Impaired context reversal learning, but not cue reversal learning, in patients with amnestic mild cognitive impairment

It has been proposed that reversal learning is impaired following damage to the orbitofrontal and ventromedial frontal cortex (OFC/VMFC) and to the medial temporal lobe (MTL), including the hippocampal formation. However, the exact characteristics of the MTL-associated reversal learning deficit are not known. To investigate this issue, we assessed 30 newly diagnosed patients with amnestic mild cognitive impairment (aMCI) and 30 matched healthy controls. All patients fulfilled the aMCI criteria of the Mayo Clinic Alzheimer's Disease Research Center and underwent head magnetic resonance imaging that confirmed MTL atrophy. Reversal learning was assessed using a novel reinforcement learning task. Participants first acquired and then reversed stimulus-outcome associations based on negative and positive feedback (losing and gaining points). Stimuli consisted of a cue (geometric shapes) and a spatial context (background color or pattern). Neuropsychological assessment included tasks related to the MTL (paired associates learning), dorsolateral prefrontal cortex (DLPFC) (extradimensional shift, One-touch Stockings of Cambridge), and OFC/VMFC (Holiday Apartment Task). Results revealed that, relative to controls, patients with aMCI exhibited a marked reversal learning deficit, which was highly selective for the reversal of context. The acquisition of stimulus-outcome associations and cue reversal learning were spared. Performance on the context reversal learning task significantly correlated with the right hippocampal volume. In addition, patients with aMCI had deficits on tests related to DLPFC but not to OFC/VMFC. However, DLPFC dysfunctions were not associated with context reversal learning. These results suggest that MTL deficits in aMCI selectively affect context reversal learning when OFC/VMFC functions are spared. This deficit is not influenced by the valence of the outcome (positive or negative feedback) and by executive dysfunctions.     

Learned equivalence in schizophrenia: novel method for the measurement of memory functions of the hippocampus and basal ganglia

The purpose of the present study was to investigate basal ganglia (BG) and medial temporal lobe (MTL) dependent learning in patients with schizophrenia. Acquired equivalence is a phenomenon in which prior training to treat two stimuli as equivalent (if two stimuli are associated with the same response) increases generalization between them. The learning of stimulus-response pairs is related to the BG, whereas the MTL system participates in stimulus generalization. Forty-three patients with DSM-IV schizophrenia and 28 matched healthy controls participated in the study. Volunteers received the acquired equivalence associative learning task (FISHES) during which associations between faces and pet fishes are learned. Results revealed that patients with schizophrenia showed a selective deficit on stimulus generalization, whereas stimulus-response learning was unaffected. The number of errors during stimulus-response learning correlated with the daily chlorpromazine-equivalent dose of antipsychotics. In conclusion, patients with schizophrenia showed deficits during MTL-dependent learning, but not during BG-dependent learning. High-dose first generation antipsychotics may disrupt BG-dependent learning by blocking dopaminergic neurotransmission in the nigro-striatal system.     

Probabilistic classification learning with corrective feedback is selectively impaired in early Huntington's disease--evidence for the role of the striatum in learning with feedback

In general, declarative learning is associated with the activation of the medial temporal lobes (MTL), while the basal ganglia (BG) are considered the substrate for procedural learning. More recently it has been demonstrated the distinction of these systems may not be as absolute as previously thought and that not only the explicit or implicit nature of the memory task alone is important for the distinction of MTL or BG systems. Nevertheless, patients with BG dysfunction - such as patients with Parkinson's disease (PD) or Huntington's disease (HD) - are considered to be impaired at implicit learning. However, a more recent study demonstrated that one implicit learning task, probabilistic classification learning (examples include the weather prediction (WPT) and Mr. Potato Head tasks) is only impaired in PD when it involves learning with corrective feedback (FB) but not when it involves learning in a paired associate (PA) manner, without feedback. Therefore, it has been argued that the presence of feedback rather than the implicit nature of these tasks determines whether or not the BG are recruited. As patients with HD as well as those with PD, have also been shown to be impaired on the standard FB based version of probabilistic classification learning, the question remains as to whether or not there is a similar selective deficit in FB but not PA based probabilistic classification learning in HD. 18 patients with early HD and 18 healthy controls completed FB and PA versions of the WPT task. Relative to controls, HD patients were selectively impaired at WPT learning with feedback. These findings are consistent with previous evidence from studies of probabilistic classification learning in PD. Unlike PD, selective deficits in WPT learning in HD cannot be attributed to the effects of dopaminergic medication and must be directly related to BG dysfunction; for instance even in early HD, only 50% of the neurons in the medial head of caudate remain. We conclude that the striatum is important for WPT learning with feedback. Our findings are consistent with imaging evidence showing recruitment of the caudate during FB based WPT learning, while the MTL is associated with PA based learning.     

