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.

  • Article
  • Published:

Neddylation inhibition impairs spine development, destabilizes synapses and deteriorates cognition

Abstract

Neddylation is a ubiquitylation-like pathway that controls cell cycle and proliferation by covalently conjugating Nedd8 to specific targets. However, its role in neurons, nonreplicating postmitotic cells, remains unexplored. Here we report that Nedd8 conjugation increased during postnatal brain development and is active in mature synapses, where many proteins are neddylated. We show that neddylation controls spine development during neuronal maturation and spine stability in mature neurons. We found that neddylated PSD-95 was present in spines and that neddylation on Lys202 of PSD-95 is required for the proactive role of the scaffolding protein in spine maturation and synaptic transmission. Finally, we developed Nae1CamKIIα-CreERT2 mice, in which neddylation is conditionally ablated in adult excitatory forebrain neurons. These mice showed synaptic loss, impaired neurotransmission and severe cognitive deficits. In summary, our results establish neddylation as an active post-translational modification in the synapse regulating the maturation, stability and function of dendritic spines.

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

Access options

Rent or buy this article

Prices vary by article type

from$1.95

to$39.95

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

Figure 1: Nedd8 is the most abundantly expressed UBL in neurons and is broadly expressed in the mouse brain.
Figure 2: Inhibition of neddylation blocks spine maturation.
Figure 3: Neddylation regulates spine maturation in vivo.
Figure 4: Neddylation controls spine stability.
Figure 5: Ubc12-C111S impairs spine maintenance in vivo.
Figure 6: Molecular characterization of PSD-95 neddylation.
Figure 7: Neddylation of Lys202 is required for the proactive role of PSD-95 in spine maturation and mEPSCs.
Figure 8: Conditional inactivation of neddylation in principal neurons of the forebrain leads to spine loss, decreased synaptic activity and cognitive deficits.

Similar content being viewed by others

Accession codes

Accessions

GenBank/EMBL/DDBJ

References

  1. Kawabe, H. & Brose, N. The role of ubiquitylation in nerve cell development. Nat. Rev. Neurosci. 12, 251–268 (2011).

    Article  CAS  PubMed  Google Scholar 

  2. Yang, Y., Kim, A.H. & Bonni, A. The dynamic ubiquitin ligase duo: Cdh1-APC and Cdc20-APC regulate neuronal morphogenesis and connectivity. Curr. Opin. Neurobiol. 20, 92–99 (2010).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  3. van der Veen, A.G. & Ploegh, H.L. Ubiquitin-like proteins. Annu. Rev. Biochem. 81, 323–357 (2012).

    Article  CAS  PubMed  Google Scholar 

  4. Kumar, S., Tomooka, Y. & Noda, M. Identification of a set of genes with developmentally down-regulated expression in the mouse-brain. Biochem. Biophys. Res. Commun. 185, 1155–1161 (1992).

    Article  CAS  PubMed  Google Scholar 

  5. Xirodimas, D.P. Novel substrates and functions for the ubiquitin-like molecule NEDD8. Biochem. Soc. Trans. 36, 802–806 (2008).

    Article  CAS  PubMed  Google Scholar 

  6. Petroski, M.D. & Deshaies, R.J. Function and regulation of cullin-RING ubiquitin ligases. Nat. Rev. Mol. Cell Biol. 6, 9–20 (2005).

    Article  CAS  PubMed  Google Scholar 

  7. Soucy, T.A. et al. An inhibitor of NEDD8-activating enzyme as a new approach to treat cancer. Nature 458, 732–736 (2009).

    Article  CAS  PubMed  Google Scholar 

  8. Choo, Y.S. et al. Regulation of parkin and PINK1 by neddylation. Hum. Mol. Genet. 21, 2514–2523 (2012).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  9. Djagaeva, I. & Doronkin, S. Dual regulation of dendritic morphogenesis in Drosophila by the COP9 signalosome. PLoS ONE 4, e7577 (2009).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  10. Stadler, C. et al. Immunofluorescence and fluorescent-protein tagging show high correlation for protein localization in mammalian cells. Nat. Methods 10, 315–323 (2013).

    Article  CAS  PubMed  Google Scholar 

  11. Wada, H., Yeh, E.T. & Kamitani, T. A dominant-negative UBC12 mutant sequesters NEDD8 and inhibits NEDD8 conjugation in vivo. J. Biol. Chem. 275, 17008–17015 (2000).

    Article  CAS  PubMed  Google Scholar 

  12. Ziv, N.E. & Smith, S.J. Evidence for a role of dendritic filopodia in synaptogenesis and spine formation. Neuron 17, 91–102 (1996).

    Article  CAS  PubMed  Google Scholar 

  13. Fiala, J.C., Feinberg, M., Popov, V. & Harris, K.M. Synaptogenesis via dendritic filopodia in developing hippocampal area CA1. J. Neurosci. 18, 8900–8911 (1998).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  14. Yuste, R. & Bonhoeffer, T. Genesis of dendritic spines: insights from ultrastructural and imaging studies. Nat. Rev. Neurosci. 5, 24–34 (2004).

    Article  CAS  PubMed  Google Scholar 

  15. Lohmann, C., Finski, A. & Bonhoeffer, T. Local calcium transients regulate the spontaneous motility of dendritic filopodia. Nat. Neurosci. 8, 305–312 (2005).

    Article  CAS  PubMed  Google Scholar 

  16. Ehrlich, I., Klein, M., Rumpel, S. & Malinow, R. PSD-95 is required for activity-driven synapse stabilization. Proc. Natl. Acad. Sci. USA 104, 4176–4181 (2007).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  17. Yoshihara, Y., De, R.M. & Muller, D. Dendritic spine formation and stabilization. Curr. Opin. Neurobiol. 19, 146–153 (2009).

    Article  CAS  PubMed  Google Scholar 

  18. Verpelli, C., Schmeisser, M.J., Sala, C. & Boeckers, T.M. Scaffold proteins at the postsynaptic density. Adv. Exp. Med. Biol. 970, 29–61 (2012).

    Article  CAS  PubMed  Google Scholar 

  19. Bayés, A. & Grant, S.G. Neuroproteomics: understanding the molecular organization and complexity of the brain. Nat. Rev. Neurosci. 10, 635–646 (2009).

    Article  CAS  PubMed  Google Scholar 

  20. Chan, Y. et al. DEN1 deneddylates non-cullin proteins in vivo. J. Cell Sci. 121, 3218–3223 (2008).

    Article  CAS  PubMed  Google Scholar 

  21. Colledge, M. et al. Ubiquitination regulates PSD-95 degradation and AMPA receptor surface expression. Neuron 40, 595–607 (2003).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  22. Tsai, N.P. et al. Multiple autism-linked genes mediate synapse elimination via proteasomal degradation of a synaptic scaffold PSD-95. Cell 151, 1581–1594 (2012).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  23. Opazo, P., Sainlos, M. & Choquet, D. Regulation of AMPA receptor surface diffusion by PSD-95 slots. Curr. Opin. Neurobiol. 22, 453–460 (2012).

    Article  CAS  PubMed  Google Scholar 

  24. Han, K. & Kim, E. Synaptic adhesion molecules and PSD-95. Prog. Neurobiol. 84, 263–283 (2008).

    Article  CAS  PubMed  Google Scholar 

  25. Shepherd, J.D. & Huganir, R.L. The cell biology of synaptic plasticity: AMPA receptor trafficking. Annu. Rev. Cell Dev. Biol. 23, 613–643 (2007).

