Learning and cognitive flexibility: frontostriatal function and monoaminergic modulation

https://doi.org/10.1016/j.conb.2010.01.007Get rights and content

Learning in a constant environment, and adapting flexibly to a changing one, through changes in reinforcement contingencies or valence-free cues, depends on overlapping circuitry that interconnects the prefrontal cortex (PFC) with the striatum and is subject to several forms of neurochemical modulation. We present evidence from recent studies in animals employing electrophysiological, pharmacological and lesion techniques, and neuroimaging, neuropsychological and pharmacological investigations of healthy humans and clinical patients. Dopamine (DA) neurotransmission in the medial striatum and PFC is critical for basic reinforcement learning and the integration of negative feedback during reversal learning, whilst orbitofrontal 5-hydroxytryptamine (5-HT) likely mediates this type of low level flexibility, perhaps by reducing interference from salient stimuli. The role of prefrontal noradrenaline (NA) in higher order flexibility indexed through attentional set-shifting has recently received significant empirical support, and similar avenues appear promising in the field of task switching.

Introduction

The ability to learn, or acquire associations between stimuli, actions and environmental outcomes, and flexibly adapt ongoing behaviour according to salient changes in the environment or our current intentions, carries survival value and is arguably fundamental to what cognitive agents perceive as voluntary action. Research on neural mechanisms subserving learning and flexibility over the last decade has focused on the prefrontal cortex (PFC) and basal ganglia, particularly the striatum, and their interactions through their multiple serial and parallel loops [1, 2], subject to neuromodulatory inputs by the monoamines (dopamine (DA), serotonin (5-hydroxytryptamine, 5-HT) and noradrenaline (NA)).

We review developments in recent years that have contributed to the state of the science, progressing through a conceptual gradient of cognitive processes and paradigms. First, we address basic reinforcement learning, where behaviour is critically dependent on environmental signals or (reward) feedback. Subsequently, we focus on reversal learning, reflecting the ability to switch responding to a previously non-reinforced exemplar, which relies on feedback but critically incorporates flexibility. We proceed to attentional set-shifting, which also indexes the ability to adapt behaviour flexibly following feedback, but pertains to broader stimulus dimensions rather than a specific exemplar. Finally, we address task switching, a relatively purer form of cognitive flexibility uncontaminated by learning and feedback processing.

The literature cited represents a corresponding gradient in spatiotemporal scale as a function of the complexity of cognitive paradigms, and the feasibility of interspecies translation. Basic neuronal and molecular mechanisms of reinforcement learning and low level flexibility emerge predominantly from animal studies employing invasive techniques such as single unit recordings, microdialysis and selective gene knock-out. More cognitively complex forms of flexibility are tapped mostly through systemic pharmacology (across species), lesion studies in animals and functional neuroimaging in humans.

Section snippets

Cortico-striatal anatomy

Before addressing the cognitive neuroscience of cortico-striatal interactions, we note developments in the field of anatomy. The landmark proposal of anatomically and functionally segregated cortico-striatal loops by Alexander and colleagues is gradually becoming superseded by evidence for more complex cross-talk and convergence enabling integration of processing across these circuits [3, 4]. Thus, functional connectivity network effects that may underpin learning and flexible cognition are

Reinforcement learning

For some years it has been clear that both the striatum and the frontal cortex play key roles in learning, being active during the expectation of reward and following its delivery, as well as coding prediction error during learning [8, 9]. Beyond the postulated role of synaptic plasticity [10], monkey cell recording suggests that changes in dynamic network states of caudate and lateral PFC neurons may also be critical for learning [11]. These regions exhibit sustained activity during simple

Reversal learning

Deficits in reversal learning, the ability to switch responding to a previously non-reinforced stimulus, are present not just in chronic illness but also in the early stages of disorders such as Parkinson's disease (PD) [20], schizophrenia and bipolar disorder [21•, 22•]. Catecholamines, specifically cortico-striatal DA, and indoleamines, in particular orbitofrontal (OFC) 5-HT, are considered critical for this type of flexibility. Reversal learning is impaired not only following infusions of a

Attentional set-shifting

In contrast to reversal learning which pertains to a specific exemplar, attentional set-shifting (extra-dimensional shifting (EDS)) refers to switching between higher order modalities (e.g. from lines to shapes, from texture to odour) on the basis of feedback. The neural correlates of attentional set-shifting in humans were elucidated using fMRI [36••], associating OFC with reversal learning and ventrolateral PFC (vlPFC) with EDS. In the rat, a well-controlled lesion study of strategy shifting

