Skip to main content

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

  • Brief Communication
  • Published:

Cerebellar Purkinje cell activity drives motor learning

Nature Neuroscience volume 16, pages 1734–1736 (2013)Cite this article

Subjects

Abstract

The climbing fiber input to the cerebellar cortex is thought to provide instructive signals that drive the induction of motor skill learning. We found that optogenetic activation of Purkinje cells, the sole output neurons of the cerebellar cortex, can also drive motor learning in mice. This dual control over the induction of learning by climbing fibers and Purkinje cells can expand the learning capacity of motor circuits.

This is a preview of subscription content, access via your institution

Access options

Buy this article

  • Purchase on Springer Link
  • Instant access to full article PDF
Buy now

Prices may be subject to local taxes which are calculated during checkout

Figure 1: Purkinje cell activation induced associative motor learning.
Figure 2: Climbing fiber activation induced motor learning with different timing than Purkinje cells.

Similar content being viewed by others

References

  1. Ito, M. Annu. Rev. Neurosci. 12, 85–102 (1989).

    Article  CAS  Google Scholar 

  2. Mauk, M.D., Steinmetz, J.E. & Thompson, R.F. Proc. Natl. Acad. Sci. USA 83, 5349–5353 (1986).

    Article  CAS  Google Scholar 

  3. Ke, M.C., Guo, C.C. & Raymond, J.L. Nat. Neurosci. 12, 1171–1179 (2009).

    Article  CAS  Google Scholar 

  4. Catz, N., Dicke, P.W. & Thier, P. Curr. Biol. 15, 2179–2189 (2005).

    Article  CAS  Google Scholar 

  5. Miles, F.A. & Lisberger, S.G. Annu. Rev. Neurosci. 4, 273–299 (1981).

    Article  CAS  Google Scholar 

  6. Lisberger, S.G. & Fuchs, A.F. Brain Res. 69, 347–353 (1974).

    Article  CAS  Google Scholar 

  7. McCormick, D.A. & Thompson, R.F. Science 223, 296–299 (1984).

    Article  CAS  Google Scholar 

  8. Mauk, M.D. & Donegan, N.H. Learn. Mem. 4, 130–158 (1997).

    Article  CAS  Google Scholar 

  9. Shutoh, F., Ohki, M., Kitazawa, H., Itohara, S. & Nagao, S. Neuroscience 139, 767–777 (2006).

    Article  CAS  Google Scholar 

  10. Babalian, A.L. & Vidal, P.P. J. Neurophysiol. 84, 2514–2528 (2000).

    Article  CAS  Google Scholar 

  11. Heuer, H.W., Tokiyama, S. & Lisberger, S.G. J. Neurophysiol. 100, 1320–1331 (2008).

    Article  Google Scholar 

  12. Ito, M., Shiida, T., Yagi, N. & Yamamoto, M. Brain Res. 65, 170–174 (1974).

    Article  CAS  Google Scholar 

  13. Raymond, J.L. & Lisberger, S.G. J. Neurosci. 18, 9112–9129 (1998).

    Article  CAS  Google Scholar 

  14. Stahl, J.S. J. Neurophysiol. 91, 2066–2078 (2004).

    Article  Google Scholar 

  15. Crepel, F. & Jaillard, D. J. Physiol. (Lond.) 432, 123–141 (1991).

    Article  CAS  Google Scholar 

  16. Belmeguenai, A. et al. J. Neurosci. 30, 13630–13643 (2010).

    Article  CAS  Google Scholar 

  17. Pugh, J.R. & Raman, I.M. Neuron 51, 113–123 (2006).

    Article  CAS  Google Scholar 

  18. Menzies, J.R.W., Porrill, J., Dutia, M. & Dean, P. PLoS ONE 5, e13182 (2010).

    Article  Google Scholar 

  19. McElvain, L.E., Bagnall, M.W., Sakatos, A. & du Lac, S. Neuron 68, 763–775 (2010).

    Article  CAS  Google Scholar 

  20. Aizenman, C.D., Manis, P.B. & Linden, D.J. Neuron 21, 827–835 (1998).

    Article  CAS  Google Scholar 

  21. Barski, J.J., Dethleffsen, K. & Meyer, M. Genesis 28, 93–98 (2000).

    Article  CAS  Google Scholar 

  22. Madisen, L. et al. Nat. Neurosci. 15, 793–802 (2012).

    Article  CAS  Google Scholar 

  23. Sohal, V.S., Zhang, F., Yizhar, O. & Deisseroth, K. Nature 459, 698–702 (2009).

