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.

Synaptic pruning through glial synapse engulfment upon motor learning

Abstract

Synaptic pruning is a fundamental process of neuronal circuit refinement in learning and memory. Accumulating evidence suggests that glia participates in sculpting the neuronal circuits through synapse engulfment. However, whether glial involvement in synaptic pruning has a role in memory formation remains elusive. Using newly developed phagocytosis reporter mice and three-dimensional ultrastructural characterization, we found that synaptic engulfment by cerebellar Bergmann glia (BG) frequently occurred upon cerebellum-dependent motor learning in mice. We observed increases in pre- and postsynaptic nibbling by BG along with a reduction in spine volume after learning. Pharmacological blockade of engulfment with Annexin V inhibited both the spine volume reduction and overnight improvement of motor adaptation. These results indicate that BG contribute to the refinement of the mature cerebellar cortical circuit through synaptic engulfment during motor learning.

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

Access options

Buy article

Get time limited or full article access on ReadCube.

$32.00

All prices are NET prices.

Fig. 1: Bergmann glial engulfment of neuronal materials.
Fig. 2: BG engulf neuronal materials through PS-dependent phagocytosis.
Fig. 3: Distinct engulfment of neuronal materials by different types of glia in the adult cerebellum.
Fig. 4: Bergmann glial engulfment of neuronal materials increases after learning.
Fig. 5: Bergmann glial phagocytotic nibbling of synaptic and non-synaptic compartments.
Fig. 6: Close association between aberrant neuronal organelles and glial nibbling.
Fig. 7: Synaptic engulfment by BG increases upon learning.
Fig. 8: Synaptic nibbling is involved in spine volume reduction and in behavioral improvement.

Data availability

Data and resources are available on request from the corresponding authors. Source data are provided with this paper.

Code availability

Code used to present visual stimulation for this paper is available on request from the corresponding authors.

References

  1. Holtmaat, A. & Svoboda, K. Experience-dependent structural synaptic plasticity in the mammalian brain. Nat. Rev. Neurosci. 10, 647–658 (2009).

    Article  CAS  PubMed  Google Scholar 

  2. Caroni, P., Donato, F. & Muller, D. Structural plasticity upon learning: regulation and functions. Nat. Rev. Neurosci. 13, 478–490 (2012).

    Article  CAS  PubMed  Google Scholar 

  3. Forrest, M. P., Parnell, E. & Penzes, P. Dendritic structural plasticity and neuropsychiatric disease. Nat. Rev. Neurosci. 19, 215–234 (2018).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  4. Kasai, H., Ziv, N. E., Okazaki, H., Yagishita, S. & Toyoizumi, T. Spine dynamics in the brain, mental disorders and artificial neural networks. Nat. Rev. Neurosci. 22, 407–422 (2021).

    Article  CAS  PubMed  Google Scholar 

  5. Ito, M. Long-term depression. Annu Rev. Neurosci. 12, 85–102 (1989).

    Article  CAS  PubMed  Google Scholar 

  6. Boyden, E. S., Katoh, A. & Raymond, J. L. Cerebellum-dependent learning: the role of multiple plasticity mechanisms. Annu. Rev. Neurosci. 27, 581–609 (2004).

    Article  CAS  PubMed  Google Scholar 

  7. Kano, M., Rexhausen, U., Dreessen, J. & Konnerth, A. Synaptic excitation produces a long-lasting rebound potentiation of inhibitory synaptic signals in cerebellar Purkinje cells. Nature 356, 601–604 (1992).

    Article  CAS  PubMed  Google Scholar 

  8. Schonewille, M. et al. Purkinje cell-specific knockout of the protein phosphatase PP2B impairs potentiation and cerebellar motor learning. Neuron 67, 618–628 (2010).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  9. Kim, Y. G. & Kim, S. J. Decreased intrinsic excitability of cerebellar Purkinje cells following optokinetic learning in mice. Mol. Brain 13, 136 (2020).

    Article  PubMed  PubMed Central  Google Scholar 

  10. Gutierrez-Castellanos, N. et al. Motor learning requires Purkinje cell synaptic potentiation through activation of AMPA-receptor subunit GluA3. Neuron 93, 409–424 (2017).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  11. Wang, W. et al. Distinct cerebellar engrams in short-term and long-term motor learning. Proc. Natl Acad. Sci. USA 111, E188–E193 (2014).

    CAS  PubMed  Google Scholar 

  12. Aziz, W. et al. Distinct kinetics of synaptic structural plasticity, memory formation, and memory decay in massed and spaced learning. Proc. Natl Acad. Sci. USA 111, E194–E202 (2014).

