Abstract
A core principle of cerebellar learning theories is that climbing fibers from the inferior olive convey error signals about movement execution to Purkinje cells in the cerebellar cortex. These inputs trigger synaptic changes, which are purported to drive progressive adjustment of future movements. Individually, binary complex spike signals lack information about the sign and magnitude of errors which presents a problem for cerebellar learning paradigms exhibiting fast adaptation. Here, using a newly developed behavioral paradigm in mice, we introduced sensorimotor perturbations into a simple joystick-pulling behavior and found parasagittal bands of Purkinje cells with reciprocal modulation of complex spike activity, along with rapid adaptation of the behavior. Whereas complex spiking showed little modulation in the unperturbed condition, alternating bands were activated or inhibited when the perturbation was introduced and this modulation encoded the sign and magnitude of the resulting sensorimotor mismatch. These findings provide important insight about how the cerebellum uses supervised learning to quickly adapt motor behavior in response to perturbations.
This is a preview of subscription content, access via your institution
Access options
Access Nature and 54 other Nature Portfolio journals
Get Nature+, our best-value online-access subscription
$32.99 / 30 days
cancel any time
Subscribe to this journal
Receive 12 print issues and online access
$259.00 per year
only $21.58 per issue
Buy this article
- Purchase on SpringerLink
- Instant access to the full article PDF.
USD 39.95
Prices may be subject to local taxes which are calculated during checkout
Similar content being viewed by others
Data availability
Data that support the findings of this study are available via Zenodo at https://doi.org/10.5281/zenodo.17160426 (ref. 44). Source data are provided with this paper.
Code availability
Code used to analyze the data and generate the results presented here are available via GitHub at https://github.com/StPeres-Cerebellum/NguyenGrosStell-CS-Analysis.
References
Albus, J. S. A new approach to manipulator control: the cerebellar model articulation controller (CMAC). J. Dyn. Syst. Meas. Control 97, 220–227 (1975).
Ito, M., Sakurai, M. & Tongroach, P. Climbing fibre induced depression of both mossy fibre responsiveness and glutamate sensitivity of cerebellar Purkinje cells. J. Physiol. 324, 113–134 (1982).
Marr, D. & Brindley, G. S. A theory for cerebral neocortex. Proc. R. Soc. Lond. B 176, 161–234 (1970).
Medina, J. F. & Lisberger, S. G. Links from complex spikes to local plasticity and motor learning in the cerebellum of awake-behaving monkeys. Nat. Neurosci. 11, 1185–1192 (2008).
Yang, Y. & Lisberger, S. G. Purkinje-cell plasticity and cerebellar motor learning are graded by complex-spike duration. Nature 510, 529–532 (2014).
Wolpert, D. M., Miall, R. C. & Kawato, M. Internal models in the cerebellum. Trends Cogn. Sci. 2, 338–347 (1998).
Gilbert, P. F. C. & Thach, W. T. Purkinje cell activity during motor learning. Brain Res. 128, 309–328 (1977).
Baizer, J. S., Kralj-Hans, I. & Glickstein, M. Cerebellar lesions and prism adaptation in macaque monkeys. J. Neurophysiol. 81, 1960–1965 (1999).
Martin, T. A., Keating, J. G., Goodkin, H. P., Bastian, A. J. & Thach, W. T. Throwing while looking through prismsI. Focal olivocerebellar lesions impair adaptation. Brain 119, 1183–1198 (1996).
Najafi, F. & Medina, J. F. Beyond ‘all-or-nothing’ climbing fibers: graded representation of teaching signals in Purkinje cells. Front. Neural Circuits https://doi.org/10.3389/fncir.2013.00115 (2013).
Rowan, M. J. M. et al. Graded control of climbing-fiber-mediated plasticity and learning by inhibition in the cerebellum. Neuron 99, 999–1015 (2018).
Najafi, F., Giovannucci, A., Wang, S. S.-H. & Medina, J. F. Coding of stimulus strength via analog calcium signals in Purkinje cell dendrites of awake mice. Elife https://doi.org/10.7554/eLife.03663 (2014).