    Article  CAS  Google Scholar 

  24. Boyden, E.S. & Raymond, J.L. Neuron 39, 1031–1042 (2003).

    Article  CAS  Google Scholar 

  25. Aravanis, A.M. et al. J. Neural Eng. 4, S143–S156 (2007).

    Article  Google Scholar 

  26. Barmack, N.H. & Yakhnitsa, V. J. Neurosci. 31, 9824–9835 (2011).

    Article  CAS  Google Scholar 

  27. Yoshida, T., Funabiki, K. & Hirano, T. Eur. J. Neurosci. 25, 1467–1474 (2007).

    Article  Google Scholar 

  28. Hoebeek, F.E. et al. Neuron 45, 953–965 (2005).

    Article  CAS  Google Scholar 

  29. Lisberger, S.G. & Fuchs, A.F. J. Neurophysiol. 41, 733–763 (1978).

    Article  CAS  Google Scholar 

  30. Nagao, S. Exp. Brain Res. 77, 531–540 (1989).

    CAS  PubMed  Google Scholar 

  31. Pastor, A.M., De la Cruz, R.R. & Baker, R. Prog. Brain Res. 114, 359–381 (1997).

    Article  CAS  Google Scholar 

  32. Kahlon, M. & Lisberger, S.G. J. Neurophysiol. 84, 2945–2960 (2000).

    Article  CAS  Google Scholar 

Download references

Acknowledgements

We thank E. Knudsen, T. Moore, C. Shatz, D. Madison, G. Zhao, O. Winter, S. Umamoto, A. Adamantidis, M. Carter, H. Nguyen and R. Hemmati for discussions and assistance. This study was supported by grants from the US National Institutes of Health (RO1 DC04154, RO1 NS072406 and P01 NS053862) and the James S. McDonnell Foundation to J.L.R., from the US National Science Foundation Graduate Research Fellowship Program and US National Institutes of Health (F31DC010547) to T.D.B.N.-V., from the US National Institutes of Health and Defense Advanced Research Projects Agency to K.D., from the US National Institutes of Health (K01 NS069617) to R.R.K., and from the US National Institutes of Health (P30 DC10363 and P30 NS069375) for imaging and virus core facilities.

Author information

Author notes

  1. T D Barbara Nguyen-Vu and Rhea R Kimpo: These authors contributed equally to this work.

Authors and Affiliations

  1. Department of Neurobiology, Stanford School of Medicine, Stanford, California, USA

    T D Barbara Nguyen-Vu, Rhea R Kimpo, Jacob M Rinaldi, Arunima Kohli & Jennifer L Raymond

  2. Department of Molecular and Cellular Physiology, Stanford School of Medicine, Stanford, California, USA

    T D Barbara Nguyen-Vu

  3. Allen Institute for Brain Science, Seattle, Washington, USA

    Hongkui Zeng

  4. Department of Bioengineering, Stanford University, Stanford, California, USA

    Karl Deisseroth

  5. Department of Psychiatry, Stanford School of Medicine, Stanford, California, USA

    Karl Deisseroth

  6. Howard Hughes Medical Institute, Stanford University, Stanford, California, USA

    Karl Deisseroth

Authors
  1. T D Barbara Nguyen-Vu
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Rhea R Kimpo
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Jacob M Rinaldi
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Arunima Kohli
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Hongkui Zeng
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. Karl Deisseroth
    View author publications

    You can also search for this author in PubMed Google Scholar

  7. Jennifer L Raymond
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

T.D.B.N.-V. conducted all Purkinje cell experiments. R.R.K. conducted all climbing fiber experiments. J.M.R. conducted pilot experiments. A.K. helped with histology. K.D. and H.Z. provided reagents. J.L.R. supervised all aspects of the work.

Corresponding author

Correspondence to Jennifer L Raymond.

Ethics declarations

Competing interests

The authors declare no competing financial interests.

