Muscarinic antagonists impair multiple aspects of operant discrimination learning and performance

Abstract

Acetylcholine signaling can strengthen associations between environmental cues and reward availability. Diverse subtypes (M1-M5) of the muscarinic acetylcholine receptor (mAChR) family may have distinct roles in different learning and memory processes, such as encoding cue-reward associations and consolidating these associations in long-term memory. Using an operant discrimination learning task in which mice are trained to nose poke during a tone to receive a food reward, we found that acquisition of the task requires mAChR signaling in the central nervous system. In addition, post-session injections of a broad mAChR antagonist, scopolamine impaired consolidation of the cue-reward memory. Further, after successful learning of a cue-reward contingency across multiple training sessions, mice that received a single pre-session injection of scopolamine were unable to use the learned cue association to receive rewards. Taken together, these data demonstrate distinct roles for muscarinic signaling in acquisition, consolidation and recall of the operant discrimination learning task. Understanding mechanisms underlying natural reward-related responding may provide insight into other maladaptive forms of reward learning such as addiction.

INTRODUCTION

Forming associations between environmental stimuli and food availability is critical for survival and reproductive fitness. Cholinergic signaling via muscarinic acetylcholine receptors (mAChRs) plays a central role in developing associations between cues (e.g. tone) and reward delivery [1, 2, 3, 4, 5]. There are five distinct subtypes of mAChRs (M1-M5), all of which are G protein-coupled receptors [6, 7, 8]. Pharmacological studies have suggested that the M1 and M3 subtypes are important mediators of learning and memory processes. For example, blockade of M1/M3 mAChRs impairs Pavlovian learning such as eyeblink conditioning [9], conditioned place preference [10], cued fear conditioning [11, 12, 13] and contextual fear learning [12, 14, 15]. In contrast to Pavlovian tasks, the specific mAChR subtypes involved in encoding of an appetitive operant learning task has not been well described.

Pharmacological blockade of mAChRs using the non-selective muscarinic antagonist, scopolamine disrupts encoding of both Pavlovian and instrumental reward-related learning [1, 2, 3, 4, 5, 16]. However, mAChRs may have different roles depending on the stage of memory formation. Inhibition of mAChRs impairs encoding of contextual fear memories [14, 17, 18, 19] but not consolidation [14], a stage when the memory is stabilized into long-term storage. Conversely, other studies show that mAChRs play a role in memory consolidation of cued fear [20], and an operant odor discrimination task [21]. Whether mAChR signaling is required for consolidation of operant discrimination learning remains to be elucidated.

Using a discriminative stimulus guided operant reward learning paradigm, in which mice are trained to nose poke during a tone presentation to receive a palatable food reward, we first observe that scopolamine inhibits encoding of the operant discrimination learning task. In addition to using the non-selective antagonist, scopolamine, we used a second muscarinic receptor inhibitor, benztropine, which has been shown to have preferential affinity for the M1/M3 subtypes [22, 23, 24]. Acquisition of this task requires mAChRs in the central nervous system (CNS), as pirenzepine, which shows reduced CNS penetration, had no effect on learning. We then went on to show that scopolamine administration following each behavioral training session inhibits consolidation of the associative memory. Finally, we demonstrate that after mice successfully learn the task, a pre-session injection of scopolamine inhibits the ability to use the learned tone association to nose poke during an appropriate tone interval to receive the food reward. Understanding how mAChR signaling modulates memory processes will provide insight into cholinergic mechanisms regulating the motivational properties of natural reward- and drug-associated cues.

MATERIALS AND METHODS

Animals

All experiments were approved by Yale University Institutional Animal Care and Use Committee in compliance with the National Institute of Health’s Guide for the Care and Use of Laboratory Animals. C57BL/6J male, adult mice were purchased from Jackson Laboratory. Mice were kept single-housed at age ~5 months in a temperature-controlled animal facility on a 12 hr light/dark cycle.

