Nicotine facilitates memory consolidation in perceptual learning

Perceptual learning is a special type of non-declarative learning that involves experience-dependent plasticity in sensory cortices. The cholinergic system is known to modulate declarative learning. In particular, reduced levels or efficacy of the neurotransmitter acetylcholine were found to facilitate declarative memory consolidation. However, little is known about the role of the cholinergic system in memory consolidation of non-declarative learning. Here we compared two groups of non-smoking men who learned a visual texture discrimination task (TDT). One group received chewing tobacco containing nicotine for 1?h directly following the TDT training. The other group received a similar tasting control substance without nicotine. Electroencephalographic recordings during substance consumption showed reduced alpha activity and P300 latencies in the nicotine group compared to the control group. When re-tested on the TDT the following day, both groups responded more accurately and more rapidly than during training. These improvements were specific to the retinal location and orientation of the texture elements of the TDT suggesting that learning involved early visual cortex. A group comparison showed that learning effects were more pronounced in the nicotine group than in the control group. These findings suggest that oral consumption of nicotine enhances the efficacy of nicotinic acetylcholine receptors. Our findings further suggest that enhanced efficacy of the cholinergic system facilitates memory consolidation in perceptual learning (and possibly other types of non-declarative learning). In that regard acetylcholine seems to affect consolidation processes in perceptual learning in a different manner than in declarative learning. Alternatively, our findings might reflect dose-dependent cholinergic modulation of memory consolidation. This article is part of a Special Issue entitled 'Cognitive Enhancers'.     

Cigarette smoking and state-dependent memory

Two experiments were carried out to investigate whether cigarette smoking could produce state-dependent learning (SDL) in humans. The first experiment was concerned with the methodological issue of choosing an appropriate control cigarette for use in an SDL design. A low nicotine content (0.2 mg) cigarette was chosen as it did not appear to affect the physiological arousal of the subjects. In Experiment 2, it was shown that cigarette smoking can produce state-dependent memory effects. The most likely basis for the results is the arousal produced by the nicotine content of the cigarette.     

Nicotinic receptor activation in perirhinal cortex and hippocampus enhances object memory in rats

The perirhinal cortex (PRh) and its cholinergic inputs are implicated in object recognition memory. Conversely, the hippocampus (HPC) may be involved in spatial recognition processes that are not essential to the recognition of objects per se. Systemic nicotine has been shown to facilitate both object and spatial memory. The current study compared the roles of perirhinal and hippocampal nicotinic acetylcholine receptors (nAChRs) in object and spatial recognition memory using spontaneous object recognition (SOR) and object-location (OL) tasks for rats. Systemic pre-sample (acquisition) nicotine administration dose-dependently facilitated SOR and OL performance compared to vehicle conditions in which performance was at chance with a 72-h retention delay between the sample and choice phases. Subsequently, pre-sample intra-PRh infusions of nicotine significantly facilitated SOR; somewhat surprisingly, pre-sample intra-HPC nicotine also enhanced object recognition memory. Further experiments indicated facilitative effects on OL performance caused by pre-sample intra-PRh or intra-HPC nicotine administration. These results not only demonstrate that nAChR activation facilitates performance on object recognition and object-location memory tasks, but suggest that these effects can be produced by nAChR action in either PRh or HPC. Thus, although PRh and HPC are not required for OL or SOR task performance, respectively, nAChR-mediated enhancement of neural function in either of these temporal lobe regions appears capable of promoting stronger memory encoding and/or consolidation in either task. These findings further support the supposed interactive relationship between the HPC and PRh in object information processing and highlight the potential therapeutic value of nicotinic receptor activation in amnesic disorders.     

Separate and combined effects of scopolamine and nicotine on retrieval-induced forgetting