    Article  CAS  PubMed  Google Scholar 

  13. Inoshita, T. & Hirano, T. Occurrence of long-term depression in the cerebellar flocculus during adaptation of optokinetic response. eLife 7, e36209 (2018).

    Article  PubMed  PubMed Central  Google Scholar 

  14. Kakegawa, W. et al. Optogenetic control of synaptic AMPA receptor endocytosis reveals roles of LTD in motor learning. Neuron 99, 985–998 (2018).

    Article  CAS  PubMed  Google Scholar 

  15. Chung, W. S. et al. Astrocytes mediate synapse elimination through MEGF10 and MERTK pathways. Nature 504, 394–400 (2013).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  16. Schafer, D. P. et al. Microglia sculpt postnatal neural circuits in an activity and complement-dependent manner. Neuron 74, 691–705 (2012).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  17. Hong, S. et al. Complement and microglia mediate early synapse loss in Alzheimer mouse models. Science 352, 712–716 (2016).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  18. Morizawa, Y. M. et al. Reactive astrocytes function as phagocytes after brain ischemia via ABCA1-mediated pathway. Nat. Commun. 8, 28 (2017).

    Article  PubMed  PubMed Central  Google Scholar 

  19. Lee, J. H. et al. Astrocytes phagocytose adult hippocampal synapses for circuit homeostasis. Nature 590, 612–617 (2021).

    Article  CAS  PubMed  Google Scholar 

  20. Chung, W. S. et al. Novel allele-dependent role for APOE in controlling the rate of synapse pruning by astrocytes. Proc. Natl Acad. Sci. USA 113, 10186–10191 (2016).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  21. Rueda-Carrasco, J. et al. Microglia detect externalized phosphatidylserine on synapses for elimination via TREM2 in Alzheimer’s disease models. Preprint at bioRxiv https://doi.org/10.1101/2022.04.04.486424 (2022).

  22. Weinhard, L. et al. Microglia remodel synapses by presynaptic trogocytosis and spine head filopodia induction. Nat. Commun. 9, 1228 (2018).

    Article  PubMed  PubMed Central  Google Scholar 

  23. Tantama, M., Hung, Y. P. & Yellen, G. Imaging intracellular pH in live cells with a genetically encoded red fluorescent protein sensor. J. Am. Chem. Soc. 133, 10034–10037 (2011).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  24. Katayama, H., Kogure, T., Mizushima, N., Yoshimori, T. & Miyawaki, A. A sensitive and quantitative technique for detecting autophagic events based on lysosomal delivery. Chem. Biol. 18, 1042–1052 (2011).

    Article  CAS  PubMed  Google Scholar 

  25. Katayama, H. et al. Visualizing and modulating mitophagy for therapeutic studies of neurodegeneration. Cell 181, 1176–1187 (2020).

    Article  CAS  PubMed  Google Scholar 

  26. Tanaka, K. F. et al. Expanding the repertoire of optogenetically targeted cells with an enhanced gene expression system. Cell Rep. 2, 397–406 (2012).

    Article  CAS  PubMed  Google Scholar 

  27. Nitta, K., Matsuzaki, Y., Konno, A. & Hirai, H. Minimal Purkinje cell-specific PCP2/L7 promoter virally available for rodents and non-human primates. Mol. Ther. Methods Clin. Dev. 6, 159–170 (2017).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  28. Scott-Hewitt, N. et al. Local externalization of phosphatidylserine mediates developmental synaptic pruning by microglia. EMBO J. 39, e105380 (2020).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  29. Li, T. et al. A splicing isoform of GPR56 mediates microglial synaptic refinement via phosphatidylserine binding. EMBO J. 39, e104136 (2020).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  30. van den Eijnde, S. M. et al. Spatiotemporal distribution of dying neurons during early mouse development. Eur. J. Neurosci. 11, 712–724 (1999).

    Article  PubMed  Google Scholar 

  31. Lu, Z. et al. Phagocytic activity of neuronal progenitors regulates adult neurogenesis. Nat. Cell Biol. 13, 1076–1083 (2011).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  32. Hill, R. A., Damisah, E. C., Chen, F., Kwan, A. C. & Grutzendler, J. Targeted two-photon chemical apoptotic ablation of defined cell types in vivo. Nat. Commun. 8, 15837 (2017).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  33. Nagao, S. Role of cerebellar flocculus in adaptive interaction between optokinetic eye movement response and vestibulo-ocular reflex in pigmented rabbits. Exp. Brain Res. 77, 541–551 (1989).

