Elsevier

Neuropsychologia

Volume 46, Issue 10, August 2008, Pages 2476-2484
Neuropsychologia

Specific impairments in visuospatial working and short-term memory following low-dose scopolamine challenge in healthy older adults

https://doi.org/10.1016/j.neuropsychologia.2008.04.010Get rights and content

Abstract

Scopolamine-induced deficits in cognitive and motor processes have been widely demonstrated in animals and humans, although the role of acetylcholine in working memory is not as well understood. This study examined the role of acetylcholine neurotransmission in visuospatial short term and working memory using the Groton Maze Learning Test (GMLT). The GMLT is a computerized hidden maze learning test that yields measures of component cognitive processes such as spatial memory, working memory, and visuomotor function, as well as their integration in trial-and-error problem solving. Healthy older adults were administered scopolamine (0.3 mg subcutaneous), the acetlycholinesterase inhibitor donepezil (5 mg oral), scopolamine with donepezil, or placebo. Compared to placebo, low-dose scopolamine led to performance deficits on all measures of the GMLT. The greatest scopolamine-induced deficits were observed in errors reflecting working memory processes (e.g., perseverative errors d = −2.98, and rule-break errors d = −2.49) and these impairments remained robust when statistical models accounted for scopolamine-related slowing in visuomotor speed. Co-administration of donepezil partially ameliorated scopolamine-related impairments and this effect was greatest for measures of working memory than short-term memory. By itself, donepezil was associated with a small improvement in visuomotor function. These results suggest that scopolamine disrupts processes required for rule maintenance and performance monitoring, in combination with visuomotor slowing and sequential location learning.

Introduction

The role of the dopaminergic (DA) system in frontal cortex activity and working memory functions associated with the frontal cortex has been widely investigated. Animal experiments and recent pharmacological research in humans also suggests a role of the cholinergic system in working memory (Hasselmo & Stern, 2006). In humans, cholinergic neurotransmission has been widely implicated in sensory processing, attention, and memory encoding functions (Giocomo & Hasselmo, 2007). Although working memory and short-term memory impairments have been observed following administration of cholinergic antagonists in humans, this has been demonstrated mainly within the verbal domain (Curran, Schifano, & Lader, 1991; Ebert, Siepman, Oertel, Wesnes, & Kirch, 1998; Eddington & Rusted, 2003; Ellis et al., 2006; Mintzer & Griffiths, 2003; Tariot, Patel, Cox, & Henderson, 1996). It is not clear from these studies the degree to which working memory impairments produced by cholinergic antagonists are a consequence of attentional, sensory, or short-term memory impairments, and are consistent with spatial working memory deficits observed in animal studies. This study was designed to investigate the changes in short-term learning and memory, and working memory associated with modulation of central nervous system acetylcholine transmission in healthy older adults. We expected that administration of a muscarinic acetylcholine agonist (donepezil), antagonist (scopolamine), their co-administration, or placebo would produce different, but well-ordered performance profiles on a computerized version of a hidden pathway maze learning task designed to assess multiple aspects of change in different cognate abilities.

Executive function is often defined as a cognitive system that supports the integration of memory, planning, monitoring and motor functions towards goal-directed actions in a flexible manner (Baddeley, 1996, Goldman-Rakic, 1996; Levy & Goldman-Rakic, 2000; Postle, 2006). The maintenance of information in immediate memory, or short-term memory is distinguished from working memory, which requires the transformation, or manipulation of information above simple storage and retrieval (Baddely, 2003). Working memory may be regarded as a blanket term that encompasses operations such as memory updating, performance monitoring, inhibition of prepotent responses, and mental set-shifting, all of which support the optimal functioning of executive control (Lehto, Juujarvi, Kooistra, & Pulkkinen, 2003; Miyake, Friedman, Emerson, Witzki, & Howerter, 2000 but see Salthouse, Atkinson, & Berish, 2003 for a more ‘unitary’ account of executive function; and Koechlin & Summerfield, 2007 for a theory of functional hierarchical organization of the prefrontal cortex). While psychopharmacological investigations of working memory have focused on the role of dopaminergic neurotransmission (e.g., Ellis & Nathan, 2001; Gruber, Dayan, Gutkin, & Solla, 2006; Iversen & Iversen, 2007; Li, Sham, Owen, & He, 2006; Robbins, 2000), there is a growing body of evidence showing the necessary involvement of acetylcholine (i.e. cholinergic neurotransmission) in working memory functions (Giocomo & Hasselmo, 2007; Hasselmo & Stern, 2006; Mameli-Engvall et al., 2006).