    Article  CAS  PubMed  Google Scholar 

  26. Bingol, B. & Schuman, E.M. A proteasome-sensitive connection between PSD-95 and GluR1 endocytosis. Neuropharmacology 47, 755–763 (2004).

    Article  CAS  PubMed  Google Scholar 

  27. Ehlers, M.D. Activity level controls postsynaptic composition and signaling via the ubiquitin-proteasome system. Nat. Neurosci. 6, 231–242 (2003).

    Article  CAS  PubMed  Google Scholar 

  28. Kirkpatrick, D.S., Denison, C. & Gygi, S.P. Weighing in on ubiquitin: the expanding role of mass-spectrometry-based proteomics. Nat. Cell Biol. 7, 750–757 (2005).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  29. El-Husseini, A.E., Schnell, E., Chetkovich, D.M., Nicoll, R.A. & Bredt, D.S. PSD-95 involvement in maturation of excitatory synapses. Science 290, 1364–1368 (2000).

    CAS  PubMed  Google Scholar 

  30. Schlüter, O.M., Xu, W.F. & Malenka, R.C. Alternative N-terminal domains of PSD-95 and SAP97 govern activity-dependent regulation of synaptic AMPA receptor function. Neuron 51, 99–111 (2006).

    Article  CAS  PubMed  Google Scholar 

  31. Sturgill, J.F., Steiner, P., Czervionke, B.L. & Sabatini, B.L. Distinct domains within PSD-95 mediate synaptic incorporation, stabilization, and activity-dependent trafficking. J. Neurosci. 29, 12845–12854 (2009).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  32. Hsueh, Y.P., Kim, E. & Sheng, M. Disulfide-linked head-to-head multimerization in the mechanism of ion channel clustering by PSD-95. Neuron 18, 803–814 (1997).

    Article  CAS  PubMed  Google Scholar 

  33. Schnell, E. et al. Direct interactions between PSD-95 and stargazin control synaptic AMPA receptor number. Proc. Natl. Acad. Sci. USA 99, 13902–13907 (2002).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  34. Sainlos, M. et al. Biomimetic divalent ligands for the acute disruption of synaptic AMPAR stabilization. Nat. Chem. Biol. 7, 81–91 (2011).

    Article  CAS  PubMed  Google Scholar 

  35. Ehrlich, I. & Malinow, R. Postsynaptic density 95 controls AMPA receptor incorporation during long-term potentiation and experience-driven synaptic plasticity. J. Neurosci. 24, 916–927 (2004).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  36. Stein, V., House, D.R.C., Bredt, D.S. & Nicoll, R.A. Postsynaptic density-95 mimics and occludes hippocampal long-term potentiation and enhances long-term depression. J. Neurosci. 23, 5503–5506 (2003).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  37. Elias, G.M. et al. Synapse-specific and developmentally regulated targeting of AMPA receptors by a family of MAGUK scaffolding proteins. Neuron 52, 307–320 (2006).

    Article  CAS  PubMed  Google Scholar 

  38. Felix-Ortiz, A.C. & Tye, K.M. Amygdala inputs to the ventral hippocampus bidirectionally modulate social behavior. J. Neurosci. 34, 586–595 (2014).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  39. Chen, P.C. et al. Ubiquitin homeostasis is critical for synaptic development and function. J. Neurosci. 31, 17505–17513 (2011).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  40. Trachtenberg, J.T. et al. Long-term in vivo imaging of experience-dependent synaptic plasticity in adult cortex. Nature 420, 788–794 (2002).

    Article  CAS  PubMed  Google Scholar 

  41. Bianchetta, M.J., Lam, T.T., Jones, S.N. & Morabito, M.A. Cyclin-dependent kinase 5 regulates PSD-95 ubiquitination in neurons. J. Neurosci. 31, 12029–12035 (2011).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  42. Christopherson, K.S. et al. Lipid- and protein-mediated multimerization of PSD-95: implications for receptor clustering and assembly of synaptic protein networks. J. Cell Sci. 116, 3213–3219 (2003).

    Article  CAS  PubMed  Google Scholar 

  43. Nonaka, M., Doi, T., Fujiyoshi, Y., Takemoto-Kimura, S. & Bito, H. Essential contribution of the ligand-binding beta B/beta C loop of PDZ1 and PDZ2 in the regulation of postsynaptic clustering, scaffolding, and localization of postsynaptic density-95. J. Neurosci. 26, 763–774 (2006).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  44. Han, K. et al. Regulated RalBP1 binding to RalA and PSD-95 controls AMPA receptor endocytosis and LTD. PLoS Biol. 7, e1000187 (2009).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  45. Kim, K. et al. Synaptic removal of diacylglycerol by DGKzeta and PSD-95 regulates dendritic spine maintenance. EMBO J. 28, 1170–1179 (2009).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  46. Xie, Z. et al. Kalirin-7 controls activity-dependent structural and functional plasticity of dendritic spines. Neuron 56, 640–656 (2007).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  47. Kim, I.H. et al. Disruption of Arp2/3 results in asymmetric structural plasticity of dendritic spines and progressive synaptic and behavioral abnormalities. J. Neurosci. 33, 6081–6092 (2013).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  48. Hung, A.Y. et al. Smaller dendritic spines, weaker synaptic transmission, but enhanced spatial learning in mice lacking Shank1. J. Neurosci. 28, 1697–1708 (2008).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  49. Cahill, M.E. et al. Kalirin regulates cortical spine morphogenesis and disease-related behavioral phenotypes. Proc. Natl. Acad. Sci. USA 106, 13058–13063 (2009).

    Article  PubMed  PubMed Central  Google Scholar 

  50. Dil Kuazi, A. et al. NEDD8 protein is involved in ubiquitinated inclusion bodies. J. Pathol. 199, 259–266 (2003).

    Article  CAS  PubMed  Google Scholar 

  51. Shirogane, T., Jin, J.P., Ang, X.L. & Harper, J.W. SCF beta-TRCP controls Clock-dependent transcription via casein kinase 1-dependent degradation of the mammalian Period-1 (Per1) protein. J. Biol. Chem. 280, 26863–26872 (2005).

    Article  CAS  PubMed  Google Scholar 

  52. Kamura, T., Conrad, M.N., Yan, Q., Conaway, R.C. & Conaway, J.W. The Rbx1 subunit of SCF and VHL E3 ubiquitin ligase activates Rub1 modification of cullins Cdc53 and Cul2. Genes Dev. 13, 2928–2933 (1999).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  53. Refojo, D. et al. Glutamatergic and dopaminergic neurons mediate anxiogenic and anxiolytic effects of CRHR1. Science 333, 1903–1907 (2011).

    Article  CAS  PubMed  Google Scholar 

  54. Dymecki, S.M. Flp recombinase promotes site-specific DNA recombination in embryonic stem cells and transgenic mice. Proc. Natl. Acad. Sci. USA 93, 6191–6196 (1996).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  55. Erdmann, G., Schütz, G. & Berger, S. Inducible gene inactivation in neurons of the adult mouse forebrain. BMC Neurosci. 8, 63 (2007).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  56. Feng, G.P. et al. Imaging neuronal subsets in transgenic mice expressing multiple spectral variants of GFP. Neuron 28, 41–51 (2000).

