The cerebellum is hypothesized to refine movement through online adjustments. We examined how such predictive control may be generated using a mouse reach paradigm, testing whether the cerebellum uses within-reach information as a predictor to adjust reach kinematics. We first identified a population-level response in Purkinje cells that scales inversely with reach velocity, pointing to the cerebellar cortex as a potential site linking kinematic predictors and anticipatory control. Next, we showed that mice can learn to compensate for a predictable reach perturbation caused by repeated, closed-loop optogenetic stimulation of pontocerebellar mossy fiber inputs. Both neural and behavioral readouts showed adaptation to position-locked mossy fiber perturbations and exhibited aftereffects when stimulation was removed. Surprisingly, position-randomized stimulation schedules drove partial adaptation but no opposing aftereffects. A model that recapitulated these findings suggests that the cerebellum may decipher cause-and-effect relationships through time-dependent generalization mechanisms.
This is a preview of subscription content, access via your institution
Open Access articles citing this article.
Nature Neuroscience Open Access 20 July 2023
Access Nature and 54 other Nature Portfolio journals
Get Nature+, our best-value online-access subscription
$29.99 / 30 days
cancel any time
Subscribe to this journal
Receive 12 print issues and online access
$209.00 per year
only $17.42 per issue
Rent or buy this article
Prices vary by article type
Prices may be subject to local taxes which are calculated during checkout
Source data used to make each of the figures are provided with this paper. Raw data are available upon reasonable request to the authors.
The code for cerebellar model and custom analysis code can be found at https://github.com/dycala.
Holmes, G. The symptoms of acute cerebellar injuries due to gunshot injuries. Brain 40, 461–535 (1917).
Becker, M. I. & Person, A. L. Cerebellar control of reach kinematics for endpoint precision. Neuron 103, 335–348.e5 (2019).
Baizer, J. S., Kralj-Hans, I. & Glickstein, M. Cerebellar lesions and prism adaptation in macaque monkeys. J. Neurophysiol. 81, 1960–1965 (1999).
Tseng, Y. W., Diedrichsen, J., Krakauer, J. W., Shadmehr, R. & Bastian, A. J. Sensory prediction errors drive cerebellum-dependent adaptation of reaching. J. Neurophysiol. 98, 54–62 (2007).
Morton, S. M. & Bastian, A. J. Cerebellar contributions to locomotor adaptations during splitbelt treadmill walking. J. Neurosci. 26, 9107–9116 (2006).
Smith, M. A. & Shadmehr, R. Intact ability to learn internal models of arm dynamics in Huntington’s disease but not cerebellar degeneration. J. Neurophysiol. 93, 2809–2821 (2005).
Miall, R. & Wolpert, D. Forward models for physiological motor control. Neural Netw. 9, 1265–1279 (1996).
Kawato, M. & Gomi, H. The cerebellum and VOR/OKR learning models. Trends Neurosci. 15, 445–453 (1992).
Ohyama, T., Nores, W. L., Murphy, M. & Mauk, M. D. What the cerebellum computes. Trends Neurosci. 26, 222–227 (2003).
Shadmehr, R., Smith, M. A. & Krakauer, J. W. Error correction, sensory prediction, and adaptation in motor control. Annu. Rev. Neurosci. 33, 89–108 (2010).
Lee, K. H. et al. Circuit mechanisms underlying motor memory formation in the cerebellum. Neuron 86, 529–540 (2015).
Lavond, D. G., Knowlton, B. J., Steinmetz, J. E. & Thompson, R. F. Classical conditioning of the rabbit eyelid response with a mossy-fiber stimulation CS: II. lateral reticular nucleus stimulation. Behav. Neurosci. 101, 676–682 (1987).
Mauk, M. D., Steinmetz, J. E. & Thompson, R. F. Classical conditioning using stimulation of the inferior olive as the unconditioned stimulus. Proc. Natl Acad. Sci. USA 83, 5349–5353 (1986).
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).
Ito, M. & Kano, M. Long-lasting depression of parallel fiber-Purkinje cell transmission induced by conjunctive stimulation of parallel fibers and climbing fibers in the cerebellar cortex. Neurosci. Lett. 33, 253–258 (1982).
Blazquez, P. M., Hirata, Y., Heiney, S. A., Green, A. M. & Highstein, S. M. Cerebellar signatures of vestibulo-ocular reflex motor learning. J. Neurosci. 23, 9742–9751 (2003).
Yang, Y. & Lisberger, S. G. Interaction of plasticity and circuit organization during the acquisition of cerebellum-dependent motor learning. eLife 2, e01574 (2013).
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).
Raymond, J. L., Lisberger, S. G. & Mauk, M. D. The cerebellum: a neuronal learning machine? Science 272, 1126–1131 (1996).
Gandolfo, F., Li, C. S. R., Benda, B. J., Padoa Schioppa, C. & Bizzi, E. Cortical correlates of learning in monkeys adapting to a new dynamical environment. Proc. Natl Acad. Sci. USA 97, 2259–2263 (2000).
Li, C. S. R., Padoa-Schioppa, C. & Bizzi, E. Neuronal correlates of motor performance and motor learning in the primary motor cortex of monkeys adapting to an external force field. Neuron 30, 593–607 (2001).
Mathis, M. W., Mathis, A. & Uchida, N. Somatosensory cortex plays an essential role in forelimb motor adaptation in mice. Neuron 93, 1493–1503.e6 (2017).
Proville, R. D. et al. Cerebellum involvement in cortical sensorimotor circuits for the control of voluntary movements. Nat. Neurosci. 17, 1233–1239 (2014).
Hewitt, A. L., Popa, L. S. & Ebner, T. J. Changes in Purkinje cell simple spike encoding of reach kinematics during adaption to a mechanical perturbation. J. Neurosci. 35, 1106–1124 (2015).
Ito, M. Neural design of the cerebellar motor control system. Brain Res. 40, 81–84 (1972).
Weiler, J., Gribble, P. L. & Pruszynski, J. A. Spinal stretch reflexes support efficient hand control. Nat. Neurosci. 22, 529–533 (2019).
Low, A. Y. T. et al. Precision of discrete and rhythmic forelimb movements requires a distinct neuronal subpopulation in the interposed anterior nucleus. Cell Rep. 22, 2322–2333 (2018).
Guo, J.-Z. et al. Disrupting cortico-cerebellar communication impairs dexterity. eLife 10, e65906 (2021).
Becker, M. I., Calame, D. J., Wrobel, J. & Person, A. L. Online control of reach accuracy in mice. J. Neurophysiol. 124, 1637–1655 (2020).
Tibshirani, R. Regression shrinkage and selection via the lasso. J. R. Stat. Soc. 58, 267–288 (1996).
Popa, L. S., Streng, M. L. & Ebner, T. J. Long-term predictive and feedback encoding of motor signals in the simple spike discharge of Purkinje cells. eNeuro 4, ENEURO.0036-17.2017 (2017).
Streng, M. L., Popa, L. S. & Ebner, T. J. Climbing fibers control Purkinje cell representations of behavior. J. Neurosci. 37, 1997–2009 (2017).
Coltz, J. D., Johnson, M. T. & Ebner, T. J. Cerebellar Purkinje cell simple spike discharge encodes movement velocity in primates during visuomotor arm tracking. J. Neurosci. 19, 1782–1803 (1999).
Hewitt, A. L., Popa, L. S., Pasalar, S., Hendrix, C. M. & Ebner, T. J. Representation of limb kinematics in Purkinje cell simple spike discharge is conserved across multiple tasks. J. Neurophysiol. 106, 2232–2247 (2011).
Roitman, A. V., Pasalar, S., Johnson, M. T. V. & Ebner, T. J. Position, direction of movement, and speed tuning of cerebellar Purkinje cells during circular manual tracking in monkey. J. Neurosci. 25, 9244–9257 (2005).
Musall, S., Kaufman, M. T., Juavinett, A. L., Gluf, S. & Churchland, A. K. Single-trial neural dynamics are dominated by richly varied movements. Nat. Neurosci. 22, 1677–1686 (2019).
Herzfeld, D. J., Kojima, Y., Soetedjo, R. & Shadmehr, R. Encoding of action by the Purkinje cells of the cerebellum. Nature 526, 439–441 (2015).
Brown, S. T. & Raman, I. M. Sensorimotor integration and amplification of reflexive whisking by well-timed spiking in the cerebellar corticonuclear circuit. Neuron 99, 564–575.e2 (2018).
Person, A. L. & Raman, I. M. Purkinje neuron synchrony elicits time-locked spiking in the cerebellar nuclei. Nature 481, 502–505 (2012).
Thier, P., Dicke, P. W., Haas, R. & Barash, S. Encoding of movement time by populations of cerebellar Purkinje cells. Nature 405, 72–76 (2000).
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).
Kimpo, R. R., Rinaldi, J. M., Kim, C. K., Payne, H. L. & Raymond, J. L. Gating of neural error signals during motor learning. eLife 3, e02076 (2014).
Avraham, G., Taylor, J. A., Breska, A., Ivry, R. B. & McDougle, S. D. Contextual effects in sensorimotor adaptation adhere to associative learning rules. eLife 11, e75801 (2022).
Kitazawa, S., Kimura, T. & Yin, P. B. Cerebellar complex spikes encode both destinations and errors in arm movements. Nature 392, 494–497 (1998).
Gaffield, M. A., Sauerbrei, B. A. & Christie, J. M. Cerebellum encodes and influences the initiation, performance, and termination of discontinuous movements in mice. eLife 11, e71464 (2022).
Chabrol, F. P., Blot, A. & Mrsic-Flogel, T. D. Cerebellar contribution to preparatory activity in motor neocortex. Neuron 103, 506–519.e4 (2019).
Miall, R. C., Keating, J. G., Malkmus, M. & Thach, W. T. Simple spike activity predicts occurrence of complex spikes in cerebellar Purkinje cells. Nat. Neurosci. 1, 13–15 (1998).
Brodal, P. & Bjaalie, J. G. Organization of the pontine nuclei. Neurosci. Res. 13, 83–118 (1992).
Sillitoe, R. V., Fu, Y. H. & Watson, C. in The Mouse Nervous System (eds Watson, C. et al.) Ch. 11 (Elsevier, 2012).
Rasmussen, A., Jirenhed, D. A., Wetmore, D. Z. & Hesslow, G. Changes in complex spike activity during classical conditioning. Front. Neural Circuits 8, 90 (2014).
Herzfeld, D. J., Hall, N. J., Tringides, M. & Lisberger, S. G. Principles of operation of a cerebellar learning circuit. eLife 9, e55217 (2020).
Mauk, M. D. & Buonomano, D. V. The neural basis of temporal processing. Annu. Rev. Neurosci. 27, 307–340 (2004).
Shimansky, Y., Wang, J. J., Bauer, R. A., Bracha, V. & Bloedel, J. R. On-line compensation for perturbations of a reaching movement is cerebellar dependent: Support for the task dependency hypothesis. Exp. Brain Res. 155, 156–172 (2004).
Albergaria, C., Silva, N. T., Darmohray, D. M. & Care, M. R. Locomotor activity modulates associative learning in mouse cerebellum. Nat. Neurosci. 21, 725–735 (2018).
Nashef, A., Cohen, C., Isreal, Z., Harel, R. & Prut, Y. Cerebellar shaping of motor cortical firing is correlated with timing of motor actions. Cell Rep. 23, 1275–1285 (2018).
White, J. J. et al. An optimized surgical approach for obtaining stable extracellular single-unit recordings from the cerebellum of head-fixed behaving mice. J. Neurosci. Methods 262, 21–31 (2016).
TestArduino (2015); https://github.com/cortex-lab/phy
Yu, B., Gabriel, D., Noble, L. & An, K. N. Estimate of the optimum cut-off frequency for the Butterworth low-pass digital filter. J. Appl. Biomech. 15, 318–329 (1999).
Pachitariu, M., Steinmetz, N., Kadir, S., Carandini, M. & Harris, K. D. Kilosort: realtime spike-sorting for extracellular electrophysiology with hundreds of channels. Preprint at bioRxiv https://doi.org/10.1101/061481
Hensbroek, R. A. et al. Identifying Purkinje cells using only their spontaneous simple spike activity. J. Neurosci. Methods 232, 173–180 (2014).
Kennedy, A. et al. A temporal basis for predicting the sensory consequences of motor commands in an electric fish. Nat. Neurosci. 17, 416–422 (2014).
Lanore, F., Cayco-Gajic, N. A., Gurnani, H., Coyle, D. & Silver, R. A. Cerebellar granule cell axons support high-dimensional representations. Nat. Neurosci. 24, 1142–1150 (2021).
Wagner, M. J. et al. Shared cortex-cerebellum dynamics in the execution and learning of a motor task. Cell 177, 669–682 (2019).
We thank the members of A.L.P.’s laboratory for critical feedback on the manuscript; the Neurotechnology Center at the University of Colorado Anschutz Medical campus for use of core facilities, including the Advanced Light Microscopy Core and the Optogenetics and Neural Engineering Core. We thank M. Spindle and E. Judd for technical assistance. Work was supported by F31 NS113395-01 to D.J.C.; and NS114430, National Science Foundation CAREER grant 1749568, and by a grant from the Simons Foundation as part of the Simons-Emory International Consortium on Motor Control to A.L.P.
The authors declare no competing interests.
Peer review information
Nature Neuroscience thanks Kamran Khodakhah and the other, anonymous, reviewer(s) for their contribution to the peer review of this work.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
a. The right hand was tracked with high-speed cameras as mice reached upwards and outwards towards a food pellet. Positional outreach trajectories from a single session viewed are shown from a lateral (left) or bottom-up (right) vantage point with traces colored by the magnitude of outward velocity. b. Mice were trained for a minimum of 15 days on the reaching task. Pellet retrieval success was tracked throughout training for each mouse, mean is shown in red. c. Quantification of success rate on day 1 of training and day 15. (p-values: (c) 6.4×10-8, paired t-test; Sample size: n = 19 animals; * indicates p-value < 0.05).
a. Cerebellar recordings using single electrodes were first anatomically targeted to cerebellar cortex. If a recorded cell had visible Cspks they were classified as PCs. Otherwise, if cells had a firing rate > 40 Hz, a median absolute difference firing rate from the median interspike interval (MAD) < 0.008, and a CV2 > 0.2, they were classified as PCs81. b. Neuropixel-recorded single units were crosscorrelated with nearby (<200 microns) low firing rate (<5 Hz) single units. If this crosscorrelation exhibited the characteristic firing rate pause seen in PC simple spikes after a Cspk, these units were classified as the simple spikes and Cspks of a single PC. If no pause was seen, cells that exhibited the same firing rate, MAD, and CV2 profile described in a were classified as PCs. c. Example simple spike pause aligned to the time of a Cspk from a Neuropixel recording. d. Embedding MAD, CV2, and FR into a two-dimensional space using tSNE shows two distinct clusters, one corresponding largely to cells that were identified using the criteria in a and b and the other corresponding to other cells (n = 1268 sorted cells). e. Three example cells from a single session showing a neuron that was classified as a PC due to the presence of complex spikes (red, left), a neuron that was classified as a PC using firing rate criteria (blue, middle), and a neuron that was classified as other (grey, right). The simple spike raster and averaged simple spike firing rate PETH are shown on the bottom and top, respectively.
a. Schematic of LASSO regressions. 23 kinematic variables were regressed against firing rate at different lags from 0 to -300 ms. The lag that minimized the mean squared error (MSE) of the regressions was selected. b. Peak modulation time of all cells across all reaches. c. Optimal lags of the LASSO regression for each cell. d. Fraction of the unique contribution to total variance explained for each regressor. e. Fraction of regressions with each variable selected (mean shown for each regressor). f. Same as in Fig. 1d but only analyzing the subset of PCs that had Cspks identified. g. PCs with Cspks show no changes in model error across the reach, consistent with the total PC dataset. h. PCs with Cspks display kinematic variables with similar relative contributions to model variance explained compared to the total PC dataset. i. Variables included in the LASSO model in PCs with Cspks are consistent with data in the total PC dataset (mean shown for each regressor). j. Same as in Fig. 1h but only analyzing PCs with Cspks. The top and bottom 50% of outward reach velocities are analyzed. k. Quantification of the simple spike suppression of the data in f. l. Time of FR suppression for the data in f. (p-values: (f) empirical vs. reach shuffle: 4.4 ×10-11, empirical vs spike shuffle: 2.4 ×10-11, Wilcoxon signed rank test (k) 3.7 ×10-2, Wilcoxon signed rank test; Sample size: (b-e) N = 11 animals, 465 cells (f-k) N = 8 animals, 59 cells; * indicates p-value < 0.05; all error bars and bands represent mean ± SEM; in box and whiskers plots box denotes median and 25th/75th percentiles, whiskers denote 10th and 90th percentiles, circle indicates mean).
a. Cspks in the 500 ms before reach onset were not associated with differences in target error as assessed with euclidean distance form session median compared to non-Cspk trials. b. No difference in peak outward velocity was observed between Cspks and non-Cspk trials. c. Simple spike firing rate in trials with early Cspk and non-Cspk trials. d. No difference in simple spike rate during outreach was seen in early Cspk trials compared with non-Cspk trials. e. No difference in simple spike rate per outward velocity was seen in early Cspk trials compared with non-Cspk trials. f. Simple spike firing aligned to the time of early Cspks compared to similarly aligned trials without early Cspk trials.g. No difference in simple spike rate in the 100 ms preceding early Cspks was seen compared to the similarly aligned previous or next trial. (p-values: (a) 0.32, Wilcoxon signed rank test (b) 0.75, paired t-test (d) 0.37, paired t-test (e) 0.29, Wilcoxon signed rank test (g) 0.47, RM one-way ANOVA, previous trial vs cspk trial: 0.97, cspk trial vs next trial: 0.42, Tukey’s multiple comparisons test; Sample size: N = 8 animals, 58 cells; all error bars and bands represent mean ± SEM).
a. Mossy fiber boutons expressing hSyn-ChR2-mCherry in the cerebellar cortex. b. Simple spike responses to mossy fiber stimulation. Left: examples of single-cell simple spike responses to mossy fiber stimulation. Right: quantification of simple spike responses to all recorded cells. Significance of differences are indicated by the color and corresponding p-value map. c. Cspk responses to mossy fiber stimulation. Left: PSTH of the population of recorded cells with Cspks binned at 50 ms. A single trace showing a Cspk after stimulation is shown above. Right: Quantification of Cspk probability in the 250 ms after stimulation and non-stimulated epochs for each cell. (p-values: (b) paired t-test, (c) 5.6 ×10-3, paired t-test Sample size: (a) 1 of 4 mice displayed, (b) N = 4 animals, 151 cells (c) N = 4 animals, 39 cells; * indicates p-value < 0.05; all error bars and bands represent mean ± SEM).
a. Histological section showing ChR2-mCherry expression at the injection site in the left pontine nuclei (Pn: pontine nuclei; RtTg: reticulotegmental nuclei; PnO: pontine reticular nuclei, oral part; PnC: pontine reticular nuclei, caudal part; 1 of 7 mice displayed). b. Contours of ChR2 expression in the pontine nuclei for mice used in behavioral experiments. c. Right cerebellum of the animal shown in a. Mossy fiber axons (grey arrow) and boutons (white arrow) can be seen expressing ChR2 in the cerebellar cortex. The approximate location of the optical fiber and recording site path are shown in white (1 of 7 mice displayed). d. Location of fiber placement in a representative section for animals used in behavioral experiments.
a. Two example mice with differing effects of stimulation on early reaches in the stimulation block. To account for diverging effects we define the direction of deviation with stimulation as positive and the opposing direction as negative. b. Summary of the relative change in upward position for the same data shown in Fig. 3e. Relative change in upward position was assessed in the 50-ms window following the end of stimulation. c. Summary of the relative change in lateral position for the same data shown in Fig. 3e. d. Summary of the relative change in outward position for in the 50-ms window before stimulation. e. Stimulating with 635-nm light did not cause deviations in position or adaptation profiles. f. Stimulating while the mouse was awake with its hand at rest on the bar produced virtually no movement. (Sample size: (b-d) N = 5 mice; 104 sessions, (e) N = 2 mice; 19 sessions, (f) N = 4 animals, 21 sessions; all error bars and bands represent mean ± SEM; in box and whiskers plots box denotes median and 25th/75th percentiles, whiskers denote 10th and 90th percentiles, circle indicates mean).
Extended Data Fig. 8 Assessing unit stability across recording sessions and population responses during fixed-position stimulation experiments.
a. Left: Waveforms templates detected on each Neuropixel electrode for a cell during baseline and during washout. Right: Histogram of waveform correlation of PCs across sessions (red) and of mismatched neighboring cells, across the session (shuffled control, grey). PCs with an across-session waveform correlation that fell below the 99th percentile of the shuffled control (dashed line) were excluded from further analysis. b. Left: Unit displacement for cells across a session. Baseline unit position is shown in grey and washout position is shown in red. Right: Histogram of unit displacement of PCs across sessions (red) and of mismatched neighboring cells, across the session (shuffled control, grey). PCs with an across-session displacement that fell below the 1st percentile of the shuffled control (dashed line) were excluded from further analysis. c. Same as data shown in Fig. 4g with the last 5 stimulated and washout reaches included. The initial stimulation and washout effects are reduced across the stimulation and washout blocks, respectively. d. Cspks analyzed during fixed-position stimulation experiments for the baseline, stimulation, and washout blocks. e. Same as the analysis shown a but only including PCs with Cspks. These cells show similar negative deflections with stimulation then adaptation upwards over the stimulation block compared to the total PC dataset. f. Quantification of simple spike firing rates in the stimulation window for the data shown in c. (p-values: (f) 5.1 ×10-5, RM one-way ANOVA, end baseline vs. first wash: 0.022, end baseline vs. first 5 wash: 0.047, first stim vs. last 5 stim: 0.043, first stim vs. first wash: 6.1 ×10-4, first stim vs. last 5 wash: 3.8 ×10-3, last 5 stim vs. first wash: 0.034, Tukey’s multiple comparisons test; Sample size: (c) n = 159 cells (d-f) n = 13 cells; * indicates p-value < 0.05; all error bars and bands represent mean ± SEM).
Extended Data Fig. 9 Temporal analysis of early washout effect for fixed-position and random-stimulation experiments.
a. Analysis of fixed-position stimulation experiment early washout reaches in 50-ms time windows across the reach. Each window is shifted 10 ms from the adjacent time window. Aftereffect emerges around the time stimulation was delivered in the stimulation block. b. Same as a but for random-position stimulation experiments. Consistent aftereffects relative to baseline reaches do not emerge in any of the analyzed windows. (Sample size: (a) N = 5 mice, 104 sessions, (b) N = 5 mice, 60 sessions; in box and whiskers plots box denotes median and 25th/75th percentiles, whiskers denote 10th and 90th percentiles, circle indicates mean).
a. Model as described in Fig. 6. In this case the number of stimulated MLIs is greater than the number of parallel fibers (bottom) leading to a net negative stimulation effect. Negative simple spike error lowers the probability of Cspks below baseline, leading to LTP (top). b. PC simple spike activity during the stimulation block and washout block of fixed-position stimulation as described in Fig. 6b. Here the stimulation reduces firing rate. c. Same as described in Fig. 6c. Here parallel fiber weight changes increase to compensate for the stimulation. Note that while not displayed the quantification of the adaptation is identical to the data displayed in Fig. 6d. d. Comparison of model output to empirical observations for fixed-position stimulus conditions (Fig. 3). Model closely matches behavioral adaptation. e. Same as b. but here the stimulation window is randomized across the reach. Note that while not displayed the quantification of the adaptation is identical to the data displayed in Fig. 6d.
About this article
Cite this article
Calame, D.J., Becker, M.I. & Person, A.L. Cerebellar associative learning underlies skilled reach adaptation. Nat Neurosci 26, 1068–1079 (2023). https://doi.org/10.1038/s41593-023-01347-y
This article is cited by
Nature Neuroscience (2023)