Neural Correlates of Reinforcement Learning in Mid-lateral Cerebellum

Sendhilnathan, Naveen; Semework, Mulugeta; Goldberg, Michael E.; Ipata, Anna E.

doi:10.1016/j.neuron.2019.12.032

Cited by 77 publications

(50 citation statements)

References 51 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We used the NEI REX-VEX system coupled with Spike2 (CED electronics) for event and neural data acquisition. We verified all recordings off-line to ensure that we had isolated Purkinje cells 4 .…”

Section: Methodsmentioning

confidence: 99%

“…Although P-cells in the mid-lateral cerebellum, encode a reinforcement learning error signal 4 , it is not clear if this activity is only correlative or causal to behavior. To test if mid-lateral cerebellum is necessary for the monkeys to perform the task, we infused 4-10 μl of a 10 mg/ml solution of muscimol, a GABA A agonist that hyperpolarizes cell bodies without affecting fibers of passage, near crus I and II of the mid-lateral cerebellum of two monkeys in the same location where we had recorded task-dependent P-cells (see methods, Fig 2a; see Fig S1 for infusion locations).…”

Section: Figurementioning

confidence: 99%

“…We recently showed that the simple spikes of P-cells in the mid-lateral cerebellum encode reinforcement error signals during visuomotor association learning 4 . Two types of P-cells report whether the monkey’s decision on the most recent trial was correct (correct-preferring neurons, cP-cells) or wrong (wrong-preferring neurons, wP-cells), and although each neuron reports this only during a brief epoch (the ‘delta epoch’), the population as a whole maintains the information from shortly after the previous outcome to shortly after the next.…”

Section: Figurementioning

confidence: 99%

“…There was no loss of function when we inactivated a more anterior region of the cerebellum that is implicated in motor control. We suggest that the mid-lateral cerebellum, which provides a reinforcement learning error signal 4 , is necessary for visuomotor association learning. Our results have implications for the involvement of cerebellum in cognitive control, and add critical constraints to brain models of non-motor learning 14,15 .…”

mentioning

confidence: 93%

“… Summary The cerebellum has long been considered crucial for supervised motor learning and its optimization 1-3 . However, new evidence has also implicated the cerebellum in reward based learning 4-8 , executive function 9-12 , and frontal-like clinical deficits 13 . We recently showed that the simple spikes of Purkinje cells (P-cells) in the mid-lateral cerebellar hemisphere (Crus I and II) encode a reinforcement error signal when monkeys learn to associate arbitrary symbols with hand movements 4 .…”

mentioning

confidence: 99%

See 4 more Smart Citations

The mid-lateral cerebellum is necessary for reinforcement learning

Sendhilnathan

Goldberg

2020

Preprint

Self Cite

View full text Add to dashboard Cite

The cerebellum has long been considered crucial for supervised motor learning 1 and its optimization 1-3 . However, new evidence has also implicated the cerebellum in reward 2 based learning 4-8 , executive function 9-12 , and frontal-like clinical deficits 13 . We recently 3 showed that the simple spikes of Purkinje cells (P-cells) in the mid-lateral cerebellar 4 hemisphere (Crus I and II) encode a reinforcement error signal when monkeys learn to 5 associate arbitrary symbols with hand movements 4 . However, it is unclear if the cerebellum 6 is necessary for any process beyond motor learning. To investigate if the mid-lateral 7 cerebellum is actually necessary for learning visuomotor associations, we reversibly 8 inactivated the mid-lateral cerebellum of two primates with muscimol while they learned to 9 associate arbitrary symbols with hand movements. Here we show that cerebellar inactivation 10 impaired the monkey's ability to learn new associations, although it had no effect on the 11 monkeys' performance on a task with overtrained symbols. A computational model 12 corroborates our results. Cerebellar inactivation increased the reaction time, but there were 13 no deficits in any motor kinematics such as the hand movement, licking or eye movement. 14 There was no loss of function when we inactivated a more anterior region of the cerebellum 15 that is implicated in motor control. We suggest that the mid-lateral cerebellum, which 16 provides a reinforcement learning error signal 4 , is necessary for visuomotor association 17 learning. Our results have implications for the involvement of cerebellum in cognitive 18 control, and add critical constraints to brain models of non-motor learning 14,15 . 19 20We trained two monkeys to associate arbitrary visual symbols with left and right hand 21 movements to earn an immediate liquid reward. On each recording session, the monkeys first 22

