Ventral Tegmental Area M5 Muscarinic Receptors Mediate Effort-Choice Responding and Nucleus Accumbens Dopamine in a Sex-Specific Manner 

Abstract

Optimization of effort-related choices is impaired in depressive disorders. Acetylcholine (ACh) and dopamine (DA) are linked to depressive disorders, and modulation of ACh tone in the ventral tegmental area (VTA) affects mood-related behavioral responses in rats. However, it is unknown if VTA ACh mediates effort-choice behaviors. Using a task of effort-choice, rats can choose to lever press on a fixed-ratio 5 (FR5) schedule for a more-preferred food or consume freely available, less-preferred food. VTA administration of physostigmine (1 μg and 2 μg/side), a cholinesterase inhibitor, reduced FR5 responding for the more-preferred food while leaving consumption of the less-preferred food intact. VTA infusion of the M5 muscarinic receptor negative allosteric modulator VU6000181 (3 μM, 10 μM, 30 μM/side) did not affect lever pressing or chow consumption. However, VU6000181 (30 μM/side) coadministration with physostigmine (2 μg/side) attenuated physostigmine-induced decrease in lever pressing in female and male rats and significantly elevated lever pressing above vehicle baseline levels in male rats. In in vivo voltammetry experiments, VTA infusion of combined physostigmine and VU6000181 did not significantly alter evoked phasic DA release in the nucleus accumbens core (NAc) in female rats. In male rats, combined VTA infusion of physostigmine and VU6000181 increased phasic evoked DA release in the NAc compared with vehicle, physostigmine, or VU6000181 infusion alone. These data indicate a critical role and potential sex differences of VTA M5 receptors in mediating VTA cholinergic effects on effort choice behavior and regulation of DA release.

SIGNIFICANCE STATEMENT

Effort-choice impairments are observed in depressive disorders, which are often treatment resistant to currently available thymoleptics. The role of ventral tegmental area (VTA) acetylcholine muscarinic M5 receptors, in a preclinical model of effort-choice behavior, is examined. Using the selective negative allosteric modulator of the M5 receptor VU6000181, we show the role of VTA M5 receptors on effort-choice and regulation of dopamine release in the nucleus accumbens core. This study supports M5 receptors as therapeutic targets for depression.

Introduction

Mood disorders, including major depressive disorder, are mental illnesses that affect a significant portion of the US population, with significant increases in reported cases during the COVID-19 pandemic (Bueno-Notivol et al., 2021). Symptoms of depression can include increases in negative affective states and a loss of positive mood, which often coexist with effort-related choice impairments (Nutt et al., 2007; Treadway et al., 2012). Moreover, impairments in effort-choice behavior are also associated with symptoms such as fatigue and exhaustion without physical strain or exertion (Stahl, 2002; Salamone and Correa, 2012; Ghanean et al., 2018). Clinically, persons suffering from major depressive disorder will exert less work and to obtain a higher effort, larger monetary reward and instead shift to a lower-effort option for a lower monetary reward compared with control subjects (Treadway et al., 2012). Furthermore, these effort-related symptoms, like fatigue, can be treatment resistant and further exacerbated by common thymoleptics such as selective serotonin uptake inhibitors (Shelton et al., 2001; Demyttenaere et al., 2005; Nutt et al., 2007; Padala et al., 2012).

Preclinical models of effort-choice have revealed a critical role of dopamine (DA), particularly in the nucleus accumbens (NAc), in the regulation of effort-choice behaviors (Berridge and Robinson, 1998; Bardgett et al., 2009; Salamone and Correa, 2012; Schultz et al., 2015; Salamone et al., 2018). Depleting DA or blockade of DA receptors in the NAc shifts effort-choice behavior, biasing it toward less-preferred, low-effort options (Salamone et al., 1991; Cousins and Salamone, 1994; Mott et al., 2009; Nunes et al., 2010, 2013a, 2013b; Yohn et al., 2016). In contrast, pharmacological manipulations that increase NAc DA attenuate effort-choice impairments and facilitate the selection of high-effort choice options (Nunes et al., 2013b; Randall et al., 2014; Yohn et al., 2016). These preclinical findings demonstrate the important role of NAc DA transmission in rodent models of effort-choice behavior. Furthermore, these findings have informed and are consistent with clinical studies, which show the crucial role of DA in effort-choice tasks in humans (Treadway et al., 2009; Wardle et al., 2011; Treadway et al., 2012; Soder et al., 2021). Therefore, brain mechanisms that regulate DA release are attractive target candidates for novel therapeutic development for effort-related impairments in depressive disorders.