    Article  CAS  PubMed  Google Scholar 

  57. Kühne, C. et al. Visualizing corticotropin-releasing hormone receptor type 1 expression and neuronal connectivities in the mouse using a novel multifunctional allele. J. Comp. Neurol. 520, 3150–3180 (2012).

    Article  CAS  PubMed  Google Scholar 

  58. Dotti, C.G., Sullivan, C.A. & Banker, G.A. The establishment of polarity by hippocampal neurons in culture. J. Neurosci. 8, 1454–1468 (1988).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  59. Allen, J.W., Mutkus, L.A. & Aschner, M. Isolation of neonatal rat cortical astrocytes for primary cultures. Curr. Protoc. Toxicol. Chapter 12 Unit 12.4 (2000).

  60. Harris, K.M. & Kater, S.B. Dendritic spines—cellular specializations imparting both stability and flexibility to synaptic function. Annu. Rev. Neurosci. 17, 341–371 (1994).

    Article  CAS  PubMed  Google Scholar 

  61. Sala, C. et al. Regulation of dendritic spine morphology and synaptic function by Shank and Homer. Neuron 31, 115–130 (2001).

    Article  CAS  PubMed  Google Scholar 

  62. Okamura, K. et al. Cadherin activity is required for activity-induced spine remodeling. J. Cell Biol. 167, 961–972 (2004).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  63. Saito, T. In vivo electroporation in the embryonic mouse central nervous system. Nat. Protoc. 1, 1552–1558 (2006).

    Article  CAS  PubMed  Google Scholar 

  64. 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  PubMed  Google Scholar 

  65. Cho, K.O., Hunt, C.A. & Kennedy, M.B. The rat-brain postsynaptic density fraction contains a homolog of the Drosophila disks-large tumor suppressor protein. Neuron 9, 929–942 (1992).

    Article  CAS  PubMed  Google Scholar 

  66. Blackstone, C.D. et al. Biochemical characterization and localization of a non-N-methyl-D-aspartate glutamate receptor in rat brain. J. Neurochem. 58, 1118–1126 (1992).

    Article  CAS  PubMed  Google Scholar 

  67. States, D.J. et al. Challenges in deriving high-confidence protein identifications from data gathered by a HUPO plasma proteome collaborative study. Nat. Biotechnol. 24, 333–338 (2006).

    Article  CAS  PubMed  Google Scholar 

  68. Alves, P. et al. Fast and accurate identification of semi-tryptic peptides in shotgun proteomics. Bioinformatics 24, 102–109 (2008).

    Article  CAS  PubMed  Google Scholar 

  69. Kleinknecht, K.R. et al. Hippocampus-dependent place learning enables spatial flexibility in C57BL6/N mice. Front. Behav. Neurosci. 6, 87 (2012).

    Article  PubMed  PubMed Central  Google Scholar 

  70. Wang, X.D. et al. Forebrain CRF1 modulates early-life stress-programmed cognitive deficits. J. Neurosci. 31, 13625–13634 (2011).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  71. Hartmann, J. et al. Fkbp52 heterozygosity alters behavioral, endocrine and neurogenetic parameters under basal and chronic stress conditions in mice. Psychoneuroendocrinology 37, 2009–2021 (2012).

    Article  CAS  PubMed  Google Scholar 

  72. Kamprath, K. & Wotjak, C.T. Nonassociative learning processes determine expression and extinction of conditioned fear in mice. Learn. Mem. 11, 770–786 (2004).

    Article  PubMed  PubMed Central  Google Scholar 

  73. Petrini, E.M. et al. Endocytic trafficking and recycling maintain a pool of mobile surface AMPA receptors required for synaptic potentiation. Neuron 63, 92–105 (2009).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

Download references

Acknowledgements

We thank S. Bauer, A. Moebus, C. Mattusch, K. Henes, D. Glass and S. Opitz for their technical assistance; the Bordeaux Imaging Center and C. Poujol for assistance with FRET experiments; and R. Malenka, P. Scheiffele, C. Sala, V.R. Wedlich-Soeldner, V. Tarabykin, D.C. Lie, W. Snider, T. Soderling, Y. Yarden and J.W. Conaway for providing plasmids. We are grateful to P. Scheiffele and R. Klein for helpful comments and advice. This work was supported by ERC grant Nano-Dyn-Syn (D.C. and A.-S.H.); the Helmholtz Alliance for Mental Health (D.M.V.-W.); BMBF grant nos. NGNFplus 01GS08174-14 (D.M.V.-W.), 01GS08151 and 01GS08155 (J.M.D.); the DFG SPP1365 grant (D.R.); and the Max Planck Society (J.M.D. and D.R.).

Author information

Authors and Affiliations

Authors

Contributions

A.M.V. and M.M.B. designed and performed the experiments and interpreted the results. S.A.G., C.A.V., J.S.R., C.A.B., N.D. and F.R. performed experiments. G.M. and C.W.T. performed mass spectrometry experiments. A.-S.H. and D.C. performed FRET and FRAP studies. V.S. designed electrophysiology experiments and helped write the manuscript. D.M.V.-W., C.T.W. and J.M.D. analyzed the data. D.R. supervised the project and developed the conceptual and experimental framework. A.M.V. and D.R. wrote the manuscript.

Corresponding author

Correspondence to Damian Refojo.

Ethics declarations

Competing interests

The authors declare no competing financial interests.

Integrated supplementary information

Supplementary Figure 1 Expression analysis of Nedd8 and Ubc12 in the adult mouse brain and primary neurons.

(a) Nedd8 is highly expressed throughout neuronal development. Real-time qRT-PCRs of Ubl mRNAs in mouse cortical neurons at DIV7, DIV14 and DIV21. Relative Ubl mRNA levels were normalized to Rpl19, n = 4, from 2 cell cultures. ND = not detectable. (b) Representative dark-field photomicrographs of radioactive in situ hybridizations (ISH) on coronal brain sections (from rostral to caudal) of wild-type C57BL/6 mice show high and widespread expression of Nedd8 and Ubc12 mRNA throughout the adult mouse brain. Scale bars, 1 mm. (c,d) Control of specificity of ISH probes against Ubc12 and Nedd8. Representative bright-field photomicrographs of X-ray films of radioactive ISH on coronal brain sections of wild-type C57BL/6 mice with anti-sense (upper panel) and sense control (lower panel) riboprobes (c). In contrast to the anti-sense probes, ISH sense probes show no specific signals. Scale bar, 1 mm. Reduced signal of Ubc12 mRNA in the brain of Ubc12+/− mice (d). Representative dark-field photomicrographs of ISH against Ubc12 on sagittal brain sections of Ubc12+/+ and Ubc12+/− (Ubc12-LacZ reporter) mice. Scale bar, 500 μm. (e) Generation of Ubc12-LacZ reporter mice. A genetrap vector, consisting of a reporter (β-Galactosidase)/selector (Neomycin-resistance) cassette was randomly integrated into intron 1 of the mouse Ubc12 gene (ENSMUSG00000005575) in E14 mouse ES cells. β-Galactosidase is spliced to exon 1 of Ubc12 and functions as a reporter of endogenous Ubc12 expression. Ubc12 Prom, Ubc12 Promoter; 1, Exon 1; 2, Exon 2; 3, Exon 3; I2, Intron 2; AdSA, Adenoviral splice acceptor; IRES, Internal ribosome entry site; Neo-R, Neomycin resistance; BGH pA, Bovine growth hormone polyadenylation signal. On the right, LacZ staining on a sagittal brain section of a heterozygous Ubc12-LacZ reporter mouse. (f) Analysis of the intracellular localization of Ubc12 and Nedd8 in mature cultured mouse hippocampal neurons, transfected with mRFP and Venus-Nedd8 or Venus-Ubc12 (left panel) and HA-Nedd8 and HA-Ubc12 (right panel) plasmids at DIV18 and fixed at DIV20. Neurons transfected with HA-tagged Nedd8 and Ubc12 were immunostained with α-HA antibodies. Tagged-Nedd8 and tagged-Ubc12 show somato-dendritic and axonal localization and both proteins are present in dendritic protrusions. Scale bars, 10 μm.