Rationale Proficiency of information processing is likely to derive from a combination of effective processing of relevant information and efficient inhibition of unwanted or irrelevant material. Compromised inhibitory processes have been associated with the memory deficits in dementia and the elderly. These deficits in inhibition could be directly related to the reduced cholinergic function evident in dementia of the Alzheimer type (DAT). Scopolamine, a cholinergic antagonist, has been associated with disinhibition. Nicotine, a cholinergic agonist, has been associated with enhanced focus and reduced intrusions in both healthy adults and people with dementia.Objective This study examines the separate and combined effects of nicotine and scopolamine on inhibitory processes using retrieval induced forgetting (RIF).Methods In two studies, minimally deprived smokers received either nicotine or no nicotine, in the second study combined with placebo, 0.3 or 0.5 mg SC scopolamine. All volunteers completed the RIF procedure providing within-subject measures of inhibition in an episodic recall paradigm.Results Nicotine did not modulate recall of practised exemplars but it did increase inhibition of unpractised exemplars. Consistent with an effect on encoding, scopolamine reduced recall for all exemplars. Scopolamine did not differentially affect practised or unpractised exemplars, relative to the control words.Conclusion Independent of its potential to influence memory at encoding, nicotine can affect task performance by inhibiting unpractised (and by implication, irrelevant) material, thereby reducing interference and benefiting the task in hand. The absence of effects of scopolamine on inhibition in the RIF paradigm argues for a more complex subdivision of "inhibitory" processes, which may be differentially influenced by cholinergic blockade.     

Evidence that mnesic states govern normal and disordered memory

Although it is recognized that retrieval may be state-dependent, only recently has a paradigm been identified that allows state-dependence in rats to be demonstrated reliably and at relevant doses of CNS agents. In humans, the effects of scopolamine constitute a valuable model of disordered memory. We used this paradigm to analyze the effects of scopolamine on different memory processes. Rats treated with either saline or 0.01-10 mg/kg doses of scopolamine learned to lever press for milk reward, and were then tested for retrieval while given the same or a different treatment. Saline-to-scopolamine as well as scopolamine-to-saline state changes produced robust failures to retrieve the response. Remarkably, the state produced by 2.5 mg/kg scopolamine, like that produced by saline, produced little intrinsic effect on learning or any other memory process (i.e. when the prevailing state was left unchanged). However, changing the implemented state from one to the other between two different processes, disrupted not only retrieval, but also learning, encoding and retention. We also determined whether the graded state changes produced by 0.01-10 mg/kg doses of scopolamine could mimic the peculiar and poorly understood temporally graded retrograde amnesia that occurs in Alzheimer's disease. In rats that had acquired a complex drug-discrimination task over a 6-month period, scopolamine-induced state changes seemed to produce dose-dependent deficits in the retrieval of recent information while preserving those abilities that had been acquired in the increasingly remote past. Beyond its role in retrieval, the findings implicate state dependence in learning, encoding and retention, and suggest that physiologically defined mnesic states govern each of these. The changes of mnesic state that are likely associated with excessively labile cholinergic neurotransmission may conceivably cause the complex disabilities of Alzheimer's disease.     

Distinguishing between attentional and amnestic effects in information processing: the separate and combined effects of scopolamine and nicotine on verbal free recall

An important issue in our understanding of cholinergic modulation of information processing is the extent to which drug-induced changes affect memory processes per se or simply the attentional processes required for effective acquisition of information. In this study, we examined the separate and combined effects of scopolamine and nicotine on verbal free recall. A single dose of nicotine improved recall performance on supraspan lists (30 words), but not on short lists (10 words). The same dose of nicotine had no effect on the scopolamine-induced recall deficits observed for both 30 and 10 word lists. The results are discussed in terms of the independence of attention and memory processes and the specificity of action of these two cholinergic compounds.     

Neonatal exposure to the brominated flame retardant 2,2',4,4',5-pentabromodiphenyl ether causes altered susceptibility in the cholinergic transmitter system in the adult mouse.

Polybrominated diphenyl ethers (PBDEs) are used as flame-retardants and have recently been shown to be increasing in the environment and in human mother's milk. We have recently reported that neonatal exposure to 2,2',4,4',5-pentaBDE (PBDE 99) can induce persistent aberrations in spontaneous behavior and also affect learning and memory functions in the adult animal. The present study indicates that the cholinergic system, in its developing stage, may be a target of and sensitive to PBDEs. Neonatal exposure of male NMRI mice on postnatal day 10, to 2,2',4,4',5-pentaBDE (8 mg/kg bw) was shown to alter the response to a cholinergic agent, nicotine, at an adult age. The nicotine-induced behavior test revealed a hypoactive response to nicotine in PBDE 99-treated animals, whereas the response of controls was an increased activity. These findings show similarities to observations made from neonatal exposure to PCBs and nicotine, compounds shown to affect cholinergic nicotinic receptors. This indicates that PBDE 99 can affect the cholinergic system and might thereby interact with other environmental toxicants.     

Effect of vitamin E on memory retention in rats: possible involvement of cholinergic system.