The role of the basal ganglia and its cortical connections in sequence learning: Evidence from implicit and explicit sequence learning in Parkinson's disease

Implicit (unconscious/incidental) and explicit (conscious/intentional) learning are considered to have distinct neural substrates. It is proposed that implicit learning is mediated by the basal ganglia (BG), while explicit learning has been linked to the medial temporal lobes (MTL). To test such a dissociation we investigated implicit and explicit sequence learning in Parkinson's disease (PD), a disorder characterized by striatal dysfunction. We studied both implicit and explicit learning of a 12-item sequence of target locations in 13 PD patients and 15 age-matched controls. In the implicit sequence learning task all participants completed 10 blocks of a probabilistic serial reaction time (SRT) task in which they were exposed to the sequence without explicit knowledge of it. Participants also completed between 1 and 10 blocks of an explicit sequence learning task in which the sequence was learned deliberately by trial-and-error. Both implicit and explicit sequence learning were significantly impaired in PD patients compared to controls. The results indicate that, in addition to playing a role in implicit sequence learning, the BG and its frontal projections are also involved in explicit sequence learning.     

How to find the way out from four rooms? The learning of "chaining" associations may shed light on the neuropsychology of the deficit syndrome of schizophrenia

Recent meta-analytic evidence suggests that clinical neuropsychological methods are not likely to uncover circumscribed cognitive impairments in the deficit syndrome of schizophrenia. To overcome this issue, we adapted a cognitive neuroscience perspective and used a new "chaining" habit learning task. Participants were requested to navigate a cartoon character through a sequence of 4 rooms by learning to choose the open door from 3 colored doors in each room. The aim of the game was to learn the full sequence of rooms until the character reached the outside. In the training phase, each stimulus leading to reward (open door in each room) was trained via feedback until the complete sequence was learned. In the probe phase, the context of rewarded stimuli was manipulated: in a given room, in addition to the correct door of that room, there also appeared a door which was open in another room. Whereas the training phase is dominantly related to basal ganglia circuits, the context-dependent probe phase requires intact medial-temporal lobe functioning. Results revealed that deficit and non-deficit patients were similarly impaired on the probe phase compared with controls. However, the training phase was only compromised in deficit patients. More severe negative symptoms were associated with more errors on the training phase. Executive functions were unrelated to performance on the "chaining" task. These results indicate that the deficit syndrome is associated with prominently impaired stimulus-response reinforcement learning, which may indicate abnormal functioning of basal ganglia circuits.     

Dissociation between medial temporal lobe and basal ganglia memory systems in schizophrenia

The purpose of this study was to investigate basal ganglia (BG) and medial temporal lobe (MTL) dependent learning in patients with schizophrenia. Acquired equivalence is a phenomenon in which prior training to treat two stimuli as equivalent (if two stimuli are associated with the same response) increases generalization between them. The learning of stimulus-response pairs is related to the BG, whereas the MTL system participates in stimulus generalization. Forty-three patients with DSM-IV schizophrenia and 28 matched healthy controls participated. Volunteers received the Rutgers acquired equivalence task (face-fish task) by [Myers, C.E., Shohamy, D., Gluck, M.A. et al., 2003. Dissociating hippocampal versus basal ganglia contributions to learning and transfer. J. Cogn. Neurosci. 15, 185-193.], the California Verbal Learning Test (CVLT), and the n-back working memory test. The Rutgers acquired equivalence task investigates BG-dependent processes (stimulus-response learning) and MTL-dependent processes (stimulus generalization) with a single test. Results revealed that patients with schizophrenia showed a selective deficit on stimulus generalization, whereas stimulus-response learning was spared. The stimulus generalization deficit correlated with the CVLT performance (total scores from trials 1-5 and long-delay recall), but not with the n-back test performance. The number of errors during stimulus-response learning correlated with the daily chlorpromazine-equivalent dose of antipsychotics. In conclusion, this is the first study to show that patients with schizophrenia exhibit deficits during MTL-dependent learning, but not during BG-dependent learning within a single task. High-dose first generation antipsychotics may disrupt BG-dependent learning by blocking dopaminergic neurotransmission in the nigro-stiratal system.     