Herzfeld, D. J., Kojima, Y., Soetedjo, R. & Shadmehr, R. Encoding of error and learning to correct that error by the Purkinje cells of the cerebellum. Nat. Neurosci. 21, 736–743 (2018).
Soetedjo, R., Kojima, Y. & Fuchs, A. Complex spike activity signals the direction and size of dysmetric saccade errors. Prog. Brain Res. 171, 153–159 (2008).
Heffley, W. et al. Coordinated cerebellar climbing fiber activity signals learned sensorimotor predictions. Nat. Neurosci. 21, 1431–1441 (2018).
Kostadinov, D., Beau, M., Blanco-Pozo, M. & Häusser, M. Predictive and reactive reward signals conveyed by climbing fiber inputs to cerebellar Purkinje cells. Nat. Neurosci. 22, 950–962 (2019).
Martin, T. A., Keating, J. G., Goodkin, H. P., Bastian, A. J. & Thach, W. T. Throwing while looking through prisms. II. Specificity and storage of multiple gaze—throw calibrations. Brain 119, 1199–1211 (1996).
Morton, S. M. & Bastian, A. J. Cerebellar contributions to locomotor adaptations during splitbelt treadmill walking. J. Neurosci. 26, 9107–9116 (2006).
Smith, M. A., Ghazizadeh, A. & Shadmehr, R. Interacting adaptive processes with different timescales underlie short-term motor learning. PLoS Biol. 4, e179 (2006).
Witter, L., Rudolph, S., Pressler, R. T., Lahlaf, S. I. & Regehr, W. G. Purkinje cell collaterals enable output signals from the cerebellar cortex to feed back to purkinje cells and interneurons. Neuron 91, 312–319 (2016).
Gaffield, M. A., Bonnan, A. & Christie, J. M. Conversion of graded presynaptic climbing fiber activity into graded postsynaptic Ca2+ signals by Purkinje cell dendrites. Neuron https://doi.org/10.1016/j.neuron.2019.03.010 (2019).
Ramirez, J. E. & Stell, B. M. Calcium imaging reveals coordinated simple spike pauses in populations of cerebellar Purkinje cells. Cell Rep. 17, 3125–3132 (2016).
Calame, D. J., Becker, M. I. & Person, A. L. Cerebellar associative learning underlies skilled reach adaptation. Nat. Neurosci. 26, 1068–1079 (2023).
Wagner, M. J. et al. A neural circuit state change underlying skilled movements. Cell 184, 3731–3747.e21 (2021).
Mathis, A. et al. DeepLabCut: markerless pose estimation of user-defined body parts with deep learning. Nat. Neurosci. 21, 1281–1289 (2018).
Bouvier, G. et al. Burst-dependent bidirectional plasticity in the cerebellum is driven by presynaptic NMDA receptors. Cell Rep. 15, 104–116 (2016).
Coesmans, M., Weber, J. T., De Zeeuw, C. I. & Hansel, C. Bidirectional parallel fiber plasticity in the cerebellum under climbing fiber control. Neuron 44, 691–700 (2004).
Lev-Ram, V., Wong, S. T., Storm, D. R. & Tsien, R. Y. A new form of cerebellar long-term potentiation is postsynaptic and depends on nitric oxide but not cAMP. Proc. Natl Acad. Sci. USA 99, 8389–8393 (2002).
Safo, P. & Regehr, W. G. Timing dependence of the induction of cerebellar LTD. Neuropharmacology 54, 213–218 (2008).
Sarkisov, D. V. & Wang, S. S.-H. Order-dependent coincidence detection in cerebellar purkinje neurons at the inositol trisphosphate receptor. J. Neurosci. 28, 133–142 (2008).
Apps, R. & Hawkes, R. Cerebellar cortical organization: a one-map hypothesis. Nat. Rev. Neurosci. 10, 670–681 (2009).
Apps, R. et al. Cerebellar modules and their role as operational cerebellar processing units. Cerebellum 17, 654–682 (2018).
Chen, L., Bao, S., Lockard, J. M., Kim, J. K. & Thompson, R. F. Impaired classical eyeblink conditioning in cerebellar-lesioned and Purkinje cell degeneration (PCD) mutant mice. J. Neurosci. 16, 2829–2838 (1996).