This study concerned effects of vitamin E and the cholinergic system on memory retention of passive avoidance learning in rats. Post-training intracerebroventricular (i.c.v.) injections were carried out in all experiments. Administrations of vitamin E (10, 25 and 50 microg/rat), nicotine (0.1 microg/rat) and pilocarpine (0.5 microg/rat), the muscarinic receptor agonist increased memory retention, while mecamylamine (0.01, 0.1 and 0.5 microg/rat), the nicotinic receptor antagonist and scopolamine (0.1, 1 and 5 microg/rat), the muscarinic receptor antagonist decreased memory retention. The combination of vitamin E with nicotine or pilocarpine showed potentiation. Effects of mecamylamine or scopolamine were attenuated by vitamin E. It is concluded that vitamin E has a close interaction with cholinergic system in memory retention process.     

Involvement of dorsal hippocampal nicotinic receptors in the effect of morphine on memory retrieval in passive avoidance task

The present study evaluated the possible role of nicotinic acetylcholine receptors of the dorsal hippocampus on morphine-induced amnesia and morphine state-dependent memory in adult male Wistar rats. The animals were bilaterally implanted with chronic cannulas in the CA1 regions of the dorsal hippocampi, trained in a step-through type passive avoidance task, and tested 24 h after training to measure step-through latency. Results indicate that post-training subcutaneous (s.c.) administration of morphine (2.5-7.5 mg/kg) dose-dependently reduced the step-through latency, showing an amnestic response. Post-training intra-CA1 microinjection of nicotine (0.5-1 microg/rat) decreased significantly the amnesia induced by post-training morphine (7.5 mg/kg). Moreover, co-treatment of mecamylamine (0.5 and 1 microg/rat, intra-CA1) with an ineffective dose of morphine (2.5 mg/kg), immediately after training, caused inhibition of memory retrieval. On the other hand, amnesia produced by post-training morphine (7.5 mg/kg) was reversed by pre-test administration of the opioid that is due to a state-dependent effect. Interestingly, pre-test intra-CA1 microinjection of nicotine (0.25 and 0.5 microg/rat) improved post-training morphine (7.5 mg/kg)-induced retrieval impairment. Moreover, pre-test administration of the same doses of nicotine in combination with a lower dose of morphine (0.5 mg/kg), which had no effects alone, synergistically improved memory performance impaired by post-training morphine. Pre-test injection of mecamylamine (0.5-2 microg/rat) prevented the restoration of memory by pre-test morphine. It is important to note that post-training or pre-test intra-CA1 administration of the same doses of nicotine or mecamylamine, alone did not affect memory retrieval. These results suggest that nicotinic acetylcholine receptors of the hippocampal CA1 regions may play an important role in morphine-induced amnesia and morphine state-dependent memory.     

Improved incidental memory with nicotine after semantic processing, but not after phonological processing

RATIONALE: A number of lines of evidence suggest that a nicotinic cholinergic system is mediating attentional processing. However, the evidence is less clear for a nicotinic system being involved in mnemonic processing. OBJECTIVES: The present study investigated the effects of nicotine on memory using a depth of processing paradigm. METHODS: A double-blind design was used with participants (n = 40) smoking either a nicotine containing cigarette (n = 20) and a denicotinized cigarette (n = 20). After smoking, each set of these participants was further subdivided into two groups (n = 10 for each). One group were presented with a series of trials each beginning with the presentation of a "decision word" which they had to say whether it represented something which was living or non-living (semantic-orienting). The second group had to say whether the word had one syllable or two syllables (phonological or non-semantic orienting condition). This decision was followed by a word in coloured ink whose colour participants were required to name as quickly as possible. On completion of the whole task the participants were given an unexpected free recall test. RESULTS: The nicotine-containing cigarette reduced the latencies for decision-making and colour naming in comparison with the denicotinized cigarette. The free recall test showed that nicotine-containing cigarette increased the number of words remembered, but only for the semantic-orienting condition and not the non-semantic condition. CONCLUSIONS: There is a nicotinic cholinergic system that mediates effortful processing. It can be deployed for attentional processing, including the associative processing required for memory encoding.     