Behavioral Training

The operant conditioning protocol was adapted from Crouse et al., 2020 [5]. All mice were weighed daily and provided standard chow (Envigo, Madison, WI) to maintain 85% body weight. Mice were handled for 3 min/day for 7 days in the same room in which behavioral training was conducted. Three days prior to nose poke training, mice were exposed to the reward (1:1 EnsurePlus Vanilla Nutrition Shake and water) in their home cages. All operant training was conducted in Med Associates test chambers (Med Associates Inc, Georgia, VT) in sound attenuating boxes. The left side of the chamber had two nose pokes (one active and one inactive port). The opposite side of the chamber had a reward receptacle connected to a syringe pump on the opposite side. The house light, tone generator (ENV-230), and speaker (ENV-224BM) were located outside the chamber but inside the sound attenuating chamber.

During the first behavioral training session (35 min) rewards were freely available, allowing mice to identify the location of reward delivery. Mice then went through 4 days of nose poke training (30 min), in which one nose poke response in the active port resulted in reward reinforcement (24 μL over 2 s). The inactive nose poke (unreinforced) was used as a locomotor control. When mice reached criteria for nose poke training (average of 30 rewards for 2 consecutive days), they moved to tone training (30 min). During tone training, mice were required to nose poke in the active port during a 10 s auditory tone (2.5–5 kHz, ~ 60 dB) window to receive an Ensure reward in the receptacle. There were no consequences if mice nose poked in the active port outside the 10 s tone (incorrect nose pokes) or in the inactive port (inactive nose pokes). Following reward retrieval, the tone was presented on a Variable intertrial Interval of 30 s (VI30; intervals from 10 – 50 s). After 4 days of tone training, mice progressed to the discrimination training stage, during which incorrect nose pokes (responses outside the 10 s tone or on the inactive nose poke) resulted in a 5 s house light (timeout) and a restart of the previous intertrial interval. During the discrimination training stage (30 min), mice were required to reach an acquisition criterion of 20 rewards (12–18 days). Mice can acquire approximately 50 rewards during the 30 min period.

Behavioral data were analyzed using SPSS and GraphPad Prism. Statistical differences between groups and interactions across training were evaluated using two-way (Figure 1 and 2) and three-way (Figure 3) repeated measures ANOVA. Geisser and Greenhouse correction was used to adjust for lack of sphericity. When appropriate effects were detected with ANOVAs, Tukey’s post hoc tests were used to make multiple comparisons.

Figure 1. Pre-session injections of mAChR blockers impair the ability of mice to acquire appropriate operant discrimination responses.

(A) Timeline for behavioral training. (B) Mice that were administered saline (n = 9) or pirenzepine dihydrochloride (mAChR blocker that shows reduced CNS penetration; n = 10) prior to discrimination training obtain significantly more rewards compared to mice that were treated with scopolamine (mAChR blocker; n = 10) or benztropine mesylate (mAChR blocker; n = 10). (C) During tone and discrimination training, incorrect nose pokes were increased in mice that were administered scopolamine or benztropine mesylate compared to saline- or pirenzepine dihydrochloride-treated mice. (D) Scopolamine-treated mice made a higher number of inactive nose pokes compared to saline controls during both tone and discrimination training stages. (E) Saline- and pirenzepine-treated mice received a higher number of tone presentations compared to benztropine- and scopolamine-injected mice. (F) Compared to saline control mice or pirenzepine dihydrochloride-treated mice, central mAChR blockade increased hit rate (# of rewards obtained/# of tones delivered) during tone training but reduced hit rate during discrimination training. (G) Benztropine-injected mice have a higher number of receptacle entries compared to controls. Main effects and interactions are for discrimination training.

Figure 2. mAChRs contribute to consolidation of operant discrimination learning.

(A) Timeline for behavioral training. (B) Mice injected with scopolamine (n = 12) immediately following each session show no impairment during tone training stage but earn significantly fewer rewards than saline-treated mice during discrimination training. when incorrect responses result in a timeout (n = 12). (C) Both groups decreased the number of incorrect nose pokes during discrimination training (note that error bars are very small and are therefore obscured by symbols). (D) Number of inactive nose pokes were higher in scopolamine treated mice during discrimination training. (E) Saline-treated mice received more tones per session during discrimination training compared to mice that were administered scopolamine, indicating that they made fewer errors and therefore had fewer timeout periods. (F) Post-session scopolamine administration significantly impaired the accuracy of nose poking when a tone was presented (hit rate: # of rewards obtained/# of tones delivered). (G) Scopolamine has no effect on the number of receptable entries. Main effects and interactions are for discrimination training.