    Article  CAS  PubMed  Google Scholar 

  34. Palay, S. L. & Chan-Palay, V. Cerebellar Cortex: Cytology and Organization (Springer, 1974).

  35. Spacek, J. & Harris, K. M. Trans-endocytosis via spinules in adult rat hippocampus. J. Neurosci. 24, 4233–4241 (2004).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  36. Khaminets, A. et al. Regulation of endoplasmic reticulum turnover by selective autophagy. Nature 522, 354–358 (2015).

    Article  CAS  PubMed  Google Scholar 

  37. McArthur, K. et al. BAK/BAX macropores facilitate mitochondrial herniation and mtDNA efflux during apoptosis. Science 359, eaao6047 (2018).

    Article  PubMed  Google Scholar 

  38. Saab, A. S. et al. Bergmann glial AMPA receptors are required for fine motor coordination. Science 337, 749–753 (2012).

    Article  CAS  PubMed  Google Scholar 

  39. Xavier, A. L. et al. Cannula implantation into the cisterna magna of rodents. J. Vis. Exp. https://doi.org/10.3791/57378 (2018).

  40. Tabula Muris, C. A single-cell transcriptomic atlas characterizes ageing tissues in the mouse. Nature 583, 590–595 (2020).

    Article  Google Scholar 

  41. Lehrman, E. K. et al. CD47 protects synapses from excess microglia-mediated pruning during development. Neuron 100, 120–134 (2018).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  42. Zhou, Z., Hartwieg, E. & Horvitz, H. R. CED-1 is a transmembrane receptor that mediates cell corpse engulfment in C. elegans. Cell 104, 43–56 (2001).

    Article  CAS  PubMed  Google Scholar 

  43. Moynault, A., Luciani, M. F. & Chimini, G. ABC1, the mammalian homologue of the engulfment gene ced-7, is required during phagocytosis of both necrotic and apoptotic cells. Biochem. Soc. Trans. 26, 629–635 (1998).

    Article  CAS  PubMed  Google Scholar 

  44. Hamon, Y. et al. ABC1 promotes engulfment of apoptotic cells and transbilayer redistribution of phosphatidylserine. Nat. Cell Biol. 2, 399–406 (2000).

    Article  CAS  PubMed  Google Scholar 

  45. Hamon, Y. et al. Cooperation between engulfment receptors: the case of ABCA1 and MEGF10. PLoS ONE 1, e120 (2006).

    Article  PubMed  PubMed Central  Google Scholar 

  46. Loov, C., Mitchell, C. H., Simonsson, M. & Erlandsson, A. Slow degradation in phagocytic astrocytes can be enhanced by lysosomal acidification. Glia 63, 1997–2009 (2015).

    Article  PubMed  PubMed Central  Google Scholar 

  47. Hayashi-Takagi, A. et al. Labelling and optical erasure of synaptic memory traces in the motor cortex. Nature 525, 333–338 (2015).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  48. Cichon, J. & Gan, W. B. Branch-specific dendritic Ca2+ spikes cause persistent synaptic plasticity. Nature 520, 180–185 (2015).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  49. Holler, S., Kostinger, G., Martin, K. A. C., Schuhknecht, G. F. P. & Stratford, K. J. Structure and function of a neocortical synapse. Nature 591, 111–116 (2021).

    Article  CAS  PubMed  Google Scholar 

  50. Martin, S. J. & Morris, R. G. New life in an old idea: the synaptic plasticity and memory hypothesis revisited. Hippocampus 12, 609–636 (2002).

    Article  CAS  PubMed  Google Scholar 

  51. Shutoh, F., Ohki, M., Kitazawa, H., Itohara, S. & Nagao, S. Memory trace of motor learning shifts transsynaptically from cerebellar cortex to nuclei for consolidation. Neuroscience 139, 767–777 (2006).

    Article  CAS  PubMed  Google Scholar 

  52. Boele, H. J., Koekkoek, S. K., De Zeeuw, C. I. & Ruigrok, T. J. Axonal sprouting and formation of terminals in the adult cerebellum during associative motor learning. J. Neurosci. 33, 17897–17907 (2013).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  53. Wulff, P. et al. Synaptic inhibition of Purkinje cells mediates consolidation of vestibulo-cerebellar motor learning. Nat. Neurosci. 12, 1042–1049 (2009).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  54. Chen, S. X., Kim, A. N., Peters, A. J. & Komiyama, T. Subtype-specific plasticity of inhibitory circuits in motor cortex during motor learning. Nat. Neurosci. 18, 1109–1115 (2015).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  55. Becker, N., Wierenga, C. J., Fonseca, R., Bonhoeffer, T. & Nagerl, U. V. LTD induction causes morphological changes of presynaptic boutons and reduces their contacts with spines. Neuron 60, 590–597 (2008).