The role of dopamine in cognitive sequence learning: evidence from Parkinson's disease

Electrophysiological and computational studies suggest that nigro-striatal dopamine may play an important role in learning about sequences of environmentally important stimuli, particularly when this learning is based upon step-by-step associations between stimuli, such as in second-order conditioning. If so, one would predict that disruption of the midbrain dopamine system--such as occurs in Parkinson's disease--may lead to deficits on tasks that rely upon such learning processes. This hypothesis was tested using a "chaining" task, in which each additional link in a sequence of stimuli leading to reward is trained step-by-step, until a full sequence is learned. We further examined how medication (L-dopa) affects this type of learning. As predicted, we found that Parkinson's patients tested 'off' L-dopa performed as well as controls during the first phase of this task, when required to learn a simple stimulus-response association, but were impaired at learning the full sequence of stimuli. In contrast, we found that Parkinson's patients tested 'on' L-dopa performed better than those tested 'off', and no worse than controls, on all phases of the task. These findings suggest that the loss of dopamine that occurs in Parkinson's disease can lead to specific learning impairments that are predicted by electrophysiological and computational studies, and that enhancing dopamine levels with L-dopa alleviates this deficit. This last result raises questions regarding the mechanisms by which midbrain dopamine modulates learning in Parkinson's disease, and how L-dopa affects these processes.     

Functional imaging of sequence learning in Parkinson's disease

Sequence learning, a cognitive task linked to cortico-striatal function, is impaired in Parkinson's disease (PD). We chose this task as a behavioral paradigm to study the functional architecture of PD in treated and untreated conditions. In our studies, participants were scanned with H(2)(15)O while performing a kinematically controlled motor sequence learning task and a matching motor baseline task. Experiments revealed that a specific sequence learning network predicts learning in normal subjects, and in independent cohorts of early and advanced PD patients. The analysis of the relationship of network activity to learning performance revealed diverging influences of dopaminergic therapy and deep brain stimulation (DBS). DBS of the internal GP and of STN increased network activity and task performance, while levodopa decreased both measures. In separate studies, we investigated the role of dopaminergic modulation on brain activation during sequence learning. In healthy subjects dopamine transporter (DAT) binding correlated with learning-related brain activation in prefrontal, premotor and cingulate cortices, and in the thalamus. By contrast, in PD most of these regional relationships were lost. Only ventral and dorsolateral prefrontal cortex activation correlated with caudate dopaminergic input. In a final set of studies, we found a significant decline in learning performance in early stage PD patients followed over the course of 2 years. Longitudinal declines in learning-related activation were found in parietal areas, while concomitant increases were localized to the left hippocampus. These observations support hypotheses on disease-stage and task-specific effects within the different cortico-striato-pallido-thalamocortical loops and the mesocortical system in PD.     

Sequence learning is preserved in individuals with cerebellar degeneration when the movements are directly cued

Cerebellar pathology is associated with impairments on a range of motor learning tasks including sequence learning. However, various lines of evidence are at odds with the idea that the cerebellum plays a central role in the associative processes underlying sequence learning. Behavioral studies indicate that sequence learning, at least with short periods of practice, involves the establishment of effector-independent, abstract spatial associations, a form of representation not associated with cerebellar function. Moreover, neuroimaging studies have failed to identify learning-related changes within the cerebellum. We hypothesize that the cerebellar contribution to sequence learning may be indirect, related to the maintenance of stimulus-response associations in working memory, rather than through processes directly involved in the formation of sequential predictions. Consistent with this hypothesis, individuals with cerebellar pathology were impaired in learning movement sequences when the task involved a demanding stimulus-response translation. When this translation process was eliminated by having the stimuli directly indicate the response location, the cerebellar ataxia group demonstrated normal sequence learning. This dissociation provides an important constraint on the functional domain of the cerebellum in motor learning.     