Acetylcholine signaling, via nicotinic acetylcholine receptors (nAChRs) and muscarinic acetylcholine receptors (mAChRs) in the ventral tegmental area (VTA), is a critical regulator of DA release in the NAc (Galaj and Ranaldi, 2021). Moreover, increased cholinergic tone in humans and rodents is linked to the etiology of depression (Janowsky et al., 1974; Addy et al., 2015; Small et al., 2016; Dulawa and Janowsky, 2019). Thus, acetylcholine signaling, in part through regulation of the DA system, mediates mood-related behaviors. Indeed, increasing cholinergic tone in the VTA with the cholinesterase inhibitor physostigmine produces prodepressive and anxiogenic-like behavioral responses in female and male rats as measured by the sucrose preference test (SPT), elevated plus maze (EPM), and forced swim test (FST) (Addy et al., 2015; Small et al., 2016; Nunes et al., 2020). In contrast, blockade of VTA mAChRs and nAChRs produces antidepressant and anxiolytic-like behavioral responses in rodent models (Addy et al., 2015). We have also demonstrated the critical role of the mAChR M5 subtype in mediating behavioral responses on the SPT, EPM, and FST. Specifically, administration of the M5 mAChR subtype negative allosteric modulator (NAM) VU6000181 attenuates the ability of physostigmine in the VTA to produce prodepressive and anxiogenic behavioral responses in female and male rats (Nunes et al., 2020). These findings suggest that VTA M5 receptors are mediating the prodepressive and anxiogenic-like behaviors in the context of increased cholinergic tone in female and male rats.

Despite the evidence supporting VTA cholinergic regulation of depressive and anxiety-like behaviors, as measured by the SPT, EPM, and FST, the role of VTA AChRs in regulating behavioral tasks of effort-choice are unknown. Based on our previous data demonstrating the role of VTA acetycholine (ACh) signaling, specifically M5 mAChRs, on the behavioral responses on the SPT, EPM, and the FST, we hypothesized that increasing VTA ACh would induce effort-choice impairments, biasing rats toward a lower-effort, less-desirable choice. Further, we hypothesized that this effect would be mediated by VTA M5 receptor mechanisms with downstream regulation of NAc (NAc) DA release. Our data supports our hypothesis of VTA M5 receptors mediating effort-choice responding in male and female rats. We also reveal sex differences in the VTA M5 mAChRs regulation of effort-choice behavior and DA release in the NAc.

Materials and Methods

Animals

Adult male (n = 123) and female (n = 124) Sprague-Dawley rats (Charles River Laboratories, Wilmington, MA), weighing 259–279 g and 179–199 g, respectively, upon arrival, were used for the behavioral and voltammetry experiments outlined below. Rats were pair housed in standard home cages under a 12-hour light/dark cycle with climate control between 68 and 74°F. Behavioral training began after animals acclimated to the housing facility for 5–7 days. Animals were food restricted the day before their training sessions and provided supplemental chow (∼20 g/day for male rats and ∼16 g/day for female rats) in their home cages after the sessions to maintain a positive growth curve. Weights were taken throughout the behavioral training (Supplemental Fig. 2). Following intracranial surgery and throughout the behavioral experiments, any animals with cannula misplacements or clogged cannulas were excluded from the final analyses (male rats n = 10, female rats n = 17). All animal handling and procedures were performed per ethical guidelines and were approved by the Institutional Animal Care and Use Committee at Yale University (#2022-11366).

Behavioral Training

Behavioral training and assessments were conducted in 30-minute sessions in operant conditioning chambers (28 × 23 × 23 cm; Med Associates Inc., Fairfax, VT) with an active lever press paired with delivery of 45-mg, high-carbohydrate, nutritionally complete pellets (Bio-serv, Flemington, NJ). Rats were trained to lever press on a fixed-ratio 1 schedule(7–10 training sessions, 5 days/week) then moved to a fixed-ratio 5 (FR5) schedule (15–20 training sessions, 5 days/week). Once lever pressing stabilized (<15% variability to prior three training trials) on the FR5 schedule, 15–20 g of nutritionally complete laboratory chow (5P00 Laboratory Diet; Prolab or RHM 3000; Purina Mills, Gray Summit, MO) was made freely available in the operant chamber to introduce the choice between chow and 45 mg high-carbohydrate pellets. Rats were trained on the concurrent FR5/chow-feed procedure (15–20 training sessions) until stable lever pressing and chow intake were achieved (<15% variability to prior three training sessions). Chow consumption was calculated as the difference in weight of the chow before and immediately after each training session.

Intracranial Cannula Placement Surgery

In preparation for surgery, rats were anesthetized with ketamine HCl (100 mg/kg i.p.; Covetrus, Portland, ME) and xylazine (10 mg/kg i.p.; Henry Schein, Queens, NY) and placed in a stereotaxic frame (David Kopf Instruments, Tujunga, CA). Anesthesia was followed by subcutaneous administration of the long-acting nonsteroidal anti-inflammatory drug carprofen (5 mg/kg s.c.; Covetrus, Portland, ME). After anesthetic and analgesic administration, the scalp was shaved and swabbed with alternating 70% ethanol and iodine solution applications (3×) and then incised with a sterile scalpel. Following surgical incision, small holes were drilled in the skull, and bilateral cannulas (P1 technologies, Roanoke, VA), with the cannulas spaced 1 mm apart, were inserted 1 mm above the VTA [A/P: −5.2 to −5.5 mm (adjusted for weight) from bregma; ML: +/− 0.5 mm from bregma; and DV: −8.1 mm from the skull]. All coordinates were obtained from the rat brain atlas (Paxinos and Watson, 2007). To secure the cannula in place, three to four skull screws (Gexpro, High Point, NC) were placed around the cannula, and dental cement (Dentsply, Milford, DE) was applied over the incision area enclosing the cannula. Postsurgery, rats were single housed and given daily carprofen (5 mg/kg s.c.) for 48 hours. All rats were given 5–7 days for full recovery before behavioral training resumed.