Task switching

Task switching addresses cognitive flexibility divorced from the ability to adapt ongoing behaviour to feedback, hence unconfounding it from the learning mechanisms that underlie attentional set-shifting. It involves switching between rules that determine well-learned stimulus and response (S–R) mappings, on the basis of cues. Task switching deficits are seen in neurodegenerative disease such as PD [45] and psychiatric illness, for example, OCD [46], since frontoparietal areas and the basal

Conclusions

Recent advances in understanding the contributions of neurons in the striatum, orbital and prefrontal regions of the frontal cortex to learning and flexibility reflect developments in methods and techniques in animal and human experimentation. The roles of cortico-striatal DA and orbitofrontal 5-HT in basic forms of learning and flexibility have received growing support, and their contribution to different cognitive subprocesses has been relatively well demarcated. Conversely, the NA-ergic

References and recommended reading

Papers of particular interest, published within the period of review, have been highlighted as:

  • • of special interest

  • •• of outstanding interest

Acknowledgements

This work was supported by a Wellcome Trust Programme Grant (076274/Z/04/Z) to TWR, BJ Everitt, AC Roberts and BJ Sahakian and was completed at the University of Cambridge Behavioural and Clinical Neuroscience Institute supported by a joint award from the Medical Research Council and the Wellcome Trust (G00001354). AAK holds an Isaac Newton fellowship and is additionally supported by the Parkinson's Disease Society. GKM is supported by the Medical Research Council and NARSAD (National Alliance

References (56)

  • B. Draganski et al.

    Evidence for segregated and integrative connectivity patterns in the human basal ganglia

    J Neurosci

    (2008)
  • S.N. Haber et al.

    Reward-related cortical inputs define a large striatal region in primates that interface with associative cortical connections, providing a substrate for incentive-based learning

    J Neurosci

    (2006)
  • F. Kasanetz et al.

    Functional integration across a gradient of corticostriatal channels controls UP state transitions in the dorsal striatum

    Proc Natl Acad Sci U S A

    (2008)
  • A. Di Martino et al.

    Functional connectivity of human striatum: a resting state FMRI study

    Cereb Cortex

    (2008)
  • R.B. Postuma et al.

    Basal ganglia functional connectivity based on a meta-analysis of 126 positron emission tomography and functional magnetic resonance imaging publications

    Cereb Cortex

    (2006)
  • J.P. O’Doherty et al.

    Temporal difference models and reward-related learning in the human brain

    Neuron

    (2003)
  • J.P. Pfister et al.

    Triplets of spikes in a model of spike timing-dependent plasticity

    J Neurosci

    (2006)
  • M.H. Histed et al.

    Learning substrates in the primate prefrontal cortex and striatum: sustained activity related to successful actions

    Neuron

    (2009)
  • M.E. Bach et al.

    Transient and selective overexpression of D2 receptors in the striatum causes persistent deficits in conditional associative learning

    Proc Natl Acad Sci U S A

    (2008)
  • C. Kellendonk et al.

    Transient and selective overexpression of dopamine D2 receptors in the striatum causes persistent abnormalities in prefrontal cortex functioning

    Neuron

    (2006)
  • S.A. Castner et al.

    Reversal of antipsychotic-induced working memory deficits by short-term dopamine D1 receptor stimulation

    Science

    (2000)
  • H.C. Tsai et al.

    Phasic firing in dopaminergic neurons is sufficient for behavioral conditioning

    Science

    (2009)
  • G.D. Stuber et al.

    Reward-predictive cues enhance excitatory synaptic strength onto midbrain dopamine neurons

    Science

    (2008)
  • W. Shen et al.

    Dichotomous dopaminergic control of striatal synaptic plasticity

    Science

    (2008)
  • D.A. Peterson et al.

    Probabilistic reversal learning is impaired in Parkinson's disease

    Neuroscience

    (2009)
  • V.C. Leeson et al.

    Discrimination learning, reversal, and set-shifting in first-episode schizophrenia: stability over six years and specific associations with medication type and disorganization syndrome

    Biol Psychiatry

    (2009)
  • R. Dias et al.

    Dissociation in prefrontal cortex of affective and attentional shifts

    Nature

    (1996)
  • H.F. Clarke et al.

    Cognitive inflexibility after prefrontal serotonin depletion is behaviorally and neurochemically specific

    Cereb Cortex

    (2007)
  • Cited by (319)

    • Human creativity: Functions, mechanisms, and social conditioning

      2024, Advances in Experimental Social Psychology
    View all citing articles on Scopus
    View full text