Data are presented as mean ± s.e.m. Images in b,c,d,e and f represent data from at least two experiments.

Supplementary Figure 2 Neddylation controls spine development.

(a,b) Validation of shRNAs against Ubc12 (a) and Nedd8 (b) in Neuro-2a cells transiently transfected with the indicated plasmids for 72 h. The efficiency of shRNA constructs to downregulate 3xFLAG-Ubc12 (a) and 3xFLAG-Nedd8 (b) was assessed by Western blotting with α-FLAG antibodies. The shRNA vector #3 against the mouse Ubc12 sequence and the shRNA vector #3 targeting the mouse Nedd8 sequence displayed highest knock-down efficiencies. AKT or Actin served as loading controls. (c) MLN-4924 blocks neddylation in mature neurons. At DIV22 rat cortical neurons were treated with Vehicle (DMSO) or 1 μM MLN-4924 for 24h. Cell extracts were subjected to immunoblotting with α-Nedd8, α-Cullin1 and α-Tubulin antibodies. The arrow indicates the (slightly overexposed) bands corresponding to the neddylated cullins (~ 100 kDa). (d) Rescue of spine development in Ubc12 and Nedd8 shRNA-expressing neurons by shRNA-resistant forms of Nedd8 and Ubc12. At DIV12 mouse hippcoampal neurons were transfected with the indicated plasmids and GFP as a volume marker. Spine and filopodia formation was analyzed at DIV18 (one-way ANOVA; P < 0.0001; Bonferroni post-hoc test; **P < 0.01 vs. sh-RNA Ctrl; Number of spines per 20 μm dendrite segment: sh-RNA Ctrl: 12.91 ± 0.55; sh-Ubc12: 1.62 ± 0.29; sh-Ubc12 + Ubc12-RES: 10.71 ± 0.70; sh-RNA Ctrl: 12.66 ± 0.75; sh-Nedd8: 2.47 ± 0.38; sh-Nedd8 + Nedd8-RES: 9.19 ± 0.42; Number of fiolopodia per 20 μm dendrite segment: sh-RNA Ctrl: 0.66 ± 0.16; sh-Ubc12: 12.13 ± 0.79; sh-Ubc12 + Ubc12-RES: 2.84 ± 0.53; sh-RNA Ctrl: 1.03 ± 0.23; sh-Nedd8: 11,18 ± 0.65; sh-Nedd8 + Nedd8-RES: 3.15 ± 0.57; n = 15 neurons). Scale bars, 5 μm. On the right: Western blot validation of knock-down efficiencies of shRNAs and re-expression of shRNA-resistant forms of Nedd8 and Ubc12 in Neuro-2a cells. (e,f) Down regulation of Cullin-RING ligases does not reproduce the effects of Nedd8 inhibition on spine maturation. (e) Rat hippocampal neurons were transfected with Venus and Control, Ubc12-C111S or dominant-negative Cullin (Cul1-dn, Cul2-dn, Cul3-dn, Cul4A-dn, Cul4-dn and Cul5-dn) constructs at DIV13 and spine development was analyzed at DIV19. Scale bar, 2.5 μm. (f) Upper panel: Mouse hippocampal neurons were transfected with Venus and Control or Rbx1-C42S/C45S dominant negative constructs at DIV13. Lower panel: Mouse hippocampal neurons were transfected with mRFP and shRNA Control, shRNA Rbx1, shRNA Ubc12 or shRNA Nedd8 at DIV13. Spine formation was analyzed at DIV19 ((e): one-way ANOVA; P < 0.0001; Bonferroni post-hoc test; **P < 0.01 vs. Control or shRNA Control; n = 12 neurons; Number of spines per 10 μm dendrite segment: Control: 9.52 ±0.87; Ubc12: 2.03 ± 0.78; Cul1-dn: 10.95 ± 1.27; Cul2-dn: 9.81 ± 1.41; Cul3-dn: 7.82 ± 1.48; Cul4A-dn: 10.02 ± 0.72; Cul4B-dn: 8.16 ± 0.57; Cul5-dn: 9.14 ± 0.64; (f): one-way ANOVA; P < 0.0001; Bonferroni post-hoc test; **P < 0.01 vs. shRNA Control or shRNA Rbx1; n = 12 neurons; Number of spines per 10 μm dendrite segment: shRNA Control: 9.13 ± 0.82; shRNA Rbx1: 10.60 ± 0.81; shRNA Ubc12: 4.27 ± 0.67; shRNA Nedd8: 2.88 ± 0.74; - two-tailed unpaired Student's t-test; P = 0.654; n = 12 neurons; Number of spines per 10 μm dendrite segment: Control: 10.29 ± 1.01; Rbx1-C42S/C45S: 9.61 ± 1.11). (g) Validation of shRNAs against Rbx1 in Neuro-2a cells transiently transfected with the indicated plasmids for 72 h. The shRNA vector #4 targeting the mouse Rbx1 sequence showed the highest knock-down efficiency in Western blotting. (h) Decreased total number of excitatory synapses but increased number of excitatory dendritic shaft synapses of neddylation-deficient neurons. DIV10 mouse hippocampal neurons were transfected with GFP. At DIV12, 14 and 16 neurons were treated with Vehicle (DMSO) or 0.5 μM MLN-4924; at DIV18 neurons were fixed and immunostained with α-VGLUT1 antibodies. The number of excitatory VGLUT1-positive synapses on dendritic shafts and the total number of excitatory VGLUT1-positive synapses were counted (two-tailed unpaired Student's t-test; shaft synapses: **P < 0.0001; Vehicle: 0.84 ± 0.17; MLN-4924: 3.75 ± 0.35; total synapses: **P =0.0008; Vehicle: 10.07 ± 0.93; MLN-4914: 5.49 ± 0.72; n = 12 neurons). The arrowheads indicate excitatory shaft synapses. Scale bars, 2.5 μm.

Data are presented as mean ± s.e.m. Images and quantifications represent data from three experiments. Individual t-values, degrees of freedom and F values are available in the Supplementary Methods Checklist.

Supplementary Figure 3 Inhibition of neddylation blocks spine maturation in developing neurons and reduces spine density in mature neurons.