Pharmacological Agents

To increase VTA cholinergic tone, the cholinesterase inhibitor physostigmine (VMR, Bridgeport, NJ), prepared in a solution of 21.1% β-cyclodextrin, 5% DMSO, and saline, was infused via the bilateral cannulas. The physostigmine doses selected were previously shown to alter rodent behavior when infused into the VTA (Zhou et al., 2007; Addy et al., 2015; Small et al., 2016). The M5 NAM VU6000181, ((S)-9b-(4-methoxy-3-methylphenyl)-1-(3,4,5-trifluorobenzoyl)-1,2,3,9b-tetrahydro-5H-imidazo[2,1-a]isoindol-5-one, first described in Kurata et al. (2015), was dissolved in a 21.1% β-cyclodextrin and 5% DMSO and saline mix and infused into the VTA. VU6000181 has high selectivity at high nanomolar and low micromolar concentrations (rat IC₅₀ = 516 nM) and remains selective for the M5 receptor at concentrations up to 30 μM (Kurata et al., 2015). Therefore, VU6000181 was administered up to 30 μM/side. For coadministration experiments, VU6000181 and physostigmine were mixed as a single solution.

Drug Administration

To allow for intracranial VTA drug administration, animals were held by a trained experimenter, and a bilateral internal injector (for behavioral experiments) extending 1 mm beyond the guide cannula (∼9 mm ventral from the skull) was inserted. A total volume of 0.3 µL of the drug solution per injection site was infused over 2 minutes via a microinfusion pump and syringe (25 gauge; Hamilton Syringe, Reno, NE). The internal cannula was left in place for an additional 1 minute to allow for complete drug diffusion into the VTA.

Combined Infusion and Stimulation with Fast-Scan Cyclic Voltammetry

Evoked DA release in the NAc core was measured using fast-scan cyclic voltammetry (FSCV) in anesthetized female and male rats. A stainless steel bipolar stimulating electrode attached to a unilateral cannula (P1 technologies, Roanoke, VA) was used to stimulate the VTA and allowed for simultaneous drug infusion at the stimulation site. For drug infusion, an internal injector that extended 1 mm beyond the guide cannula (∼9 mm ventral from the skull) was inserted, and the drug was infused into the VTA. A carbon fiber microelectrode was used to measure evoked DA release in the NAc core. Carbon fiber microelectrodes were prepared using glass capillaries (Stoelting, Wood Dale, IL) with carbon fiber (T650, Amoco, Greenville, SC) aspirated inside and pulled with an electrode puller (Narishige Group, Tokyo, Japan). The microelectrodes had diameters of 7 µm at the tip, and the protruding carbon fiber was cut to a length of ∼75–100 µm beyond the tip under a light microscope (Olympus, Center Valley, PA) with a scalpel blade. A reference electrode was prepared from a silver wire coated with chloride. New carbon fiber electrodes, VTA stimulation electrodes, and reference electrodes were prepared for each animal. To prepare for voltammetric recordings, rats were anesthetized with 1.5 g/kg urethane (i.p.) (Sigma-Aldrich, St. Louis, MO) and placed in a stereotaxic frame (David Kopf Instruments, Tujunga, CA). Four holes were drilled into the skull to place the stimulating electrode (A/P −5.2 to −5.5 mm, M/L +0.5 to +0.7 from bregma), the working electrode (A/P +1 mm, M/L +1.3 to +1.4 mm from bregma), the reference electrode (on the contralateral side of the brain at a distance from both the working and stimulating electrode), and a screw to secure the reference electrode in place. A silver lead wire was inserted into the carbon fiber microelectrode following a KCl solution injection to lubricate the microelectrode and allow for a better connection between the lead wire and the carbon fiber. The carbon fiber microelectrode was initially lowered about −4.5 mm into the brain from where the electrode’s waveform was first observed. A triangular waveform (voltage going from −0.4 V to 1.3 V and back to −0.4 V, applied every 100 milliseconds at a rate of 400 V/s) was used at a frequency of 60 Hz for 15–20 minutes to increase the electrode’s sensitivity to dopamine through oxidative etching of the carbon fiber surface (Takmakov et al., 2010). Then, the waveform was changed to 10 Hz and lowered into the NAc core (D/V −5.5 to −6.2). The cyclic voltammograms were visualized and analyzed using Demon Voltammetry and Analysis software (Yorgason et al., 2017). The stimulation input was relayed via a biphasic stimulus isolator (DS4, Digitimer, Hertfordshire, England) to the stimulating electrode lowered into the VTA (D/V −7.0 to −8.0 mm). A 300-µA, 24 biphasic pulse stimulation was applied at a frequency of 60 Hz to stimulate burst DA firing in the VTA. The dorsal to ventral positions of the stimulating and carbon fiber working electrode were optimized until a time-locked spike in current was observed at the start of the stimulation along with the characteristic cyclic voltammogram of dopamine (an oxidation peak at 0.6 V and a reduction peak at −0.3 V) at the peak current measured. Baseline recordings of maximally evoked DA release, corresponding to peak evoked DA, were collected 30 minutes before VTA drug infusion. DA traces taken after drug infusion into the VTA (VTA stimulation every 3 minute) were collected for 60 minutes.

