7) or the selective M₄ mAChR toxin (MT-3). Individuals with AUD had decreased M₄ mAChRs within the putamen (B_max= 27.8 fmol/mg protein control vs. 20.2 fmol/mg protein AUD; F_(1,76) = 5.46, p = 0.022, Figure 2C), but not caudate (B_max= 23.0 fmol/mg protein control vs. 24.8 fmol/mg protein AUD; F_(1,76) = 0.315, p = 0.576, Figure 2C). No significant differences were observed for the M₁ receptor in the caudate (B_max= 28.7 fmol/mg protein control vs. 27.6 fmol/mg protein AUD; F_(1,76) = 0.033, p = 0.856), or putamen (B_max= 35.5 fmol/mg protein control vs. 27.7 fmol/mg protein AUD; F_(1,76) = 2.06, p = 0.155 (data not shown). Together these data suggest that M₄ mRNA and protein is specifically downregulated within the putamen of individuals with AUD.

Figure 2. Molecular characterisation of the caudate vs putamen of individuals with AUD.

(A) RT qPCR validation was conducted independently in the caudate and putamen of healthy controls and individuals with AUD to determine region-specific dysregulation. In line with RNA sequencing data CNR1 (t = 2.25, df = 18, p = 0.037), CHRM4 (t = 2.21, df = 18, p = 0.040), DRD1 (t = 2.12, df = 18, p = 0.048) and GPR88 (t = 2.46, df = 18, p = 0.0246) were downregulated, while DRD2 (t = 0.650, df = 18, p = 0.52) and DRD3 (t = 0.772, df = 18, p = 0.45) mRNA expression were not altered in individuals with AUD. Further, this downregulation was specific to the putamen of individuals with AUD. To further explore the downregulation of corresponding M₄ mAChR receptor expression, we conducted [³H]NMS saturation binding assays to determine the (B) total mAChR and (C) M₄ mAChR specific binding in the caudate and putamen of healthy controls and individuals with AUD. Total mAChR binding was decreased in the putamen (F_(1,76) = 8.153, p = 0.006), but not caudate (F_(1,76) = 0.776, p = 0.381). In line with CHRM4 mRNA expression M4 mAChR binding was specifically decreased in the putamen (F_(1,76) = 5.46, p = 0.022), but not caudate (F_(1,76) = 0.315, p = 0.576) of individuals with AUD. Data expressed as mean ± SEM; n=10/group, *p < 0.05, **p < 0.01 (F-test); n.s. = not significant.

Long-term alcohol consumption/abstinence leads to dysregulation of GPCRs in the rodent striatum

The translation between rodent and human models is a major hurdle in the development of CNS drugs(1, 2). Therefore, to further examine M₄ mAChR dysregulation we reverse-translated our findings into a rodent model of long-term voluntary alcohol consumption/abstinence and assessed mRNA expression in rat DMS (caudate) and DLS (putamen) (n=8/group/region). Rats consumed an average of 7.5 ± 0.5 g/kg/session with a 71 ± 2.2 % preference for alcohol over water (Figure 3A). RT-qPCR revealed long-term alcohol consumption/abstinence lead to a reduction in Chrm4 mRNA in the DLS (F_(2,21)= 6.955, p = 0.0048; Bonferroni post-hoc, naïve vs. abstinence, p = 0.0064; alcohol vs. abstinence, p = 0.0279, Figure 3C), but not DMS (Figure 3B), AcbC or AcbSh (Supplementary Figure S2). Cnr1 (F_(2,21)= 3.936, p = 0.0353; naïve vs. abstinence group, p = 0.0398) and Gpr88 (F_(2,21)= 15.33, p < 0.0001; naïve vs. abstinence group, p < 0.0001; alcohol vs. abstinence, p = 0.0014) were downregulated in rodent DLS in accordance with the human data (Figure 3C).

Figure 3. Molecular characterisation of the rodent dorsomedial and dorsolateral striatum following long-term alcohol consumption.

