APCCdh1 mediates EphA4-dependent downregulation of AMPA receptors in homeostatic plasticity

Journal name:
Nature Neuroscience
Volume:
14,
Pages:
181–189
Year published:
DOI:
doi:10.1038/nn.2715
Received
Accepted
Published online

Abstract

Homeostatic plasticity is crucial for maintaining neuronal output by counteracting unrestrained changes in synaptic strength. Chronic elevation of synaptic activity by bicuculline reduces the amplitude of miniature excitatory postsynaptic currents (mEPSCs), but the underlying mechanisms of this effect remain unclear. We found that activation of EphA4 resulted in a decrease in synaptic and surface GluR1 and attenuated mEPSC amplitude through a degradation pathway that requires the ubiquitin proteasome system (UPS). Elevated synaptic activity resulted in increased tyrosine phosphorylation of EphA4, which associated with the ubiquitin ligase anaphase-promoting complex (APC) and its activator Cdh1 in neurons in a ligand-dependent manner. APCCdh1 interacted with and targeted GluR1 for proteasomal degradation in vitro, whereas depletion of Cdh1 in neurons abolished the EphA4-dependent downregulation of GluR1. Knockdown of EphA4 or Cdh1 prevented the reduction in mEPSC amplitude in neurons that was a result of chronic elevated activity. Our results define a mechanism by which EphA4 regulates homeostatic plasticity through an APCCdh1-dependent degradation pathway.

At a glance

Figures

  1. Activation of EphA4 reduces synaptic strength.
    Figure 1: Activation of EphA4 reduces synaptic strength.

    (ac) Cortical neurons (16 DIV) were treated with Fc, ephrin-A1 or ephrin-B1 (5 μg ml−1) for 4 h, and the mEPSC was recorded. (a) Representative mEPSC traces. (b) Compared to the Fc-treated control, the mean frequency of mEPSCs decreased significantly after treatment with ephrin-A1 but not ephrin-B1. Data are expressed as mean ± s.e.m.; *P < 0.05, ANOVA with Student-Newman-Keuls test (4 experiments; >10 neurons recorded from each experiment). (c) Cumulative amplitude distribution of mEPSCs in cortical neurons after treatment with Fc, ephrin-A1 or ephrin-B1 for 4 h. There is a leftward shift of cumulative amplitude distribution of mEPSC upon ephrin-A1 treatment. (d) Knockdown of EphA4 by shRNA abolished the reduction in mEPSC amplitude triggered by ephrin-A1. Cortical neurons (12–14 DIV) were transfected with GFP and pSUPER-EphA4 shRNA (shEphA4) or pSUPER vector (Control). Neurons at 19–21 DIV were treated with ephrin-A1 for 16 h. The mEPSC was recorded for neurons that expressed GFP. Data are presented as mean ± s.e.m. from three experiments (n > 10 neurons from each experiment; *P < 0.05, ANOVA followed by Student-Newman Keuls test).

  2. Ephrin-A1 downregulates the expression of GluR1 at synapses through activation of EphA4.
    Figure 2: Ephrin-A1 downregulates the expression of GluR1 at synapses through activation of EphA4.

    (a) Confocal images show punctate staining of GluR1 and PSD-95 in hippocampal neurons. GluR1 clusters co-localized with PSD-95 (arrows) decreased after ephrin-A1 treatment; PSD-95 clusters that lacked significant GluR1 immunoreactivity (arrowheads) were frequently found in ephrin-A1-treated neurons. Scale bars, 10 μm. (b) Quantification of PSD-95 and GluR1 clusters. ***P < 0.005; ANOVA with Mann-Whitney Rank Sum Test. (c) Quantification of synaptic localization of GluR1 and GluR2, as indicated by the percentage of PSD-95 that co-localized with GluR clusters. ***P < 0.005; Student's t-test. (d) Time-lapse imaging of hippocampal neurons expressing GluR1-GFP after treatment with ephrin-A1 for 4–6 h. Representative images show that some GluR1 clusters (arrowheads) were more stable, whereas others (arrows) disappeared during the imaging period. (e) Quantification of the loss of GluR1-GFP clusters (n = 11 for Fc, n = 12 for ephrin-A1; *P < 0.05, Student's t-test). (f) Ephrin-A1-mediated reduction in total GluR1 expression depends on EphA4. Cortical neurons prepared from EphA4−/− mice or EphA4+/+ littermates were treated with Fc or ephrin-A1 for 24 h. Different amounts of protein (5–15 μg) were loaded. (g) Quantitative analysis of GluR1 protein level (3 experiments; *P < 0.05, Student's t-test). (h) Ephrin-A1 reduced both surface and total GluR1. (i) Increased GluR1 in synaptosomes of EphA4−/− mouse brains. Crude synaptosomal fractions of whole brains from adult EphA4+/+ (+/+) or EphA4−/− (−/−) mice were prepared, and western blot analysis was performed.

