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.

  • Letter
  • Published:

Rho GTPase complementation underlies BDNF-dependent homo- and heterosynaptic plasticity

Nature volume 538, pages 104–108 (2016)Cite this article

Subjects

Abstract

The Rho GTPase proteins Rac1, RhoA and Cdc42 have a central role in regulating the actin cytoskeleton in dendritic spines1, thereby exerting control over the structural and functional plasticity of spines2,3,4,5 and, ultimately, learning and memory6,7,8. Although previous work has shown that precise spatiotemporal coordination of these GTPases is crucial for some forms of cell morphogenesis9, the nature of such coordination during structural spine plasticity is unclear. Here we describe a three-molecule model of structural long-term potentiation (sLTP) of murine dendritic spines, implicating the localized, coincident activation of Rac1, RhoA and Cdc42 as a causal signal of sLTP. This model posits that complete tripartite signal overlap in spines confers sLTP, but that partial overlap primes spines for structural plasticity. By monitoring the spatiotemporal activation patterns of these GTPases during sLTP, we find that such spatiotemporal signal complementation simultaneously explains three integral features of plasticity: the facilitation of plasticity by brain-derived neurotrophic factor (BDNF), the postsynaptic source of which activates Cdc42 and Rac1, but not RhoA; heterosynaptic facilitation of sLTP, which is conveyed by diffusive Rac1 and RhoA activity; and input specificity, which is afforded by spine-restricted Cdc42 activity. Thus, we present a form of biochemical computation in dendrites involving the controlled complementation of three molecules that simultaneously ensures signal specificity and primes the system for plasticity.

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: The Rho GTPases Rac1 and Cdc42 convey postsynaptic BDNF–TrkB signalling across both homosynaptic and heterosynaptic domains.
Figure 2: BDNF–TrkB–Rac1 signalling is required for synaptic crosstalk.
Figure 3: Inhibition of signal spreading of Rac1 and RhoA prevents synaptic crosstalk.
Figure 4: Signal spreading provides additive activation of high-threshold signals during synaptic crosstalk.

Similar content being viewed by others

References

  1. Tashiro, A., Minden, A. & Yuste, R. Regulation of dendritic spine morphology by the rho family of small GTPases: antagonistic roles of Rac and Rho. Cereb. Cortex 10, 927–938 (2000)

    Article  CAS  Google Scholar 

  2. Dietz, D. M. et al. Rac1 is essential in cocaine-induced structural plasticity of nucleus accumbens neurons. Nat. Neurosci. 15, 891–896 (2012)

    Article  CAS  Google Scholar 

  3. Murakoshi, H., Wang, H. & Yasuda, R. Local, persistent activation of Rho GTPases during plasticity of single dendritic spines. Nature 472, 100–104 (2011)

    Article  CAS  ADS  Google Scholar 

  4. Okamoto, K., Nagai, T., Miyawaki, A. & Hayashi, Y. Rapid and persistent modulation of actin dynamics regulates postsynaptic reorganization underlying bidirectional plasticity. Nat. Neurosci. 7, 1104–1112 (2004)

    Article  CAS  Google Scholar 

  5. Nakayama, A. Y., Harms, M. B. & Luo, L. Small GTPases Rac and Rho in the maintenance of dendritic spines and branches in hippocampal pyramidal neurons. J. Neurosci. 20, 5329–5338 (2000)

    Article  CAS  Google Scholar 

  6. Haditsch, U. et al. A central role for the small GTPase Rac1 in hippocampal plasticity and spatial learning and memory. Mol. Cell. Neurosci. 41, 409–419 (2009)

    Article  CAS  Google Scholar 

  7. Lai, K.-O. O. et al. TrkB phosphorylation by Cdk5 is required for activity-dependent structural plasticity and spatial memory. Nat. Neurosci. 15, 1506–1515 (2012)

    Article  CAS  Google Scholar 

  8. Il Hwan, K., Hong, W., Scott, H. S. & Ryohei, Y. Loss of Cdc42 leads to defects in synaptic plasticity and remote memory recall. eLife 3, e02839 (2014)