In rodents, the relationship between cholinergic neurotransmission and spatial working memory has been studied thoroughly using maze learning paradigms such as the Morris Water Maze, Barnes Maze, and the Olton Radial Arm Maze (Douglas & Isaacson, 1966; Drew, Miller, & Baugh, 1973; Ennaceur & Meliani, 1992; Komater et al., 2005; Watts, Stevens, & Robinson, 1981). Blockade of nicotinic or muscarinic acetylcholine receptors causes impairment in performance on all of these tasks (Blockland, 1996, Gold, 2003; Power, Vazdarjanova, & McGaugh, 2003; Sarter & Bruno, 1997; Sarter, Bruno, & Givens, 2003). Furthermore, co-administration of compounds that stimulate the release of acetylcholine or block the enzymatic breakdown of acetylcholine can overcome the impairments in visuospatial learning, working memory, and executive functions (Friedman, 2004; Padlubnaya, Galizio, Pitts, & Keith, 2005; Sahakian et al., 1993; Wezenberg, Verkes, Sabbe, Ruigt, & Hulstijn, 2005).

In humans, the cognitive effects caused by modulation of cholinergic neurotransmission have also been well studied, mainly through blockade of muscarinic receptors with the antagonist scopolamine (Ellis et al., 2006; Wesnes, Simpson, & Kidd, 1988). However, despite the wealth of animal data on the effects of scopolamine on spatial working memory, most studies examining the effects of scopolamine on learning and memory, or working memory in humans have utilized auditory–verbal testing paradigms (e.g., effects on word list learning, paragraph learning, or verbal memory span; Curran et al., 1991; Ebert et al., 1998; Ellis et al., 2006; Mintzer & Griffiths, 2003; Tariot et al., 1996). When spatial cognition or working memory has been challenged in humans with scopolamine, the tasks chosen (e.g., Trail Making Test Part-B, delayed spatial memory, visual spatial short-term memory and delayed matching-to-sample) have required only passive storage and retrieval of spatial information (Flicker, Serby, & Ferris, 1990; Koller et al., 2003; Kopelman & Corn, 1988; Robbins et al., 1997; Rusted & Warburton, 1988; Tariot et al., 1996). Thus, while scopolamine disrupts auditory–verbal short term and working memory, and spatial short-term memory in humans, the nature or magnitude of the effect of scopolamine on ‘active’ functions of spatial working memory is unclear.