Caffeine attenuates scopolamine-induced memory impairment in humans

Caffeine consumption can be beneficial for cognitive functioning. Although caffeine is widely recognized as a mild CNS stimulant drug, the most important consequence of its adenosine antagonism is cholinergic stimulation, which might lead to improvement of higher cognitive functions, particularly memory. In this study, the scopolamine model of amnesia was used to test the cholinergic effects of caffeine, administered as three cups of coffee. Subjects were 16 healthy volunteers who received 250 mg caffeine and 2 mg nicotine separately, in a placebo-controlled double-blind cross-over design. Compared to placebo, nicotine attenuated the scopolamine-induced impairment of storage in short-term memory and attenuated the scopolamine-induced slowing of speed of short-term memory scanning. Nicotine also attenuated the scopolamine-induced slowing of reaction time in a response competition task. Caffeine attenuated the scopolamine-induced impairment of free recall from short- and long-term memory, quality and speed of retrieval from long-term memory in a word learning task, and other cognitive and non-cognitive measures, such as perceptual sensitivity in visual search, reading speed, and rate of finger-tapping. On the basis of these results it was concluded that caffeine possesses cholinergic cognition enhancing properties. Caffeine could be used as a control drug in studies using the scopolamine paradigm and possibly also in other experimental studies of cognitive enhancers, as the effects of a newly developed cognition enhancing drug should at least be superior to the effects of three cups of coffee.     

Acute and chronic nicotine effects on working memory in aged rats

Acute and chronic nicotine administration has been repeatedly been found in our laboratory to improve working memory performance of normal adult rats in the radial-arm maze. The current study was conducted to determine if acute or chronic nicotine administration would improve working memory performance in aged rats. Sixteen young adult (3–7 months) and 32 aged (24–28 months) male Sprague-Dawley rats were trained on an eight-arm radial maze. A significant age-related choice deficit was seen during the 21 sessions of training. After training, half of the rats in each age group were implanted with nicotine-containing osmotic minipumps and the other half implanted with vehicle-containing pumps. Consistent with previous work, the young adult rats given chronic nicotine (approximately 5 mg/kg per day as measured as nicotine base) showed a significant improvement in working memory performance. In contrast, the aged rats did not show a significant effect of this dose of chronic nicotine. After a 2 week withdrawal period the remaining rats underwent a series of acute drug challenges with nicotinic and muscarinic agonists and antagonists as well as the dopaminergic antagonist haloperidol. Mecamylamine and haloperidol impaired the memory performance of the young adult rats, whereas the aged rats showed no effect. In contrast, scopolamine impaired performance of both young adult and aged rats in a similar manner. Both pilocarpine and nicotine improved the memory performance of the aged rats, but did not improve the young adult rats, possibly due to a ceiling effect on performance. During the cholinergic agonist drug phase, the aged rats which had previously been given chronic nicotine infusions showed better performance than those which had not. The resistance of the aged rats to chronic nicotine-induced working memory improvements and acute mecamylamine-induced working memory deficits may have resulted from the decline in nicotinic receptors seen with aging. Chronic co-administration of the nicotinic antagonist mecamylamine in a previous study was found to abolish the chronic nicotine-induced working memory improvement. The aged rats were resistant to haloperidol-induced deficits which may have resulted from the decrease in dopaminergic receptors seen with aging. Interestingly, acute cholinergic agonists including nicotine did improve working memory performance in the aged rats and previous chronic nicotine infusion was beneficial during the period of acute cholinergic agonist challenge. This suggests that nicotinic treatment may be of use for treating age associated memory impairments but that special dosing regimens may be required.This work was carried out in partial fulfillment of a BSc (Hons) in Biochemistry at the University of Bath (UK) by D.T.     

Neonatal exposure to 2,2',5,5'-tetrachlorobiphenyl causes increased susceptibility in the cholinergic transmitter system at adult age

Polychlorinated biphenyls (PCBs) constitute a large category of chlorinated hydrocarbons that can have developmental neurotoxic effects. We have recently reported that neonatal exposure to a lightly chlorinated ortho-substituted PCB, 2,2',5,5'-tetrachlorobiphenyl, can induce persistent aberrations in spontaneous behaviour and also affect learning and memory functions in the adult animal. Furthermore, in the adult animal the cholinergic nicotinic receptors in cerebral cortex were affected. The present study presents further evidence that the developing cholinergic system can be sensitive to lightly chlorinated ortho-substituted PCBs. Neonatal exposure of male NMRI mice on postnatal day 10 to 2,2',5,5'-tetrachlorobiphenyl (4.1 mg/kg body weight, per os) was shown to alter the response to a cholinergic agent, nicotine, at an adult age of 4 months, but not to d-amphetamine, an agent known to affect the dopaminergic system.     

Nitric oxide in the nucleus accumbens is involved in retrieval of inhibitory avoidance memory by nicotine

In the present study, the possible effect of nitric oxide agents injected into the nucleus accumbens (NAc) in the presence or absence of nicotine on morphine state-dependent memory in adult male Wistar rats was investigated. As a model of memory, a step-through type inhibitory avoidance task was used. Post-training injection of morphine (4 and 6 mg/kg) dose dependently induced the impairment of memory retention. Administration of morphine (4 and 6 mg/kg) before retention induced state-dependent retrieval of the memory acquired under post-training morphine (6 mg/kg) influence. Injection of nicotine before retention (0.25 and 0.5 mg/kg) alone and nicotine (0.1, 0.25 and 0.5 mg/kg) plus an ineffective dose of morphine (2 mg/kg) reversed the post-training morphine-induced memory impairment. The amnesia elicited by morphine (6 mg/kg) was also prevented by pre-retention intra-NAc administration of a nitric oxide synthase (NOS) inhibitor, l-NAME (0.24 μg/rat, intra-NAc). Interestingly, an ineffective dose of nicotine (0.1 mg/kg) in combination with low doses of l-NAME (0.06 and 0.12 μg/rat, intra-NAc) synergistically improved memory performance impaired by morphine given after training. It is important to note that intra-NAc administration of l-NAME before retention impaired memory retrieval by itself. In contrast, pre-retention administration of l-arginine, a nitric oxide (NO) precursor (0.25 and 0.5 μg/rat, intra-NAc), which had no effect alone, prevented the nicotine reversal of morphine effect on memory. The results suggest a possible role for nitric oxide of nucleus accumbens in the improving effect of nicotine on the morphine-induced amnesia and morphine state-dependent memory.     

Influence of potassium channel modulators on morphine state-dependent memory of passive avoidance

In a step-down passive avoidance task, the pre-training injection of 1.25-10 mg/kg of morphine impaired memory. This was restored when injection of the same dose of morphine (pre-test treatment) was repeated 24 h later (morphine state-dependent learning: morphine St-D). ATP-dependent potassium (K(ATP)) channels have been reported to be involved in several actions of morphine following mu-receptor stimulation. We have studied the effect of K(ATP) modulators and naloxone in the restoration of memory by morphine in mice. To investigate the part played by cholinergic systems in the effects of a K(ATP) antagonist (glibenclamide) on morphine St-D, we administered low doses of atropine before glibenclamide administration. Locomotor activity was also studied. Naloxone (0.06-1 mg/kg) reversed the effect of pre-test morphine administration. The effects of the K(ATP) channel blocker glibenclamide (2-18 mg/kg) were similar to those of the pre-test administration of morphine. Pre-test co-administration of glibenclamide and morphine showed no potentiation of the morphine effect. Glibenclamide alone or in combination with morphine did not affect locomotor activity. Pre-test administration of different doses of diazoxide (15-60 mg/kg), a K(ATP)-channel opener, had no effect on restoration of memory when used alone or in combination with morphine. In both cases, the locomotor activity was significantly reduced. Diazoxide blocked the effect of glibenclamide on memory recall. Low-dose atropine also prevented glibenclamide enhancement of memory recall, suggesting that this action of glibenclamide is through the cholinergic system.     

The selective 5-HT3 receptor antagonist,WAY100289, enhances spatial memory in rats with ibotenate lesions of the forebrain cholinergic projection system

The effects of three doses (0.003, 0.03 and 1.0 mg/kg sc) of the 5-HT₃ receptor antagonist, WAY 100289, on spatial learning and memory in the water maze were examined in rats before and after ibotenate lesions to the nucleus basalis and medial septal brain regions at the source of cholinergic projections to cortex and hippocampus. The representative cholinergic nicotinic and muscarinic receptor agonists nicotine (0.1 mg/kg) and arecoline (1.0 mg/kg) were also tested for comparison. Both arecoline and nicotine improved initial acquisition in rats before lesioning, in terms of latency to find a hidden platform and accuracy of search strategy. WAY100289 did not affect the performance of normal rats significantly, apart from some non-significant trends towards improvement with the highest dose. However, in animals showing transient navigational deficits in retention and relearning after lesioning, WAY100289 improved performance at all three doses, though ameliorative effects of nicotine and arecoline were more marked also in lesioned rats. These results show that WAY100289 improved spatial learning in animals impaired after lesions to cholinergic projection nuclei, which may reflect an interaction with cholinergic transmission to enhance cognitive function. However, in the present study, WAY100289 appeared to be less effective than direct cholinergic agonists.     