Alcohol preferring iP rats underwent long-term intermittent alcohol consumption in a two-bottle choice where they consumed (A) high levels of 20% alcohol (g/kg) across >57 sessions. Rats were randomly divided into two groups, one of which were culled 24 h after their final session (alcohol, n=8) and another which were culled 14 days after their last session (abstinence, n=8). A third group of age matched alcohol naïve rats were also included (n=8). The dorsomedial (DMS; caudate equivalent) and dorsolateral (DLS; putamen equivalent) striatum were assessed for molecular adaptations following long-term alcohol consumption and abstinence. RT-qPCR analysis was conducted in the (B) DMS and (C) DLS. In line with human post-mortem tissue from individuals with AUD, long-term alcohol consumption/abstinence in rodents decreased Chrm4 mRNA (ANOVA, F_(2,21) = 6.955, p = 0.0048; Bonferroni post-hoc, naïve vs abstinence, p = 0.0064; alcohol vs. abstinence, p = 0.0279), Cnr1 (ANOVA, F_(2,21) = 3.936, p = 0.0353; Bonferroni post-hoc, naïve vs. abstinence group, p = 0.0398) and Gpr88 (ANOVA, F_(2,21) = 15.33, p < 0.0001; Bonferroni post-hoc, naïve vs. abstinence group, p < 0.0001; alcohol vs. abstinence, p = 0.0014) specifically in the DLS, without altering DMS mRNA expression. A trend towards DRD1 downregulated in the DMS was also observed (F _(2,20) = 3.04, p = 0.0704; Bonferroni post hoc naïve vs. alcohol p = 0.0780). (D) M₄ mAChRs are expressed on both direct pathway MSNs and cholinergic interneurons, therefore, to determine if Chrm4 mRNA is downregulated on direct MSNs (Drd1+) or cholinergic interneurons (Chat+) fluorescent in situ hybridization of Chrm4 expression was assessed. In line with RT-qPCR analysis, Chrm4 expression was specifically downregulated in the dorsolateral striatum, restricted to Drd1+ neurons (treatment x region interaction, F_(2,841) = 13.96, p < 0.0001; Bonferroni post hoc analysis, DLS naïve vs. alcohol, p = 0.0001; naïve vs. abstinence, p = 0.0001; alcohol vs. abstinence, p = 0.0591). A ~16% reduction on Chrm4 expression was observed on Drd1+ neurons in the DLS following long-term alcohol consumption, while a ~24% reduction was observed after abstinence. No differences were observed in Chrm4 expression on Drd1+ cells in the DMS or Chat+ cells in either the DMS or DLS. Data expressed as mean ± SEM; n=70-90 Chat+ and 140-150 Drd1+ cell from 5-6 rats/treatment, *p < 0.05, **p < 0.01, ***p < 0.001.

Chrm4 mRNA is specifically downregulated on Drd1+ cells in the DLS following long-term alcohol/abstinence

In the DLS, M₄ mAChRs are expressed postsynaptically on D₁ MSNs and cholinergic (ChAT+) interneurons. To determine the impact of long-term alcohol consumption on Chrm4 expression on specific populations in the dorsal striatum we quantified Chrm4 expression on Chat and Drd1 expressing cells in the DMS and DLS using RNAscope (n=5-6 group/region)(34). Rats consumed an average of 6.2 ± 0.5 g/kg/session with a 61 ± 4.8 % preference for alcohol over water. Chrm4 expression on Chat+ interneurons differed between the DMS and DLS (main effect of region, F_(1,480)= 30.95, p < 0.0001), but no differences were observed between naïve, alcohol or abstinence treatment groups (F_(2,480)= 0.5364, p < 0.5850). In contrast, a significant reduction in Chrm4 expression on Drd1+ cells was observed only in the DLS (main effect of region, F_(1,841)= 83.54, p < 0.0001; treatment, F_{(2, 841)}= 18.27, p < 0.0001; and interaction, F_(2,841)= 13.96, p < 0.0001; DLS naïve vs. alcohol, p = 0.0001; naïve vs. abstinence, p = 0.0001; alcohol vs. abstinence, p = 0.0591; Figure 3D). Chrm4 expression was reduced by ~16% on Drd1+ neurons in the DLS following long-term alcohol consumption, and ~24% following abstinence after long-term alcohol. In line with our observations in individuals with AUD, these data suggest long-term alcohol consumption/abstinence leads to a downregulation of M₄ mRNA expression in the rat DLS, restricted to D₁ expressing dMSNs, with no difference observed in CINs.