    Article  CAS  PubMed  Google Scholar 

  56. Li, Z. et al. Caspase-3 activation via mitochondria is required for long-term depression and AMPA receptor internalization. Cell 141, 859–871 (2010).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  57. Jiao, S. & Li, Z. Nonapoptotic function of BAD and BAX in long-term depression of synaptic transmission. Neuron 70, 758–772 (2011).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  58. Erturk, A., Wang, Y. & Sheng, M. Local pruning of dendrites and spines by caspase-3-dependent and proteasome-limited mechanisms. J. Neurosci. 34, 1672–1688 (2014).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  59. Gyorffy, B. A. et al. Local apoptotic-like mechanisms underlie complement-mediated synaptic pruning. Proc. Natl Acad. Sci. USA 115, 6303–6308 (2018).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  60. Xu, Z. X. et al. Caspase-2 promotes AMPA receptor internalization and cognitive flexibility via mTORC2–AKT–GSK3β signaling. Nat. Commun. 10, 3622 (2019).

    Article  PubMed  PubMed Central  Google Scholar 

  61. Nicolas-Avila, J. A. et al. A network of macrophages supports mitochondrial homeostasis in the heart. Cell 183, 94–109 (2020).

    Article  CAS  PubMed  Google Scholar 

  62. Bishop, D. L., Misgeld, T., Walsh, M. K., Gan, W. B. & Lichtman, J. W. Axon branch removal at developing synapses by axosome shedding. Neuron 44, 651–661 (2004).

    Article  CAS  PubMed  Google Scholar 

  63. Han, C. Z. & Ravichandran, K. S. Metabolic connections during apoptotic cell engulfment. Cell 147, 1442–1445 (2011).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  64. Ashley, J. et al. Retrovirus-like Gag protein Arc1 binds RNA and traffics across synaptic boutons. Cell 172, 262–274 (2018).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  65. Aida, T. et al. Cloning-free CRISPR/Cas system facilitates functional cassette knock-in in mice. Genome Biol. 16, 87 (2015).

    Article  PubMed  PubMed Central  Google Scholar 

  66. Sakatani, T. & Isa, T. PC-based high-speed video-oculography for measuring rapid eye movements in mice. Neurosci. Res. 49, 123–131 (2004).

    Article  PubMed  Google Scholar 

  67. Mathis, A. et al. DeepLabCut: markerless pose estimation of user-defined body parts with deep learning. Nat. Neurosci. 21, 1281–1289 (2018).

    Article  CAS  PubMed  Google Scholar 

Download references

Acknowledgements

This research was supported by JSPS KAKENHI grants 16K18388, 18K06457, 18J00603 (to Y.M.M.), 18H05110, 19H03338, 20H05046, 20H05896, 22H02713 (to K.M.),16H06280, 18H05120, 19K22469 and 19H03323 (to Y.F.) by The Naito Foundation Grant (to Y.M.M.), by the Satoshi Ohmura Foundation (to Y.M.M.), by Technical support platforms for promoting research of Advanced Bioimaging Support JP16H06280 (to Y.M.M. and N.Ohono.), by Opto-Science and Technology, the NOVARTIS Foundation, Takeda Science Foundation (to K.M), by The Joint Usage/Research Program of Medical Research Institute (to T.A., K.T. and K.M.), by AMED under the Brain/MINDS project: JP21dm0207111 (to H.H.) and by a Columbia University Target of Opportunity award. We thank T. Maegawa (Fukui University), A. Imai and N. Hattori (NIPS) for their helpful technical assistance in EM data acquisition, N. Hosoi, A. Konno (Gunma University), K. Chen (Columbia University), T. Osakada and L. Weinhard (New York University) for their helpful technical discussion, WB. Gan (Peking University Shenzhen Graduate School) for critical reading of the manuscript and all the members of the Matsui lab.

Author information

Authors and Affiliations

Authors

Contributions

Y.M.M. and K.M. conceived and designed the study. Y.M.M. performed most of the experiments and analysis. Y.F. performed FIB-SEM data acquisition. Y.N. and N.E. contributed to the FIB-SEM analysis. M.M. and N. Ohno. performed SBF-SEM data acquisition and analysis. Y.M.M, Y.F. and N.Ohno interpreted EM data. H.H. provided AAV. T.A., K.T., K.B., S.M., H.I., N. Osumi. and K.F.T. generated transgenic animals. E.S. and S.K. contributed gene-knockout experiments, G.Y. assisted in the experiments and H.I. wrote the HOKR program. Y.M.M. and K.M. wrote the manuscript. All authors discussed the results and commented on the manuscript.