Rule-based categorization deficits in focal basal ganglia lesion and Parkinson's disease patients

Patients with basal ganglia (BG) pathology are consistently found to be impaired on rule-based category learning tasks in which learning is thought to depend upon the use of an explicit, hypothesis-guided strategy. The factors that influence this impairment remain unclear. Moreover, it remains unknown if the impairments observed in patients with degenerative disorders such as Parkinson's disease (PD) are also observed in those with focal BG lesions. In the present study, we tested patients with either focal BG lesions or PD on two categorization tasks that varied in terms of their demands on selective attention and working memory. Individuals with focal BG lesions were impaired on the task in which working memory demand was high and performed similarly to healthy controls on the task in which selective-attention demand was high. In contrast, individuals with PD were impaired on both tasks, and accuracy rates did not differ between on and off medication states for a subset of patients who were also tested after abstaining from dopaminergic medication. Quantitative, model-based analyses attributed the performance deficit for both groups in the task with high working memory demand to the utilization of suboptimal strategies, whereas the PD-specific impairment on the task with high selective-attention demand was driven by the inconsistent use of an optimal strategy. These data suggest that the demands on selective attention and working memory affect the presence of impairment in patients with focal BG lesions and the nature of the impairment in patients with PD.     

Impaired dimensional selection but intact use of reward feedback during visual discrimination learning in Parkinson's disease

It has been suggested that Parkinson's disease (PD) impairs the ability to learn on the basis of reward or reinforcing feedback i.e., by trial-and-error. In many learning tasks, particular 'dimensions' of stimulus information are relevant whilst others are irrelevant; therefore, efficient performance depends on identifying the dimensions of these 'compound' stimuli and selecting the relevant dimension for further processing. We investigated the ability of patients with PD, as well as patients with Huntington's disease and patients with frontal or temporal lobe lesions, to learn visual discriminations which required either a number of associations to be learned concurrently (the 'eight-pair' task) or the selection of information from compound stimuli (the 'five-dimension' task), both tasks being learned by trial-and-error. None of the basal ganglia disorder patient groups was impaired on the eight-pair task, militating against a crucial role for these brain structures in trial-and-error learning per se. Patients with mild, medicated PD, but not unmedicated PD patients, were impaired at identifying all five feature dimensions in the five-dimension task, implying dopaminergic 'overdosing' of the ability to analyse compound stimuli in terms of their component dimensions. Temporal lobe lesion patients performed similarly, suggesting that the temporal lobe may be the site of the medication overdose effect. Patients with severe, medicated PD were impaired at compound discrimination learning on the five-dimension task in the absence of an underlying impairment in identifying component stimulus dimensions; this pattern resembled that seen in Huntington's disease and frontal lobe lesion patients, implying that fronto-striatal circuitry is involved in the formation of rules based upon selected stimulus dimensions.     

The implicit sequence learning deficit in patients with Parkinson's disease: a matter of impaired sequence integration?

Despite the wealth of research investigating the serial reaction time (SRT) learning abilities of people with Parkinson's disease (PD), the role of the basal ganglia in implicit sequence learning remains largely unclear. The present research sought to examine the ability of people with PD to implicitly learn simultaneously operating sequences and integrate patterned information from each sequence dimension. Using a version of the SRT which reduced motor demands, the present experiment investigated the implicit learning of a spatial sequence, a stimulus-response sequence, and an integrated spatial/stimulus-response sequence, all of which are usually confounded in the standard SRT task. Whereas both PD and control groups demonstrated robust learning for the individual spatial and response sequences, only control participants evidenced learning for the integrated sequence. Further, unlike implicit learning for the spatial and object sequences, impaired integrated sequence acquisition was specifically related to the severity of patients' PD symptomatology. The implicit learning deficits of PD patients are discussed with regard to the role played by the basal ganglia in integrative sequence learning in the SRT.     

Spatial and temporal sequence learning in patients with Parkinson's disease or cerebellar lesions

The functional role of different subcortical areas in sequence learning is not clear. In the current study, Parkinson's patients, patients with cerebellar damage, and age-matched control participants performed a serial reaction time task in which a spatial sequence and a temporal sequence were presented simultaneously. The responses were based on the spatial sequence, and the temporal sequence was incidental to the task. The two sequences were of the same length, and the phase relationship between them was held constant throughout training. Sequence learning was assessed comparing performance when both sequences were present versus when the dimension of interest was randomized. In addition, sequence integration was assessed by introducing phase-shift blocks. A functional dissociation was found between the two patient groups. Whereas the Parkinson's patients learned the spatial and temporal sequences individually, they did not learn the relationship between the two sequences, suggesting the basal ganglia play a functional role in sequence integration. In contrast, the cerebellar patients did not show any evidence of sequence learning at all, suggesting the cerebellum might play a general role in forming sequential associations.     