  3. Chronic elevation of synaptic activity reduces GluR1 expression in an EphA4-dependent manner.
    Figure 3: Chronic elevation of synaptic activity reduces GluR1 expression in an EphA4-dependent manner.

    (a) Bicuculline-induced tyrosine phosphorylation of EphA4 (p-EphA4). (b) Quantification analysis of p-EphA4 (3 experiments; *P < 0.05, **P < 0.01, ***P < 0.005; ANOVA with Student-Newman-Keuls test). (c) Cortical neurons were pretreated with EphA4-Fc for 0.5 h, and then treated with bicuculline (Bic) for 1 h (3 experiments). (d) Fold change in p-EphA4 (**P < 0.05; ANOVA with Student-Newman-Keuls test). (e) EphA4 was required for the reduction in total GluR1 in response to chronic treatment with bicuculline. Cortical neurons from EphA4+/+ (n = 5) and EphA4−/− mice (n = 3; from 3 experiments) were treated with bicuculline for 24 h. (f) Quantification analysis for total GluR1 (P < 0.05; Student's t-test). (gi) Bicuculline treatment reduced the number of GluR1 clusters. (g,h) Hippocampal neurons were treated with bicuculline for 24 h, after which neurons were subjected to immunocytochemical analysis using GluR1 and PSD-95 antibodies. (g) Representative confocal images showed punctate staining of GluR1 and PSD-95. (h) The number of PSD-95 and GluR1 clusters was reduced after bicuculline treatment (3 experiments; *P < 0.05, ***P < 0.001; ANOVA with Mann-Whitney Rank Sum test). (i) EphA4 is required for the reduction of mEPSC amplitude induced by chronic bicuculline treatment. The mEPSC amplitude in shEphA4-transfected cortical neurons after bicuculline treatment (3 experiments; *P < 0.05; ANOVA, with Student-Newman-Keuls test). (j) Co-treatment with bicuculline and ephrin-A1 does not further reduce the mEPSC amplitude. Cortical neurons were treated with ephrin-A1 (A1), bicuculline or bicuculline with ephrin-A1 (4 experiments; ***P < 0.005; ANOVA with Student-Newman-Keuls test).

  4. Ephrin-A-EphA4 signaling reduces GluR1 expression by a proteasome-dependent pathway.
    Figure 4: Ephrin-A-EphA4 signaling reduces GluR1 expression by a proteasome-dependent pathway.

    (a,b) EphA4-dependent degradation of GluR1 is mediated by a proteasome-dependent pathway. (a) HEK293T cells were transfected with GluR1 and EphA4, and were then treated with MG132 (10 μM), lactacystin (Lac, 10 μM), chloroquine (CHQ, 50 μM) or ammonium chloride (NH4Cl, 20 mM) for 5 h. (b) Inhibition of proteasome-mediated degradation abolished the downregulation of GluR1 by ephrin-A1. Cortical neurons were treated with MG132 for 0.5 h before stimulation by ephrin-A1 for 7–16 h in the presence of the inhibitor. (c) Quantitative analysis of GluR1 (n = 4; *P < 0.05, ANOVA with Student-Newman Keuls test). (d) Ephrin-A1 reduced mEPSC amplitude through proteasome-mediated degradation. Cortical neurons were treated with MG132 for 0.5 h and then with ephrin-A1 for 16 h. The mEPSC amplitude was measured (≥3 experiments; *P < 0.05, ANOVA with Student-Newman Keuls test). (e) Ephrin-A1 induced polyubiquitination of GluR1 in neurons. Cultured cortical neurons were treated with MG132 for 0.5 h before treated with ephrin-A1 for 6 h. Whole-cell lysate was immunoprecipitated with GluR1 antibodies, followed by immunoblotting with anti-ubiquitin (Fk2) antibody which recognizes both mono- and polyubiquitinated proteins (n = 4). Similar results were observed when anti-polyubiquitin (Fk1) antibodies were used for immunoblotting (data not shown). More polyubiquitinated protein was immunoprecipitated by GluR1 antibodies from neurons treated with ephrin-A1 than from control neurons treated with Fc. (f) Ubiquitinated GluR1 was detected in synaptosomes. GluR1 protein was immunoprecipitated from the synaptosomal fractions, followed by western blot analysis using the Fk2 antibody (n = 3).