Administration of a selective M₄ mAChR positive allosteric modulator decreases alcohol self-administration and cue-induced reinstatement of alcohol seeking in rats

Recent development of subtype selective allosteric modulators have enabled specific examination of mAChRs(37). We used the selective M₄ PAM, VU0467154, to understand the role of M₄ mAChRs in home-cage two-bottle choice and operant alcohol consumption and seeking behaviours. For home cage consumption, VU0467154 was administered 2.5 hours prior to dark phase onset and alcohol/water consumption measured 2, 4, 12 and 24 hours subsequently. Systemic administration of VU0467154 (30 mg/kg, p.o.) significantly reduced alcohol consumption during the dark phase [main effect of time, F_(2,34) = 72.04, p < 0.0001; time x treatment interaction, F_(2,34) = 4.18, p = 0.024; but effect of treatment, F_(1,17) = 1.83 p = 0.195]. Bonferroni post-hoc analysis revealed a significant difference at 12 (p = 0.0137), but not 2 (p = 0.999) or 4 h (p = 0.999) into the dark phase (Supplementary Figure S3D). No difference was observed in alcohol consumption in the light phase (12-24 h) [t = 0.696, df = 17, p = 0.496], or water consumption/alcohol preference within 24 h (Supplementary Figure S3E & F).

During operant self-administration sessions, rats pressed an average of 115 ± 3 (0.66 g/kg/20 min session) on the active lever and 3.5 ± 0.4 times on the inactive lever (Supplementary Figure S4). VU0467154 (10 mg/kg, p.o) failed to reduce operant self-administration (RM two-way ANOVA; main effect of lever, F_(1,7)= 18.88, p < 0.0032; a trend toward main effect of treatment, F_(1,7)= 4.79, p = 0.065; and interaction, F_(1,7)= 0.793, p = 0.079, n=9, data not shown). An increased dose of VU0467154 (30 mg/kg, p.o.) decreased operant responding for ethanol (main effect of lever, F_(1,7)= 62.74, p < 0.0001; treatment, F_(1,7)= 12.16, p = 0.010; and interaction, F_(1,7)= 17.09, p = 0.004; Bonferroni post hoc analysis, vehicle vs. VU0467154, p = 0.0015; n=7, Figure 4B), breakpoint for ethanol on a progressive ratio (t = 2.47, df = 9, p = 0.035; n=10, Figure 4C) and cue-induced reinstatement of alcohol seeking (main effect of lever (F_(1,6)= 16.76, p < 0.0009), treatment (F_(2,12)= 36.53, p = 0.0003) and interaction (F_(2.12)= 13.91, p = 0.0007; extinction vs. vehicle, p = 0.0001, vehicle vs VU0467154, p = 0.0013; n=8, Figure 4D). VU0467154 did not alter latency to first lever response (Supplementary Figure S4), had no effect on sucrose self-administration (main effect of lever, F_(1,14)= 147.6, p < 0.0001; no main effect of treatment, F_(1,14)= 0.418, p = 0.529; or interaction F_(1,14)= 0.295, p = 0.596; n=8, Figure 4E) and did not cause motor impairments at 30 mg/kg (students t-test, t = 0.234, df = 11, p = 0.819, n=6-7/group, Figure 4F). A higher dose of VU0467154 (100 mg/kg, p.o) decreased the time rats spent on the rotarod (t = 2.57, df = 16, p = 0.020, n=9/group, Supplementary Figure S5), therefore we did not test this dose further. Our operant data suggest the M₄ PAM, VU0467154 has the greatest effect in decreasing the incentive properties of alcohol consumption and seeking following long-term alcohol consumption. These results also suggest this does not generalize across all appetitive substances, nor does it alter normal fluid consumption. Finally, these results suggest a therapeutic window for the M₄ PAM to selectively modulate alcohol intake/seeking.

Figure 4. Systemic pharmacological manipulation of the M₄ muscarinic acetylcholine receptor in alcohol consumption and seeking.