Selective impairments in implicit learning in Parkinson's disease

The basal ganglia are thought to participate in implicit sequence learning. However, the exact nature of this role has been difficult to determine in light of the conflicting evidence on implicit learning in subjects with Parkinson's disease (PD). We examined the performance of PD subjects using a modified form of the serial reaction time task, which ensured that learning remained implicit. Subjects with predominantly right-sided symptoms were trained on a 12-element sequence using the right hand. Although there was no evidence of sequence learning on the basis of response time savings, the subjects showed knowledge of the sequence when performance was assessed in terms of the number of errors made. This effect transferred to the left (untrained) hand as well. Thus, these data demonstrate that PD patients are not impaired at implicitly learning sequential order, but rather at the translation of sequence knowledge into rapid motor performance. Furthermore, the results suggest that the basal ganglia are not essential for implicit sequence learning in PD.     

Sequence learning in Parkinson's disease: a comparison of spatial-attention and number-response sequences

The serial reaction time (SRT) task has been frequently used to assess procedural learning of sequences. Patients with Parkinson's disease (PD) have been reported to show deficits on this task, but it is as yet unclear whether this impairment reflects a general sequencing deficit or a deficit in the sequencing of motor-output responses. In order to examine this issue, PD patients and controls were administered an SRT task which allowed the simultaneous and independent assessment of the procedural learning of spatial regularities and the learning of motor-response regularities. PD patients were unimpaired at learning a sequence of spatial locations, but showed a deficit at learning a stimulus-to-motor-response sequence. The results suggest that sequencing impairments in PD are not general, but specific to the type of sequential information inherent in a task.     

Intact artificial grammar learning in patients with cerebellar degeneration and advanced Parkinson's disease

In an artificial grammar learning task, subjects were asked to memorise short lists of letter strings formed according to complex rules for letter order. After an interval they were unexpectedly asked to discriminate new grammatical strings from strings which used the same letters but violated the sequential constraints of the grammar. Artificial grammar learning can be mastered successfully by amnesic patients and is considered to be an implicit learning task independent of declarative learning and memory mechanisms. In this study, 10 patients with cerebellar degeneration (CD), 21 Parkinson's disease (PD) and 15 control subjects were tested on artificial grammar learning. Additionally PD patients with advanced disease were examined under adequate medication and dopaminergic withdrawal. All patient groups showed intact artificial grammar learning. Neither cerebellar damage nor basal ganglia dysfunction nor dopaminergic medication impairs or affects artificial grammar learning. Although the patients showed significant executive dysfunction, implicit learning remains intact. The conclusion is that cerebellar and basal ganglia circuits play no essential part in this kind of implicit learning. The results suggest that artificial grammar learning is a cortically mediated function comparable to the mechanism of visual priming.     

Aberrant learning in Parkinson's disease: A neurocomputational study on bradykinesia

Parkinson's disease (PD) is a neurodegenerative disorder characterized by a progressive decline in motor functions, such as bradykinesia, caused by the pathological denervation of nigrostriatal dopaminergic neurons within the basal ganglia (BG). It is acknowledged that dopamine (DA) directly affects the modulatory role of BG towards the cortex. However, a growing body of literature is suggesting that DA‐induced aberrant synaptic plasticity could play a role in the core symptoms of PD, thus recalling for a “reconceptualization” of the pathophysiology. The aim of this work was to investigate DA‐driven aberrant learning as a concurrent cause of bradykinesia, using a comprehensive, biologically inspired neurocomputational model of action selection in the BG. The model includes the three main pathways operating in the BG circuitry, that is the direct, indirect and hyperdirect pathways, and use a two‐term Hebb rule to train synapses in the striatum, based on previous history of rewards and punishments. Levodopa pharmacodynamics is also incorporated. Through model simulations of the Alternate Finger Tapping motor task, we assessed the role of aberrant learning on bradykinesia. The results show that training under drug medication (levodopa) provides not only immediate but also delayed benefit lasting in time. Conversely, if performed in conditions of vanishing levodopa efficacy, training may result in dysfunctional corticostriatal synaptic plasticity, further worsening motor performances in PD subjects. This suggests that bradykinesia may result from the concurrent effects of low DA levels and dysfunctional plasticity and that training can be exploited in medicated subjects to improve levodopa treatment.     

