When the cerebellum holds the starting gun
2021; Cell Press; Volume: 109; Issue: 14 Linguagem: Inglês
10.1016/j.neuron.2021.06.027
ISSN1097-4199
AutoresJimena Laura Frontera, Clément Léna,
Tópico(s)Neural dynamics and brain function
ResumoHow the cerebellum affects movement onset is poorly understood. In this issue of Neuron, Dacre et al., 2021Dacre J. Colligan M. Clarke T. Ammer J.J. Schiemann J. Chamosa-Pino V. Claudi F. Harston J.A. Eleftheriou C. Pakan J.M.P. et al.A cerebellar-thalamocortical pathway drives behavioral context-dependent movement initiation.Neuron. 2021; 109: 2326-2338Abstract Full Text Full Text PDF PubMed Scopus (21) Google Scholar establish that in the context of operant conditioning, the transient activation of the cerebello-thalamo-cortical pathway to the motor cortex is sufficient to initiate the conditioned movement. How the cerebellum affects movement onset is poorly understood. In this issue of Neuron, Dacre et al., 2021Dacre J. Colligan M. Clarke T. Ammer J.J. Schiemann J. Chamosa-Pino V. Claudi F. Harston J.A. Eleftheriou C. Pakan J.M.P. et al.A cerebellar-thalamocortical pathway drives behavioral context-dependent movement initiation.Neuron. 2021; 109: 2326-2338Abstract Full Text Full Text PDF PubMed Scopus (21) Google Scholar establish that in the context of operant conditioning, the transient activation of the cerebello-thalamo-cortical pathway to the motor cortex is sufficient to initiate the conditioned movement. The mammalian cerebellum is a kind of paradox: it is a considerable neural device, hosting around two-thirds of the brain’s neurons and participating in a large spectrum of brain functions such as motor, sensory, cognitive, and affective processing, and yet the communication from the cerebellar cortex to the forebrain goes through the narrow bottleneck of the cerebellar nuclei. Indeed, in humans, the cerebellar dentate nucleus hosts barely more neurons than, for example, the dorso-lateral geniculate thalamus, which only relays visual information to the cortex. Hence, regardless of the complexity of cerebellar computations performed to optimize the wide spectrum of brain functions, the result of these computations must be propagated to the forebrain through a narrow channel and is thus most likely concisely represented and limited to key parameters of the forebrain operations. The work of Dacre et al., 2021Dacre J. Colligan M. Clarke T. Ammer J.J. Schiemann J. Chamosa-Pino V. Claudi F. Harston J.A. Eleftheriou C. Pakan J.M.P. et al.A cerebellar-thalamocortical pathway drives behavioral context-dependent movement initiation.Neuron. 2021; 109: 2326-2338Abstract Full Text Full Text PDF PubMed Scopus (21) Google Scholar in this issue of Neuron provides strong support to a possible answer to this paradox: the cerebellum would serve as a trigger to start the action (Figure 1). Decades of study of the cerebellar cortex have established that the cerebellar circuitry is endowed with mechanisms allowing the generation of precisely timed signals; indeed, cerebellar lesions may disrupt the fine temporal coordination of muscle contraction or more simply affect the reaction time in basic sensory-motor paradigms. However, the neurophysiological processes by which the cerebellum adjusts the timing of actions via its projections to the forebrain have remained unresolved. Recently, a series of studies showed in primates that the phasic activity immediately preceding reaching movement onset is found in motor cortex neurons targeted by the cerebello-thalamo-cortical pathway and that this activity depends on cerebellar inputs (see refs in Nashef et al., 2021Nashef A. Mitelman R. Harel R. Joshua M. Prut Y. Area-specific thalamocortical synchronization underlies the transition from motor planning to execution.Proc. Natl. Acad. Sci. USA. 2021; 118 (e2012658118)Crossref PubMed Scopus (5) Google Scholar). Since the pyramidal-tract neurons are known to receive cerebello-thalamo-cortical inputs (Shinoda et al., 1982Shinoda Y. Yamazaki M. Futami T. Convergent inputs from the dentate and the interpositus nuclei to pyramidal tract neurons in the motor cortex.Neurosci. Lett. 1982; 34: 111-115Crossref PubMed Scopus (17) Google Scholar), and the cerebellar nuclei neurons also show phasic responses before movement onset (references in Dacre et al., 2021Dacre J. Colligan M. Clarke T. Ammer J.J. Schiemann J. Chamosa-Pino V. Claudi F. Harston J.A. Eleftheriou C. Pakan J.M.P. et al.A cerebellar-thalamocortical pathway drives behavioral context-dependent movement initiation.Neuron. 