Figure 3. Pre-session scopolamine administration impairs performance of the operant discrimination task in trained mice.

(A) Timeline for behavioral training. Following 18 days of discrimination training (post-training) with post-session saline or scopolamine injections, mice were split into 4 groups (n = 6/group): 1) post-session saline during post-training/pre-session saline during performance evaluation; 2) post-session scopolamine during post-training /pre-session saline during performance evaluation; 3) post-session saline during post-training /pre-session scopolamine during performance evaluation; 4) post-session scopolamine during post-training/pre-session scopolamine during performance evaluation. (B) Pre-session injection of saline preserved the consolidation impairment observed in the group that received post-session scopolamine during discrimination training. Pre-session injections of scopolamine significantly disrupted performance in both groups. (C, D) Pre-session scopolamine administration increased the number of incorrect and inactive nose pokes in both groups. (E, F) Hit rate is impaired in mice that received scopolamine 30 min before the task. (G) Pre-session injections of scopolamine increased the number of reward receptacle entries. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

Pharmacological Manipulations

To evaluate acquisition of the operant discrimination learning task, mice were injected intraperitoneally (i.p) with either saline, 0.5 mg/kg scopolamine (competitive mAChR antagonist; Millipore Sigma), 10 mg/kg benztropine mesylate (mAChR antagonist; Millipore Sigma) or 10 mg/kg pirenzepine dihydrochloride (mAChR antagonist that has reduced CNS penetration; Millipore Sigma) approximately 30 min prior to each tone (4 days) or discrimination training (12–18 days). To evaluate consolidation of operant discrimination learning, mice were injected (i.p) with either saline or 0.5 mg/kg scopolamine immediately after tone and discrimination training trials.

RESULTS

Pre-session injections of mAChR blockers impair the ability of mice to acquire appropriate operant discrimination responses

Approximately 30 min before each tone and discrimination training trial, mice received an i.p. injection of saline (n = 9), 0.5 mg/kg scopolamine (n =10), 10 mg/kg benztropine mesylate (n = 10) or 10 mg/kg pirenzepine dihydrochloride (n = 10). During tone training, pre-session injections of mAChR blockers increased hit rate (# of rewards/# of tones), incorrect and inactive nose pokes, but not the number of rewards earned, likely because mice nose poked at a higher rate overall and therefore coincidentally poked during the time the tone was delivered (Figure 1, all statistical outcomes shown in Table 1). Across 12 days of discrimination training, scopolamine and benztropine significantly reduced the number of rewards obtained compared to saline controls (Figure 1B, Table 1). During discrimination training, incorrect responses led to a greater number of timeouts. Timeouts result in onset of the house light (5 s) and restarting of the intertrial interval. We therefore analyzed the ‘number of tones’ as a measure of opportunities to make the association, and the ‘hit rate’ (# of rewards/# of tones) to provide information about whether mice nose poke for reward during tone presentations or if they ignore it. We observed that scopolamine and benztropine increased the number of incorrect nose pokes (Figure 1C, Table 1), and decreased the number of tone presentations (Figure 1E, Table 1). mAChR inhibition also reduced hit rate (# of rewards obtained/# of tones delivered; Figure 1F, Table 1). Together, the lack of a pirenzepine effect demonstrates that peripheral mAChRs are not likely to be involved in the acquisition of operant discrimination learning, whereas impairment in mice administered either scopolamine or benztropine identifies a role for mAChRs during the acquisition stage of this instrumental task (Figure 1, Table 1).

Table 1.

Two-way repeated measures ANOVA used to analyze effects of pre-session mAChR blockade on different measures of the operant discrimination task.