  5. E3 ubiquitin ligase complex APCCdh1 interacts with EphA4.
    Figure 5: E3 ubiquitin ligase complex APCCdh1 interacts with EphA4.

    (a) EphA4 interacted with APC2. Expression constructs encoding EphA4 and full-length APC2 were overexpressed in HEK293T cells separately. EphA4 was immunoprecipitated using EphA4 antibody and then pulled down by proteinG-Sepharose beads (IgG served as the control). The beads were then incubated with APC2-expressing cell lysate (400 μg or 800 μg as indicated). APC2 protein pull down by EphA4 was examined by western blot analysis. (b) Ephrin-A1 increased the interaction between APC2 and EphA4 in neurons. Cortical neurons (14 DIV) were treated with Fc or ephrin-A1 for indicated durations. Lysate was immunoprecipitated with APC2 antibody and immunoblotted with antibodies to EphA4. Similar amounts of protein were subjected to immunoprecipitation for different treatments, as indicated by immunoblotting the lysate (INP) with antibodies to EphA4 and APC2. (c) APC2 associated with EphA4 and GluR1 in rat brain in vivo. Rat brain homogenate (postnatal day (P)7, P30 and adult (Ad)) was immunoprecipitated with antibodies to APC2 and immunoblotted with antibodies to EphA4, GluR1 or APC2. (d) Adult rat brain fractions separated by differential centrifugation and extraction were subjected to western blot analysis for APC2, Cdh1, GluR1 and EphA4. Synaptophysin (SYN) served as the negative control for the different PSD fractions. P1: total brain lysates; S3: cytosolic fraction; SPM: synaptic plasma membrane. SPM was further extracted by Triton X-100 once (PSD 1T), twice (2T), or with Triton X-100 followed by Sarkosyl (1T+S).

  6. Polyubiquitination and degradation of GluR1 requires APCCdh1.
    Figure 6: Polyubiquitination and degradation of GluR1 requires APCCdh1.

    (a) Polyubiquitination of GluR1 required Cdh1. HEK293T cells were transfected with plasmids as indicated (n = 3). (b) Cdh1 interacted with the N-terminal region of GluR1. HEK293T cells were transfected with Cdh1 or GluR1 fragment (N-GluR1, amino acids 1–404 or C-GluR1, amino acids 405–907). Cdh1 protein was pulled down by Cdh1 antibody, and was then incubated with cell lysate expressing GluR1 fragments (200 or 500 μg). Input, 10 μg. (c) Degradation motifs on GluR1 are important for GluR1-Cdh1 interaction. HEK293T cells were transfected with Cdh1 and GluR1 or its 3M mutant. (d) Cdh1 reduced surface and total GluR1 in HEK293T cells. (e) Blockade of proteasome-dependent degradation partially inhibits the Cdh1-dependent degradation of GluR1. HEK293T cells overexpressing GluR1 or GluR1 with Cdh1 were pulse-labeled with [35S]-methionine and chased for 2 h. (f) The integrity of D-box motifs on GluR1 was required for Cdh1-dependent degradation of GluR1 in HEK293T cells; GluR1 (WT) or its single point mutants (D-43M, D-126M or A-159M). (g) Expression of ubiquitin mutant (K48R) inhibited the Cdh1-dependent degradation of GluR1. (h) Simultaneous mutation of lysine residues 831, 837, 840 and 886 to arginine (4KR) could not stabilize the expression of GluR1 when co-transfected with Cdh1. (i) Knockdown of Cdh1 inhibited the ephrin-A1-dependent reduction of GluR1. Cortical neurons were transfected with Cdh1 siRNA and then treated with ephrin-A1. The reduced expression of Cdh1 after EphA4 activation may be due to auto-degradation as Cdh1 is a target of APCCdh148.

  7. Reduction of synaptic strength by ephrin-A1 or chronic treatment of bicuculline depends on APCCdh1-dependent proteasome degradation pathway.
    Figure 7: Reduction of synaptic strength by ephrin-A1 or chronic treatment of bicuculline depends on APCCdh1-dependent proteasome degradation pathway.