Histology

Following behavioral experiments, rats were given a lethal dose of sodium pentobarbital (≥150 mg/kg i.p.). Prior to brain removal, injectors were reinserted into the guide cannula, and 0.3 μL of Chicago blue dye (Sigma-Aldrich, St. Louis, MO) was infused into the VTA. Brains were then removed from the skull and placed in 4% paraformaldehyde. Coronal sections (40 μM) were collected on a freezingstage microtome (Leica Microsystems, Bannockburn, IL) and mounted on glass microscope slides for observation under a light microscope. Rats with a misplaced cannula outside of VTA or significant tissue damage were excluded from statistical analyses of the behavioral data. For voltammetry experiments in anesthetized rats, carbon fiber microelectrode placements were verified by applying a stepwise 10-μA current to the carbon fiber microelectrode, causing a small lesion at the recording site. Rats were then given a lethal dose of pentobarbital (≥150 mg/kg i.p.), and the brain was removed and stored in 3.2% formalin (Sigma-Aldrich, St. Louis, MO) for 48–72 hours. Next, brain tissue was sliced into 50-μM sections on a stage-freezing microtome (Leica Microsystems, Wetzlar, Germany), and lesions were assessed under a light microscope.

Statistical Analysis

Statistical tests were performed using Graphpad Prism 9 (San Diego, CA) and SPSS Version 26 (IBM, Armonk, NY). For all behavioral measures (lever pressing and chow consumption), the two sessions prior to drug infusion were averaged to provide baseline data to compare with post-VTA infusion. Thus, data were normalized to baseline for comparison. Dose analyses for physostigmine and VU6000181 were analyzed using a one-way ANOVA. If a significant main effect was observed, Bonferroni post hoc tests were used to compare different treatment groups. In experiments where both drugs were infused into the VTA, a factorial ANOVA was used, where each drug was an independent categorical factor (physostigmine and VU6000181). Main effects and interactions were used to compare different treatment groups, and when significant results were found, Bonferroni post hoc comparisons were used for additional comparisons between groups. FSCV electrochemical recordings were analyzed using a linear mixed model. The covariate structure was explored, and the best-fit model was chosen using the lowest Akaike information criterion value. The best-fit covariate structure was Toeplitz: heterogeneous for the female FSCV data, and the best fit for the male FSCV data was unstructured. When significant interactions were found, Bonferroni post hoc comparisons were used for additional comparisons between groups. For phasic evoked DA release in female and male rats, time was the repeated factor, and the VTA drug served as another factor.

Results

Increased VTA Cholinergic Tone Decreased FR5 Lever Pressing While Leaving Chow Consumption Intact in Female and Male Rats

Female and male rats were trained on an FR5/chow choice task where they had the option to lever press for a more-preferred food pellet or consume the freely and concurrently available chow present in the chamber (Fig. 1, A and B). During the training sessions, a statistically significant interaction was observed between sex, lever pressing [F(47,3290) = 27.60; P < 0.0001; Supplemental Fig. 1A], and chow consumption [F(20,1400) = 12.78; P < 0.0001; Supplemental Fig. 1B]. Male rats significantly lever pressed and consumed more chow during training compared with female rats. Intra-VTA administration of physostigmine decreased lever pressing in female rats as demonstrated by a significant main effect of drug treatment [F(2,30) = 15.30; P = 0.0001; Fig. 1C]. Post hoc comparisons revealed a significant difference in lever pressing in rats that received a 1-µg or 2-µg physostigmine dose compared with vehicle (t test with a Bonferroni correction, P = 0.03, P = 0.008). VTA physostigmine did not affect consumption of the freely available laboratory chow in female rats as revealed by a lack of a main effect drug [F(2,30) = 1.58; P = 0.15, not significant (n.s.); Fig. 1D]. Representative placement of cannula injection sites Fig. 1E. Next, male rats were trained on an FR5/chow choice task as described in Behavioral Training (Fig. 2, A and B). A significant main effect of the drug on lever pressing was also observed in male rats given intra-VTA administration of physostigmine [F(2,26) = 5.30; P = 0.031; Fig. 2C]. Post hoc comparisons revealed a significant difference in lever pressing in rats that received a 2-µg physostigmine dose (t test with a Bonferroni correction, P = 0.030) compared with vehicle. Consistent with the female data, VTA physostigmine did not affect consumption of the freely available laboratory chow as demonstrated by a nonsignificant main effect of drug [F(2,26) = 1.65; P > 0.211, n.s.; Fig. 2D]. Our findings demonstrate that increasing cholinergic tone in the VTA with physostigmine selectively reduced lever pressing for the preferred food in male and female rats without reducing intake of the concurrent and freely available food.