2021; 109: 2326-2338Abstract Full Text Full Text PDF PubMed Scopus (21) Google Scholar, see also Tanaka et al., 2021Tanaka M. Kunimatsu J. Suzuki T.W. Kameda M. Ohmae S. Uematsu A. Takeya R. Roles of the Cerebellum in Motor Preparation and Prediction of Timing.Neuroscience. 2021; 462: 220-234Crossref PubMed Scopus (15) Google Scholar for eye saccades), the cerebellum could thus hold the starting gun and trigger movements by directly hitting the cortical motoneurons. Yet, the clear demonstration for this phenomenon was still lacking; many dots needed to be connected. Moreover, to what extent the cerebellum instructs which specific movement should be initiated remained entirely unclear. This set the stage for the elegant work of the team of Ian Duguid and collaborators. Dacre et al. used a simple operant conditioning task in head-fixed animals, requiring mice to push a lever during a brief auditory cue in order to retrieve a reward. With training, the animals learn to perform the task for increasingly short auditory cues, down to 2 s-long tones. To analyze the involvement of the cerebello-thalamo-cortical system in this task, the authors first mapped this network and identified the relay between the cerebellum and the forelimb primary motor cortex, called the caudal forelimb area (CFA). The thalamic regions showing the stronger overlap between glutamatergic terminals from the medial and lateral cerebellar nuclei and the soma of CFA-projecting thalamic neurons was found in a medial compartment of the motor thalamus. In this thalamic region, all of the CFA-projecting thalamic neurons exhibit somatic appositions of cerebellar terminals, indicating that they receive direct cerebellar excitatory inputs. Are all these brain areas involved in the task studied? Indeed, inactivation of the medial and lateral cerebellar nuclei, the motor thalamus, or the CFA by the local infusion of muscimol, a GABAergic antagonist, decreased proper movement initiation in response to the cue. This was not due to a disengagement of the mouse from the task (increased arousal of the mice at the cue onset is clearly visible) and thus likely reflected an inability to trigger and/or perform the movement associated to the cue. The authors then set out to determine if the thalamic region relaying cerebellar inputs to the CFA exhibits any task-related activity. Using deep brain 2-photon calcium imaging through a gradient index (GRIN) lens, they found that almost all neurons exhibit a movement-related change in activity, and about half of the neurons exhibit an increased activity preceding movement onset. This increase did not consist of ramps from cue onset to movement onset (which is more typical of preparatory activities, e.g., Chabrol et al., 2019Chabrol F.P. Blot A. Mrsic-Flogel T.D. Cerebellar Contribution to Preparatory Activity in Motor Neocortex.Neuron. 2019; 103: 506-519Abstract Full Text Full Text PDF PubMed Scopus (64) Google Scholar; Gao et al., 2018Gao Z. Davis C. Thomas A.M. Economo M.N. Abrego A.M. Svoboda K. De Zeeuw C.I. Li N. A cortico-cerebellar loop for motor planning.Nature. 