Variable	Training stage	Effect	F value	p value	Post hoc analysis
Rewards	Tone training	Day	26.66	7.17 × 10−12	No sig. differences between saline and scop or saline and benz
		Drug	2.75	0.06
		Day by Drug	1.31	0.25
	Discrimination training	Day	30.79	5.14 × 10−15	Saline is sig. different from scop on days 1– 2, 4–12 (p < 0.05) and from benz on days 1–2, 7–12 (p < 0.05)
		Drug	11.48	2.20 × 10−5
		Day by Drug	7.08	2.79 × 10−8
Incorrect NPs	Tone training	Day	9.84	1.28 × 10−4	No sig. differences between saline and scop or saline and benz
		Drug	1.60	0.21
		Day by Drug	2.34	0.02
	Discrimination training	Day	74.21	4.35 × 10−32	Saline is sig. different from scop on days 1–4, 6, 7, 9–12 and from benz on days 1–12 (p < 0.05)
		Drug	20.96	5.99 × 10−8
		Day by Drug	1.90	2.5 × 10−3
Inactive NPs	Tone training	Day	1.08	0.34	No sig. differences between saline and scop or saline and benz
		Drug	3.72	0.02
		Day by Drug	2.44	0.02
	Discrimination training	Day	3.83	0.02	Saline is sig. different from scop on day 6
		Drug	3.17	0.04
		Day by Drug	1.47	0.05
Tones	Tone training	Day	5.84	1.38 × 10−3	No sig. differences between saline and scop or saline and benz
		Drug	0.02	1.0
		Day by Drug	1.58	0.13
	Discrimination training	Day	17.00	2.23 × 10−14	Saline is sig different from scop and on days 1–5, 7–12 and from benz on days 1–3, 8
		Drug	17.82	3.43 × 10−7
		Day by Drug	1.19	0.28
Hit rate	Tone training	Day	25.86	1.20 × 10−10	Saline is sig. different from scop on day 1 (p < 0.05)
		Drug	3.57	0.02
		Day by Drug	1.16	0.34
	Discrimination training	Day	7.36	4.70 × 10−8	Saline is sig. different from scop on days 8–12 (p < 0.05) and from benz on days 8 and 10–12 (p < 0.05)
		Drug	3.79	0.02
		Day by Drug	4.89	2.25 × 10−10
Receptacle Entries	Tone training	Day	3.99	0.02	Saline is sig different from benz (p < 0.001) on day 1
		Drug	3.00	0.04
		Day by Drug	2.33	0.02
	Discrimination training	Day	1.51	0.19	Saline is sig different from benz (p < 0.05) on day 4
		Drug	4.05	0.01
		Day by Drug	1.76	7.00 × 10−3

Blockade of mAChRs impairs consolidation of operant discrimination learning

To evaluate a role for mAChRs in memory consolidation in the operant discrimination task, mice were administered saline (i.p.; n = 12) or scopolamine (0.5 mg/kg i.p; n = 12) immediately after each session during the tone and discrimination training phases. In contrast to the previous experiment in which pre-session injections of mAChR blockers were administered for 12 days during discrimination training and acquisition of the task was impaired (Figure 1), in the second experiment mice received injections of scopolamine immediately after the session (post-session) over 18 days of discrimination training. Across 18 days of discrimination training (non-cued responses resulted in timeouts), post-session injections of scopolamine reduced the number of rewards earned, number of tone presentations, and hit rate, and also increased the number of incorrect nose pokes (Figure 2, Table 2). Thus, blockade of mAChRs after discrimination training impaired consolidation of the cue-reward memory association.

Table 2.

Two-way repeated measures ANOVA used to analyze effects of post-session mAChR blockade on different measures of the operant discrimination task.