Integrated supplementary information

Supplementary Figure 1 Selective expression of ChR2 in Purkinje cells in the cerebellar flocculus

To selectively activate Purkinje cells, we expressed channelrhodopsin (ChR2) under the control of the Purkinje cell-specific L7/pcp-2 promoter. (a) Coronal diagram of the mouse cerebellum showing the site of optogenetic stimulation. An optical fiber delivering 473nm (blue) light was sealed to prevent light emission except for where the tip penetrated the brain. Fl, flocculus; PFl, paraflocculus. (b) Targeting of ChR2-EYFP expression to Purkinje cells of the cerebellar flocculus using viral and transgenic approaches. ChR2-EYFP (green) expression in coronal sections corresponding to grey box in (panel a) of an L7-Cre mouse injected with pAAV-EF1α-double floxed-hChR2(H134R)-EYFP (left), and a transgenic mouse resulting from a cross between L7-Cre and Ai32 ChR2 mouse lines (right). With both approaches, a substantial fraction of Purkinje cells in the flocculus expressed ChR2-EYFP. (c) Confocal image of the flocculus of a transgenic mouse (box in panel b, right) shows the extensive overlap (orange) between the membrane bound protein, ChR2-EYFP (green) and antibodies to calbindin (red), which labels the Purkinje cells. Blue, cell bodies stained with DAPI; ML, molecular layer; Pk, Purkinje cell layer.

Supplementary Figure 2 Optogenetic activation of Purkinje cells in the cerebellar flocculus

(a) Optrode recordings in awake mice verified that illumination with blue light successfully elicited spiking in Purkinje cells of the flocculus, the region of the cerebellum necessary for VOR learning. In vivo, extracellular optrode recording from an awake animal showing Purkinje cell activation with trains of blue light (trains of 5 ms pulses at 50 Hz for 420 ms every 1 second; ≤ 3 mW/mm2 intensity). Rasters show the spikes elicited by all the stimulus trains delivered during six 5-min blocks of stimulation (40 seconds between blocks); histogram shows the effect of stimulation over the entire 30-min period (below). Bin size 5 ms. (b) Expanded view of red box in a, and example spontaneous (black) and ChR2-elicted (cyan) Purkinje cell simple spike waveforms (black dot), which differ from the spontaneous complex spike waveform (orange dot) recorded from the same cell. ChR2-mediated depolarization of Purkinje cells elicited action potentials resembling spontaneous simple spikes, rather than the complex spikes elicited by climbing fiber input, in that they occurred at high rates and did not have the Ca2+ spikelets characteristic of spontaneous or optogenetically elicited complex spikes. Bin size 5 ms. (c) Histograms of the simple spike inter-spike intervals (ISIs) of two, representative Purkinje cells immediately before the start of stimulation (open bars), and during optogenetic stimulation (cyan bars; trains of 5 ms pulses at 50 Hz for 420 ms every 1 second; ≤ 3 mW/mm2 intensity). Optogenetic stimulation increased the mean firing rate of the Purkinje cells, as shown by the leftward shift in the ISI distribution, with no evidence of bursting. In some cells, optogenetic stimulation induced some entrainment of firing at the frequency of stimulation (50 Hz/ 20 ms ISI) (right). Data in the left panel are from the same Purkinje cell as in panel a. Bin size 500 μs. (d) Blue light stimulation elevated Purkinje cell firing rates (Light ON; 100.1 ± 7.8 Hz) above the spontaneous levels recorded in between the stimulus trains (Light OFF; 45.9 ± 2.6 Hz) (t(8) = 7.322, P < 0.0001, Paired t-test, n = 9 cells from 8 mice; circles, transgenic; diamond, virus injected). The increase in Purkinje cell firing rate achieved with optogenetic stimulation was comparable to what has been reported during visually and vestibularly driven eye movements. Mean ± s.e.m.

Supplementary Figure 3 Unilateral Purkinje cell activation induced motor learning

In a separate cohort of animals to that shown in Figure 1, training was done with stimulation of floccular Purkinje cells in only one hemisphere during the contraversiveor ipsiversivephase of the vestibular stimulus. (a) Schematic of the unilateral optogenetic training paradigms for modifying the VOR. A sinusoidal vestibular stimulus was used for training and testing (black trace, angular velocity of the head). Unilateral optogenetic stimulation of Purkinje cells was paired with the contraversive phase of the vestibular stimulus (red, light on when the head was rotating away from the side of Purkinje cell stimulation) or the ipsiversive phase of the vestibular stimulus (blue, light on when the head was rotating toward the side of stimulation). (b) Motor learning induced by Purkinje cell activation depended on the training condition (F(2, 36) = 3.513, P = 0.04, two-way repeated measures ANOVA). Purkinje cell activation paired with contraversive vestibular stimulation (red circles) was different than the vestibular-alone control (black open squares; *P = 0.025, Fisher test; n = 13 mice). The learning induced by the unilateral training paradigm was approximately half that induced by bilateral training (compare with Figure 1b, red circles). The effects on the behavior were associative, since the timing of the Purkinje cell activity relative to the vestibular stimulus was critical to the induction of VOR learning (Purkinje cell contra vs. ipsistim, *P = 0.030, Fisher test; n = 13 mice).There was no significant difference between training with Purkinje cell stimulation during the ipsiversive phase of the vestibular stimulus (blue squares) and the vestibular-alone control (open black squares; P = 0.94, Fisher test; n = 13 mice). This is consistent with the previous finding that electrical stimulation of the flocculus induced no learning when paired with stimulation of the ipsilateral vestibular nerve to mimic an ipsiversive vestibular stimulus (Babalian & Vidal, 2000, J Neurophysiol 84: 2514). Thus, neither the Purkinje cell activation itself, nor any calcium influx or non-associative plasticity it may have caused were sufficient to alter the behavior. Mean ± s.e.m.