    Article  Google Scholar 

  9. Machacek, M. et al. Coordination of Rho GTPase activities during cell protrusion. Nature 461, 99–103 (2009)

    Article  CAS  ADS  Google Scholar 

  10. McAllister, A. K. Biolistic transfection of neurons. Sci. STKE 2000, pl1 (2000)

    Article  CAS  Google Scholar 

  11. Stoppini, L., Buchs, P. A. & Muller, D. A simple method for organotypic cultures of nervous tissue. J. Neurosci. Methods 37, 173–182 (1991)

    Article  CAS  Google Scholar 

  12. Yasuda, R. et al. Supersensitive Ras activation in dendrites and spines revealed by two-photon fluorescence lifetime imaging. Nat. Neurosci. 9, 283–291 (2006)

    Article  CAS  Google Scholar 

  13. Matsuzaki, M., Honkura, N., Ellis-Davies, G. C. & Kasai, H. Structural basis of long-term potentiation in single dendritic spines. Nature 429, 761–766 (2004)

    Article  CAS  ADS  Google Scholar 

  14. Honkura, N., Matsuzaki, M., Noguchi, J., Ellis-Davies, G. C. & Kasai, H. The subspine organization of actin fibers regulates the structure and plasticity of dendritic spines. Neuron 57, 719–729 (2008)

    Article  CAS  Google Scholar 

  15. Harward, S. C. et al. Autocrine BDNF–TrkB signalling within a single dendritic spine. Nature http://dx.doi.org/10.1038/nature19766 (2016)

  16. Lu, W. et al. Subunit composition of synaptic AMPA receptors revealed by a single-cell genetic approach. Neuron 62, 254–268 (2009)

    Article  CAS  Google Scholar 

  17. Luikart, B. W. et al. TrkB has a cell-autonomous role in the establishment of hippocampal Schaffer collateral synapses. J. Neurosci. 25, 3774–3786 (2005)

    Article  CAS  Google Scholar 

  18. Miyamoto, Y., Yamauchi, J., Tanoue, A., Wu, C. & Mobley, W. C. TrkB binds and tyrosine-phosphorylates Tiam1, leading to activation of Rac1 and induction of changes in cellular morphology. Proc. Natl Acad. Sci. USA 103, 10444–10449 (2006)

    Article  CAS  ADS  Google Scholar 

  19. Harvey, C. D. & Svoboda, K. Locally dynamic synaptic learning rules in pyramidal neuron dendrites. Nature 450, 1195–1200 (2007)

    Article  CAS  ADS  Google Scholar 

  20. Harvey, C. D., Yasuda, R., Zhong, H. & Svoboda, K. The spread of Ras activity triggered by activation of a single dendritic spine. Science 321, 136–140 (2008)

    Article  CAS  ADS  Google Scholar 

  21. Chen, X. et al. A chemical-genetic approach to studying neurotrophin signaling. Neuron 46, 13–21 (2005)

    Article  Google Scholar 

  22. Huber, K. M., Sawtell, N. B. & Bear, M. F. Brain-derived neurotrophic factor alters the synaptic modification threshold in visual cortex. Neuropharmacology 37, 571–579 (1998)

    Article  CAS  Google Scholar 

  23. Gao, Y., Xing, J., Streuli, M., Leto, T. L. & Zheng, Y. Trp(56) of rac1 specifies interaction with a subset of guanine nucleotide exchange factors. J. Biol. Chem. 276, 47530–47541 (2001)

    Article  CAS  Google Scholar 

  24. Korobova, F. & Svitkina, T. Molecular architecture of synaptic actin cytoskeleton in hippocampal neurons reveals a mechanism of dendritic spine morphogenesis. Mol. Biol. Cell 21, 165–176 (2010)

    Article  CAS  Google Scholar 

  25. Zhong, H. et al. Subcellular dynamics of type II PKA in neurons. Neuron 62, 363–374 (2009)

    Article  CAS  Google Scholar 

  26. Seoh, M. L., Ng, C. H., Yong, J., Lim, L. & Leung, T. ArhGAP15, a novel human RacGAP protein with GTPase binding property. FEBS Lett. 539, 131–137 (2003)