2018; 563: 113-116Crossref PubMed Scopus (153) Google Scholar) but was tightly time-locked to the movement onset. Interestingly, a concomitant study in a reaching task in primates demonstrated a very similar thalamic premotor activity, which tightly co-varied with the activity of the corresponding motor cortex (Nashef et al., 2021Nashef A. Mitelman R. Harel R. Joshua M. Prut Y. Area-specific thalamocortical synchronization underlies the transition from motor planning to execution.Proc. Natl. Acad. Sci. USA. 2021; 118 (e2012658118)Crossref PubMed Scopus (5) Google Scholar). What happens then in the motor cortex? CFA neurons exhibit a transient increase in firing rate before movement onset, but this does not provide a clear timing of the onset of the subthreshold depolarization in the neurons. The authors thus accomplished the feat of obtaining intracellular recordings in the layer 5B neurons during the task; they found that both intratelencephalic (IT; Figure 1) and pyramidal-tract (PT) neurons underwent a depolarization preceding movement onset, with a time course similar to the change observed in the motor thalamus. Therefore, this depolarization could result, at least in part, from the activation of thalamic inputs preceding movement onset. Indeed, the inhibition of the motor thalamus largely suppressed the depolarization observed in layer 5B cortical neurons. This indicates that the thalamic inputs are essential to establish the cortical activation preceding movement onset. Now comes the most striking result of the study. The authors then reasoned that the pre-movement discharge of the thalamo-cortical afferents to CFA could act as a trigger of the forelimb movement. Indeed, they found that replacing the sound cue by an optogenetic stimulation of the motor thalamus neurons, their axonal terminals in the CFA, or the cerebellar afferents to the motor thalamus was sufficient to trigger movements of lever pushing. These stimulations and the cue similarly recruited an overlapping population of neurons in the deep layers of the CFA. Strikingly, when the animals were tested in a different head-fixed context where the lever was absent and the limb rested on the floor, the stimulations failed to produce any consistent forelimb movement. This suggests that in the learned context of the task, the animal is continuously ready to push the lever and only requires a “go” signal to initiate the action. Interestingly, the stimulations did not elicit the subsequent licking behavior that would correspond to reward acquisition, suggesting that the effect was limited to triggering the conditioned forelimb movement rather than completing the operant task. This impressive set of experiments now raises a number of interesting questions. In this study, a global activation of the cerebello-thalamo-cortical pathway to the CFA suffices to trigger the movement in absence of the cue, suggesting that the cue is not integrated in the cortical representation of the movement. What would be the result of the stimulation of this pathway in an operant task where two concurrent forelimb movements are associated to different cues (e.g., one tone for pushing and one tone for pulling)? How could the thalamo-cortical go-cue operate to trigger a context-dependent movement? Studies in the primate suggest the existence of a clear transition of the neuronal activity in the motor cortex between the preparatory and execution phases (Elsayed et al., 2016Elsayed G.F. Lara A.H. Kaufman M.T. Churchland M.M. Cunningham J.P. Reorganization between preparatory and movement population responses in motor cortex.Nat. Commun. 2016; 7: 13239Crossref PubMed Scopus (133) Google Scholar); the context could thus induce a preparatory activity in the motor (and premotor) cortices, and the thalamo-cortical activation could thus help precipitate the transition between theses phases (Nashef et al., 2021Nashef A. Mitelman R. Harel R. Joshua M. Prut Y. Area-specific thalamocortical synchronization underlies the transition from motor planning to execution.Proc. Natl. Acad. Sci. USA. 2021; 118 (e2012658118)Crossref PubMed Scopus (5) Google Scholar). Is the fine cerebellar control of the forelimb kinematics (Becker and Person, 2019Becker M.I. Person A.L. Cerebellar Control of Reach Kinematics for Endpoint Precision.Neuron. 