    (a,b) Expression of Cdh1 ΔWD40 mutant abolished the reduction in synaptic strength after ephrin-A1 treatment. Cortical neurons (14–16 DIV) were transfected with Cdh1 ΔWD40 and GFP, and then treated at 20–22 DIV with ephrin-A1 for 16–24 h. (a) Representative traces. (b) Quantification of mEPSC amplitude after ephrin-A1 treatment (from three independent experiments; *P < 0.05, ANOVA with Student-Newman Keuls test). (c,d) Expression of shRNA-resistant Cdh1 in Cdh1-depleted neurons restored the ephrin-A1-dependent reduction in mEPSC amplitude. Cortical neurons were transfected with shCdh1 together with the Cdh1-WT or shRNA-resistant Cdh1 expression constructs, then treated with ephrin-A1. (c) Representative traces. (d) Quantification of mEPSC amplitude (from three experiments; *P < 0.05, ANOVA with Student-Newman Keuls test). (eh) Inhibition of Cdh1 or knockdown of Cdh1 in neurons abolished the reduction in mEPSC amplitude after treatment with bicuculline for 16–24 h. Cortical neurons were transfected with Cdh1 ΔWD40 and GFP (e,f) or pSUPER-Cdh1 shRNA (shCdh1) or pSUPER vector (Control) together with GFP (g,h), and then treated with bicuculline (Bic) for 16–24 h. (e,g) Representative traces. (f,h) Quantification of mEPSC amplitude upon bicuculline treatment. Data are expressed as mean ± s.e.m. (>10 neurons recorded from each experiment, ≥3 experiments; *P < 0.05, ANOVA with Dunn's test).