    Article  CAS  Google Scholar 

  27. Tanaka, J. et al. Protein synthesis and neurotrophin-dependent structural plasticity of single dendritic spines. Science 319, 1683–1687 (2008)

    Article  CAS  ADS  Google Scholar 

  28. Korte, M. et al. Hippocampal long-term potentiation is impaired in mice lacking brain-derived neurotrophic factor. Proc. Natl Acad. Sci. USA 92, 8856–8860 (1995)

    Article  CAS  ADS  Google Scholar 

  29. Bosch, M. et al. Structural and molecular remodeling of dendritic spine substructures during long-term potentiation. Neuron 82, 444–459 (2014)

    Article  CAS  Google Scholar 

  30. Chrostek, A. et al. Rac1 is crucial for hair follicle integrity but is not essential for maintenance of the epidermis. Mol. Cell. Biol. 26, 6957–6970 (2006)

    Article  CAS  Google Scholar 

  31. Yasuda, R. Imaging spatiotemporal dynamics of neuronal signaling using fluorescence resonance energy transfer and fluorescence lifetime imaging microscopy. Curr. Opin. Neurobiol. 16, 551–561 (2006)

    Article  CAS  Google Scholar 

  32. Murakoshi, H., Lee, S.-J. & Yasuda, R. Highly sensitive and quantitative FRET-FLIM imaging in single dendritic spines using improved non-radiative YFP. Brain Cell Biol. 36, 31–42 (2008)

    Article  Google Scholar 

  33. Pologruto, T. A., Sabatini, B. L. & Svoboda, K. ScanImage: flexible software for operating laser scanning microscopes. Biomed. Eng. Online 2, 13 (2003)

    Article  Google Scholar 

  34. Shaner, N. C. et al. Improved monomeric red, orange and yellow fluorescent proteins derived from Discosoma sp. red fluorescent protein. Nat. Biotech. 22, 1567–1572 (2004)

    Article  CAS  Google Scholar 

  35. Pédelacq, J. D., Cabantous, S., Tran, T. & Terwilliger, T. C. Engineering and characterization of a superfolder green fluorescent protein. Nat. Biotech. 24, 79–88 (2005)

    Article  Google Scholar 

  36. He, X.P. et al.Conditional deletion of TrkB but not BDNF prevents epileptogenesis in the kindling model. Neuron 43, 31–42 (2004)

    Article  CAS  Google Scholar 

Download references

Acknowledgements

This work was supported by grants from NIMH (R01MH080047 (R.Y.), R01NS068410 (R.Y.)), NINDS (F31NS078847 (S.C.H.), R01NS05621 (J.O.M.), DP1NS096787 (R.Y.)), JSPS KAKENHI (H.M.), JST PRESTO (H.M.), the Wakeman Fellowship at Duke University (S.C.H.) and Max Planck Florida Institute (R.Y.). We thank A. E. West, S. Soderling, and F. Wang for guidance and valuable discussion, and Yasuda and McNamara laboratory members for discussion. We also thank R. Puranam for isolating MTBD. Finally, we thank D. Kloetzer for managing the laboratory.

Author information

Author notes

  1. Nathan G. Hedrick

    Present address: †Present address: Neurobiology Section, Center for Neural Circuits and Behavior, and Department of Neurosciences, University of California, San Diego, 9500 Gilman Drive, La Jolla, California 92093, USA.,

  2. Nathan G. Hedrick and Stephen C. Harward: These authors contributed equally to this work.

Authors and Affiliations

  1. Neurobiology Department, Duke University Medical Center, Research Drive, Durham, 27710, North Carolina, USA

    Nathan G. Hedrick, Stephen C. Harward, Charles E. Hall, James O. McNamara & Ryohei Yasuda

  2. National Institute for Physiological Science, Myodaiji, Okazaki, 444-8585, Aichi, Japan

    Hideji Murakoshi

  3. Max Planck Florida Institute for Neuroscience, 1 Max Planck Way, Jupiter, 33458, Florida, USA

    Ryohei Yasuda

Authors
  1. Nathan G. Hedrick
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Stephen C. Harward
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Charles E. Hall
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Hideji Murakoshi
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. James O. McNamara
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. Ryohei Yasuda
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

N.G.H., S.C.H., J.O.M. and R.Y. designed the experiments. N.G.H., S.C.H. and H.M. developed the sensor and inhibitors; N.G.H. and S.C.H. collected the data with assistance from C.E.H.; N.G.H., S.C.H., C.E.H. and R.Y. analysed the data; N.G.H., S.C.H., J.M.O. and R.Y. wrote the paper. All authors discussed the results and comments on this manuscript.

Corresponding author

Correspondence to Ryohei Yasuda.

Additional information

Reviewer Information

Nature thanks B. Bingol, H. Zhang and the other anonymous reviewer(s) for their contribution to the peer review of this work.

Extended data figures and tables

Extended Data Figure 1 Design and characterization of the Rac1 sensor.

a, Measurements of the affinity between sfGFP–Rac1 or sfGFP–Cdc42 and the p21-activated-kinase-derived acceptor construct, PAK2(65–117)R71C,S78A. The binding fraction was measured using 2pFLIM (see Methods) across several concentrations of the acceptor construct. The dissociation constant was obtained by fitting the data (red) with a Michaelis–Menten function (grey). b. Representative fluorescence lifetime images of Rac1 sensor variants in HEK293T cells. Cells were transfected with a 1:2 donor:acceptor ratio of wild-type, dominant-negative (Rac1T17N), substrate-binding dead (Rac1Y40C), or constitutively active (Rac1Q61L) variants of Rac1 with mCh–PBD2R71C,S78A–mCh. Some experiments included the addition of an additional construct (the Rac1 guanine nucleotide exchange factor (GEF) Tiam1, or GTPase-activating protein (GAP) ARHGAP15) in a 1:2:1 donor:acceptor:GEF/GAP ratio. Cells were imaged 12–36 h after transfection in a warmed solution containing 30 mM Na-HEPES, pH 7.3, 130 mM NaCl, 2.5 mM KCl, 1 mM CaCl2, 1 mM MgCl2, 2 mM NaHCO3, 1.25 mM NaH2PO4 and 25 mM glucose. Warmer colours indicate a lower fluorescence lifetime value/higher binding fraction of eGFP–Rac1 to the acceptor construct. Scale bars, 50 μm. c, Basal binding fraction of Rac1–PBD2R71C, S78A for the conditions listed in b. Error bars represent s.e.m. *P < 0.0001 compared to wild-type Rac1 control condition, ANOVA followed by post-hoc tests using the least significant difference. d, Time course of Rac1 activation in HEK293T cells upon application of 100 μg ml−1 epidermal growth factor (EGF). Control experiment corresponds to eGFP–Rac1 (donor) plus mCh–PBD2R71C,S78A–mCh (acceptor) expression alone (n = 39 cells/8 plates), Rac1T17N (n = 33 cells/4 plates) and Rac1Y40C (n = 40 cells/5 plates) corresponds to the expression of the donor variant with the acceptor in the same ratio as controls, and Rac1plus W56 corresponds to the expression of the wild-type donor with the acceptor and the Rac1 inhibitory peptide W56.

Extended Data Figure 2 Characterization of the Rac1 sensor during single spine structural plasticity in hippocampal slices.

a, Left, comparison of the signal spreading of Rac1, RhoA, Cdc42 and TrkB, normalized to spine activity from 1–2 min after stimulus onset. Error bars represent s.e.m. Data correspond to that shown in Fig. 1. Right, comparison of the spreading indices of the four sensors (spreading index = activity in dendrite (1–5 μm) from 1–2 min post-stimulation/maximum spine activation). Error was estimated by bootstrapping. *P < 0.05, independent-samples t-test. b, Same as a, but for 15–20 min after stimulation. c, Analysis of the diffusion time constants of constitutively active (Q61L) and wild-type Rac1 using photo-activatable GFP (paGFP). Left, representative 2p images of paGFP–Rac1Q61L in a single dendritic spine after photo-activation. While active Rac1 shows a slightly slower diffusion than wild-type Rac1 (middle), both Rac1Q61L and wild-type Rac1 diffuse away from the spine within approximately 10 s, similarly to Cdc42 and RhoA (ref. 3) (right two plots; wild type: n = 41 spines/4 cells; Rac1Q61L: 67 spines/7 cells). d, Pharmacological characterization of Rac1 signal during sLTP. Rac1 activation in control (n = 102 cells/121 spines) conditions during 2p-glutamate uncaging, in the presence of the NMDAR blocker APV (100 μM; n = 6 cells/13 spines), and in the presence of the cell-permeable CaMKII inhibitory peptide tatCN21 (CN21; 10 μM; n = 5 cells/11 spines). Error bars represent s.e.m. e, Time course of spine volume change for experiments (d). Error bars represent s.e.m. f, Summary of effect of AP5 and CN21 on the transient (1–2 min after stimulation; control = 0.049 ± 0.003; AP5 = 0.00 ± 0.01; CN21 = 0.01 ± 0.01) and sustained (>10 min after stimulation; control = 0.033 ± 0.002; AP5 = 0.006 ± 0.007; CN21 = 0.026 ± 0.05) phases of Rac1 activation during sLTP. Error bars represent s.e.m. *P < 0.05, independent-samples t-test. g. Effects of near-physiological temperature of Rac1 sensor activation. Perfusion was warmed with a heating block holding the ACSF container, and the temperature measured at the back of the perfusion chamber (room temperature (RT), n = 102 cells/121 spines; for 30–32 °C, n = 11 cells/13 spines). Error bars represent s.e.m. h, Variability of unstimulated spine volume changes after the induction of sLTP in a nearby spine in Rac1 sensor-overexpressing neurons. Data shown are time courses of unstimulated spines close to the site of sLTP induction. While there is no average change in nearby spine volume with this stimulus (Fig. 1c), there is occasional enlargement or shrinkage. Data correspond to 100 randomly selected spines from the total of 777 nearby spines measured for the average depicted in Fig. 1c. i, Transient (1–2 min) change in volume of nearby spines as a function of distance from the stimulated spine. Data correspond to all nearby spines measured for the experiments in which the spatial profile of Rac1 was measured (n = 56 cells/79 experiments/218 nearby spines; Fig. 1e). Inset equation corresponds to the linear model of best fit. j, Same as i, but for the sustained (10–20 min) change in volume of nearby spines.

Extended Data Figure 3 Estimation of endogenous Rac1 concentration and effects of Rac1 sensor overexpression.

a, Left, representative western blot used to analyse endogenous Rac1 expression (endo-Rac1) from the CA1 region of hippocampal slices cultures compared to known concentrations of purified polyhistidine-tagged Rac1 (His-Rac1). Right, quantification of protein expression level for His-Rac1 from western blot shown in a averaged over three experiments. The concentration of endogenous Rac1 was estimated by measuring the intersection of the intensity of Rac1 protein from CA1 on the established calibration curve. b–e, Effect of concentration of the Rac1 sensor on the observed change in binding fraction or volume of the spine with glutamate uncaging (b or b, respectively), the basal binding fraction of the sensor (d), or the change in binding fraction in the dendrite in response to uncaging (e). Rac1 sensor concentration was estimated by making a standard curve of the intensity values known concentrations of eGFP at a range of imaging powers. The estimated endo-Rac1 concentration is plotted with a dashed line for comparison.

Extended Data Figure 4 Effect of Rac1 inhibition on sLTP.

a. The effect of a high concentration of NSC-23766 (120 μM) on uncaging-evoked spine volume change. Experiments were performed in eGFP-expressing rat hippocampal CA1 pyramidal neurons. Error bars represent s.e.m. n = 5 cells/9 control spines; n = 5 cells/15 NSC-23766 spines. b, Effect of single-cell Rac1 knockout on spine sLTP. Rac1fl/fl slices were transfected with either eGFP alone (Cre (−); black curve, n = 5 cells/5 spines) or eGFP + tdTomato-Cre recombinase (Cre (+); red curve; n = 7 cells/7 spines). Error bars represent s.e.m. c, Summary of data in a and b. For NSC-23766 experiments, control ΔVtransient = 319 ± 38%; control ΔVsustained = 67 ± 13%; NSC-23766 ΔVtransient = 104 ± 48%; NSC-23766 ΔVsustained = 20 ± 13%. For Rac1 knockout experiments, Cre(−) ΔVtransient = 420 ± 130%; Cre(−) ΔVsustained = 90 ± 22%; Cre(+) ΔVtransient = 23 ± 9%; Cre(+) ΔVsustained = 8 ± 8%. Error bars represent s.e.m. *P < 0.05, independent-samples t-test. All data are mean ± s.e.m.

Extended Data Figure 5 Characterization of Rho-GTPase dependence on BDNF–TrkB signalling.

a, b, Dependence of sLTP (Δvolume) in Rho GTPase sensor-expressing neurons on postsynaptic BDNF. Data correspond to the volume data from experiments shown in Fig. 1f. Black denotes Cre−, red denotes Cre+. b, Summary of data from a. Bars represent the average spine volume change from 10 min after stimulation to the end of the experiment. c, d, Dependence of Rac1 on extracellular BDNF. c, Left, Rac1 (ctrl: n = 7 cells/11 spines; TrkB-Ig: n = 8 cells/14 spines), activity in the presence of 2 mg ml−1 TrkB-Ig. Right, spine volume change in Rac1 sensor-overexpressing cells in the two conditions. Grey bars indicate duration of uncaging bout. d, Summary of data in c. e, f, Dependence of Rac1 on postsynaptic TrkB. e, Left, Rac1 (Cre(−): n = 3 cells/8 spines; Cre(+): n = 3 cells/8 spines), activity in the presence or absence of Cre recombinase in Trkbfl/fl mouse slices. Right, spine volume change in Rac1 sensor-overexpressing cells in the two conditions. Grey bars indicate duration of uncaging bout. f, Summary of data in e. g, h, Dependence of Cdc42 on extracellular BDNF. Left, Cdc42 (n = 7 cells/12 spines, 5 cells, 12 TrkB-Ig spines), activity in the presence of 2 mg ml−1 TrkB-Ig. Right, spine volume change in Rac1 sensor-overexpressing cells in the two conditions. Grey bars indicate duration of uncaging bout. h, Summary of data in g. i, j, Dependence of Cdc42 on postsynaptic TrkB. i, Left, Cdc42 (Cre(−): n = 5 cells/12 spines; Cre(+): n = 6 cells/16 spines), activity in the presence or absence of Cre recombinase in Trkbfl/fl mouse slices. Right, spine volume change in Cdc42 sensor-overexpressing cells in the two conditions. Grey bars indicate duration of uncaging bout. j, Summary of data in i. All data are mean ± s.e.m. *P < 0.05, two-tailed independent-samples t-test.

Extended Data Figure 6 Characterization of the effects of weak pharmacological inhibition of BDNF–TrkB signalling on sLTP and Rac1 activity spreading.

a. Individual data points for crosstalk experiments under low [TrkB-Ig] exposure. Each plot contains the average values of the experiments (thick, black curve) along with the corresponding individual experiments for the suprathreshold spine (left) and the subthreshold spine (right). Inset figures correspond to a close-up view of the data distributions for >10 min (the values used to calculate the sustained volume changes shown in Fig. 2). b, Same as a, but for crosstalk experiments using low [1NMPP1]. c, Same as a and b, but for crosstalk experiments using low [NSC-23766]. d, Effect of 0.125 μg ml−1 TrkB-Ig on Rac1 signal spreading. Each plot represents a specific time epoch after glutamate uncaging onset. Curves represent control (red = spine; black = dendrite) and +TrkB-Ig (blue = spine; green = dendrite) conditions plotted as a function of distance from the stimulated spine (n = 5 cells/6 control spines; n = 5 cells/9 +TrkB-Ig spines). e, Summary of data in d. Bars represent averages of the indicated temporal window across 1–5 μm of the dendrite. f, Same as d, but with the absence (red/black) and presence (blue/green) of 0.125 μM 1NMPP1 in TrkbF616A slices (control: n = 5 cells/8 spines; +1NMPP1: n = 5 cells/11 spines). g, Summary of data in f. Bars represent averages of the indicated temporal window across 1–5 μm of the dendrite. h, Same as d and f, but in the absence (red/black) and presence (blue/green) of the Rac1 inhibitor, 15 μM NSC-23766 (control: n = 6 cells/8 spines; NSC-23766: n = 8 cells/13 spines). i, Summary of data in h. Bars represent averages of the indicated temporal window across 1–5 μm of the dendrite. *P < 0.05, t-test.

Extended Data Figure 7 Effect of BDNF application on Rho GTPase signalling.

a, Effect of bath application of 20 ng ml−1 exogenous BDNF on Rac1 (n = 8 cells), Cdc42 (n = 10 cells), and RhoA (n = 5 cells) sensors. *P < 0.05, statistically different from zero, one-sample t-test. b, Summary of data in a. c, Effects of BDNF application on spine volume in Rac1 (blue), Cdc42 (orange), and RhoA (green) sensor-expressing cells. Data are from the experiments in a and b. d, Summary of data in c. ANOVA determined that there was no significant difference between the sensor conditions. None of the conditions showed a significant difference from zero as determined by a one-sample t-test. e, Response of Rac1 to a subthreshold stimulus in the presence of 20 ng ml−1 BDNF (control n = 15 cells/15 spines; +BDNF n = 6 cells/6 spines) f, Same as e, but for RhoA (control n = 14 cells/18 spines; +BDNF n = 7 cells/8 spines) g, Same as e and f, but for Cdc42 (control n = 17 cells/20 spines; +BDNF n = 7 cells/7 spines).

Extended Data Figure 8 Characterization of the dendritic Rac1 inhibitor approach.

a, Representative two-photon images illustrating the filamentous distribution HEK293T cells (top row), consistent with microtubule localization, and the largely dendrite-specific localization in CA1 neurons (bottom two rows) of W56–mCh–MTBD in comparison to an eGFP cell fill (first column). HEK293T cells were imaged at 1,050 nm to increase specificity of excitation for mCh versus eGFP. All images were acquired 2–5 days after transfection. b, Comparison of the expression levels of scr–MTBD and W56–MTBD from a subset of cells used for synaptic crosstalk experiments. Expression was estimated by acquiring intensity values in the red channel of an ~10 μm section of a secondary dendrite using ImageJ. Each point represents a single cell. c, Average time course of Rac1 activation for cells expressing only the Rac1 sensor (control, black curve; n = 105 spines), the Rac1 sensor plus W56–mCh–MTBD (red curve; n = 21 cells/33 spines), and the Rac1 sensor plus W56–mC (green curve; n = 6 cells/13 spines). Error bars represent s.e.m. d, Comparison of the effect of W56–mCh–MTBD (left; n = 21 cells/33 spines) to untargeted W56–mCh (right; n = 6 cells/13 spines) on the spatial profile of Rac1 activation. Data are mean ± s.e.m. Control corresponds to the data in Fig. 1d. e, Effect of expression of W56–mCh–MTBD on RhoA sensor activation. Left, spatial profile of RhoA activation during sTLP induction in the presence of scr–mCh–MTBD control (black curve) and W56–mCh–MTBD (red curve). Right, representative 2pFLIM images of RhoA activation in the presence of W56–mCh–MTBD. White circle indicates the targeted spine. f, Schematic of the design of the GAP-based Rac1 dendritic inhibitor. A Rac1-specific GTPase-activating protein (ARHGAP15) replaced W56 in the general dendritic inhibitor construct (see Fig. 3a). g, Representative 2pFLIM images of the effect of GAP–mCh–MTBD on Rac1 signal spreading for the indicated time windows. White circle indicates the targeted spine. h, Quantification of the effects of expression of GAP–mCh–MTBD on Rac1 signal spreading after glutamate uncaging. Data are depicted as the change in binding fraction in the dendrite as a function of distance from the stimulated spine (change in binding fraction in spine plotted on y axis) (n = 8 cells/11 spines). i, Summary of the data depicted in g and h. Also shown are data for Rac1 spreading from scr–MTBD (see Fig. 3d) for comparison. *P < 0.05, independent-samples t-test. j. Effects of GAP–mCh–MTBD expression on synaptic crosstalk (n = 4 cells/8 crosstalk experiments). k, Summary of the data in j. *P < 0.05, ANOVA and post-hoc test using the Tukey–Kramer method.