(a,b) Normal filopodia dynamics in neddylation-deficient neurons. (a) Extension and retraction of filopodia of immature mouse hippocampal neurons, transfected at DIV10 with myr-Venus and Control or Ubc12-C111S plasmids and imaged at DIV12 with a time interval of 3 min. The filopodia of Ubc12-C111S-expressing neurons displayed a motile behavior, extending and retracting similar to filopodia structures of control neurons. The white asterisks indicate a filopodia in stable contact with a putative presynaptic axon; the white arrows highlight a filopodia which transiently contacts a potential axon collateral; the pink arrowheads mark retracting filopodia; and the yellow arrowheads indicate growing filopodia structures. Scale bars, 2.5 μm. Representative images for Control and Ubc12-C111S are shown. (b) Motility of filopodia of neddylation deficient neurons is not impaired. Immature DIV8 mouse hippocampal neurons were transfected with Control or Ubc12-C111S, Lifeact-GFP (green) and MAP2-RFP (red) plasmids. At DIV10 confocal time lapse movies (60 min) were recorded with a time interval of 60 seconds and average filopodia motility was measured (two-tailed unpaired Student's t-test; P = 0.46; average filopodia motility (μm/min): Control: 0.34 ± 0.03; Ubc12-C111S: 0.37 ± 0.02; n = 12 filopodia, from 3 neurons). For Control and Ubc12-C111S, 15 frames of a representative filopodium are shown from the 60 min movies. Scale bars, 1 μm.

(ce) Inducible inhibition of neddylation in mature neurons impairs spine stability. (c) Transfection-free inducible control of gene expression in neurons: Schematic outline depicting the experimental procedure. At E13.5, cortical neurons of mouse embryos from CamKIIαCreERT2 x CD1 breedings were in utero electroporated with a tamoxifen-inducible Ubc12-C111S (CAG-floxedSTOP-FLAG-Ubc12-C111S-IRES-EGFP) plasmid and a constitutively expressed CAG-mRFP construct. At E17.5 dissociated primary cortical neurons were prepared from electroporated cortices, which were identified by the expression of mRFP under a fluorescence stereomicroscope. The embryos were additionally genotyped for CamKIIαCreERT2 to confirm a mixed culture of Wild-type and CamKIIαCreERT2-expressing neurons. At DIV17, neurons were treated with 4-OH-tamoxifen to induce the expression of Ubc12-C111S and EGFP in CamKIIαCreERT2-positive neurons. At DIV20 neurons were fixed and analyzed. (d) Primary neurons from an in utero electroporation described in (c) were fixed at the indicated times. The number of spines was counted per 10 μm dendrite segment at the indicated DIVs. As shown, in our culture conditions, spine density reached a plateau at DIV17 (n=12; DIV7: 0.53 ± 0.12; DIV13: 2.41 ± 0.25; DIV15: 6.19 ± 0.53; DIV17: 7.00 ± 0.35; DIV20: 7.24 ± 0.41). (e) Mixed primary cortical neurons were prepared from Wild-type and CamKIIαCreERT2 in utero electroporated embryos as described in (c). At DIV17, neurons were treated with 0.5 μM 4-OH-tamoxifen to induce the expression of Ubc12-C111S in CamKIIαCreERT2 neurons. Neurons were fixed at DIV20 and analyzed for the presence (CamKIIαCreERT2 neurons, induced Ubc12-C111S expression) or absence (Wild-type neurons, no expression of Ubc12-C111S) of EGFP signal. The number of spines was counted per 10 μm dendrite segment (two-tailed unpaired Student's t-test; **P < 0.0001; Wild-type: 6.94 ± 0.29; CamKIIαCreERT2: 4.15 ± 0.22; n = 30 neurons). Scale bars, 10 μm and 2.5 μm in the higher magnification pictures.

Data are presented as mean ± s.e.m and box-and-whisker min. and max. plots. Images and quantifications in a,b and e represent data from three experiments. The data in d was obtained in two experiments. Individual t-values and degrees of freedom are available in the Supplementary Methods Checklist.

Supplementary Figure 4 The Nedd8 pathway controls spine stability in hippocampal and cortical principal neurons in vivo.

(a) The inducible expression of the control backbone vector containing the β-Galactosidase reporter cDNA in mature neurons did not result in changes in spine density. At E14.5 mouse embryos from CamKIIαCreERT2 x CD1 breedings were microinjected with inducible LacZ control plasmids together with inducible EGFP and constitutive mRFP constructs and electroporated to target the precursors of hippocampal CA1 pyramidal neurons and granule neurons of the dentate gyrus (DG). After birth, offspring were genotyped by PCR. At P35 tamoxifen was applied to CamKIIαCreERT2 animals and Wild-type littermates. At P50, spines were analyzed on electroporated apical and basal dendrites of CA1 pyramidal neurons and on dendrites of DG neurons (two-tailed unpaired Student's t-test; CA1 apical: P = 0.08; Wild-type: 16.27 ± 0.43; Cre: 15.07 ± 0.49; CA1 basal: P = 0.53; Wild-type: 13.93 ± 0.32; Cre: 14.29 ± 0.47; DG: P = 0.20, Wild-type: 13.36 ± 0.75; Cre: 14.50 ± 0.43; n = 15 neurons from 4 brains for CA1 apical and basal; n = 12 neurons from 4 brains for DG). (b) The inducible inhibition of neddylation in mature cortical neurons leads to a decrease in spine density. At E13.5, mouse embryos from CamKIIαCreERT2 x CD1 breedings were microinjected with inducible Ubc12-C111S-IRES-EGFP and constitutive mRFP constructs and electroporated to target precursors of layer II/III cortical neurons. After birth, offspring were genotyped by PCR. At P35, tamoxifen was applied to CamKIIαCreERT2 animals and Wild-type littermates and brains were analyzed at P50. As shown also in Fig. 5 in hippocampal neurons, the cortical expression of EGFP (and Ubc12-C111S) was induced only in CamKIIαCreERT2 animals demonstrating that the inducible constructs used, are not leaky. Spines were counted on apical dendrites of electroporated cortical neurons in layer II and III per 10 μm of dendrite segment (two-tailed unpaired Student's t-test; **P < 0.0001; Wild-type: 12.23 ± 0.63; Cre: 8.41 ± 0.50; n = 15 neurons from 4 brains). Scale bars, 100 µm and 2 µm in the higher magnification pictures.

Data are presented as mean ± s.e.m. and box-and-whisker min. and max. plots. Images and quantifications represent data from two experiments. Individual t-values and degrees of freedom are available in the Supplementary Methods Checklist.

Supplementary Figure 5 Screening of neddylation of post-synaptic scaffolding proteins in neurons reveals PSD-95 as a target of Nedd8.

(ad) Mouse cortical neurons were nucleofected with 6xHis-BIO-Nedd8 and the indicated post-synaptic scaffold protein constructs, including MAGUK (a and Fig. 6a), Homer (b), GKAP (c) and Shank (d) proteins. At DIV18 neurons were treated with Vehicle (DMSO) or 1 μM MLN-4924 for 6 h. Cell lysates were purified with Streptavidin beads under denaturing conditions and analyzed by immunoblotting for the indicated proteins. Of note, the only post-synaptic scaffold protein found neddylated was PSD-95 (Fig. 6a). The predominant Nedd8-positive band in Streptavidin-Nedd8 pull downs corresponds to neddylated cullins. (eg) Neddylation of PSD-95 is blocked by MLN-4924 treatment. HEK293 cells were transiently transfected with the indicated plasmids for 24 h and then treated with Vehicle (DMSO) or 1 µM MLN-4924 for 6 h. Cell extracts were immunoprecipitated with α-PSD-95 antibodies under non-denaturing conditions (e) or purified with Streptavidin beads under denaturing conditions (f,g). The eluted samples were analyzed by immunoblotting with the indicated antibodies. The arrowheads indicate neddylated PSD-95 bands; the additional band around 100 kDa in (e), marked by the asterisk, represents co-precipitating proteins associated with PSD-95.