References

  1. Turrigiano, G.G. The self-tuning neuron: synaptic scaling of excitatory synapses. Cell 135, 422435 (2008).
  2. Turrigiano, G.G., Leslie, K.R., Desai, N.S., Rutherford, L.C. & Nelson, S.B. Activity-dependent scaling of quantal amplitude in neocortical neurons. Nature 391, 892896 (1998).
  3. Shepherd, J.D. et al. Arc/Arg3.1 mediates homeostatic synaptic scaling of AMPA receptors. Neuron 52, 475484 (2006).
  4. Stellwagen, D. & Malenka, R.C. Synaptic scaling mediated by glial TNF-α. Nature 440, 10541059 (2006).
  5. Bingol, B. & Schuman, E.M. Synaptic protein degradation by the ubiquitin proteasome system. Curr. Opin. Neurobiol. 15, 536541 (2005).
  6. Yi, J.J. & Ehlers, M.D. Ubiquitin and protein turnover in synapse function. Neuron 47, 629632 (2005).
  7. Haglund, K., Di Fiore, P.P. & Dikic, I. Distinct monoubiquitin signals in receptor endocytosis. Trends Biochem. Sci. 28, 598603 (2003).
  8. Ciechanover, A. Intracellular protein degradation: from a vague idea thru the lysosome and the ubiquitin-proteasome system and onto human diseases and drug targeting. Cell Death Differ. 12, 11781190 (2005).
  9. Zhao, Y., Hegde, A.N. & Martin, K.C. The ubiquitin proteasome system functions as an inhibitory constraint on synaptic strengthening. Curr. Biol. 13, 887898 (2003).
  10. Patrick, G.N., Bingol, B., Weld, H.A. & Schuman, E.M. Ubiquitin-mediated proteasome activity is required for agonist-induced endocytosis of GluRs. Curr. Biol. 13, 20732081 (2003).
  11. Jordan, B.A. et al. Identification and verification of novel rodent postsynaptic density proteins. Mol. Cell. Proteomics 3, 857871 (2004).
  12. Li, K.W. et al. Proteomics analysis of rat brain postsynaptic density. Implications of the diverse protein functional groups for the integration of synaptic physiology. J. Biol. Chem. 279, 9871002 (2004).
  13. Ehlers, M.D. Activity level controls postsynaptic composition and signaling via the ubiquitin-proteasome system. Nat. Neurosci. 6, 231242 (2003).
  14. Burbea, M., Dreier, L., Dittman, J.S., Grunwald, M.E. & Kaplan, J.M. Ubiquitin and AP180 regulate the abundance of GLR-1 glutamate receptors at postsynaptic elements in C. elegans. Neuron 35, 107120 (2002).
  15. Juo, P. & Kaplan, J.M. The anaphase-promoting complex regulates the abundance of GLR-1 glutamate receptors in the ventral nerve cord of C. elegans. Curr. Biol. 14, 20572062 (2004).
  16. van Roessel, P., Elliott, D.A., Robinson, I.M., Prokop, A. & Brand, A.H. Independent regulation of synaptic size and activity by the anaphase-promoting complex. Cell 119, 707718 (2004).
  17. Colledge, M. et al. Ubiquitination regulates PSD-95 degradation and AMPA receptor surface expression. Neuron 40, 595607 (2003).
  18. Zhang, D. et al. Na,K-ATPase activity regulates AMPA receptor turnover through proteasome-mediated proteolysis. J. Neurosci. 29, 44984511 (2009).
  19. Pasquale, E.B. Eph-Ephrin bidirectional signaling in physiology and disease. Cell 133, 3852 (2008).
  20. Klein, R. Bidirectional modulation of synaptic functions by Eph/ephrin signaling. Nat. Neurosci. 12, 1520 (2009).
  21. Deininger, K. et al. The Rab5 guanylate exchange factor Rin1 regulates endocytosis of the EphA4 receptor in mature excitatory neurons. Proc. Natl. Acad. Sci. USA 105, 1253912544 (2008).
  22. Murai, K.K., Nguyen, L.N., Irie, F., Yamaguchi, Y. & Pasquale, E.B. Control of hippocampal dendritic spine morphology through ephrin-A3/EphA4 signaling. Nat. Neurosci. 6, 153160 (2003).
  23. Fu, W.Y. et al. Cdk5 regulates EphA4-mediated dendritic spine retraction through an ephexin1-dependent mechanism. Nat. Neurosci. 10, 6776 (2007).
  24. Penzes, P. et al. Rapid induction of dendritic spine morphogenesis by trans-synaptic ephrinB-EphB receptor activation of the Rho-GEF kalirin. Neuron 37, 263274 (2003).
  25. Henkemeyer, M., Itkis, O.S., Ngo, M., Hickmott, P.W. & Ethell, I.M. Multiple EphB receptor tyrosine kinases shape dendritic spines in the hippocampus. J. Cell Biol. 163, 13131326 (2003).
  26. Dalva, M.B. et al. EphB receptors interact with NMDA receptors and regulate excitatory synapse formation. Cell 103, 945956 (2000).
  27. Takasu, M.A., Dalva, M.B., Zigmond, R.E. & Greenberg, M.E. Modulation of NMDA receptor-dependent calcium influx and gene expression through EphB receptors. Science 295, 491495 (2002).
  28. Wierenga, C.J., Ibata, K. & Turrigiano, G.G. Postsynaptic expression of homeostatic plasticity at neocortical synapses. J. Neurosci. 25, 28952905 (2005).
  29. O'Brien, R.J. et al. Activity-dependent modulation of synaptic AMPA receptor accumulation. Neuron 21, 10671078 (1998).
  30. Malinow, R. AMPA receptor trafficking and long-term potentiation. Philos. Trans. R. Soc. Lond. B Biol. Sci. 358, 707714 (2003).
  31. Bredt, D.S. & Nicoll, R.A. AMPA receptor trafficking at excitatory synapses. Neuron 40, 361379 (2003).
  32. Cheng, K. et al. Pctaire1 interacts with p35 and is a novel substrate for Cdk5/p35. J. Biol. Chem. 277, 3198831993 (2002).
  33. Castro, A., Bernis, C., Vigneron, S., Labbe, J.C. & Lorca, T. The anaphase-promoting complex: a key factor in the regulation of cell cycle. Oncogene 24, 314325 (2005).
  34. Konishi, Y., Stegmuller, J., Matsuda, T., Bonni, S. & Bonni, A. Cdh1-APC controls axonal growth and patterning in the mammalian brain. Science 303, 10261030 (2004).
  35. Gieffers, C., Peters, B.H., Kramer, E.R., Dotti, C.G. & Peters, J.M. Expression of the CDH1-associated form of the anaphase-promoting complex in postmitotic neurons. Proc. Natl. Acad. Sci. USA 96, 1131711322 (1999).
  36. Kraft, C., Vodermaier, H.C., Maurer-Stroh, S., Eisenhaber, F. & Peters, J.M. The WD40 propeller domain of Cdh1 functions as a destruction box receptor for APC/C substrates. Mol. Cell 18, 543553 (2005).
  37. Kato, A., Rouach, N., Nicoll, R.A. & Bredt, D.S. Activity-dependent NMDA receptor degradation mediated by retrotranslocation and ubiquitination. Proc. Natl. Acad. Sci. USA 102, 56005605 (2005).
  38. Lim, K.L. et al. Parkin mediates nonclassical, proteasomal-independent ubiquitination of synphilin-1: implications for Lewy body formation. J. Neurosci. 25, 20022009 (2005).
  39. Matyskiela, M.E., Rodrigo-Brenni, M.C. & Morgan, D.O. Mechanisms of ubiquitin transfer by the anaphase-promoting complex. J. Biol. 8, 92 (2009).
  40. Pozo, K. & Goda, Y. Unraveling mechanisms of homeostatic synaptic plasticity. Neuron 66, 337351 (2010).
  41. Lissin, D.V. et al. Activity differentially regulates the surface expression of synaptic AMPA and NMDA glutamate receptors. Proc. Natl. Acad. Sci. USA 95, 70977102 (1998).
  42. O'Brien, R.J., Lau, L.F. & Huganir, R.L. Molecular mechanisms of glutamate receptor clustering at excitatory synapses. Curr. Opin. Neurobiol. 8, 364369 (1998).
  43. Seeburg, D.P., Feliu-Mojer, M., Gaiottino, J., Pak, D.T. & Sheng, M. Critical role of CDK5 and Polo-like kinase 2 in homeostatic synaptic plasticity during elevated activity. Neuron 58, 571583 (2008).
  44. Ang, X.L., Seeburg, D.P., Sheng, M. & Harper, J.W. Regulation of postsynaptic RapGAP SPAR by Polo-like kinase 2 and the SCFbeta-TRCP ubiquitin ligase in hippocampal neurons. J. Biol. Chem. 283, 2942429432 (2008).
  45. Helton, T.D., Otsuka, T., Lee, M.C., Mu, Y. & Ehlers, M.D. Pruning and loss of excitatory synapses by the parkin ubiquitin ligase. Proc. Natl. Acad. Sci. USA 105, 1949219497 (2008).
  46. Murthy, V.N., Schikorski, T., Stevens, C.F. & Zhu, Y. Inactivity produces increases in neurotransmitter release and synapse size. Neuron 32, 673682 (2001).
  47. Burrone, J., O'Byrne, M. & Murthy, V.N. Multiple forms of synaptic plasticity triggered by selective suppression of activity in individual neurons. Nature 420, 414418 (2002).
  48. Listovsky, T. et al. Mammalian Cdh1/Fzr mediates its own degradation. EMBO J. 23, 16191626 (2004).
  49. Sala, C. et al. Regulation of dendritic spine morphology and synaptic function by Shank and Homer. Neuron 31, 115130 (2001).
  50. Sala, C. et al. Inhibition of dendritic spine morphogenesis and synaptic transmission by activity-inducible protein Homer1a. J. Neurosci. 23, 63276337 (2003).

Download references

Author information

Affiliations

  1. Department of Biochemistry, State Key Laboratory of Molecular Neuroscience and Molecular Neuroscience Center, The Hong Kong University of Science and Technology, Clear Water Bay, Hong Kong, China.

    • Amy K Y Fu,
    • Kwok-Wang Hung,
    • Wing-Yu Fu,
    • Chong Shen,
    • Yu Chen,
    • Jun Xia,
    • Kwok-On Lai &
    • Nancy Y Ip

Contributions

N.Y.I. supervised the project. A.K.Y.F., K.-W.H., W.-Y.F., K.-O.L. and N.Y.I. designed the experiments. K.-W.H., W.-Y.F., Y.C. and K.-O.L. conducted the majority of experiments. A.K.Y.F., K.-W.H., W.-Y.F., Y.C., K.-O.L. and N.Y.I. did the data analyses. J.X. designed and did the data analyses on the electrophysiology experiment and C.S. performed electrophysiology experiment. A.K.Y.F., K.-O.L. and N.Y.I. wrote the manuscript.

Competing financial interests

The authors declare no competing financial interests.

Corresponding author

Correspondence to:

Author details

Supplementary information

PDF files

  1. Supplementary Text and Figures (1M)

    Supplementary Figures 1–8

Additional data