Corresponding authors

Correspondence to Yosuke M. Morizawa or Ko Matsui.

Ethics declarations

Competing interests

The authors declare no competing interests.

Peer review

Peer review information

Nature Neuroscience thanks Won-Suk Chung, Cagla Eroglu 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 No leak expression of pHRed in BGs.

a, Representative images of pHRed expression and PV-IHC in PV-tTA; tetO-pHRed mice (12 weeks old; 4 weeks after Dox off). Most PCs expressed pHRed. We have done 1 independent experiment (n = 3 mice). b, PV-tTA mouse was crossed with tetO-FRET-sensor (PV-Venus) mouse. This mouse was generated in the same way as the tetO-pHRed mouse. As with the pHRed in PV-pHRed mouse, Venus expression was observed in PCs and interneurons. Venus was not found in BGs, Iba1 positive microglia, and CD31 positive endothelial cells. This result suggests that the gene expression is induced selectively in neurons with PV-tTA. pHRed found in glia is due to internalization of pHRed into the glia. We have done 1 independent experiment (n = 2 mice). c, Red and green fluorescence merged images from WT, 4w Dox off, and no Dox of PV-pHRed mice (all with 3PGDH staining for BGs). Dashed square areas are enlarged in inset with BG outline shown with yellow line. Quantification of the red fluorescence signal is shown on the right. pHRed gene expression in PC and pHRed internalization into BG are increased after Dox off (n = 17, 25, 29, 37, 29 PCs from 4, 3, 4, 3, 3 mice; 35, 45, 52, 42, 30 BGs from 5, 3, 4, 5, 3 mice). One-way ANOVA followed by Dunnett’s multiple comparison test (F (4, 13) = 80.93, p < 0.0001; **** p < 0.0001 for PC; F(4, 15) = 10.51, p = 0.0003; * p = 0.02, *** p = 0.0004 for BG). d, Magnified images of the dashed white square in a (Arrowhead: BG soma; dashed white circles: pHRed signals in BG; Asterisk: PC soma). Right graph, line profiles of the normalized fluorescence intensities at the yellow line shown in the left panel image. pHRed signals are found, but no PV-IHC signals are found in BG soma. e, PC and BG soma images from a 12-week-old, 4 weeks off Dox PV-pHRed mouse. No Calbindin-IHC signals in BG. f, Images from a 6-week-old, no Dox PV-FRET mouse. Venus (GFP antibody)-IHC signals were seen only in PC soma and not in BG. g, No membrane GFP-IHC signal in BG in PV-tTA; tetO-LckE2GFP mice (12 weeks old, 4 weeks off Dox). h, Images from PV-tTA; tetO-pHRed; tetO-LckE2GFP trigenic mice (12 weeks old, 3 weeks off dox). Punctate pHRed signals were found in BG soma, but no GFP-IHC signals in BG. i, Representative images from WT, Mlc1-tTA; tetO-pHRed, PV-tTA; tetO-pHRed mice. Yellow dashed lines and white dashed lines demarcate PC and BG soma, respectively. pHRed signals were found in BGs of the Mlc1-pHRed mouse and in BGs of the PV-pHRed mouse, but pHRed signals were not found in PCs of the Mlc1-pHRed mouse. This shows that inter-cell material transfer is unidirectional between PC and BG (n = 19, 31, 23, 32 PCs, 28, 37, 23, 30 BGs from 4 mice). One-way ANOVA followed by Dunnett’s multiple comparison test (F (3, 12) = 26.46, p < 0.0001; **** p < 0.0001 for PC; F(3, 12) = 23.23, p < 0.0001; * p = 0.040, **** p < 0.0001 for BG; NS p > 0.05). Data are shown as mean ± s.e.m.

Source data

Extended Data Fig. 2 Engulfment of neuron-derived pHRed signal in microglia.

Representative images of microglial engulfment of pHRed in the cerebellar white matter of the PV-pHRed mouse. pHRed is expressed in Purkinje cell axons. Condensed pHRed signals are highlighted by dashed circles. These pHRed signals are contained in the microglial process and soma. We have done 1 independent experiments (n = 3 mice).