Representative images from three experiments are shown. Full-length blots are presented in Supplementary Figure 10.

Supplementary Figure 6 Identification and molecular characterization of PSD-95 neddylation.

(a) Endogenous synaptic PSD-95 is neddylated. Pure synaptosomal fractions, purified from mouse forebrain tissue, were immunoprecipitated with α-PSD-95 antibodies and analyzed by Western blotting. The arrowhead indicates the specific neddylated PSD-95 band which was corroborated by mass spectrometry. The additional band around 100 kDa, marked by the asterisk, represents co-precipitating proteins associated with PSD-95. (bd) Neddylation does not control the stability of PSD-95 in neurons. At DIV14, mouse hippocampal neurons were treated with Vehicle (DMSO) or 1 µM MLN-4924 for 4, 8, 24 and 36 h. Total cell lysates were subjected to immunoblotting with α-PSD-95 antibodies; Tubulin served as loading control (two-tailed unpaired Student's t-test; 4 h: P = 0.85; DMSO: 1.00 ± 0.14; MLN: 0.97 ± 0.10; 8h: P = 0.65; DMSO: 1.00 ± 0.02; MLN: 1.11 ± 0.22; 24 h: P = 0.44; DMSO: 1.00 ± 0.05; MLN: 0.92 ± 0.08; 36 h: P = 0.10; DMSO: 1.00 ± 0.05; MLN: 0.85 ± 0.05; n = 4 samples from 2 cultures) (b). Primary mouse hippocampal neurons, cultured on filter plates, were treated with Vehicle (DMSO) or 1 µM MLN-4924 for 8 and 24 h at DIV16. Dendrite-enriched protein fractions were subjected to immunoblotting (c). Inhibition of neddylation does not alter PSD-95 degradation. Mouse hippocampal neurons (DIV16) were treated with 20 μg/ml CHX and DMSO or 1 μM MLN-4924 for the indicated time points. Total cell lysates were subjected to immunoblotting with α-PSD-95 antibodies; β-Actin served as loading control (two-way ANOVA; no significant treatment x time interaction; interaction P = 0.68; treatment P = 0.49; time = P < 0.0001; DMSO: 0 h: 1.00 ± 0.08; 2 h: 0.72 ± 0.14; 4 h: 0.61 ± 0.01; 12 h: 0.78 ± 0.19; 24 h: 0.41 ± 0.05; 36 h: 0.36 ± 0.04; MLN-4924: 0 h: 1.00 ± 0.09; 2 h: 0.88 ± 0.13; 4 h: 0.75 ± 0.08; 12 h: 0.59 ± 0.05; 24 h: 0.56 ± 0.01; 36 h: 0.39 ± 0.06; n = 4 samples from 2 cultures) (d). (eg) Identification of PSD-95 neddylated lysine residues by tandem mass spectrometry analysis. Generation of MS-Nedd8 mutants (e). 3xFLAG-tagged MS-Nedd8 constructs were generated using site-directed mutagenesis by insertion of an Ala (-AGG) or Val (-VGG) at position 74, or exchanging Leu at position 73 with Arg at position 74 (-LGG) in Nedd8 wild-type cDNA (-RGG). HEK293 were transiently transfected with the indicated plasmids. Total cell lysates were subjected to immunoblotting using α-FLAG antibodies. Expression of 3xFLAG-Nedd8-AGG and 3xFLAG-Nedd8-LGG resulted in similar neddylation patterns as the wild-type Nedd8 construct. In contrast, the 3xFLAG-Nedd8-VGG displayed reduced conjugation levels and was discarded for further experiments. Flow-chart showing the experimental procedure (f). PSD-95 was immunoprecipitated from primary cortical neurons or from transiently transfected HEK293 cells using α-PSD-95 antibodies. Isolated proteins were separated by SDS-PAGE. PSD-95 bands at ~100 and ~130 kDa were cut and subjected to in-gel tryptic digestion. Peptides were analyzed using LC-ESI-MS-MS. Nedd8- and Ubiquitin-modified lysine residues were identified using Mascot search engine (reference sequences: Uniprot, DLG4_MOUSE, Q62108 and DLG4_RAT, P31016). The following three protocols were conducted. Protocol 1: HEK293 cells were transiently transfected with the MS-3xFLAG-Nedd8-AGG and MS-3xFLAG-Nedd8-LGG constructs (Nedd8 mutants) and mouse PSD-95. After 24-36 h, cells were treated with 20 μM MG-132 for 6 h. In this set of experiments the protein sequence coverage was 86% and included 31 of the 38 PSD-95 lysine residues (residues K10, K11, K165, K168, K503, K588, K705 were not represented). Protocol 2: HEK293 cells were transiently transfected with 3xFLAG-Nedd8 (wild-type) and mouse PSD-95 constructs. 24-36 h later cells were treated with Vehicle (DMSO) or 1 μM MLN-4924 for 6 h. In order to increase the stringency of the experiment MG-132 treatment was omitted. For the first experimental condition (Vehicle) 87% coverage of the PSD-95 protein sequence was achieved and 28 of the 38 lysine residues were detected (residues K10, K11, K165, K168, K403, K428, K503, K588, K679, K705, were not represented). For the second experimental condition (1 μM MLN-4924) 85% of the PSD-95 sequence were covered and 27 of 38 lysine residues were identified (K10, K11, K165, K168, K211, K403, K428, K503, K588, K672, K705 were not represented). Protocol 3: Rat primary cortical neurons were treated with 20 μM MG-132 at DIV18 for 6 h. 81.5% protein sequence coverage and 32 of 38 lysine residues were identified (residues K11, K165, K245, K403; K503; K705 were not represented). ESI-MS-MS spectra of the modified PSD-95 peptide by GG and by GGL at K202 (g). In case of the modification by the GG tag, the y-ion series from y4 to y11 and the b-ion series from b9 to b11 undergo a +114.04 Da mass shift. In case of the modification by the GGL-tag, the y-ion series from y4 to y9 undergo a +227 Da mass shift.

Data are presented as mean ± s.e.m. Representative images from three experiments are shown in a. Images and quantifications in be represent data from two experiments. Each experimental protocol presented in f and g was repeated at least twice. Individual t-values, degrees of freedom and F-values are available in the Supplementary Methods Checklist. Full-length blots are presented in Supplementary Figure 10.

Supplementary Figure 7 PSD-95 K202R shows reduced neddylation and impairs spine maturation.