Identification of the effects of cholinergic neurotransmission on attention, short-term memory, and working memory has been difficult to determine using standard neuropsychological tests for a number of reasons. First, most studies investigating the effects of compounds that act on the cholinergic system have been based on performance across batteries of neuropsychological tests that may differ in sensitivity to detect cognitive impairment or change (Collie, Maruff, Faletti, Silbert, & Darby, 2002; Osterberg, Orbaek, Karlson, Bergendorf, & Seger, 2000). Theoretical inferences regarding the presence and relative magnitude of impairment in domain, or process-specific functioning may be confounded by the different psychometric properties of tests used for comparison (Lewis, Maruff, & Silbert, 2005; Lewis, Maruff, Silbert, Evered, & Scott, 2006). Second, identifying the underlying processes which lead to impaired performance on complex tests of short-term memory or working memory are difficult to establish from methods that produce single outcome measures (such as speed or accuracy). The disruption to executive processes (e.g., working memory, goal maintenance), supporting functions (e.g., sensory processing, attention), or the domain itself (e.g., language, visuospatial or numerical processing) can produce sub-optimal performance on tests of purported executive functions (Chan, Shum, Toulopoulou, & Chen, 2008; Royall et al., 2002). The problem of inferring process- or domain-specific activity from variably complex tasks has been noted by many researchers (Burgess, 1997; Conway, Kane, & Engle, 2003; Hamilton, Coates, & Heffernan, 2003; Jacoby, 1991; Merian, 2005). Of course it is also difficult to dissociate completely verbal and visuospatial executive functions in humans. When faced with difficult visual tasks, humans generally develop verbal strategies to guide performance on spatial tasks even in the absence of linguistic stimuli, performance rules expressed verbally, or requirements for spoken responses (Gillett, 2007). These issues aside, it is increasingly recognized that in complex tasks, it is important to identify the sources of performance variability within and between groups and individuals (Leite, Ratcliff, & Hale, 2007; Ratcliffe, Spieler, & McKoon, 2000; Verhagen, Cerella, Semenec, Leo, & Steitz, 2002), and to use appropriate statistical methods for characterizing the parameters of change or impairment (Collie et al., 2002, Maruff et al., 2006). One approach that can accommodate these imperatives is to use tasks that can measure component processes that contribute to performance on the same test, and from which valid comparisons of relative cognitive impairments can be inferred using appropriate statistical methods. Such an approach need not assume that the underlying processes are independent or organized hierarchically but can investigate associations between measures of the different component processes.

In a recent study, we used performance on a hidden pathway maze learning task (the Groton Maze Learning Test; GMLT) to investigate pharmacodynamic response to acute doses of scopolamine and donepezil, alone and together in healthy older people (Snyder, Bednar, Cromer, & Maruff, 2005). Older participants were recruited because of the known decline in cholinergic function with aging, hence the sensitivity of this population to compounds which enhance or impair the cholinergic system (Ellis & Nathan, 2001). The GMLT was chosen for this study because of its demonstrated sensitivity to drug effects, and its demonstrated neuroscientific validity, brevity, repeatability, and utility across groups according to age, clinical, or cultural status (e.g., Maruff et al., 2006; Pietrzak, Cohen, & Snyder, 2007; Schroder, Snyder, Seilski, & Mayes, 2004; Snyder, Maruff, Pietrzak, Cromer, & Snyder, 2008). The GMLT is adapted from the Milner “stepping stone” maze (Milner, 1965) and the Austin maze (Morrison & Gates, 1988). The GMLT is administered via touchscreen computer monitor, and individuals are required to identify, one step at a time, a pathway that is hidden in a 10 × 10 matrix. There are four rules of the task, which aid the discovery of the pathway by reducing the possible number of correct locations at each decision point to a maximum of three possibilities. That is, subjects are instructed not to move diagonally, not to go backwards on the path unless they make an error, not to ‘jump’ tiles, and to return to the last correct location after an error (see Fig. 1 for an image of the appearance of the task to a subject). The GMLT software measures the latency of each response and classifies it as correct or incorrect according to error type. The pathway is never revealed in its entirety, and the subject sees only the current correct location. Once the subject has worked their way through to the end of the pathway they must return to the start and find the same pathway again. Successful performance on the GMLT requires the individual to hold the task rules in working memory and use this information to guide trial-and-error learning in accordance with moment-to-moment performance feedback and the developing representation in memory of the actual pathway location.

Analysis of the responses made in learning the maze can indicate the integrity of different cognitive operations after pharmacological challenge (Snyder et al., in press). For example, assuming that subjects adhere to the task rules, errors made in finding the pathway (i.e. exploratory errors) reflect the ability to build a short-term spatial representation of the pathway in memory. The number of rule-break errors provides an index of the ability of individuals to hold and use the general rules in working memory, and to monitor ongoing performance. Perseverative errors (or the same error made in succession) provide an index of the individual's tendency to automatically repeat errors they have already made. Hence, while exploratory errors reflect spatial working memory processes, the combined measures of perseverative and rule-break errors provide a measure of working memory functions such as error/performance monitoring (Pietrzak et al., 2007; Pietrzak, Maruff, Mayes, Roman, Sosa, & Snyder, in press). The number of correct speeded responses made on a simple reaction time version of the maze (Chase test) can provide an index of visuomotor function. This can be used to establish the extent to which changes in working memory due to modulation of cholinergic neurotransmission are an indirect consequence of the reduced alertness, sedation, and changes in sensory processing that can occur with muscarinic antagonists (e.g., Callaway, Halliday, Naylor, & Schechter, 1985; Cohen, Gross, Semple, Nordahl, & Sunderland, 1994; Curran et al., 1991; Eddington & Rusted, 2003; Robbins et al., 1997; Sipos, Burchnell, & Galbicka, 1999; Wesnes & Warburton, 1984).