The cerebellum and cognition: cerebellar lesions impair sequence learning but not conditional visuomotor learning in monkeys

Claims that the cerebellum contributes to cognitive processing in humans have arisen from both functional neuroimaging and patient studies. These claims challenge traditional theories of cerebellar function that ascribe motor functions to this structure. We trained monkeys to perform both a visuomotor conditional associative learning task and a visually guided sequence task, and studied the effects of bilateral excitotoxic lesions in the lateral cerebellar nuclei. In the first experiment three operated monkeys showed a small impairment in post-operative retention of a visuomotor associative task (A) but were then not impaired in learning a new task (B). However, the impairment on A could have been due to a problem in making the movements themselves. In a second experiment we therefore gave the three control animals a further pre-operative retest on both A and B and then tested after surgery on retention of both tasks. Though again the animals showed motor problems on task A, they reached criterion, and at this stage could clearly make both movements satisfactorily. The critical test was then retention of task B, and they were not impaired. In the final experiment (serial reaction time task) the monkeys response times on a repeating visuomotor sequence were compared with those for a pseudo-random control sequence. After bilateral nuclei lesions they were slow to execute the pre-operatively learned sequence but were still faster on this than on the control task. However, when they were then given a new repeating sequence to learn, they never performed the sequence as quickly as they had on retention of the first sequence. We conclude that the cerebellum is not essential for the learning or recall of stimulus-response associations but that it is crucially involved in the process by which motor sequences become automatic with extended practice.     

Parallel contributions of cerebellar, striatal and M1 mechanisms to motor sequence learning

When learning a new motor sequence, we must execute the correct order of movements while simultaneously optimizing sensorimotor parameters such as trajectory, timing, velocity and force. Neurophysiological studies in animals and humans have identified the major brain regions involved in sequence learning, including the motor cortex (M1), basal ganglia (BG) and cerebellum. Current models link these regions to different stages of learning (early vs. late) or different components of performance (spatial vs. sensorimotor). At the same time, research in motor control has given rise to the concept that internal models at different levels of the motor system may contribute to learning. The goal of this review is to develop a new framework for motor sequence learning that combines stage and component models within the context of internal models. To do this, we review behavioral and neuroimaging studies in humans and neurophysiological studies in animals. Based on this evidence, we present a model proposing that sequence learning is underwritten by parallel, interacting processes, including internal model formation and sequence representation, that are instantiated in specific cerebellar, BG or M1 mechanisms depending on task demands and the stage of learning. The striatal system learns predictive stimulus-response associations and is critical for motor chunking. The role of the cerebellum is to acquire the optimal internal model for sequence performance in a particular context, and to contribute to error correction and control of on-going movement. M1 acts to store the representation of a learned sequence, likely as part of a distributed network including the parietal lobe and premotor cortex.     

Understanding failure of visual paired associate learning in amnestic mild cognitive impairment

Impairment in visual paired associate learning occurs often in with amnestic mild cognitive impairment (aMCI), a condition considered to be an early stage of Alzheimer's disease (AD). To date, studies of aMCI have characterized impaired visual paired associate learning only in terms of summary scores such as total errors or total trials to criterion. The aim of this study was to determine the nature and magnitude of errors made on a continuous paired associate learning (CPAL) task designed to allow analysis of the component processes involved in paired associate learning. Twenty-one individuals with aMCI and 54 healthy age-matched older adults (HC) performed the CPAL task in which they had to learn six pattern-location pairings over six trials. Results suggested that aMCI patients performed significantly worse on the CPAL, both in learning rate and in error accumulation. Qualitative analyses of CPAL performance revealed that in aMCI there were significantly more errors on all indices except perseverative errors. When expressed as a percentage of total errors, abnormalities occurred only for within-search and exploratory errors. These findings suggest that poor performance on visual paired associate learning tasks in aMCI reflects impairments in both learning and executive function.