Variable	Training stage	Effect	F value	p value	Post hoc analysis
Rewards	Tone training	Day	6.38	1.39 × 10−4	No sig. differences between scop and saline groups
		Drug	0.025	0.88
		Day by Drug	4.40	7.00 × 10−3
	Discrimination training	Day	63.30	2.24 × 10−19	Saline is sig. different from scop on days 11, 12, 14–16 (p < 0.05)
		Drug	9.94	4.60 × 10−3
		Day by Drug	5.72	2.00 × 10−3
Incorrect NPs	Tone training	Day	6.68	9.87 × 10−4	No sig. differences between scop and saline groups
		Drug	0.21	0.65
		Day by Drug	1.81	0.15
	Discrimination training	Day	61.57	2.46 × 10−26	Saline is sig. different from scop on days 15 and 17 (p < 0.05)
		Drug	0.28	0.60
		Day by Drug	5.69	2.74 × 10−4
Inactive NPs	Tone training	Day	14.68	1.20 × 10−5	No sig. differences between scop and saline groups
		Drug	0.27	0.61
		Day by Drug	0.14	0.90
	Discrimination training	Day	5.25	5.83 × 10−4	No sig. differences between scop and saline groups
		Drug	0.22	0.65
		Day by Drug	1.27	0.28
Tones	Tone training	Day	1.45	0.24	No sig. differences between scop and saline groups
		Drug	0.42	0.52
		Day by Drug	1.22	0.31
	Discrimination training	Day	22.91	2.13 × 10−14	Saline is sig. different from scop on days 15–17 (p < 0.05)
		Drug	5.49	0.03
		Day by Drug	5.65	2.14 × 10−4
Hit rate	Tone training	Day	5.95	2.35 × 10−3	No sig. differences between scop and saline groups
		Drug	0.04	0.85
		Day by Drug	4.98	3.50 × 10−3
	Discrimination training	Day	40.32	2.24 × 10−22	No sig. differences between scop and saline groups
		Drug	7.58	0.01
		Day by Drug	1.71	0.04
Receptacle Entries	Tone training	Day	6.74	2.10 × 10−3	No sig. differences between scop and saline groups
		Drug	1.60	0.22
		Day by Drug	0.77	0.52
	Discrimination training	Day	9.73	2.55 × 10−21	No sig. differences between scop and saline groups
		Drug	1.34	0.26
		Day by Drug	0.46	0.97

Pre-session administration of scopolamine impairs performance of the operant discrimination task in trained mice

To determine whether pre-session scopolamine injections impaired performance of the operant task after learning the cue-reward contingency, mice that had undergone post-training (discrimination training) saline or scopolamine injections after each trial were semi-randomized into new groups that received either saline (Pre-Sal; n = 6) or scopolamine (Pre-Scop; n = 6) BEFORE the performance task (Perf; Figure 3A). The performance task had timeouts and was performed in the same way as discrimination training. First, we tested whether scopolamine could disrupt performance after mice successfully acquired the memory. Post-session saline mice (post-training (Sal)) received a pre-session injection of saline (Perf (Pre-Sal); n = 6) or scopolamine (Perf (Pre-Scop); n = 6) and were evaluated for operant discrimination performance (Perf; Figure 3A, Table 3). Mice that received saline AFTER day 18 of discrimination training and BEFORE the performance test earned a similar number of rewards between the two days (p > 0.05; Figure 3B, Table 3). However, post-training (Sal) control mice earned significantly fewer rewards when they received a scopolamine injection 30 min BEFORE the performance test (Perf (Pre-Scop)) as indicated by a post hoc analysis (p < 0.05; Figure 3B, Table 3).

Table 3.

Three-way ANOVA was used to analyze effects of pre-session scopolamine on task performance.