Ruigrok, T. J. H., Pijpers, A., Goedknegt-Sabel, E. & Coulon, P. Multiple cerebellar zones are involved in the control of individual muscles: a retrograde transneuronal tracing study with rabies virus in the rat. Eur. J. Neurosci. 28, 181–200 (2008).
Singla, S., Dempsey, C., Warren, R., Enikolopov, A. G. & Sawtell, N. B. A cerebellum-like circuit in the auditory system cancels responses to self-generated sounds. Nat. Neurosci. 20, 943–950 (2017).
Oscarsson, O. Functional units of the cerebellum—sagittal zones and microzones. Trends Neurosci. 2, 143–145 (1979).
Brooks, J. X. & Cullen, K. E. The primate cerebellum selectively encodes unexpected self-motion. Curr. Biol. 23, 947–955 (2013).
Lang, E. J., Sugihara, I. & Llinas, R. GABAergic modulation of complex spike activity by the cerebellar nucleoolivary pathway in rat. J. Neurophysiol. 76, 255–275 (1996).
Zhang, Y. et al. Fast and sensitive GCaMP calcium indicators for imaging neural populations. Nature 615, 884–891 (2023).
Pachitariu, M. et al. Suite2p: beyond 10,000 neurons with standard two-photon microscopy. Preprint at bioRxiv https://doi.org/10.1101/061507 (2017).
Santosa, F. & Symes, W. W. Linear inversion of band-limited reflection seismograms. SIAM J. Sci. Stat. Comput. 7, 1307–1330 (1986).
Tibshirani, R. Regression shrinkage and selection via the Lasso. J. R. Stat. Soc. Ser. B 58, 267–288 (1996).
Friedrich, J., Zhou, P. & Paninski, L. Fast online deconvolution of calcium imaging data. PLoS Comput. Biol. 13, e1005423 (2017).
Nguyen, V., Gros, C. & Stell, B. Imaging and behavior data for Rapid Motor Adaptation via Population-level Modulation of Cerebellar Error Signals. Zenodo https://doi.org/10.5281/zenodo.17160426 (2025).
Acknowledgements
We are grateful to the following for discussions throughout the project and/or comments on the manuscript: P. Ascher, C. Auger, I. Llano, A. Marty and, in particular, B. Barbour. Funding was from Agence nationale de la recherche (grant no. ANR-19-CE37-0011-01, awarded to B.M.S.).
Author information
Authors and Affiliations
Contributions
Conceptualization: B.M.S. Methodology: B.M.S. Investigation: V.N., C.G. and B.M.S. Visualization: V.N. and B.M.S. Funding acquisition: B.M.S. Project administration: B.M.S. Supervision: B.M.S. Writing—original draft: V.N. and B.M.S. Writing—review and editing: V.N. and B.M.S.
Corresponding author
Ethics declarations
Competing interests
The authors declare no competing interests.
Peer review
Peer review information
Nature Neuroscience thanks Reza Shadmehr and the other, anonymous, reviewer(s) for their contribution to the peer review of this work.
Additional information
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Extended data
Extended Data Fig. 1 Analysis of various behavior parameters.
A. Joydecel normalized to the first pull for group data (105 recordings across 13 mice). The similar evolution across the first 250 pulls (blue) and the last 250 pulls (red) within a session shows that learning is the same at the beginning and end of a recording session. B. Maximum velocity of the limb/joystick for the example recording in Fig. 1 (left) and for all animals (right). C. The maximum position of the joystick in the 100 ms following pull initiation for the example recording in Fig. 1 (left) and for all animals (right).
Extended Data Fig. 2 Behavior tracking for all conditions.
Mean lateral position as distance from the mouth for the recording session shown in Fig. 5. All positions tracked from high-speed video recordings with DeepLabCut. Note that prior to the initiation of the pull, limb positions are static. The passive lick-port condition does not elicit any systematic behavior. Licks in each condition are denoted as points in A to show the timing of all licks in the session. A. Lick-port position vs time since pull threshold. B. Joystick position vs time since pull threshold. C. Lateral position of the left paw vs time since pull threshold. D. Same as C for the right paw. E. PSTH of mean lick counts per pull vs time since pull threshold.