(a,b) PSD-95 K202R shows reduced neddylation under non-denaturing (a) and denaturing (b) conditions. HEK293 cells were transiently transfected with the indicated plasmids for 24 h and then treated with DMSO or 1 μM MLN-4924 as indicated for 6 h. Cell extracts were immunoprecipitated under non-denaturing conditions with α-PSD-95 antibodies and analyzed by immunoblotting for the indicated proteins (a). Neddylation of PSD-95 K202R is reduced by ~60% compared to PSD-95 WT. The ratio of signal intensity of neddylated PSD-95 (indicated by the asterisk) was divided by the signal intensity of total precipitated PSD-95 (middle panel) (one-way ANOVA; P = 0.0003; Bonferroni post-hoc test; **P < 0.01 vs. PSD-95 WT; PSD-95 WT: 1.00 ± 0.08; PSD-95 WT MLN-4924: 0.05 ± 0.05; PSD-95 K202R: 0.46 ± 0.09; n = 3 samples from 3 cell cultures). The asterisk indicates the specific neddylated PSD-95 band, which was further corroborated by mass spectrometry analysis. The other bands around 100 kDa represent co-precipitating proteins associated with PSD-95. Reduced neddylation of PSD-95 (PDZ1+2+3) K202R construct (b). Cell extracts were purified under denaturing conditions with Ni-NTA beads and analyzed by immunoblotting for the indicated proteins. (c) Molecular replacement of endogenous PSD-95 with PSD-95 K202R impairs spine maturation. Mouse hippocampal neurons were transfected with mRFP and the indicated plasmids at DIV12–14. For molecular replacement, shRNA-resistant versions of PSD-95 WT-EGFP and PSD-95 K202R-EGFP were expressed. The number of spines was analyzed at DIV18–19 (one-way ANOVA; P < 0.0001; Bonferroni post-hoc test; **P < 0.01 vs. Control and shRNA Control; ##P < 0.01 vs. shRNA PSD-95 + PSD-95 WT; Number of spines per 10 μm dendrite segment: Control: 9.24 ± 0.43; shRNA Control: 9.74 ± 0.30; shRNA PSD-95: 5.92 ± 0.28; shRNA PSD-95 + PSD-95 WT: 10.21 ± 0.33; sRNA PSD-95 + PSD-95 K202R: 6.47 ± 0.46; n = 12 neurons).

Data are presented as mean ± s.e.m. Images and quantifications represent data from three experiments. F-values are available in the Supplementary Methods Checklist.

Supplementary Figure 8 PSD-95 K202R shows normal molecular behavior in synaptic localization, clustering, interaction with stargazing and ubiquitylation.

(a) PSD-95 K202R is trafficked to and clusters within dendritic spines similar to PSD-95 WT. Mouse hippocampal neurons were transfected with mRFP and PSD-95-WT-EGFP or PSD-95-K202R-EGFP plasmids at DIV14 and analyzed at DIV15. The insets show higher magnifications of spines containing PSD-95 puncta (green). Scale bar, 5 μm. (b) The diffusion properties and trafficking of PSD-95 K202R to the spine are preserved. Analysis of the mobility properties of PSD-95-WT and K202R-EGFP in DIV14 rat hippocampal neurons by FRAP. PSD-95-WT-EGFP, immobile fraction 25.13 ± 0.25%, halftime recovery after photobleaching 3.77 ± 0.49 s, n = 4 cells (12 dendritic regions + 24 spines); PSD-95-K202R-EGFP, immobile fraction 21.46 ± 0.22%, halftime recovery after photobleaching 3.03 ± 0.38 s, n = 6 cells (21 dendritic regions + 42 spines). The recovery after photobleaching in dendrite shaft regions and dendritic spines for the two conditions were compared by calculating the area under the curve for each region of interest. Areas under the curves were compared using two-tailed unpaired Student's t-test (dendrite shaft P = 0.54; dendritic spine P = 0.15). (c) PSD-95 K202R clusters ion channels in COS-7 cells similar to PSD-95 WT. Clustering of PSD-95 and the potassium channel subunit Kv1.4 was analyzed in COS-7 cells transfected with the indicated plasmids. The ratio of the total area of clusters divided by the total area of the cell was calculated (one-way ANOVA; P < 0.0001; Bonferroni post-hoc test; **P < 0.01 for PSD-95 + ctrl vs. PSD-95 WT + Kv1.4; PSD-95 + ctrl vs. PSD-95 K202R + Kv1.4; Kv1.4 + ctrl vs. PSD-95 WT + Kv1.4; Kv1.4 + ctrl vs. PSD-95 K202R + Kv1.4; Cluster area / cell area: PSD-95 + ctrl: 0.006 ± 0.002; Kv1.4 + ctrl: 0.009 ± 0.007; PSD-95 WT + Kv1.4: 0.219 ± 0.027; PSD-95 K202R + Kv1.4: 0.243 ± 0.046; n = 24 cells). Scale bar, 25 μm. (d) The interaction of PSD-95 K202R with Stargazin is preserved. COS-7 cells were transiently transfected with the indicated plasmids for 36 h. Whole cell extracts were immunoprecipitated with α-PSD-95 antibodies and analyzed by immunoblotting with α-Stargazin and α-PSD-95 antibodies. (e) No effect of K202R mutation on the binding of PSD-95 to Stargazin. On the left: Scheme illustrating the FRET system between Stargazin and PSD-95. mCherry (fluorescence acceptor) was inserted at position −21 in the cytosolic tail of Stargazin, thus preserving the PDZ domain-binding motif. EGFP (fluorescence donor) was introduced between the second and third PDZ domains of PSD-95. In order to reduce interference with the native interaction, the two fluorescent proteins were incorporated in regions that lack secondary structure. Adapted from Sainlos et al., 2011. On the right: The interaction between PSD-95-WT-EGFP and PSD-95-K202R-EGFP and Stargazin-mCherry was analyzed in DIV13 rat hippocampal neurons by FRET. The K202R mutation does not affect GFP lifetime neither at extrasynaptic nor at synaptic sites (two-tailed unpaired Student's t-test; Extrasynaptic: P = 0.12; GFP lifetime (ns): PSD-95 WT: 2.42 ± 0.01; PSD-95 K202R: 2.46 ± 0.02; Synaptic: P = 0.26; GFP lifetime (ns): PSD-95 WT: 2.17 ± 0.02; PSD-95 K202R: 2.24 ± 0.06; PSD-95-WT-EGFP: n = 9 cells, 89 spines; PSD-95-K202R-EGFP: n = 7 neurons, 72 spines). (f) Ubiquitylation of PSD-95 is not affected by MLN-4924 or the K202R mutation. Cell lysates from transiently transfected HEK293 cells were immunoprecipitated with α-PSD-95 antibodies and analyzed by Western blotting as indicated. The arrowhead indicates ubiquitylated PSD-95 bands; the asterisk marks an unspecifically bound ubiquitylated protein.

Data are presented as mean ± s.e.m. and box-and-whisker min. and max. plots. Images and quantifications represent data from three experiments. Individual t-values, degrees of freedom and F-values are available in the Supplementary Methods Checklist.

Supplementary Figure 9 Generation and characterization of conditional Nae1CamKIIα-CreERT2 mice.