Supplementary Figure 4 Optogenetic stimulation of Purkinje cells elicits eye movements

(a) Schematic of the VOR circuit. The inhibitory Purkinje cells (PC) are only two synapses from the oculomotor neurons (MN). GC, granule cell; PF, parallel fiber; VN, vestibular nucleus. (b) Optogenetic Purkinje cell stimulation elicited eye movement responses at short latencies. When brief (5 ms) light pulses were delivered at a relatively low frequency (20 Hz), the eye movement elicited by each pulse could be resolved, illustrating the effectiveness of Purkinje cell activity to affect online motor performance.

Supplementary Figure 5 Distinct readouts of Purkinje cell activity for motor performance and motor learning

The effects of Purkinje cell activation on the induction of motor learning could be dissociated from its immediate effects on motor performance. (a) Examples illustrating the variation across mice in the effects of Purkinje cell stimulation on eye movement performance during training. In different mice, Purkinje cell stimulation during the contraversive phase of the vestibular stimulus could produce an immediate decrease (left, purple) or increase (right, orange) in eye movement amplitude, compared with the VOR response in the absence of Purkinje cell stimulation (grey). In both animals, 30-min of training with Purkinje cell activation during the contraversive phase of the vestibular stimulus induced a learned increase in the VOR, as measured in the absence of Purkinje cell stimulation (corresponding color symbols in panel b). (b) Within a given experiment, there was no significant correlation between the immediate effect of Purkinje cell stimulation on the eye movement performance during training and the learning it induced (R(25) = −0.17, P = 0.42, Pearson correlation; n = 25 mice). Each point represents data from an individual experiment in a different mouse. Learning was measured as the percent change in VOR amplitude at the end of training compared to pre-training, tested in the absence of Purkinje cell stimulation. The effect on performance was calculated as the percent change in eye movement amplitude observed immediately, when Purkinje cells were stimulated during the vestibular stimulus, as compared with responses made to the vestibular stimulus alone before training. The immediately evoked eye movements could increase (n = 9, right), decrease (n = 10, left), or have no significant effect (n = 6, points lying on vertical axis, including 4 mice with evoked vertical eye movements) on the amplitude of the on-going VOR during training. This variable immediate effect on VOR performance mirrors the heterogeneity of the Purkinje cell population within the flocculus, observed in single unit recordings. Variable expression levels of ChR2 and/or the placement of the optical fiber within the flocculus could differentially activate subpopulations of Purkinje cells driving ipsiversive or contraversive eye movements to yield an immediate increase or decrease in eye movement, or no change if the two populations were activated equally. Despite variable immediate effects on the on-going eye movement performance, there was a remarkably consistent effect of Purkinje cell activation on VOR-increase learning (points above solid horizontal axis, which shows mean for vestibular-alone control). Notably, the immediate effect of Purkinje cell stimulation on eye movement performance was similar at the end versus the beginning of training (1st 5-min block versus 6th 5-min block, t(24)= 1.978, P = 0.06, paired t-test, n = 25; data not shown), suggesting that the effectiveness of optogenetic stimulation and Purkinje cell excitability were similar throughout training. (c) When Purkinje cell stimulation was paired with an ipsiversive vestibular stimulus, there was no significant effect on learning, and there was no significant correlation between the effect of Purkinje cell stimulation on the immediate eye movement performance and learning (R(24) = −0.26, P = 0.22, Pearson correlation; n = 25 mice). (d) Unsupported model: the results in panels b and c indicate that the effect of Purkinje cell activity on learning is not a secondary consequence of the eye movements present during training. (e) Our data support the model that there are independent read outs of the Purkinje cell activity for the control of movement on different time scales: immediate performance versus learning. Different subpopulations of Purkinje cells may make different contributions to eye movement performance, but their contribution to learning appears to be more uniform.