Extended Data Fig. 3 Kinetics and amplitudes of spontaneous CS versus those that were evoked by the passive lick-port condition.
A. Individual recordings (circles) and box-plot distributions of amplitude (A) of spontaneous and evoked CS across all cells (black), those that increased (red), decreased (blue) or showed no change (gray) in firing rate in response to the lick-port movement. B. Same as A for the CS decay times (τ). C. The difference in amplitude between evoked and spontaneous CS for the same conditions as above. D. Same as C for the decay times. Dashed lines indicate no change.
Extended Data Fig. 4 The effect of passive lick-port movement and gain adaptation on CS rate for the clusters show in Fig. 4.
Same data shown in Fig. 4 with an additional panel showing the CS rate elicited by passive lick-port movement to allow comparison. There does not appear to be a consistent relationship between the effect in the adaptation and recovery phases of the behavior sessions and the passive lick-port movement, which never over- or under-shoots the mouth.
Extended Data Fig. 5 CS rate changes in different time windows near pull initiation.
The same data shown in Fig. 4 with additional time windows drawn 200 ms prior to and following the initiation of the pull. The correlation of CS rate with the behavior only appears during the 100 ms following the pull.
Extended Data Fig. 6 Behavior modification related to CS error.
A. Mean image from the session shown throughout the figure with k-means clusters of ROIs that increase (gold) or decrease (green) CS rate during adaptation to the sensorimotor perturbation. B. Mean lateral position of the left forelimb (blue) and CS rate of the adaptation-inhibited cluster (green; same trace as in A). 0 mm denotes the approximate mouth position. Inset: mean left forelimb position in the 100 ms analysis window (gray box) (n = 7 mice, p < 0.001, paired two-sided t-test). C. Same as B for the right forelimb (n = 7 mice, p = 0.678, paired two-sided t-test). D. Histogram of mean lick rate in 10 ms bins throughout the pull. Inset: sum of the bins in the gray box shows no difference between the adaptation and recovery conditions (n = 7 mice, p = 0.771, paired two-sided t-test). E. CS rate in the adaptation-inhibited cluster during rewarded and reward-omitted pulls. Inset: group data of 10 adaptation-inhibited clusters from 3 mice showing no significant effect of reward (p = 0.833, paired two-sided t-test). F. The relationship between the mean CS rate of the cluster and the motor error for all 528 joystick pulls in the same session shown throughout the figure. Pearson’s r = -0.19, p < 0.001 G. The relationship between the change in motor error and the mean CS rate of the cluster for all 528 joystick pulls. Pearson’s r = 0.03, p = 0.6. H. Box plots of Pearson’s correlation coefficients between behavioral parameters and cluster CS rates for all pulls of all recordings; n = 10 mice. Shading in B-G represents 95% CI.
Supplementary information
Supplementary Video 1
Upper panel shows mouse pulling joystick in the control gain condition where the joystick and water port movement are linked with gain = 1. Lower panel same mouse in the adaptation condition with the gain = 2.
Supplementary Video 2
Upper panel shows the mouse pulling the joystick with the control gain. In the lower panel the joystick is retracted and the mean water port position that was recorded during the control trials is ‘replayed’ back to the mouse.
Source data
Source Data Fig. 1
Statistical source data.
Source Data Fig. 2
Statistical source data.
Source Data Fig. 3
Statistical source data.
Source Data Fig. 4
Statistical source data.
Source Data Fig. 5
Statistical source data.
Source Data Extended Data Fig. 2
Statistical source data.
Source Data Extended Data Fig. 3
Statistical source data.
Source Data Extended Data Fig. 4
Statistical source data.
Source Data Extended Data Fig. 5
Statistical source data.
Source Data Extended Data Fig. 6
Statistical source data.
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Nguyen, V., Gros, C. & Stell, B.M. Rapid motor skill adjustment is associated with population-level modulation of cerebellar error signals. Nat Neurosci 29, 136–146 (2026). https://doi.org/10.1038/s41593-025-02126-7
Received:
Accepted:
Published:
Version of record:
Issue date:
DOI: https://doi.org/10.1038/s41593-025-02126-7