Fig. 1.

Effect of intra-VTA physostigmine reduces FR5 responding on a task of effort-choice in female rats. (A) Representative lever pressing training data; lever pressing decreases when chow is introduced. (B) Representative chow consumption training data; x-axis starts on training day when chow is first introduced in the chamber. (C) Intra-VTA physostigmine on lever pressing (0 µg, n = 11; 1 µg, n = 11; 2 µg, n = 11) expressed as (%) change of baseline from mean of last two training sessions plus or minus standard error of the mean. **Significant difference from control (P < 0.05); ***significant difference from control (P < 0.01). (D) Intra-VTA physostigmine on chow consumption expressed as (%) change of baseline from mean of last two training sessions plus or minus standard error of the mean. (E) Representative placement of injection sites.

Fig. 2.

Effect of intra-VTA physostigmine reduces FR5 responding on a task of effort-choice in male rats. (A) Representative lever pressing training data; lever pressing decreases when chow is introduced. (B) Representative chow consumption training data; x-axis starts on training when chow is first introduced in the chamber. (C) Intra-VTA physostigmine on lever pressing (0 µg, n = 9; 1 µg, n = 9; 2 µg, n = 11) expressed as (%) change of baseline from mean of last two training sessions plus or minus standard error of the mean. **Significant difference from control (P < 0.05). (D) Intra-VTA physostigmine on chow consumption expressed as (%) change of baseline from mean of last two training sessions plus or minus standard error of the mean.

Administration of the Negative Allosteric Modulator of the M5 Receptor VU6000181 into the VTA Did Not Affect Lever Pressing and Chow Consumption in Female and Male Rats

A separate cohort of female and male rats were trained on the FR5/chow choice task as described above. To assess the role of VTA M5 receptors on effort-choice behavior, the M5 selective NAM VU6000181 was infused immediately before the testing session in both female and male rats. VU6000181 did not significantly alter female behavior as revealed by a lack of a significant main effect of drug on lever pressing [F(3,30) = 1.97; P = 0.212, n.s.; Fig. 3A] and chow intake [F(3,30) = 1.73; P = 0.319, n.s.; Fig. 3B]. Consistent with females, data from the males showed a nonsignificant main effect of drug on lever pressing [F(3,33) = 2.50; P = 0.113 n.s.; Fig. 3C] and chow intake [F(3,33) = 1.454; P = 0.325, n.s.; Fig. 3D]. Taken together, this data demonstrates that VTA infusion of VU6000181 did not affect lever pressing or chow consumption on the FR5/chow choice task.

Fig. 3.

Effect of intra-VTA VU6000181 in female and male rats. (A) Intra-VTA VU6000181 on lever pressing in female rats (0 µM, n = 9; 3 µM, n = 9; 10 µM, n = 8; 30 µM, n = 8) expressed as (%) change of baseline from mean of last two training sessions plus or minus standard error of the mean. (B) Intra-VTA VU6000181 on chow consumption in female rats expressed as (%) change of baseline from mean of last two training sessions plus or minus standard error of the mean. (C) Intra-VTA VU6000181 on lever pressing in male rats (0 µM, n = 9; 3 µM, n = 10; 10 µM, n = 8; 30 µM, n = 10) expressed as (%) change of baseline from mean of last two training sessions plus or minus standard error of the mean. (D) Intra-VTA VU6000181 on chow consumption in male rats expressed as (%) change of baseline from mean of last two training sessions plus or minus standard error of the mean.

VU6000181 Coadministered with Physostigmine into the VTA Attenuated the Decrease in FR5 Lever Pressing in Female and Male Rats

Next, we examined the effect of coadministration of physostigmine and VU6000181 on effort-related choice to test our hypothesis that M5 receptors mediate the effort-related behavioral effects of VTA physostigmine. A separate cohort of female and male rats were trained on the FR5/chow choice task as described above. In female rats, there was a significant main effect of physostigmine [F(1,32) = 33.85; P = 0.0007; Fig. 4A] but no significant main effect of VU6000181 [F(1,32) = 1.08; P = 0.310, n.s.; Fig. 4A] on lever pressing. There was also a significant interaction between physostigmine and VU6000181 on lever pressing [F(1,32) = 7.570; P = 0.001; Fig. 4A]. Post hoc comparisons revealed a significant difference between physostigmine 2 µg versus vehicle (P = 0.011), replicating the decrease in lever pressing following physostigmine administration seen in Fig. 1C. In addition, post hoc analyses revealed a significant difference in lever pressing between physostigmine (2 µg) and VU6000181 (30 µM) alone, as well as physostigmine (2 µg) plus VU6000181 (30 µM) (P = 0.013 for both), demonstrating the ability of VU6000181 to attenuate physostigmine-induced decreases in lever pressing in female rats. For chow consumption, there was a nonsignificant main effect of physostigmine [F(1,32) = 1.56; P = 0.401, n.s.; Fig. 4B] or VU 6000181 [F(1,32) = 0.30; P = 0.811, n.s.; Fig. 4B]. In male rats, there was a significant main effect of physostigmine [F(1,34) = 39.45; P = 0.011; Fig. 4C] but not VU6000181 [F(1,34) = 1.01; P = 0.941, n.s.; Fig. 4C] on lever pressing. There was also a significant interaction of physostigmine and VU6000181 on lever pressing [F(1,34) = 41.710, P = 0.010; Fig. 4C]. Post hoc comparisons revealed a significant difference between physostigmine (2 µg) versus vehicle (P = 0.001), replicating the decrease in lever pressing following physostigmine administration seen in Fig. 2C. Also, like the female data in Fig. 4A, post hoc comparisons showed a significant difference between physostigmine (2 µg) and physostigmine (2 µg) plus VU6000181 (30 uM) (P = 0.032), demonstrating the ability of VU6000181 to attenuate physostigmine-induced decreases in lever pressing in male rats as well. Post hoc comparisons also revealed a difference between vehicle versus physostigmine (2 µg) plus VU6000181 (30 µM) (P = 0.007), demonstrating the ability of coadministration of physostigmine plus VU6000181 to increase further FR5 responding compared with vehicle in male but not female rats. For chow consumption, there was a lack of a significant main effect of physostigmine [F(1,34) = 0.94; P = 0.602; n.s.] or VU 6000181 [F(1,34) = 0.35; P = 0.913, n.s.; Fig. 4D].