2019; 103: 335-348Abstract Full Text Full Text PDF PubMed Scopus (39) Google Scholar) mediated through this cerebello-thalamo-cortical pathway, and if yes, how is this function articulated with the trigger function demonstrated here? In Dacre et al., 2021Dacre J. Colligan M. Clarke T. Ammer J.J. Schiemann J. Chamosa-Pino V. Claudi F. Harston J.A. Eleftheriou C. Pakan J.M.P. et al.A cerebellar-thalamocortical pathway drives behavioral context-dependent movement initiation.Neuron. 2021; 109: 2326-2338Abstract Full Text Full Text PDF PubMed Scopus (21) Google Scholar, the involvement of the cerebello-thalamo-cortical system is demonstrated as the trigger required for the fast initiation of movement (before the cue terminates); would that hold at longer timescales? In paradigms of self-timed eye saccade, the cerebellum appears essential for well-timed saccades only at short timescales (a few hundred milliseconds), while the basal ganglia seem to determine the timing of saccade onset at longer timescales, which are not affected by cerebellar inactivations (Tanaka et al., 2021Tanaka M. Kunimatsu J. Suzuki T.W. Kameda M. Ohmae S. Uematsu A. Takeya R. Roles of the Cerebellum in Motor Preparation and Prediction of Timing.Neuroscience. 2021; 462: 220-234Crossref PubMed Scopus (15) Google Scholar). Finally, a central question is what causes cerebello-thalamic activation? In simple forms of associative motor learning, like in eyeblink conditioning, the cerebellum may trigger the movement (independently from the motor cortex) based on the time elapsed since a cue onset (Jirenhed and Hesslow, 2016Jirenhed D.A. Hesslow G. Are Purkinje Cell Pauses Drivers of Classically Conditioned Blink Responses?.Cerebellum. 2016; 15: 526-534Crossref PubMed Scopus (32) Google Scholar). Alternatively, the cerebellum also receives inputs from the collaterals of the pyramidal-tract cortical neurons via the pontine nuclei and feeds back to the motor cortex via the thalamus to form functional “closed” loops (Léna and Popa, 2015Léna C. Popa D. Cerebrocerebellar Loops in the Rodent Brain.in: Heck D. The Neuronal Codes of the Cerebellum. Academic Press-Elsevier, Waltham2015: 135-153Google Scholar; Figure 1). Such loops between the cerebellum and the premotor cortex have been shown to sustain ramping activity during motor preparation (Chabrol et al., 2019Chabrol F.P. Blot A. Mrsic-Flogel T.D. Cerebellar Contribution to Preparatory Activity in Motor Neocortex.Neuron. 2019; 103: 506-519Abstract Full Text Full Text PDF PubMed Scopus (64) Google Scholar; Gao et al., 2018Gao Z. Davis C. Thomas A.M. Economo M.N. Abrego A.M. Svoboda K. De Zeeuw C.I. Li N. A cortico-cerebellar loop for motor planning.Nature. 2018; 563: 113-116Crossref PubMed Scopus (153) Google Scholar). An intriguing possibility is thus that the phasic cortical activity before movement onset results from reverberating activation of the cerebello-cortical loop. Therefore, the beautiful work of Dacre et al., 2021Dacre J. Colligan M. Clarke T. Ammer J.J. Schiemann J. Chamosa-Pino V. Claudi F. Harston J.A. Eleftheriou C. Pakan J.M.P. et al.A cerebellar-thalamocortical pathway drives behavioral context-dependent movement initiation.Neuron. 2021; 109: 2326-2338Abstract Full Text Full Text PDF PubMed Scopus (21) Google Scholar will certainly inspire future research to further clarify how the cerebellar signals are integrated in cortical activity to control actions. A cerebellar-thalamocortical pathway drives behavioral context-dependent movement initiationDacre et al.NeuronJune 18, 2021In BriefDacre et al. show the contribution of a cerebellar-thalamocortical pathway to movement initiation. Using gain- and loss-of-function manipulations they demonstrate that output from dentate/interpositus nuclei, via the thalamus, shapes cortical activity dynamics necessary for learned behavioral context-dependent movement initiation. Full-Text PDF Open Access
Referência(s)