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Figurementioning

confidence: 99%

Section: Figurementioning

confidence: 99%

mentioning

confidence: 93%

mentioning

confidence: 99%

See 3 more Smart Citations

The mid-lateral cerebellum is necessary for reinforcement learning

Sendhilnathan

Goldberg

2020

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

Cerebellar impulsivity–compulsivity assessment scale

Lin

Amokrane

Chen

et al. 2022

Ann Clin Transl Neurol

View full text Add to dashboard Cite

Objective The cerebellum has been identified as the key brain region that modulates reward processing in animal models. Consistently, we recently found that people with cerebellar ataxia have impulsive and compulsive behaviors (ICBs), the main symptoms related to abnormal reward processing. Due to the lack of a validated scale to quantitatively measure ICBs in cerebellar disorders, we aim to develop and validate a new scale, Cerebellar Impulsivity–Compulsivity Assessment (CIA). Methods We recruited 62 cerebellar ataxia cases, categorized into those with ICBs and those without. We developed a preliminary version of CIA, containing 17 questions. We studied the internal consistency, test–retest reliability, and inter‐rater reliability to formulate the final version of CIA, which constitutes only 10 questions. The receiver operating characteristic curve (ROC) was generated to assess the sensitivity and specificity of CIA. Results Cerebellar ataxia cases with ICBs have threefold higher total preliminary CIA scores than those without ICBs (12.06 ± 5.96 vs. 4.68 ± 3.50, p = 0.038 ). Cronbach's alpha revealed good internal consistency across all items ( α > 0.70 ). By performing the test–retest reliability and inter‐rater reliability on the preliminary version of CIA, we excluded seven questions ( r < 0.70) and generated the final version of CIA. Based on the ROC, a score of 8.0 in CIA was chosen as the cut‐off for ICBs in individuals with cerebellar ataxia with 81% sensitivity and 81% specificity. Interpretation CIA is a novel tool to assess ICBs in cerebellar ataxia and broaden our understanding of the cerebellum‐related cognitive and behavioral symptoms.

show abstract

Cerebellum neuropathology and motor skill deficits in fragile X syndrome

Chen

Guo

Han

et al. 2022

Intl J of Devlp Neuroscience

View full text Add to dashboard Cite

Fragile X syndrome (FXS) is a leading form of inherited intellectual disability and single‐gene cause of autism spectrum disorder (ASD) and is characterized by core deficits in cognitive flexibility, sensory sensitivity, emotion, and social interactions. Motor deficits are a shared feature of FXS and autism. The cerebellum has emerged as one of the target brain areas affected by neurodevelopmental diseases. Alterations in the cerebellar structure, circuits, and function may be the key drivers of impaired fine and gross motor skills in FXS and fragile X‐associated tremor/ataxia syndrome (FXTAS). In this review, we briefly examined recent findings in FXS and present a discussion on the literature supporting motor skill deficits in FXS. Subsequently, we focused on neuropathological alterations in the cerebellum in FXS and FXTAS. We highlight studies that have directly examined the function of fragile X mental retardation protein and related epigenetic variations in the cerebellum. Overall, we obtained considerable supporting evidence for the hypothesis that cerebellar dysfunction is evident in FXS and FXTAS; however, compared with studies on other ASD models, studies on motor skills related to fragile X disorders are particularly rare and inconclusive. Hence, future research should address FXS‐related motor and behavioral trajectories and examine the underlying mechanisms at both the cell and circuit levels.

show abstract

Neural Correlates of Reinforcement Learning in Mid-lateral Cerebellum

Cited by 77 publications

References 51 publications

The mid-lateral cerebellum is necessary for reinforcement learning

The mid-lateral cerebellum is necessary for reinforcement learning

Cerebellar impulsivity–compulsivity assessment scale

Cerebellum neuropathology and motor skill deficits in fragile X syndrome

Contact Info

Product

Resources

About