There is now overwhelming evidence to suggest, as anticipated by David Marr (1969), that for the neocortex to be fully operative all neural signals must loop through the cerebellum during and after learning (e.g., Buckner 2013; Chabrol, Mrsic-Flogel et al. 2019; Gao, De Zeeuw et al. 2018; Hasanbegović 2024; Schmahmann 1991; Mariën et al. 2017). The readiness potential as generated at the frontal lobes of humans occurs well in advance of any conscious awareness that a movement is about to happen (Libet 1985; Soon et al. 2008). Following damage of the cerebello-neocortical fibres, the readiness potential of the frontal lobes is abolished (Ikeda et al. 1994).
The dentate nucleus, which is innervated by the mediolateral cerebellum, was found to exhibit preparatory activity prior to reward delivery, which was comparable to the activity of the antero-lateral motor cortex (Chabrol, Mrsic-Flogel et al. 2019). Mice were required to locomote on a track wheel such that they were trained to come to an abrupt stop once they confronted a specific visual stimulus (i.e., large black and white checkerboards) after which a drop of water was provided as reward. Three types of neurons were identified in the dentate nucleus: those that increased in activity gradually in anticipation of reward (1-2 seconds before the reward), those that discharged abruptly before the reward (1-0 seconds before), and finally those that started firing at the time of reward delivery (Fig. 2 of Chabrol, Mrsic-Flogel et al. 2019). A similar group of neurons was identified in the antero-lateral motor cortex with comparable timing characteristics (Fig. 1 of Chabrol, Mrsic-Flogel et al. 2019). All recordings were made using silicon probes. It is noteworthy that there was an enhancement of complex spike at the Purkinje cells that subserved the dentate nucleus under investigation at the time of reward delivery, which fits with the notion that motivational signals are transmitted via the olivary nucleus to anchor the operant response. Furthermore, optogenetic activation of the Purkinje cells that innervate and suppress the dentate nucleus was found to eliminate the preparatory activity in the anterolateral motor cortex as well as induce a momentary interruption of locomotion (Fig. 3 of Chabrol et al. 2019). The authors conclude that preparatory activity is controlled by a ‘learned’ decrease in Purkinje cells firing in advance of the reward.
The receptive-field remapping paradigms of Goldberg and Bruce (1990) are all controlled by reward delivery (also see: Duhamel et al. 1992). Rhesus monkeys were trained to generate saccadic eye movements to various locations in the visual field as visual stimuli were flashed in the receptive field of neurons (Tolias et al. 2001; Zirnsak et al. 2014). It was discovered that immediately before the execution of a saccade the receptive field of neocortical neurons (of area V4 and the frontal eye fields) were biased in the direction of the endpoint of the saccadic eye movement. Neurons located in lobule VI (including the midline vermis) and in the subjacent fastigial nucleus are tuned to the direction of saccadic eye movements using a firing rate-code (Noda and Fujikoda 1987ab), which is characteristic of oculomotor neurons of the brain stem (Schiller and Tehovnik 2015; Tehovnik, Patel, Tolias et al. 2021). We predict that visually-guided saccades will be interrupted (by altering the saccadic endpoint) and the receptive-field bias abolished following the silencing of the cerebellar circuits (see: Barash et al. 1999; Kojima et al. 2010; Robinson et al. 1993; Sato and Noda 1992; Takagi et al. 1998). To avoid introducing position habits (e.g., Sommer and Wurtz 2004), we would suggest performing bilateral inactivation of the cerebellum to test the hypothesis.
- Badges
- Science topic
- Similar topics
- Engineering
- Materials Engineering
- Powders