Memory access pathway: role of adrenergic versus cholinergic neurons

A temporary depletion of brain norepinephrine in rats produced by injection of a dopamine beta-hydroxylase inhibitor, diethyldithiocarbamate (DDC), 30 min prior to testing, prevented performance of a trained passive avoidance response 1, 3, 5 or 7 days after training. Subsequent recovery in performance indicated that the memory itself was not destroyed, but rather that the process of memory retrieval was affected. Anticholinesterase treatment produced a similar retrieval amnesia, but the effect was dependent upon the age of the memory at the time of drug injection [11]. In both cases, when the animals were presented a recall trial prior to injection, the normally observed amnesia was blocked. Animals treated with DDC up to 3 hr before training were capable of learning the passive avoidance task and of avoidance performance for a few minutes after training. However, these animals failed to produce a long-term memory of the trained response. Anticholinesterase treatment had no effect on memory formation. These results suggest different roles for adrenergic and cholinergic neurons in a pathway associated with memory storage and retrieval.     

Cholinergic modulation of memory in rats

Central cholinergic systems have long been implicated in the modulation of learning and memory processes in animals and man. Drugs that affect the central cholinergic system have been found either to enhance or to hinder performance in tests of learning and memory. Few studies have evaluated the effects of different cholinergic drugs within a single experimental paradigm and with a relatively wide dose range. The studies reported here investigated the effects of cholinergic drugs with diverse modes of action on the retention of a passive avoidance response. Physostigmine, arecoline, oxotremorine, nicotine, and 4-aminopyridine were administered IP immediately following the acquisition of a one-trial passive avoidance task. All of the drugs were found to enhance 72-h retention of passive avoidance; however, the effective doses were different for each of the drugs studied.     

Influence of central administration ATP-dependent K+ channel on morphine state-dependent memory of passive avoidance

Pre-training injection of a moderate dose of morphine (5-10 mg/kg) in a step-down passive avoidance task induced state-dependent learning with impaired memory retrieval on the test day. The impairment of memory was restored after the pre-test administration of the same dose of the drug. We have studied the effect of intracerebroventricular administration of naloxone and K(ATP) channel modulators (glibenclamide and diazoxide) on the test day on restoration of memory by morphine in mice. The effect of scopolamine on restoration of memory on the test-day by glibenclamide was studied as well. Naloxone pretreatment (0.006, 0.025 and 0.1 microg/mouse) reversed the effect of pre-test morphine administration. The K(ATP) channel blocker, glibenclamide (0.1, 0.5 and 1 microg/mouse), showed effects similar to those of pre-test administration of morphine. Glibenclamide tended to potentiate the morphine response. Scopolamine (0.15 and 0.30 microg/mouse) prevented the effect of glibenclamide on the restoration of memory. The pre-test administration of different doses of diazoxide (1.7, 5 and 15 microg/mouse), a K(ATP) channel opener, showed no effect on restoration of memory when used alone but decreased morphine state-dependence. Diazoxide blocked the effects of glibenclamide on memory restoration. It is concluded that K(ATP) channel modulators may be involved, at least in part, in morphine state dependence through a cholinergic system mechanism.     

Differential effects of scopolamine on working and reference memory depend upon level of training.

Controversy exists whether the cholinergic system in the brain is involved in working memory (WM) selectively or in both WM and reference memory (RM). Rats were trained to obtain food from four baited arms of an eight-arm radial maze. The remaining arms were never baited. Three types of errors were recorded: entry into unbaited arms (RM errors), reentry into baited arms (WM errors), and reentry into unbaited arms (WRM errors). There were no differences among three control conditions: methyl scopolamine, physiological saline, and uninjected. Scopolamine increased WM but not RM errors. When rats were trained to a higher criterion of learning, however, both WM and RM were impaired. It appears that when baseline error rate is sufficiently low RM errors under scopolamine become observable. The results suggest that the cholinergic system is involved in both WM and RM, and the selective involvement of WM is the result of insufficient training. The controversy in the literature over the involvement of the cholinergic system in WM and RM was addressed.