Variable	Effect	F value	p value	Post hoc analysis
Rewards	Test type	13.57	4.22 × 10−3	Post-training (Sal) is significantly different from Perf (Pre-Scop) and Post-training (Scop) is significantly different from Perf (Pre-Scop) (p < 0.05)
	Drug by Post-training	2.23	0.17
	Drug by Perf	107.2	1.16 × 10−6
	Drug by Post-training by Perf	3.84	0.079
Incorrect NPs	Test type	30.22	2.63 × 10−4	Post-training (Sal) is significantly different from Perf (Pre-Scop) and Post-training (Scop) is significantly different from Perf (Pre-Scop) (p < 0.05)
	Drug by Post-training	0.94	0.35
	Drug by Perf	55.41	2.20 × 10−5
	Drug by Post-training by Perf	0.08	0.77
Inactive NPs	Test type	8.02	0.018	Post-training (Scop) is significantly different from Perf (Pre-Scop) (p < 0.05)
	Drug by Post-training	9.07	0.013
	Drug by Perf	0.09	0.77
	Drug by Post-training by Perf	0.62	0.45
Tones	Test type	14.86	2.20 × 10−3	Post-training (Sal) is significantly different from Perf (Pre-Scop) and Post-training (Scop) is significantly different from Perf (Pre-Scop) (p < 0.05)
	Drug by Post-training	2.55	0.14
	Drug by Perf	43.24	6.26 × 10−5
	Drug by Post-training by Perf	0.11	0.74
Hit rate	Test type	5.50	0.04	Post-training (Sal) is significantly different from Perf (Pre-Scop) (p < 0.05)
	Drug by Post-training	0.53	0.48
	Drug by Perf	9.75	0.01
	Drug by Post-training by Perf	1.11	0.32
Receptacle entries	Test type	41.50	7.42 × 10−5	Post-training (Sal) is significantly different from Perf (Pre-Scop) and Post-training (Scop) is significantly different from Perf (Pre-Scop) (p < 0.05)
	Drug by Post-training	126.8	0.63
	Drug by Perf	25.46	5.02 × 10−5
	Drug by Post-training by Perf	0.08	0.77

Next, we tested whether the effects of post-session scopolamine on memory consolidation were distinct from those of pre-session scopolamine on the performance test. On day 18 of discrimination training, mice that had received post-session injections of scopolamine (post-training (Scop)) were impaired in obtaining rewards compared to controls (post-training (Sal); Figure 2, Table 2). Mice that had a consolidation impairment due to post-session scopolamine injections (post-training (Scop)) obtained a similar number of rewards when they were injected with saline prior to the performance test (p > 0.05; Figure 3B, Table 3), suggesting that the impairment in memory consolidation across 18 days of discrimination training remained apparent when the animals were challenged with pre-session saline (Perf (Pre-Sal)). However, when the post-training (Scop) group was administered a scopolamine injection BEFORE the performance task (Perf (Pre-Scop) animals were further impaired (p < 0.05; Figure 3B, Table 3). Thus, the disrupting effects of pre- and post-session scopolamine are distinct from each other.

The impairing effects of pre-session scopolamine on performance remained consistent across other aspects of the task, such that mice that received scopolamine before the performance test had a higher number of incorrect (p < 0.05; Figure 3C; Table 3) and inactive (p < 0.05; Figure 3D; Table 3) nose pokes regardless of whether they were previously injected with saline or scopolamine during discrimination training. During the performance test, pre-session scopolamine reduced the number of tone presentations, indicating that mice were making more errors and had more timeout periods (p < 0.05; Figure 3E, Table 3), and significantly reduced hit rate (p < 0.05; Figure 3F; Table 3). Further, injections of scopolamine before the performance test led the mice to visit the reward receptacle significantly more often than pre-session saline control mice (p < 0.05; Figure 3G, Table 3). The effects of mAChR blockade on performance were persistent over 4 days (Supplementary Figure 1). The impairment in task performance was not due to a general inability to nose poke or to a decrease in motivation, since pre-session scopolamine increased the number of nose pokes. While it is possible that the increase in incorrect nose pokes was due to a change in locomotion, we believe this was unlikely since there was no effect on summed correct and incorrect nose pokes, supporting the possibility of a learning impairment.

DISCUSSION

The experiments presented here demonstrate that mAChR signaling is involved in both acquisition and consolidation of a cue-reward association in a discriminative stimulus guided operant reward learning task. This learning paradigm involves several components including attention to the tone, associative learning, and precise locomotor activity, and mAChR antagonists may disrupt all of these task parameters. Pre-session injections of the broad mAChR inhibitors scopolamine or benztropine, impair the mouse’s ability to nose poke during tone delivery to receive a palatable reward, whereas post-session antagonism of mAChRs impairs consolidation of cue-reward associations. Finally, pre-session administration of scopolamine after learning the task impairs performance of the operant discrimination task.