In a previous paper, we reported that scopolamine (0.3 mg s.c.) induced deficits in processing speed and learning efficiency that were greatest 2.5 h after scopolamine administration (Snyder et al., 2005). Administration of donepezil (5 mg oral) 3 h prior to scopolamine ameliorated these deleterious effects. Interestingly, administration of donepezil by itself actually improved performance on the GMLT above baseline levels. However, the original study was designed to use the maze to measure the pharmacodynamic effects of scopolamine and donepezil administered alone and together. Consequently, only summary measures of maze performance were analyzed. There was no investigation of the interaction between the measures of the different cognitive operations necessary for optimal problem solving on the maze. Thus, while the earlier findings provide prima facie evidence that working memory is affected by modulation of cholinergic neurotransmission, we believe that a more detailed analysis is required to identify the associations between different cognitive operations pertaining to working and short-term memory on the GMLT.

The current paper presents that analysis, with the goal of decomposing performance on the hidden maze learning task into the component operations of spatial and working memory and examining how these operations are modulated by acetylcholine neurotransmission. We hypothesized that scopolamine would interfere with all of the component cognitive processes necessary for maze learning and that scopolamine-related impairment would be ameliorated by donepezil. We investigated the extent to which modulation of acetylcholine affected visuomotor speed, short term and working memory components of maze learning. These relationships were examined by computing standardized measures of change for the different outcome measures and comparing these between treatment conditions.

Section snippets

Subjects

The subject group consisted of 32 healthy older (mean age = 71 years, range = 65–90 years) men (N = 12) and women (N = 20). All individuals satisfied rigorous inclusion and exclusion criteria (summarized below), and were both willing and able to provide written informed consent. All subjects had normal or corrected to normal visual and auditory acuity, were in good general health or without any clinically significant abnormalities, had a normal ECG, and a Mini-Mental State Exam Score ≥28/30. Women were

Results

Two subjects were excluded from analysis due to missing data, resulting in 15 subjects in each treatment group. Group means and standard deviations for each of the GMLT outcome measures at baseline and 5.5 h after baseline under each treatment condition are shown in Table 1. The group means for the standardized changes from baseline scores are shown in Fig. 2 for each of the maze outcome measures. Fig. 2 indicates that, relative to baseline, the placebo group showed subtle improvement from

Discussion

The current study examined the extent to which modulation of cholinergic neurotransmission altered the different component cognitive processes of visuospatial executive function in older adults. As observed previously, blockage of muscarinic receptors with scopolamine induced large impairments in all aspects of performance on the hidden pathway maze learning task (Snyder et al., 2005). However, the magnitude of this deleterious effect was greater for working memory processes (i.e. rule-break

References (93)

  • M. Hasselmo et al.

    Mechanisms underlying working memory for novel information

    Trends in Cognitive Science

    (2006)
  • S.D. Iversen et al.

    Dopamine: 50 years in perspective

    Trends in Neurosciences

    (2007)
  • L.L. Jacoby

    A process dissociation framework: Separating automatic from intentional uses of memory

    Journal of Memory and Language

    (1991)
  • R.P.C. Kessels et al.

    Varieties of human spatial memory: A meta-analysis of the effects of hippocampal lesions

    Brain Research Reviews

    (2001)
  • E. Koechlin et al.

    An information theoretical approach to prefrontal executive function

    Trends in Cognitive Sciences

    (2007)
  • V.A. Komater et al.