Extended Data Fig. 3 BG engulf Purkinje cell materials.

a, b, Purkinje cell specific promoter-tTA mediated pHRed expression. AAV2/9 L7-4-tTA was injected in the lobule IV/V of the tetO-pHRed mouse. pHRed expression was observed in the Purkinje cell layer (yellow arrowheads). pHRed expression was exclusive in PC (Asterisks in b; Calbindin: PC marker) with a punctate pHRed signal also found in BG (Arrowheads and dotted circle area: BG). We have done 2 independent experiments (n = 6 mice). c, d, AAV uninjected and injected lobules of the tetO-pHRed mice were compared. In the injected lobule, pHRed signals in BG were observed with higher intensity than the background (in WT and uninjected lobule; n = 23, 46, 31 BGs from 4 WT, 6 injected mice). One-way ANOVA followed by Tukey’s multiple comparison test (F (2, 13) = 20.59, p < 0.0001; *** p = 0.0005 for WT vs inj.; ***p = 0.0002 for uninj. vs inj.). Data are shown as mean ± s.e.m.

Source data

Extended Data Fig. 4 Reduction of mGluR1a signals after HOKR training.

a, Representative image of mGluR1a punctate signals in the molecular layer (ML) of the cerebellar cortex. b, High magnified images from Fig. 4f of the mGluR1a punctate signals localized on PC spines (Arrowheads: spine-like localization). VGLUT1 positive parallel fiber boutons appose these mGluR1a puncta. Arrowheads: spine-like localization. We have done more than 4 independent experiments. c, Cumulative frequencies of distal ML of mGluR1a IHC signal intensity. Significant reduction of the mGluR1a intensity was observed in HOKR trained mice (n = 6 mice). Kolmogorov-Smirnov test (**** P < 0.0001). d, The reduction of the mGluR1a immunostaining signal intensity after HOKR training was not dependent on the depth from the slice surface (n = 3 mice). Two-sided unpaired t-test (* p = 0.05, 0.037, 0.024, 0.023, 0.016, 0.011, 0.0012, 0.014 from 0 to 2.31 µm with 0.33 µm step). e, The intensity of mGluR1a in Pfl was not different between sham and HOKR trained mice (n = 6 mice). Two-sided unpaired t-test (p = 0.88). Data are shown mean ± s.e.m.

Source data

Extended Data Fig. 5 Phagocytic features in microglia were unchanged after HOKR training.

a, Sparse distribution of microglia (Iba1) in the ML (Sham, above) and no accumulation of microglia into Fl after HOKR training (below). b, Representative images of Iba1, PV-pHRed, and CD68 in ML with (sham) or without training (HOKR, 24hrs). c, No significant differences in the pHRed expression in PC, pHRed internalization into microglia (MG), CD68, and Iba1 expression level in ML without (sham) or with HOKR training (n = 67, 85 PCs, 14, 18 Microglia from 3 mice). Two-sided unpaired t-test (p = 0.65 for PC pHRed; p = 0.31 for MG pHRed; p = 0.97 for CD68; p = 0.53 for Iba1). Data are shown as mean ± s.e.m.

Source data

Extended Data Fig. 6 Definition of glial engulfment structure.

a, Representative image of glial engulfment structures of an axon and a spine head (same example as shown in Fig. 5d). Gray values from different ROIs demarcated by the yellow dashed circles are shown on the right. Using serial image reconstruction (not shown), each segmentation can be tracked to its origins. Two engulfed structures shown were tracked back to their axon and spine head origin, respectively. These engulfed invaginations did not contain organelles and were less electric dense compared to their origins. Although these engulfed structures are still in continuum with their origin, such character of lower electron density appearance suggests that these invaginations are on their way to become independent from their origins and are becoming less viable components. Gray values of randomly picked engulfed invagination with neither organelle nor lamellar bodies were quantified (middle panel, n = 15 invaginations from 2 HOKR mice). Two-sided paired t-test (**** p < 0.0001). The line profile of the yellow line in the left panel is shown on the right panel. The distance between the membrane surrounding the engulfed structure and the membrane of the engulfed structure was wide. This indicates that the normal tight-membrane interaction between two different cells is lost in the engulfment structure. b, Spinules appear to be different from engulfed structures. Spinules are short vesicular or long vermiform evaginations that emerge from neuronal compartments, which are enwrapped by other cells. Thus, basically, spinules have similar shapes with the engulfed structures. However, in spinules, the electron density of the origin and the leading protrusion of the spinule is similar. The membrane interaction between cells appears to be intact with a constant tight distance between the spinule and the surrounding membrane. We defined the structure shown on the upper pannel as spinule and the lower panel as engulfment, respectively. In this case, aberrant organelles were found in the structure inside of the invagination, and tight membrane interaction with its surrounding was lost. In the presented example, both the spinule and the engulfed structure originate from the same spine head. Right, original serial sections in 10 nm z-step. Data are shown as mean. We have done 1 independent experiment.