To functionally test the role of M₄ mAChRs in alcohol consumption and seeking we systemically administered a selective M₄ positive allosteric modulator, VU0467154, as outlined in (A). Systemic administration of VU0467154 (30 mg/kg, p.o) significantly reduced (B) active lever responding (n=8) in FR3 self-administration (treatment x lever interaction, F_(1,7) = 17.09, p = 0.004; Bonferroni post hoc analysis, vehicle vs. VU0467154, p = 0.0015), (C) Breakpoint (n=10) in the progressive ratio paradigm (t = 2.47, df = 9, p = 0.035), (D) active lever responding (n=8) in cue-induced reinstatement of alcohol seeking (treatment x lever interaction, F_(2.12) = 13.91, p = 0.0007; Bonferroni post hoc, extinction vs. vehicle, p = 0.0001, vehicle vs VU0467154, p = 0.0013). No effect was observed on natural reward consumption during (E) active lever responding (n=8) in sucrose self-administration (F_(1,14) = 0.418, p = 0.529), or (F) ataxia (n=13) determined via performance on the rotarod (t = 0.234, df = 11, p = 0.819). Data expressed as mean ± SEM, *p < 0.05, **p < 0.01, ***p < 0.001.

Dorsolateral striatum M₄ mAChRs regulate alcohol consumption and cue-induced reinstatement of alcohol seeking

Our molecular data suggest the DLS is the main locus of M₄ mAChR dysregulation. Therefore, to determine if DLS M₄ mAChR’s are involved in mediating alcohol self-administration and cue-induced reinstatement of alcohol seeking we directly administered VU0467154 into the DLS. During operant self-administration sessions, rats pressed an average of 85 ± 4 (0.51 ± 0.03 g/kg/20 min session) on the active lever and 2 ± 0.2 times on the inactive lever (Supplementary Figure S5). Intra-DLS microinjections of VU0467154 (600 pmol/hemisphere) had no effect on alcohol self-administration (main effect of lever, F_(1,7)= 65.93, p < 0.00001; but no effect of treatment, F_(1,7)= 0.017, p = 0.897; or interaction F_(1,7)= 0.011, p = 0.918, n=8, data not shown). An increased dose of VU0467154 (3 μmol/hemisphere) decreased alcohol self-administration (main effect of lever, F_(1,11)= 91.45, p < 0.00001; treatment, F_(1,11)= 6.69, p = 0.0253; interaction F_(1,11)= 5.67, p = 0.0364; Bonferroni post hoc, active lever vehicle vs VU0467154, p = 0.0086; n=14, Figure 5B). Peak responding was significantly reduced (main effect of time, F_(3,33)= 8.74, p = 0.0002; treatment F_(1,11)= 6.138, p = 0.0307; and interaction F_(3,33)= 8.52, p = 0.0002. vehicle vs VU0467154 at 5 min, p <0.0001, n=7, Figure 5B) without decreasing latency to first active lever press (main effect of lever, F_(1,11)= 64.96, p < 0.0001; effect of treatment F_(1,11)= 5.459, p = 0.039; but no interaction F_(1,11)= 2.189, p = 0.1671).

Figure 5. Dorsolateral striatum pharmacological manipulation of the M₄ muscarinic acetylcholine receptor in alcohol consumption and seeking.

To determine the role of dorsolateral (DLS) M₄ mAChRs in alcohol consumption and seeking we microinjected the selective M₄ mAChR PAM, VU0467154, into the DLS prior to self-administration (n=12) and cue-induced reinstatement (n=7) as outlined in (A). Intra-DLS VU0467154 (3 μmol/hemisphere) reduced (B) active lever responding in alcohol self-administration (treatment x lever interaction F_(1,11) = 5.67, p = 0.0364; Bonferroni post hoc, active lever vehicle vs VU0467154, p = 0.0086). This reduction occurred during peak responding (treatment x time interaction F_(3,33) = 8.52, p = 0.0002. Bonferroni post hoc analysis, vehicle vs VU0467154 at 5 min, p <0.0001). (C) Intra-DLS VU0467154 (3 μmol/hemisphere) also reduced active lever responding during cue-induced reinstatement of alcohol seeking (treatment x lever, F_(2,12) = 11.22, p = 0.0018; Bonferroni post hoc, extinction vs. vehicle, p = 0.0004, vehicle vs. VU0467154, p = 0.0016). Decreased lever pressing was significantly reduced during peak responding (treatment x time interaction F_(3,18) = 3.729, p = 0.0303. Bonferroni post hoc analysis, treatment vs. vehicle at 5 min, p = 0.0067). (D) Shows the injection sites for VU0467154 in self-administration (light blue) and cue-induced reinstatement (dark blue)(36). Data are expressed as mean ± SEM, **p < 0.01, ****p < 0.0001.

Intra-DLS VU0467154 also decreased cue-induced reinstatement (main effect of lever, F_(1,6)= 122.7, p < 0.00001; treatment, F_(2,12)= 10.34, p = 0.0025; interaction F_(2,12)= 11.22, p = 0.0018; extinction vs. vehicle, p = 0.0004, vehicle vs VU0467154, p = 0.0016; Figure 5C). Peak responding was decreased (main effect of time, F_(3,18) = 3.94, p = 0.0253; treatment F_(1,6)= 7.139, p = 0.0369; and interaction F_(3,18)= 3.729, p = 0.0303, treatment vs. vehicle at 5 min, p = 0.0067, Figure 5C) without decreasing latency to first active lever press (main effect of lever, F_(1,6)= 32.97, p = 0.0012; no effect of treatment F_(1,6)= 0.00007, p = 0.9932; or interaction F_(1,6)= 3.632, p = 0.1053; Supplementary Figure S6).

Discussion

Acamprosate was the last FDA approved medication to treat AUD over 15 years ago. Consequently, there is great need for novel AUD treatments that target core pathophysiology induced by long-term alcohol consumption. Here we employed a cross-species back-translational approach to establish that a history of AUD leads to GPCR dysregulation in the putamen of humans and corresponding DLS of rats. Altered expression of GPCRs implicated preclinically and/or clinically in AUD included metabotropic glutamate 5 receptor (GRM5)(45, 46), cannabinoid 1 receptor (CNR1)(47-49), D₁ dopamine receptor (DRD1)(50, 51), GPR88(52, 53) and M₄ mAChR (CHRM4)(28).

The M₄ mAChR is expressed within mesocorticolimbic circuitry(54, 55). Dopaminergic and cholinergic systems dynamically interact to gate and potentiate sensory inputs to the striatum(56). Endogenous cholinergic activity controls extracellular dopamine levels through a complex convergence of neurotransmitter/neuromodulator systems. Tonically active cholinergic interneurons modulate dopamine output through activation of both mAChR and nAChRs on dopamine terminals(14, 57, 58) and through the corelease of glutamate(59, 60).

Further, within the striatum, M₄ mAChRs are necessary for muscarinic regulation of dopamine release(14), and the M₄ PAM, VU0467154 modulates DA release indirectly through M₄ receptors on D₁ expressing dMSNs(16). Based on this, we pursued M₄ mAChRs to further understand their role in AUD. We validated and expanded our RNA sequencing findings to demonstrate CHRM4 mRNA and M₄ mAChR binding were specifically downregulated in the putamen of individuals with AUD. The putamen/DLS is a key structure executing behavioural control after long-term alcohol consumption in both rodents(10) and humans(11). Chronic exposure to ethanol alters the DLS network, strengthening its control over behaviour(61, 62). In humans, cue-induced drug craving is associated with activation of, and dopamine release in, the DLS(9, 63, 64). Further, DLS dopamine regulates alcohol seeking and an inability to disengage the DLS is associated with vulnerability to develop aberrant alcohol seeking behaviour(65). Our data in combination with our previous study(19) where microinjections of the M₅ NAM, ML375, or α₄β₂ nAChR partial agonist, varenicline, decreased alcohol self-administration within the DLS, but not DMS, suggest this dysregulation predominates in the DLS.

Humans with AUD show persistent aberrant motivation towards alcohol consumption(66-68). The cholinergic system within the dorsal striatum has been linked to motivation for reward(12) and modulating activity of the central dopaminergic system(13). To probe the role of the M₄ mAChR in alcohol consumption and seeking, we back-translated our findings to a rodent model of long-term voluntary alcohol use/abstinence. M₄ mAChR downregulation was recapitulated in the corresponding DLS. Within the striatum, M₄ mAChR are expressed on cholinergic interneurons and direct pathway MSNs (dMSNs). Cholinergic M₄ mAChRs act via negative feedback to regulate cholinergic neuron output, while M₄ mAChRs expressed post-synaptically on D₁ dMSNs modulate their activity(13) as well as indirectly modulating presynaptic dopamine release(16). M₄ mAChRs on these populations exert opposing functions, with deletion from cholinergic interneurons altering drug associated learning tasks, while deletion from D₁-positive neurons enhanced drug seeking behaviour (42). We observed specific downregulation of Chrm4 expression on D₁ dMSNs in the DLS following long-term alcohol use that persisted during abstinence. Repeated cycles of excessive alcohol intake and withdrawal potentiate glutamatergic signalling onto D₁ dMSNs, which in turn drives greater alcohol seeking(69). Therefore, indirectly reducing signalling of D₁ dMSNs via M₄ mAChR activation may provide a nuanced mechanism to re-tune striatal function in relation to aberrant alcohol seeking.

Recent development of subtype selective allosteric modulators has allowed detailed examination of specific behavioural roles of discrete mAChRs in affective disorders (19, 43). A selective M₄ PAM, VU0467154, possesses both cognitive enhancing and antipsychotic-like effects, without tolerance upon repeated dosing in rodents(37, 43). VU0467154 acts via a D₁ dMSN and CB₂ receptor-dependent mechanism to reduce dopamine signalling and mediate antipsychotic-like effects(16). VU0467154 administered either systemically or directly within the DLS reduced alcohol consumption, motivation for alcohol and cue-induced alcohol seeking (relapse). VU0467154 caused no alterations in water or natural reward consumption, motor control or latency to respond. Interestingly, the effect size of intra-DLS administration of VU0467154 to reduce cue-induced reinstatement of alcohol seeking was greater than that on self-administration of alcohol. This suggests a more robust role for striatal M₄ mAChRs in regulating relapse behaviour, consistent with the pronounced downregulation of Chrm4 observed following abstinence after long-term alcohol consumption.

Together our data show that long-term alcohol consumption/AUD downregulates CHRM4 expression in both human and rodent, specifically on D₁ MSNs in the DLS. Enhancement of M₄ mAChR signalling reduces alcohol consumption and seeking, mediated at least in part through the DLS. Overall, these findings provide a strong rationale to further explore M₄ mAChR PAMs as candidates for novel AUD medications. Moreover, allosteric modulation of M₄ mAChRs may provide a strategy that restores cholinergic function in AUD and helps promote abstinence. VU0467154 is not a desirable candidate for clinical development with limited potency at the human M₄ mAChR (70). Attempts to develop M₄ PAMs with desirable efficacy and potency at the human M₄ receptor are ongoing(71, 72). A number of pharmaceutical companies have active M₄ PAM/agonist programs in relation to psychosis, Karuna therapeutics (KarXT), Sosei Heptares (HTL-0016878), Pfizer (Undisclosed PAM; PF-06852231) and Lundbeck/VCNDD (New generation VU PAMs). Together this opens up a genuine avenue for the treatment of AUDs by repurposing a class of molecules for which there are agents in clinical development.

KEY RESOURCES TABLE

Resource Type	Specific Reagent or Resource	Source or Reference	Identifiers	Additional Information
Add additional rows as needed for each resource type	Include species and sex when applicable.	Include name of manufacturer, company, repository, individual, or research lab. Include PMID or DOI for references; use “tdis paper” if new.	Include catalog numbers, stock numbers, database IDs or accession numbers, and/or RRIDs. RRIDs are highly encouraged; search for RRIDs at https://scicrunch.org/resources.	Include any additional information or notes if necessary.
Chemical Compound or Drug	VU0467154 (5-amino-3,4-dimethyl-N-(4-((trifluoromethyl)sulfonyl)benzyl)thieno[2,3-c]- pyridazine-6-carboxamide)	Provided by Craig Lindsley (Vanderbilt University)	N/A
Commercial Assay Or Kit	RNeasy Plus Mini Kit	Qiagen	Cat No. 74136
	SureSelect Strand - Specific RNA Library Preparation kit for Illumina	Aligent	Cat No. G9691A
	SuperScriptII qRT-PCR kit	Invitrogen	Cat No. 10928042
	SYBR Powerup PCR kit	Qiagen	Cat No. 4368702
	RNAscope Multiplex Fluorescent Reagent Kit	Advanced Cell Diagnostics	Cat No. 320850
Deposited Data; Public Database	Basic Local Alignment Search Tool	https://blast.ncbi.nlm.nih.gov/Blast.cgi	N/A
Organism/Strain	Male inbred Indiana preferring (iP) rats	Parental stock from the late Professor T.K. Li (while at Indiana University, USA).	N/A
Sequence-Based Reagent	Primers for RT-qPCR, see Table S6	This paper	N/A
Sequence-Based Reagent	drd1 RNAscope probe	Advanced Cell Diagnostics	Cat No. 317031
Sequence-Based Reagent	chat RNAscope probe	Advanced Cell Diagnostics	Cat No. 430111-C3
Sequence-Based Reagent	chrm4 RNAscope probe	Advanced Cell Diagnostics	Cat No. 456671-C2
Software; Algorithm	Subjunc aligner	- http://subread.sourceforge.net/	version 1.4.4 - http://subread.sourceforge.net/
Software; Algorithm	EdgeR	https://www.bioconductor.org/packages/release/bioc/html/edge.html	version edgeR_3.4.2 in R studio, R version 3.0.2 -
Software; Algorithm	BioMart	https://bioconductor.org/packages/release/bioc/html/biomaRt.html	version 3.10
Software; Algorithm	Partek Genomics suite	https://www.partek.com/partek-genomics-suite/	version 6.16.0419
Software; Algorithm	qbase+	https://www.qbaseplus.com/	version 3.0
Software; Algorithm	Primer3	http://bioinfo.ut.ee/primer3-0.4.0/	version 0.4.0
Software; Algorithm	GraphPad Prism	https://www.graphpad.com/	version 7 & 8
Software; Algorithm	Image J	NIH	version 2.0.0
Software; Algorithm	MEDPC-IV	Med Associates	version 4

Abbreviation list:

AcbC

nucleus accumbens core

AcbSh

nucleus accumbens shell

ACh

acetylcholine

AUD

alcohol use disorder

BMI

body mass index

CHAT

Choline Acetyl Transferase

CHRM4

M₄ muscarinic acetylcholine receptor mRNA

CNR1

cannabinoid receptor 1 mRNA

DEG

differentially regulated genes

DLS

dorsolateral striatum

DMS

dorsomedial striatum

dMSN

direct pathway medium spiny neurons

DMSO

dimethyl sulfoxide

DRD1

dopamine D₁ receptor mRNA

DRD2

dopamine D₂ receptor mRNA

DRD3

dopamine D₃ receptor mRNA

EIF2A

Eukaryotic Translation Initiation Factor 2A

FISH

fluorescent in situ hybridisation

FR3

fixed ratio 3

GPCR

G-protein coupled receptor

GPR88

G-protein coupled receptor 88 mRNA

GPR4

G-protein coupled receptor 4 mRNA

GPR6

G-protein coupled receptor 6 mRNA

HMBS

hydroxymethylbilane synthase

HSP90

heatshock protein 90

IPA

Ingenuity pathway analysis

iP rat

Indiana preferring rat

qRT-PCR

quantitative Real Time PCR

mAChR

muscarinic acetylcholine receptor

MT-3

muscarinic toxin – 3

MT-7

muscarinic toxin – 7

nAChR

nicotinic acetylcholine receptor

NMS

N-methylscopolamine

PAM

positive allosteric modulator

PMI

post-mortem interval

PPIA

Peptidylprolyl Isomerase A

RIN

RNA integrity number

RM

repeated measures

TBP

TATA binding box protein

Data Availability

Materials, data sets and protocols are available upon reasonable request to the corresponding authors.