    Effects of histamine H3 receptor antagonists in two models of spatial learning

    Behavioral Brain Research

    (2005)
  • M.S. Lewis et al.

    Detection of postoperative cognitive decline in choronary artery bypass graft surgery is affected by the number of neuropsychological tests in the assessment battery

    Annals of Thoracic Surgery

    (2006)
  • M. Mameli-Engvall et al.

    Hierarchical control of dopamine neuron-firing patterns by nicotinic receptors

    Neuron

    (2006)
  • B. Milner

    Visually-guided maze learning in man: Effects of bilateral hippocampal, bilateral frontal, and unilateral cerebral lesions

    Neuropsychologia

    (1965)
  • A. Miyake et al.

    The unity and diversity of executive functions and their contributions to complex “frontal lobe” tasks: A latent variable analysis

    Cognitive Psychology

    (2000)
  • M. Petrides et al.

    Deficits on subject-ordered tasks after frontal and temporal-lobe lesions in man

    Neuropsychologia

    (1982)
  • R.H. Pietrzak et al.

    Spatial learning efficiency and error monitoring in normal aging: An investigation using a novel hidden maze learning test

    Archives of Clinical Neuropsychology

    (2007)
  • B. Postle

    Working memory as an emergent property of the mind and brain

    Neuroscience

    (2006)
  • A.E. Power et al.

    Muscarinic cholinergic influences in memory consolidation

    Neurobiology of Learning and Memory

    (2003)
  • M.E. Ragozzino

    Acetylcholine actions in the dorsomedial striatum support the flexible shifting of response patterns

    Neurobiology of Learning and Memory

    (2003)
  • M.E. Ragozzino et al.

    Involvement of the dorsiomedial striatum in behavioral flexibility: Role of muscarinic receptors

    Brain Research

    (2002)
  • J.L. Rogers et al.

    Cholinergic modulation of the hippocampus during encoding and retrieval

    Neurobiology of Leaning and Memory

    (2003)
  • M. Sarter et al.

    Cognitive functions of cortical acetylcholine: Toward a unifying hypothesis.

    Brain Research Reviews

    (1997)
  • M. Sarter et al.

    Attentional functions of cortical cholinergic inputs: What does it mean for learning and memory?

    Neurobiology of Learning and Memory

    (2003)
  • M.D. Schroder et al.

    Impaired performance of children exposed in utero to cocaine on a novel test of visuospatial working memory

    Brain and Cognition

    (2004)
  • P.J. Snyder et al.

    Reversal of scopolamine-induced deficits with a single dose of donepezil, an acetylcholinesterase inhibitor

    Alzheimer's & Dementia

    (2005)
  • C.M. Thiel

    Cholinergic modulation of learning and memory in the human brain as detected with functional neuroimaging

    Neurobiology of Learning and Memory

    (2003)
  • J. Watts et al.

    Effects of scopolamine on radial maze performance in rats

    Physiology and Behavior

    (1981)
  • A. Baddeley

    Exploring the central executive

    The Quarterly Journal of Experimental Psychology

    (1996)
  • A. Baddely

    Working memory: Looking back and looking forward

    Nature Reviews Neuroscience

    (2003)
  • J.M. Bland et al.

    Multiple significance tests: The Bonferroni method

    British Medical Journal

    (1995)
  • J.M. Bland et al.

    Measurement error

    British Medical Journal

    (1996)
  • A. Blockland

    Acetylcholine: A neurotransmitter for learning and memory?

    Brain Research Reviews

    (1996)
  • E. Bullmore et al.

    Practice and difficulty evoke anatomically and pharmacologically dissociable brain activation dynamics

    Cerebral Cortex

    (2003)
  • P.W. Burgess

    Theory and methodology in executive function research

  • S.M. Cabrera et al.

    Selective lesions of the nucleus basalis magnocellularis impairs cognitive flexibility

    Behavioral Neuroscience

    (2006)
  • E. Callaway et al.

    Effects of oral scopolamine in human stimulus evaluation

    Psychopharmacology

    (1985)
  • J. Cohen

    Statistical power analysis for the behavioral sciences

    (1988)
  • R.M. Cohen et al.