(a) Scheme of the targeting strategy. A gene trap cassette, consisting of an engrailed 2 splice acceptor (En2 SA)-IRES-LacZ sequence followed by a pGK promoter-neomycin-resistance gene construct, was inserted by homologous recombination in JM8A3.N1 ES cells into intron 3 of mouse Nae1 gene, leading to disruption of the Nae1 gene and to splicing of exon 3 of Nae1 to the SA-IRES-LacZ sequence. Correct targeting of the Nae1 locus was confirmed by long range PCRs. The whole gene trap cassette is flanked by frt sites for FLP-mediated excision, resulting in the floxed Nae1 allele, in which the critical exon 4 of Nae1 is flanked by loxP sites. Cre-mediated excision of the floxed exon 4 results in a frameshift and a premature stop codon in exon 5. The inducible conditional KO of Nae1 in principal neurons of the forebrain was generated by breeding the floxed Nae1 mice to CamKII αCreERT2 mice. Genotyping primers are indicated as arrows in the construct. (bd) Molecular characterization of Nae1CamKIIα-CreERT2 mice. Genotyping PCRs on genomic DNA for the Nae1 locus with Nae1-for and Nae1-rev primers (left panel) and for CamKIIαCreERT2 with iCre-for1, iCre-rev1 and 2 primers (right panel) (b). Scheme illustrating the tamoxifen-inducible deletion of Nae1 in adult mice (c). For all experiments, Nae1Control and Nae1CamKIIα-CreERT2 mice received tamoxifen-food pellets during postnatal week 7/8 and the analysis was performed 3–4 weeks later during post-natal week 11/12. Confirmation of the inducible and forebrain-specific deletion of exon 4 in Nae1CamKIIα-CreERT2 mice after tamoxifen on the genomic DNA level (d). (e) Tamoxifen application, as described in (c), results in a forebrain-specific inducible KO of Nae1 and in reduced neddylation patterns in Nae1CamKIIα-CreERT2 mice. Protein extracts from brain tissues from Nae1Control and Nae1CamKIIα-CreERT2 mice were subjected to immunoblotting with α-Nae1 and α-Nedd8 antibodies (two-tailed unpaired Student's t-test; Nae1: Cortex: **P = 0.0069; Hippocampus: **P < 0.0001; Cerebellum: P = 0.474; Nedd8: Cortex: **P < 0.0001; Hippocampus: **P = 0.0018; Cerebellum: P = 0.677; Nae1: Control cortex: 1.00 ± 0.09; KO cortex: 0.51 ± 0.09; Control hippocampus: 1.00 ± 0.05; KO hippocampus: 0.37 ± 0.04; Control cerebellum: 1.00 ± 0.04; KO cerebellum: 1.05 ± 0.05; Nedd8: Control cortex: 1.00 ± 0.02; KO cortex: 0.54 ± 0.05; Control hippocampus: 1.00 ± 0.07; KO hippocampus: 0.48 ± 0.07; Control cerebellum: 1.00 ± 0.04; KO cerebellum: 1.03 ± 0.06; n = 4 animals per group). (f) Nae1CamKIIα-CreERT2 mice do not show any gross dendritic morphological alterations. DAPI-stained brain sections of Thy1EGFP-Nae1Control and Thy1EGFP-Nae1CamKIIα-CreERT2 mice show no obvious differences in dendritic architecture. Scale bars, 50 μm. (g) Characterization of basal synaptic transmission in Nae1CamKIIα-CreERT2 mice. No changes are observed in the excitability (two-way ANOVA; genotype x FV interaction P = 0.977; genotype P = 0.149; FV P < 0.0001; Control: 0.05 mV: 1.69 ± 0.28; 0.10 mV: 2.56 ± 0.32; 0.15 mV: 3.53 ± 0.42; 0.20 mV: 4.64 ± 0.58; 0.25 mV: 4.98 ± 0.66; 0.30 mV: 5.40 ± 1.50; KO: 0.05 mV: 1.72 ± 0.15; 0.10 mV: 2.84 ± 0.31; 0.15 mV: 4.07 ± 0.44; 0.20 mV: 5.33 ± 0.64; 0.25 mV: 5.44 ± 1.05; 0.30 mV: 6.35 ± 1.37; n = 10 slices from 4 Control animals; n = 11 slices from 4 CKO animals). Only recordings with a detectable response were analyzed, excluding a bias of unhealthy preparations. (hl) Nae1CamKIIα-CreERT2 mice show specific impairments in learning and memory paradigms. Object memory (24 h inter-trial interval) (h) is impaired in Nae1CamKIIα-CreERT2 mice. General locomotor activity, analyzed in the Y-maze test (i) was not changed. The social novelty paradigm (j), freezing behavior induced by a new environment exposure (k) as well as anxiety-related behavior (Dark-light box test) (l) are not affected in Nae1CamKIIα-CreERT2 mice ((hj and l) two-tailed unpaired Student's t-test; (h) Object memory: percentage novel object time: **P = 0.0058; total interaction time: P = 0.481; (i) Y-Maze: P = 0.07; (j) Social novelty: percentage interaction time with novel mouse: P = 0.131; total interaction time: P = 0.621; (l) Dark-light box: Number of entries to lit comp.: P = 0.662; Time in the lit comp.: P = 0.355; Distance travelled in lit comp.: P = 0.451; (k) two-way ANOVA repeated measures; New environment-induced freezing: genotype x time interaction P = 0.403; genotype P = 0.308; time P = 0.214; two-tailed unpaired Student's t-test; New environment-induced freezing: P = 0.253; (h) Distance travelled (m): Control: 13.86 ± 0.96; KO: 11.70 ± 0.59; (i) Novel object time (percentage): Control: 63.45 ± 3.23; KO: 49.57 ± 3.20; Total interaction time with objects (s): Control: 32.57 ± 4.08; KO: 36.88 ± 4.43; (j) Percentage interaction time with novel mouse: Control: 68.65 ± 3.35; KO: 59.32 ± 4.92; Total interaction time (s): Control: 225.90 ± 26.59; KO: 241.30 ± 15.09; (k) Percentage freezing: Control: 1.35 ± 0.77; KO: 0.40 ± 0.26; (l) Number of entries to lit comp.: Control: 12.42 ± 0.99; KO: 11.40 ± 2.08; Time in the lit comp. (s): Control: 71.89 ± 6.11; KO: 58.78 ± 12.46; Distance travelled in lit comp. (m): Control: 5.93 ± 0.59; KO: 4.95 ± 1.13; n = 12 animals per group).

Data are presented as mean ± s.e.m. and box-and-whisker min. and max. plots. Images and quantifications in b and dg represent data from at least three experiments. Individual t-values, degrees of freedom and F-values are available in the Supplementary Methods Checklist. Full-length gels are presented in Supplementary Figure 10.

Supplementary Figure 10 Uncropped full-length pictures of Western blotting membranes and DNA gels.

Uncropped full-length pictures of Western blotting membranes and DNA gels presented in the main figures and supplementary figures. Membranes were often cut to enable blotting for multiple antibodies and to enable blotting for Nedd8-modified targets.

Supplementary information

Supplementary Text and Figures

Supplementary Figures 1–10 and Supplementary Tables 1 and 2 (PDF 10109 kb)

Supplementary Methods Checklist (PDF 801 kb)

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Vogl, A., Brockmann, M., Giusti, S. et al. Neddylation inhibition impairs spine development, destabilizes synapses and deteriorates cognition. Nat Neurosci 18, 239–251 (2015). https://doi.org/10.1038/nn.3912

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1038/nn.3912

This article is cited by

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