Author: Jasmine Sarkar
Updated on May 2nd 2022
Our reward circuitry has been known to be controlled predominantly by the midbrain dopaminergic neurons in the ventral tegmental area (VTA) and substantia nigra pars compacta (SNc). They mostly respond to external rewards and enhanced goal-directed decision-making processes by projecting out to the cortex. However, recent studies have reported the involvement of cerebellar neurons in operant and associative behavioural learning and reward circuitry and also their connection to the VTA and SNc. This has been largely researched and published, reviewing several studies conducted over the past years.
Cerebellar Circuit:
The cerebellum comprises two innervating glutamatergic neurons: Mossy fibers from the spinal cord, brainstem, and pontine nuclei and climbing fibers from the inferior olivary nucleus. They innervate the cerebellum into the Purkinje cells and transmit sensory and motor signals. An external stimulus during learning processes activates some climbing fibers and suppresses others causing both depression and potentiation in distinct cerebellar zones.
Numerous theories of learning are based on performance error signals which are primarily transmitted by climbing fibers and this concept led to the discovery of reward signals conveyed by these cerebellar neurons.
Reward-systems through Mossy and Climbing fibres:
Cerebellar granule cells dominate most of the brain neurons and have been largely studied by researchers. Studies conducted on these cells have found high activity (high simple spikes) during reward delivery and prediction over the course of learning tasks keeping movement signals constant. Similar studies have also suggested that these signals are conditional and the Mossy-fibre granule cells can combine such signals with motor movements and behaviours.
Climbing fibers and Purkinje dendrites have also shown a significant association between sensorimotor signals and reward prediction. Experiments conducted on mice have shown cerebellar signaling through climbing fibers to be associated with rewards, reward prediction or expectation, and a violation of reward-related expectation. Experiments on monkeys have indicated that these fibers can also predict the reward size. All these reward-related signals are usually combined with sensory signals.
When animals learn the type of sensory or motor task that leads to rewards, certain reward signals emerge simultaneously. These exact reward signals or responses are suppressed when rewards are predicted.
Simultaneous activation of both mossy and climbing fiber reward inputs leads to depression (reduced synaptic strength) in parallel fibers whereas activation in one of the fibers (particularly climbing fibers) leads to long-term potentiation (stronger synaptic connection).
Researchers have used electrophysiological methods to study the difference between reward-related simple (mossy fiber-granule cell pathway) and complex spikes (climbing fiber input). Monkeys performing a visual pursuit task, where a cue is introduced to signal reward size followed by a smooth pursuit (eye movement wherein eyes remain fixated on a moving object) showed slight variance in simple spikes with different cues and reward sizes. This marked the significance of simple spikes in reward signals. Researchers demonstrated that simple spikes encode reinforcement error signals while learning new visuomotor tasks but their timing does not correlate to complex spikes indicating that predictive complex spikes function differently than error-elated complex spikes.
Origin of Reward-related signals:
Now, the study of the functioning of these reward systems largely depends on the understanding of the origin of these signals. Afferent inputs to the olivary nucleus arise from the projections from the neocortex and the inferior olive has been reported to be associated with VTA that supports the reward signal system through climbing fibers. The subthalamic nucleus that projects to pontine nuclei provides adequate evidence of reward-related signals transmitted through mossy fibers. The pre-frontal cortex (PFC) that controls goal-directed behaviour also shows strong connectivity with the cerebellum regarding reward-related signals.
Cerebellum and Basal Ganglia:
Researchers have proposed an association between basal ganglia and cerebellum systems. Basal ganglia are responsible for reward (reinforcement) learning, and the cerebellum is responsible for performance-based (supervised) learning. The dopaminergic reward prediction errors determine the actions that seek maximum rewards and the sensory prediction error (climbing fiber) determines the modification process of the motor commands to execute correct actions. Cerebellum employs both performance-based learning by climbing fiber pathways and reward-based learning by the dopaminergic system.
All these studies have supported the fact that the cerebellum is capable of controlling functions beyond motor activities. It can not only mediate a range of cognitive processes including behaviors, learning, and reinforcement but can also function in combination with the pathways of the cerebral cortex and basal ganglia. Any disruption in its activity can cause impairments in cognitive behaviours. Recent studies also established that the cerebellum can help with selecting and executing the most valuable action.
An explicit understanding of how the sensory, motor, behavioural, and reward systems are presented and expressed in different stages and areas of the cerebellum circuit and lobules, will be a fundamental objective in the future.
Reviewer: Dr. Jitendra Kumar Sinha
Illustration(s): Jasmine Sarkar
Editor: Dr. Shampa Ghosh
References:
Sinha J.K., Pathak S., Rana N., & Ghosh S. (2021) Cerebellum. In: Vonk J., Shackelford T. (eds) Encyclopedia of Animal Cognition and Behavior. Springer, Cham. https://doi.org/10.1007/978-3-319-47829-6_1338-1
Kostadinov, D., & Häusser, M. (2022). Reward signals in the cerebellum: Origins, targets, and functional implications. Neuron, 110(8), 1290–1303. https://doi.org/10.1016/j.neuron.2022.02.015
Apps, R., & Garwicz, M. (2005). Anatomical and physiological foundations of cerebellar information processing. Nature reviews. Neuroscience, 6(4), 297–311. https://doi.org/10.1038/nrn1646
Carta, I., Chen, C. H., Schott, A. L., Dorizan, S., & Khodakhah, K. (2019). Cerebellar modulation of the reward circuitry and social behavior. Science (New York, N.Y.), 363(6424), eaav0581. https://doi.org/10.1126/science.aav0581