Cholinergic signaling involved in operant discrimination learning

In addition to its role in associative learning [1, 2, 3, 4, 5, 14, 16, 17, 25], mAChR signaling affects motor activity [26, 27, 28], motivation [29, 30, 31], and attention [32, 33, 34, 57], which are critical components for performance of the operant discrimination task. The significant scopolamine-induced impairment on performance of the operant discrimination task raises the question whether mAChR antagonism may impair motor function or motivation rather than associative memory. During tone training, when the task is simpler and mice are not penalized for non-cued responses, scopolamine- and benztropine-treated mice made a larger number of incorrect responses in the active and inactive nose ports, and had a higher hit rate compared to controls, suggesting that the impairing effects of scopolamine are not due to reduced activity or motivation. One reason for the high number of nose pokes may be that mice repetitively nose poke to obtain rewards during nose poke training, which occurs immediately before the tone training stage. During nose poke training, mice do not receive any injections of mAChR antagonists and by day 4 are nose poking for approximately 100 rewards in a session. During tone training, mice received injections of mAChR blockers and continued to nose poke repetitively as they did in the previous nose poke training, suggesting they are unable to switch their behavior to new task requirements [72,73]. mAChR inhibitors increased activity during discrimination training as well, in which, mice that poke in the active nose port outside the 10 s tone window (incorrect nose pokes) get a higher number of timeouts and a reduced number of tone presentations. In addition, mice that received the mAChR antagonists had a reduced hit rate (# of rewards obtained/# of tones delivered), suggesting that antagonist-treated mice do not nose poke when the tone is presented, and are therefore unable to associate the tone with reward. It is possible that mAChR antagonists may alter motor activity, however, scopolamine (0.5 mg/kg) did not significantly increase number of nose pokes in a previous study with the same task parameters or have any influence in other tests of locomotion [5]. Thus, in the current study, the enhanced nose poking induced by mAChR antagonists is likely to represent expression of a learning impairment.

Interestingly, compared to controls, pre-session scopolamine-injected mice showed increased visits to the reward receptacle, suggesting that mAChR signaling may influence the mouse’s engagement with the location of reward delivery (“goal-tracking”) and decreased attention to the tone (impaired “sign-tracking”). In sign-tracking/goal-tracking models, animals are classified in one category or the other based on their attention to the cue (sign) or reward location (goal) in an appetitive Pavlovian task [35, 36]. Compared to sign-trackers, goal-trackers may have elevated cholinergic signaling [37, 38, 39], regulated by nicotinic AChRs [37, 40]. Future work on mAChR signaling could reveal additional cholinergic mechanisms in sign- and goal-tracking.

Cholinergic receptors involved in acquisition and consolidation of memory

We showed previously that activity of mAChRs is necessary for the acquisition of this operant discrimination task [5]. In addition to replicating these findings, we demonstrate here that mAChRs in the peripheral nervous system are not required during encoding of this task [58, 59]. Pharmacological studies have shown that mAChRs contribute to both Pavlovian and instrumental reward-related learning [1, 2, 3, 4, 5, 16], but studies differ on whether muscarinic agents affect acquisition versus consolidation of these memories. Inconsistencies may be due to differences in behavioral models or studies that do not test the role of mAChRs in acquisition and consolidation in the same paradigm. Therefore, drug treatments before and after sessions in the same behavioral paradigm and doses may provide better insight on how mAChR signaling is involved in different aspects of memory formation.

Several lines of evidence have shown that development of associative memories is mediated by the basolateral amygdala (BLA) [56, 60, 62] complex, and that the BLA receives cholinergic inputs from the basal forebrain [63, 65, 66, 67]. Specifically, inputs from the nucleus basalis of Meynert to the BLA promote both Pavlovian fear conditioning [68, 69] and appetitive instrumental learning [5, 70]. Further, a recent study demonstrated that cholinergic inputs from the midbrain (pedunculopontine and laterodorsal tegmental nuclei) to the striatum regulate association of contingencies in instrumental reward behavior [71]. While studies have investigated the role of cholinergic transmission in plasticity and learning, there is still more to know about the specifics of how AChRs modulate circuits recruited for associative learning.