Extended Data Figure 9 Effect of a subthreshold stimulus on single spines of sensor-expressing CA1 neurons.

a, Plots showing the variability of the response of the Rac1 (left), Cdc42 (middle), and RhoA (right) sensors to a suprathreshold stimulus. Thick black curves correspond to the averages depicted in Fig. 4a–c (‘unpaired’ condition) of the main text. The peak responses (1–2 min after stimulation, the same points used for statistical claims in Fig. 4) were subjected to a Shapiro–Wilk test to confirm the normality of the data. All data supported the null hypothesis, illustrating that the data are Guassian distributed and justifying the use of parametric statistics. b, Plots showing the variability of the response of the Rac1 (left), Cdc42 (middle) and RhoA (right) sensors to a subtreshold stimulus. Thick black curves correspond to the averages depicted in Fig. 4a–c (‘unpaired’ condition). b, Change in spine volume with glutamate uncaging (arrow and dotted line) during an unpaired threshold (black) and subthreshold (red) stimuli for CA1 pyramidal cells expressing the Rac1 (left), Cdc42 (centre), or RhoA (right) sensors. Shaded region represents s.e.m. Data correspond to the volume curves for the data presented in Fig. 4. c, 8-Hz two-photon imaging of BDNF–SEP intensity during an unpaired suprathreshold (black) or subthreshold (red) stimulus. Left, the full time course. Glutamate uncaging stimuli are delivered at 0.5 Hz beginning at t = 0, indicated by the black arrow and dashed line. Inset shows the associated volume change (as measured from mCh cell fill) of the two conditions. Middle panel, the uncaging-triggered average of 30 16-frame bins (corresponding to each uncaging pulse) and thus shows the average response to individual glutamate uncaging events. Right panel, the average of the first point after uncaging for the indicated conditions. Both suprathreshold and subthreshold conditions show a statistically significant difference from zero, while the presence of AP5 plus NBQX or the POMC peptide eliminate this signal. Error bars represent s.e.m. (n = 28 cells/217 spines (LTP), 5/84 (Sub), 2/46 AP5+NBQX), and 2/29 (POMC)). d, Left, activation of the TrkB sensor in response to an unpaired suprathreshold (black; n = 4 cells/5 spines) or subthreshold (red; n = 6 cells/8 spines) stimulus. Right, change in spine volume in a cell expressing the TrkB sensor in response to an unpaired threshold or subthreshold stimulus. Error bars represent s.e.m.

Extended Data Figure 10 Change in volume of paired spines during synaptic crosstalk in Rho GTPase-expressing CA1 neurons.

a, Spine volume change in response to a suprathreshold (black curve; black arrow indicates stimulus initiation) and a paired subthreshold (red curve, grey arrow indicates stimulus initiation) stimulus in spines from cells expressing the Rac1 sensor (n = 6 cells/12 spine pairs). Error bars represent s.e.m. b, Same as a, but for Cdc42-expressing cells (n = 9 cells/10 spine pairs). c, Same as a and b, but for RhoA-expressing cells (n = 8 cells/12 spine pairs).

Supplementary information

Supplementary Information

This file contains Supplementary Notes, Supplementary Table 1 and additional references. (PDF 620 kb)

PowerPoint slides

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Hedrick, N., Harward, S., Hall, C. et al. Rho GTPase complementation underlies BDNF-dependent homo- and heterosynaptic plasticity. Nature 538, 104–108 (2016). https://doi.org/10.1038/nature19784

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1038/nature19784

This article is cited by

Comments

By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.

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