Fig. 4.

Effect of intra-VTA physostigmine plus VU6000181 on effort-choice in female and male rats. (A) Intra-VTA physostigmine plus VU6000181 on lever pressing in female rats (0 µg physo + 0 µM VU, n = 8; 2 µg physo + 0 µM VU, n = 10; 0 µg physo + 30 µM VU, n = 9; 2 µg physo + 30 µM VU, n = 8) expressed as (%) change of baseline from mean of last two training sessions plus or minus standard error of the mean. **Significant differences between the control group and physostigmine alone group and the physostigmine alone group and VU6000181 alone and physostigmine plus VU6000181 group. (B) Intra-VTA physostigmine plus VU6000181 on chow consumption in female rats expressed as (%) change of baseline from mean of last two training sessions plus or minus standard error of the mean. (C) Intra-VTA physostigmine plus VU6000181 on lever pressing in male rats (0 µg physo + 0 µM VU, n = 8; 2 µg physo + 0 µM VU, n = 9; 0 µg physo + 30 µM VU, n = 9; 2 µg physo + 30 µM VU, n = 9) expressed as (%) change of baseline from mean of last two training sessions plus or minus standard error of the mean. **Significant differences between the control group and physostigmine alone group. ***Significant difference between the control group and physostigmine alone group and physostigmine plus VU6000181 group. (D) Intra-VTA physostigmine plus VU6000181 on chow consumption in female rats expressed as (%) change of baseline from mean of last two training sessions plus or minus standard error of the mean. Physo, physostigmine.

Differential Effects of VTA VU6000181 and Physostigmine Administration on Evoked Dopamine Release in the NAc Core of Female and Male Rats

We examined whether a behaviorally effective dose of physostigmine and VU6000181 infused into the VTA would also regulate evoked DA release, specifically in the NAc core in female and male rats. Evoked phasic dopamine signaling in the NAc core was examined following phasic-like (300 μA, 24 pulses, 60 Hz) stimulation of the VTA. Linear mixed-model analysis revealed that VTA infusion of physostigmine (2 µg), VU6000181 (30 µM), or physostigmine (2 µg) plus VU6000181 (30 µM) did not alter phasic DA release in female rats as shown by a lack of a significant main effect of VTA drug infusion [F(3,12.169) = 3.001; P = 0.533; Fig. 5A] and time [F(20,201.23) = 1.12; P = 0.500; Fig. 5A]. Representative color plots from the scan at minute 12 are shown in Fig. 5B. In male rats, linear mixed-model analysis revealed significant main effect of VTA drug infusion [F(3,25.38) = 9.198; P = 0.008; Fig. 6A], a significant main effect of time [F(20,233.746) = 107.635; P = 0.010; Fig. 6A], and a significant VTA drug by time interaction [F(60,235.044) = 1.43; P = 0.005; Fig. 6A]. Post hoc comparisons revealed a significant difference between physostigmine (2 µg) plus VU6000181 (30 µM) versus physostigmine (2 µg) alone, VU6000181 (30 µM) alone, and vehicle infusion at time points 3–39, 54, and 60; (P = 0.001). Representative color plots from the scan at minute 12 are shown in Fig. 6B. This demonstrates that combined VTA infusion of physostigmine and VU6000181 significantly increased phasic evoked DA release in the NAc core compared with physostigmine (2 µg) alone, VU6000181 (30 µM) alone, and vehicle infusion.

Fig. 5.

In vivo fast-scan cyclic voltammetry in female rats. (A) Intra-VTA infusion of cholinergic agents and evoked DA release represented as % baseline (vehicle, n= 5; 2 µg physo n = 5; 30 µM VU, n = 5; 2 µg physo + 30 µM VU, n = 6). (B) Representative color plots from the 12-minute recording. Physo, physostigmine; VU, VU6000181.