Source data

Extended Data Fig. 7 Multiple examples of neuronal nibbling by BG.

Multiple examples of the nibbling by BG of the pre-synaptic (a-e; blue), the post-synaptic (f-i; orange), and the inhibitory interneurons (j). a-c, Typical examples of empty (a), sparse lamellar (b), and multilamellar (c) phagosomes (arrowheads) and their origins (asterisks). d, Three other representative examples of unknown aberrant organelles (red dashed circles) were frequently observed at the origin of the engulfed protrusions. e, One example of an unknown organelle connected to the expanded ER at the origin of the phagosome. f-i, Typical examples of empty (f), sparse lamellar (g), and multilamellar (h) phagosomes (arrowheads) and their origins (asterisks). i, Examples of aberrant lamellar structures in phagosomes connected to ERs in spines. j, Two other cases of the multilamellar body from herniated mitochondria (purple). All scale bars were 0.5 µm and serial images were taken at 50 nm z-steps. We have done 1 independent experiment.

Extended Data Fig. 8 mGluR1a IHC signal reduction after HOKR training was inhibited by Annexin V.

a, Representative images of Calbindin and mGluR1a IHC signals in sham, in HOKR with PBS, and in HOKR with Annexin V. mGluR1a punctate likely reflects the postsynaptic spine location. With HOKR training, mGluR1a signal decreased. Such decrease was not observed after HOKR training when Annexin V was treated to inhibit phagocytosis. We have done 3 independent experiments. b, Reduction of mGluR1a IHC signals in the flocculus (Fl) after HOKR training was inhibited with Annexin V treatment. Reduction of mGluR1a signal was not observed with HOKR training in the paraflocculus (PFl) either with PBS or with Annexin V treatment. Each dot shows data from a single animal (n = 4–10). One-way ANOVA followed by Bonferroni multiple comparison test (F (3, 20) = 2.49, p = 0.09; *p = 0.038 for PBS sham vs HOKR; p = 0.92 for Annexin V sham vs HOKR in Fl; F (3, 20) = 0.94, p = 0.44; p = 0.48 for PBS sham vs HOKR; p = 0.89 for Annexin V sham vs HOKR in PFl). Data are shown as mean ± s.e.m.

Source data

Extended Data Fig. 9 Enhancement of PS exposure in the Fl after HOKR learning.

a, PBS or Annexin V was intracisternally injected (iCM) and HOKR training was performed in a subset of animals. Annexin V was detected by immunohistochemical (IHC) staining of brain samples taken at 24 hours after the initial HOKR training on Day 1. Annexin V antibody is expected to react both to the Annexin V instrinsic to mouse and to the externally injected human recombinant Annexin V. Shown in this panel is representative images of IHC of Annexin V and Calbindin, which is expressed in PC dendrites and spines. This sample is from a HOKR trained mouse injected with Annexin V. Shown is the molecular layer of the flocculus (Fl). Numerous Annexin V signals could be localized at the proximity of dendrite-spine like structures. We have done 3 independent experiments. b, Annexin V IHC labeling was the most evident in the group of mice that was injected with Annexin V and had HOKR training. c, Quantification of Annexin V IHC signal in Fl (top) and PFl (bottom). PS exposure specifically occurred in the Fl after HOKR training and not in the PFl. Each dot shows single animal (n = 4–10 mice). One-way ANOVA followed by Bonferroni multiple comparison test (F (3, 18) = 12.98, p < 0.0001; p > 0.99 for PBS sham vs HOKR; ***p = 0.001 for Annexin V sham vs HOKR in Fl; F(3, 18) = 2.43, p = 0.09; p = 0.75 for PBS sham vs HOKR; p = 0.84 for Annexin V sham vs HOKR in PFl). Data are shown as mean ± s.e.m. Data are shown as mean ± s.e.m.

Source data

Extended Data Fig. 10 Bergmann glial ABCA1 mediate phagocytosis and HOKR learning.