    The metabolic brain of young subjects given scopolamine

    Experimental Brain Research

    (1994)
  • H.V. Curran et al.

    Models of memory dysfunction? A comparison of the effects of scopolamine and lorazepam on memory, psychomotor performance, and mood

    Psychopharmacology

    (1991)
  • M.C. Davidson et al.

    Local infusion of scopolamine into intraparietal cortex slows covert orienting in rhesus monkeys

    Journal of Neurophysiology

    (2000)
  • Cited by (44)

    • The relationship between cerebral and retinal microbleeds in cerebral amyloid angiopathy (CAA): A pilot study

      2021, Journal of the Neurological Sciences
      Citation Excerpt :

      All participants completed neuropsychological evaluation and a thorough neurological, psychological, and medical history by trained staff and verified via electronic medical record. Neuropsychological evaluation using the Cogstate Brief Battery [28,29] included measures of episodic memory (International Shopping List Test, ISLT, [30]), visuospatial learning and executive function (Groton Maze Learning Test, GMLT, [31]), visual memory (one card learning task, OCL [28],) and the working memory (one back test, OBK, [28]). Participants completed the Geriatric Depression Scale – 15 item (GDS) to assess mood and the Mini Mental Status Exam (MMSE) to assess general cognitive function.

    • Exploring the interoceptive stimulus effects of nicotine and varenicline

      2019, Pharmacology Biochemistry and Behavior
      Citation Excerpt :

      The inability to alter a response to a new stimulus-reinforcer relation is often referred to as perseverative errors or perseverative responding. Perseverative errors are commonly seen in successive reversal paradigms in which the reinforcer relations are flipped multiple times akin to an ABAB study design (Clarke et al., 2008; Prado et al., 2017; Thomas et al., 2008). In the DGT task, they manifest as the inability to withhold responding on non-reinforced sessions of the reversal groups.

    • No synergistic effect of subtherapeutic doses of donepezil and EVP-6124 in healthy elderly subjects in a scopolamine challenge model

      2019, Alzheimer's and Dementia: Translational Research and Clinical Interventions
      Citation Excerpt :

      The application of the scopolamine challenge model is the most extensively used pharmacological model of cognitive impairment [17]. Previous studies have shown that a dose of 0.5 mg intravenously induces significant cognitive deficits in healthy young volunteers, while in healthy elderly volunteers, a subcutaneous dose of 0.3 mg resulted in quantifiable and reproducible cognitive deficits [14,15,18]. Because intravenous dosing was expected to lead to a shorter duration of effect with only a slightly higher Cmax, it was decided to administer a dose of 0.3 mg scopolamine intravenously to the healthy elderly volunteers in this study [19].

    • Change in retinal structural anatomy during the preclinical stage of Alzheimer's disease

      2018, Alzheimer's and Dementia: Diagnosis, Assessment and Disease Monitoring
    • Can visuospatial measures improve the diagnosis of Alzheimer's disease?

      2018, Alzheimer's and Dementia: Diagnosis, Assessment and Disease Monitoring
    • The muscarinic system, cognition and schizophrenia

      2015, Neuroscience and Biobehavioral Reviews
      Citation Excerpt :

      This drug-induced psychotic state can be treated with antipsychotic medications or acetylcholinesterase inhibitors, in addition to discontinuing the anti-muscarinic treatment that trigged the state (Göpel et al., 2002; Granacher and Baldessarini, 1975; Nogué et al., 1991). Similarly, acute administration of anti-muscarinic agents can also produce cognitive impairments amongst otherwise healthy human subjects that mimic the deficits documented amongst schizophrenia patients (e.g. Green et al., 2005; Mintzer and Griffiths, 2003, 2007; Thomas et al., 2008). Parallel to these findings in humans, administration of muscarinic antagonists has been shown to induce psychotic-like effects in various animal models of schizophrenia (see Andrus et al., 2007; Barak, 2009; Barak and Weiner, 2011a; Klinkenberg and Blokland, 2010).

    View all citing articles on Scopus
    View full text