Fig. 6.

In vivo fast-scan cyclic voltammetry in female rats. (A) Intra-VTA infusion of cholinergic agents and evoked DA release represented as % baseline (vehicle, n = 5; 2 µg physo, n = 6; VU, n = 5; 2 µg physo + 30 µM VU, n = 5) *Significant difference between physostigmine plus VU6000181, vehicle, physostigmine alone, and VU6000181 alone. (B) Representative color plots from the 12-minute recording. Physo, physostigmine; VU, VU6000181.

Discussion

In female and male rats, intra-VTA infusion of physostigmine decreased FR5 lever pressing for preferred food while leaving consumption of the less-preferred food intact. VU6000181 infused into the VTA did not affect lever pressing or chow consumption. In female rats, when VU6000181 was coadministrated with physostigmine, it attenuated the decrease in lever pressing compared with physostigmine alone. VTA administration of either physostigmine, VU6000181, or physostigmine plus VU6000181 did not alter evoked DA release compared with vehicle. In male rats, coadministration of physostigmine with VU6000181 attenuated the behavioral effects of physostigmine alone. In contrast to females, VTA infusion of physostigmine plus VU6000181, which significantly increased lever pressing in male rats, significantly increased phasic evoked DA release in the NAc. This indicates that VTA M5 receptors mediate cholinergic regulation of effort-choice responses in female and male rats, particularly in the context of increased cholinergic tone. Furthermore, FSCV data suggests sex-specific differences of VTA cholinergic regulation of DA release in NAc.

Midbrain DA neurons receive cholinergic innervation from laterodorsal tegmental and pedunculopontine tegmental nuclei in brainstem, and cholinergic transmission in the midbrain modulates DA neuron activity and release (Forster and Blaha, 2000; Wang and Morales, 2009). Evidence demonstrates a critical role of M5 receptors, which are preferentially expressed in the midbrain (Vilaró et al., 1990; Weiner et al., 1990; Yeomans et al., 2001; Steidl et al., 2011). M5 receptors can bidirectionally regulate DA activity depending on brain region. In vitro activation of M5 receptors in the substantia nigra increased DA neuron activity, whereas activation of M5 receptors in the striatum inhibited DA release (Foster et al., 2014). Recently, we found that VTA administration of the M5 mAChR negative allosteric modulator VU6000181 attenuated the ability of physostigmine to induce depressive and anxiogenic-like behavioral responses in rats (Nunes et al., 2020). These findings led to our hypothesis that increased cholinergic signaling via M5 receptors in the VTA alter effort-choice behaviors via regulation of DA signaling in the NAc. Our observed findings of VU6000181 attenuation of physostigmine-induced decrease in effort-choice responding, in combination with our previous findings, reveals M5 receptor mechanisms mediating the effects of increased VTA cholinergic tone across various stress-, anxiety-, and effort-related behaviors. (Nunes et al., 2020). Importantly, these new data demonstrate that increased VTA cholinergic tone can produce depression-like behavior across a variety of preclinical tests, including those designed specifically to measure effort-choice behavior (Salamone and Correa, 2012; Nunes et al., 2022). Furthermore, the ability of VU6000181 to attenuate prodepressive behaviors supports the hypothesis that increased cholinergic activity on M5 receptors in the VTA promotes a prodepressive phenotype.

Effort-related choice behaviors are sensitive to DA manipulations, particularly in the NAc (Salamone and Correa, 2012). Increased extracellular DA levels in the NAc are correlated with increased effortful operant responding in rats, whereas a decrease is correlated with reduced effortful responding (Nunes et al., 2013b; Randall et al., 2014). The exertion and initiation of effortful lever pressing has been shown to be dependent on DA signaling in the NAc (Ko and Wanat, 2016). Thus, combining VTA drug infusion of physostigmine and VU6000181 with in vivo FSCV, we examined the effects on phasic evoked DA release in the NAc in female and male rats. Since infusion of physostigmine into the VTA reduced effortful responding in female and male rats, we hypothesized that physostigmine would reduce phasic evoked DA release in the NAc, consistent with prior evidence of the role of NAc DA in effort-choice behavior. Yet, we were also aware of the evidence correlating enhanced VTA acetylcholine activity with VTA DA burst firing and NAc DA release (Grace et al., 2007; Maskos, 2008). There were no significant effects of VTA physostigmine on evoked DA release in the NAc in females or males. It is plausible that infusion of physostigmine into the VTA may alter phasic evoked DA release in other terminal regions, including the anterior cingulate cortex, which is also implicated in effort-choice behavior (Hart et al., 2017; Hart et al., 2020).

Next, we examined the role of VTA M5 receptors on phasic evoked DA release in the NAc by infusing the selective M5 receptor NAM compound VU6000181 into the VTA of female and male rats. Based on the lack of a behavioral effect of VTA infusion of VU6000181, we hypothesized that VU6000181 would not affect phasic evoked DA release. Based on previous findings demonstrating substantia nigra M5 receptor activation increasing DA neuron activity in vitro (Foster et al., 2014), we were also curious if VU6000181 would reduce phasic evoked DA release in vivo. Our data demonstrates that VTA infusion of VU6000181 does not affect evoked phasic DA release in the NAc in female or male rats. Future in vivo FSCV experiments should explore M5 receptor–specific compounds such as orthostatic antagonists (e.g., VU6019650), which may produce differential effects on evoked DA release in the NAc compared with allosteric modulators (Garrison et al., 2022).

Apparent sex differences emerged in the ability of physostigmine, when coadministered with VU6000181, to increase evoked phasic DA release in males but not females. Thus, in the context of increased cholinergic tone in the VTA, VU6000181 significantly increased evoked phasic DA release in male rats. The ability of physostigmine plus VU6000181 to increase phasic DA release and enhance effort-choice behavior in male but not female rats suggests that the increase in DA in male rats could mediate the enhanced effort behavior in males. Indeed, increased NAc DA release correlates with increased lever pressing in effort-choice tasks (Segovia et al., 2011). Nevertheless, coadministration of physostigmine and VU6000181 in females is sufficient to restore effort choice behavior in females, attenuating the physostigmine-induced decrease. However, we did not observe any differences in phasic DA release in the females. Through additional investigations, one could examine whether coadministration of physostigmine and VU6000181 alters tonic-firing induced DA release or basal DA tone in females to reverse these physostigmine-induced deficits. Further experiments could also investigate whether the DA increase observed with coadministered physostigmine plus VU6000181 in males causally mediates the increase in effort-choice behavior.

The mechanisms underlying sex differences of VTA cholinergic manipulations to increase evoked DA release in male but not female rats are unknown. One possibility is differential expression of M5 muscarinic receptors between males and females. Acetylcholine receptor expression has been shown to differ between males and females, but it is unknown whether these differences occur at the level of M5 receptor expression in the VTA (Koylu et al., 1997). Future experiments should examine differences in M5 receptor protein levels in the VTA between females and males. Another explanation could be a function of the dose range of VU6000181. It is possible that in females, a higher dose of VU6000181 is required to inhibit VTA M5 mAChRs. VU6000181 as a pharmacological tool is limited to 30 µM to maintain specificity for the M5 receptor (Kurata et al., 2015). Thus, the development of selective M5 receptor compounds with a greater dose range of specificity for M5 receptors and different mechanisms of receptor inhibition (i.e., allosteric versus orthosteric) would be helpful to further elucidate these sex differences. A third potential explanation may revolve around sex-specific hormonal influences on VTA acetylcholine transmission. Estrus cycle phases have been found to influence acetylcholine transmission (Avissar et al., 1981; Hörtnagl et al., 1993) but have not been explicitly investigated in the VTA during effort-choice behavior. Further investigation into these possibilities may shed light on the sex difference observed here on DA release in the NAc.

Cholinergic regulation of VTA M5 receptors controlling aspects of effort-choice behavior in rodents is clinically relevant. M5 receptors are attractive drug targets for DA-dependent disorders due to their preferential expression on midbrain DA neurons. Thus, it may be possible to dissociate and pharmacologically target specific behaviors while leaving others intact, thereby reducing the potential for unwanted side effects. For example, VTA muscarinic receptors are known to regulate food behaviors (Sharf and Ranaldi, 2006; Sharf et al., 2006). Scopolamine, a nonselective mAChR antagonist, infused into the VTA has been shown to decrease food intake. Unpublished data from our laboratory shows that intra-VTA infusion of scopolamine decreases both lever pressing and the consumption of the freely available laboratory chow. Yet, in our current experiments with VU6000181, we did not observe an overall reduction in freely available chow, suggesting that selective inhibition of M5 receptors via negative allosteric modulation leaves consumption of food intact. Clinically, the dissociation of drug treatment to attenuate effort allocation and not general food motivation is a critical consideration. The ability to selectively target symptoms with limited side effects provides therapeutic advantages. The behavioral impact of VU6000181 on effort-choice behavior was only observed in the presence of the cholinesterase inhibitor physostigmine. Therefore, it is plausible that selective inhibition of M5 receptors via negative allosteric modulation may be helpful in the context of disease states such as depression, where increased ACh levels are correlated with depressive symptoms, but not during normal healthy conditions (Nunes et al., 2022). Thus, development and investment in muscarinic subtype-specific ligands could lead to promising treatments for effort-related choice impairments.

Abbreviations

ACh

acetylcholine

DA

dopamine

EPM

elevated plus maze

FR5

fixed-ratio 5

FSCV

fast-scan cyclic voltammetry

FST

forced swim test

mAChR

muscarinic acetylcholine receptor

NAc

nucleus accumbens core

NAM

negative allosteric modulator

n.s.

not significant

SPT

sucrose preference test

VTA

ventral tegmental area

Authorship Contributions

Participated in research design: Nunes, Kebede, Haight, Addy.

Conducted experiments: Nunes, Kebede, Haight.

Contributed new reagents or analytic tools: Foster, Lindsley, Conn.

Performed data analysis: Nunes, Kebede, Haight.

Wrote or contributed to the writing of the manuscript: Nunes, Kebede, Haight, Addy.