Acetylcholine can trigger different signaling cascades depending on the mAChR subtype it binds to. Acetylcholine binding to M1, M3, and M5 mAChRs activates phospholipase C via Gq₁₁ signaling, allowing calcium influx into the cell. Activation of M2 and M4 mAChRs results in G_i/o-induced inhibition of adenylyl cyclase [41, 42]. By triggering distinct signaling cascades, all 5 mAChR subtypes play key roles in learning processes [9, 10, 11, 12, 14, 15, 20, 33, 43, 44, 45, 46, 47, 48, 49, 50]. M1 and M3 receptors play an important role in Pavlovian and instrumental learning [9, 13, 20, 51, 54, 55]. Further, one study demonstrated that M1/M3 but not M2/M4 receptor subtypes in memory-related brain regions (hippocampus and retrosplenial cortex) are involved in encoding of associative fear memories [15]. Benztropine has been shown to have preferential affinity for M1/M3 subtypes [22, 23, 24], and in the current study, benztropine-injected mice were unable to obtain rewards compared to control mice, suggesting that this receptor subtype may be required to acquire this instrumental task successfully. This study adds to the current body of literature that M1 and/or M3 receptors may be integral for developing associative memories, however, more selective techniques such as transgenic mouse models will be needed to elucidate the precise role of mAChR subtypes in multiple memory stages of operant discrimination learning.

We demonstrate that pre-session injections of benztropine and scopolamine during tone and discrimination training disrupt acquisition of the task, however, there is a possibility that injections of mAChR blockers administered prior to each trial may also disrupt consolidation. We think this is unlikely as the half-life of scopolamine in rodents is approximately 20 min [52] and we administer scopolamine 30 minutes before the beginning of tone and discrimination training sessions. Although early consolidation processes may begin during the end of behavioral training [53], the effects of scopolamine may be declining at the start of training due to its short half-life.

While mAChR blockers may reliably inhibit acquisition of both contextual and cue-directed learning [14, 17, 25], the role of mAChRs is more complex in memory consolidation. Several studies have noted that post-training scopolamine injections do not affect context associations [14, 17, 64] but mediate cue-induced learning [25]. For example, one study showed that M1 blockade does not impair consolidation of cued fear memory, however, M1 antagonists when co-administered with β₂-adrenergic and D₅-dopaminergic receptor blockers substantially disrupt cued learning [20]. The current study suggests that consolidation of cue-reward memories may be more sensitive to muscarinic blockade than cued-fear memories. Whether these mechanisms are triggered due to the recruitment of different mAChR subtypes across phases memory formation remains to be elucidated.

In conclusion, these results demonstrate that mAChRs contribute to both acquisition and consolidation phases of operant discrimination learning. Understanding these pathways will provide further insight into disorders underlying maladaptive memories such as drug addiction.

Supplementary Material

1

Supp Figure 1. Effects of pre-session scopolamine administration are stable across multiple sessions. The impairment in consolidation resulting from post-session scopolamine administration during discrimination training, and the impairment of performance induced by pre-session administration of scopolamine are stable across 4 days of testing in: (A) number of rewards earned, (B) incorrect and (C) inactive nose pokes, (D) hit rate, (E) number of tones received during the session and, (F) number of receptacle entries.

Table 4.

Summary of results. Effect of pre- and post-session mAChR blockade on multiple aspects of operant discrimination learning task.

Behavioral Stage	Drug	Time of injection	Effect on behavior
Discrimination training	Pre-session	Benztropine	Impairs acquisition of task
		Scopolamine	Impairs acquisition of task
		Pirenzepine	Has no effect on task
Discrimination training	Post-session	Scopolamine	Impairs consolidation of cue-reward contingency
Performance test (after discrimination training)	Pre-session	Scopolamine	Impairs performance

Highlights.

• Pre-session injections of muscarinic receptor antagonists impair acquisition of the operant discrimination learning task.

• Post-session injections of muscarinic receptor inhibitors disrupt consolidation of the cue-reward contingency in the operant task.

• Acquisition of the operant discrimination task is dependent on muscarinic receptors of the central nervous system but not those of the peripheral nervous system.

• After successful learning of the operant task, one pre-session injection of a muscarinic receptor antagonist results in impaired performance of the operant discrimination learning task.