a-d, Data from Tabula Muris. a, t-Distributed Stochastic Neighbor Embedding (t-SNE) plot of individual non-myeloid cell in the brain. Colors correspond to each cell type. b, c, Abca1 was enriched mainly in oligodendrocyte precursor cell, BG, and astrocytes. d, BG expressed multiple type of PS-dependent phagocytic pathway molecules including Abca1, Megf10, Lrp1, Mertk, Bai1, Itgav, Itgb5, as well as astrocytes markers. These features were similar to the general astrocyte population. e, Immunohistochemical detection of ABCA1 in the cerebellum. ABCA1 staining showed a similar morphological pattern of BG soma and processes (arrowheads). Scale bar: 100 µm. We have done 1 independent experiment. f, Schema for conditional targeting Abca1flox/flox with human GFAP-Cre. g, Protein quantification of ABCA1 by western blotting from the cerebellum. Conditional knockout (cKO) of Abca1 gene decreased ABCA1 protein in the cerebellar tissue (n = 2 mice). We have done 2 independent experiments. h, Schema of the experimental design (top left). AAV2/9-L7-4mCMV-tTA injected into tetO-pHRed; Abca1flox/flox with or without hGFAP-Cre mice to create PCs-specific pHRed expression in ABCA1 cKO mice. Frequency of finding of pHRed puncta within BG soma was apparently less in ABCA1 cKO mice (bottom left). pHRed intensity in BG was significantly less in ABCA1 cKO (middle graph) and no such difference between the control was observed for pHRed intensity in PC (right graph), (n = 3 mice). Two-sided unpaired t-test (p = 0.0051 for BG; p = 0.75 for PC). i, Schema for conditional targeting Abca1flox/flox with GLAST-CreERT2 mice (above) and the experimental schedule (below). Tamoxifen was injected at 8 weeks for consecutive 5 days. HOKR experiments were done after three weeks of wait time to achieve efficient KO of ABCA1. j, HOKR learning curves of three types of control groups and the cKO group are shown (For control: n = 9 (Abca1fl/fl: n = 5; Abca1fl/fl + Tam: n = 2; Abca1 fl/fl, Glast-CreERT2 n = 2) For ABCA1 cKO: n = 3 mice (Abca1fl/fl, Glast-CreERT2 + Tam). k, Significant improvement in the HOKR gain was observed from the 4th training session on Day 1 to the test session on Day 2 for the control group. However, such improvement was absent in the ABCA1 cKO group (control: n = 9; ABCA1 cKO n = 3 mice). Two-sided paired t-test (4th session vs 2nd day: p = 0.0081 for control; p = 0.072 for ABCA1 cKO). Unpaired t-test (p = 0.011 control vs ABCA1 cKO). All data are shown as mean ± s.e.m.

Source data

Supplementary information

Supplementary Information

Supplementary Figs. 1–4.

Reporting Summary

Supplementary Video 1

Eye movement before and after HOKR spaced learning. Representative examples of horizontal eye movements during the first learning session on the first day and test trial on the second day. The videos were captured by an infrared-sensitive CCD camera and the example eye movements were extracted by DeepLabCut (color dots).

Supplementary Video 2

FIB-SEM image sequences of engulfment by BG. FIB-SEM image sequences of engulfment of pre- and postsynapses by BG are shown. Cerebellum samples were taken from the HOKR-trained mouse. Original images and segmented image sequences are shown along with three orthogonal views of the same sample in Fig. 5a–c.

Supplementary Video 3

FIB-SEM 3D reconstruction of the engulfment. Using data from the same sample as in Fig. 5a–c and Supplementary Video 2, an FIB-SEM 3D reconstruction of the engulfment of pre- and postsynapses by BG in a HOKR-trained mouse cerebellum was created. In the latter part of the video of the 3D reconstruction, the engulfed structures are highlighted as nontransparent (solid) protrusions surrounded by the BG processes (magenta).

Supplementary Video 4

FIB-SEM 3D reconstruction of dendritic spines and engulfed protrusions. Representative 3D-reconstructed single–PC dendrite with spines (yellow) in a HOKR-trained mouse. Engulfed protrusions are highlighted in red.

Supplementary Video 5

FIB-SEM image sequences of representative engulfment from various compartment origins. Original FIB-SEM image sequences of the structures shown in Fig. 5d. Representative examples of engulfment of the axon, the bouton, the spine head, the spine neck and the dendrite are shown. On the left are the original FIB-SEM images and on the right are the same images overlaid with segmentation shown in semi transparent colors. Engulfed protrusions are highlighted with red arrows. All scale bars are 500 nm.

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 Fig. 7

Statistical source data.

Source Data Fig. 8

Statistical source data.

Source Data Extended Data Fig. 1

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.

Source Data Extended Data Fig. 8

Statistical source data.

Source Data Extended Data Fig. 9

Statistical source data.

Source Data Extended Data Fig. 10

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.

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Morizawa, Y.M., Matsumoto, M., Nakashima, Y. et al. Synaptic pruning through glial synapse engulfment upon motor learning. Nat Neurosci 25, 1458–1469 (2022). https://doi.org/10.1038/s41593-022-01184-5

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1038/s41593-022-01184-5

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