Supplementary Figure 6 Optogenetic activation of climbing fibers in the cerebellar flocculus

The dependence of learning on which cells expressed the ChR2, the Purkinje cells vs. climbing fibers (compare Figures 1 and 2) rules out any non-specific effects of the experimental procedures, such as the presence of the blue light or general activation of the circuit. (a) Coronal diagram of the cerebellum and brainstem illustrating the site of virus injection and subsequent optogenetic activation of climbing fibers (green) with 473nm light (cyan). ChR2 was expressed in climbing fibers by injecting AAV-CaMKIIα-hChR2(H134R)-EYFP into the dorsal cap of Kooy, the subnucleus of the inferior olive that provides the climbing fiber input to the flocculus. Right, coronal sections showing ChR2-EYFP expression, visualized using immunostaining against EYFP (green), in the climbing fibers of the flocculus (top) and the corresponding injection site in the dorsal cap of Kooy (bottom). Fl, flocculus; CF, climbing fiber; IO, inferior olive; ML, molecular layer; GCL, granule cell layer; Blue, cell bodies stained with DAPI. (b) Representative example of an in vivo extracellular optrode recording from a Purkinje cell in the flocculus showing the complex spikes elicited by optogenetic activation of its climbing fiber input (left). Climbing fiber activation was elicited using 250 ms trains of three pulses of blue light (cyan; 2 ms duration, 0.3 mW/mm2) repeated at 1 s intervals, and delivered to the cerebellar flocculus. Individual waveforms show the optogenetically elicited complex spikes with the stimulus artifact subtracted. Right, overlay of all the optogenetically elicited complex spike waveforms recorded in this cell during a ∼10 minute period of stimulation (1,718 waveforms). Climbing fiber activity has been hypothesized to recruit plasticity in the cerebellar cortex or induce plasticity downstream by causing pauses in Purkinje cell simple spike firing that disinhibit their targets. (c) Histogram from a representative Purkinje cell, showing simple spike rate aligned on the time of complex spikes (t=0) elicited optogenetically using 2ms pulses of light. A typical 10–20 ms pause in simple spike firing occurred after each optogenetically elicited complex spike. Bin size, 2 ms.

Supplementary Figure 7 VOR learning induced with visual-vestibular pairing

VOR learning induced by pairing the same vestibular stimulus used in Figs. 1, 2, and Supplementary Fig. 3 with a moving visual stimulus. VOR learning was tested by briefly interrupting the pairing to measure the eye movement response to the vestibular stimulus in the absence of the visual stimulus. For comparison, the habituation induced by the vestibular stimulus alone is indicated by the dotted trace. These experiments were performed on the same cohort of mice that underwent training with unilateral optogenetic Purkinje cell stimulation (Supplementary Fig. 3). If the visual stimulus moved in the opposite direction from the head, the VOR increased during the 30-min training period (upward triangles; t(11) = 3.528, P = 0.005, one-sample t-test; n = 12). If the visual stimulus moved in the same direction as the head, the VOR decreased (downward triangles; t(7) = 8.745, P < 0.0001, one-sample t-test; n=8) below the vestibular stimulus alone condition (P = 0.004, t-test). Notably, none of our optogenetic stimulation paradigms (see Figs. 1 and 2) induced such an associative decrease in the VOR below that induced by the vestibular stimulus alone. It is possible that the mechanisms supporting the associative decrease in the VOR are partially shared with, or otherwise occluded by, the habituation observed in response to the vestibular stimulus alone. It is also possible that other, untested climbing fiber or Purkinje cell stimulation protocols would induce an associative learned decrease in the VOR. Mean ± s.e.m.

Supplementary information

Supplementary Figures

Supplementary Figures 1–7 (PDF 18320 kb)

Source data

Rights and permissions

Reprints and permissions

About this article

Cite this article

Nguyen-Vu, T., Kimpo, R., Rinaldi, J. et al. Cerebellar Purkinje cell activity drives motor learning. Nat Neurosci 16, 1734–1736 (2013). https://doi.org/10.1038/nn.3576

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1038/nn.3576

This article is cited by

Search

Advanced search

Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing