In Alzheimer’s brains, the axonal protein tau strays into dendritic spines and stirs up trouble at synapses. Researchers have called this a mislocalization, but a paper in the April 23 Journal of Neuroscience challenges this view, suggesting that tau may also perform a physiological function at the synapse. Researchers led by Alain Buisson at INSERM-Joseph Fourier University in Grenoble, France, report that tau in wild-type neurons moves into dendritic spines in response to electrical stimulation. There, the protein interacts with the actin cytoskeleton and may actively participate in the spine remodeling that underlies synaptic plasticity, the authors suggest. By contrast, in neurons treated with Aβ, tau idles in spines, and synaptic stimulation then causes them to collapse. “Somehow, Aβ disrupts the link between activity and tau translocation, as well as the plasticity potential of synapses,” Buisson told Alzforum.

The findings are surprising, said Dezhi Liao at the University of Minnesota, Minneapolis. He was not involved in the work. “The data raise a lot of new questions. What is tau doing at the synapse? How do the normal process and the disease process interact?” Future studies should look more closely at whether synaptic tau benefits or harms the cell, and how phosphorylations affect its behavior, Liao suggested.

Previous research has established that Aβ causes the migration of tau into dendritic spines. Various studies have found that this tau can dampen synaptic signaling (see Jan 2011 news story), mediate excitotoxicity through the kinase Fyn (see Jul 2010 conference story), and abolish spines (see Sep 2010 news story). On the other hand, some recent studies have hinted at a role for the protein in normal synaptic function as well. Lack of tau contributes to spine loss in cultured neurons (see Chen et al., 2012), and tau has been found to interact with synaptic proteins such as NMDA receptors in wild-type cells (see Mondragon-Rodriguez et al., 2012). 

To further investigate tau’s role at the synapse, first author Marie Lisa Frandemiche transfected primary cortical neurons from mice with fluorescently-tagged tau. Under resting conditions, tau occupied the shafts of axons and dendrites, but avoided dendritic spines. After neurons were depolarized by a pharmacological agent, the tau content of spines jumped up 50 percent (see image below). Several other synaptic proteins, including the cytoskeletal protein actin, the kinase Fyn, PSD-95, and the AMPA receptor subunit GluA1, also moved into spines. The authors saw a similar response when they induced long-term potentiation (LTP) in hippocampal slices; synaptic levels of all these proteins increased twofold.

Translocating tau: Tau (green) in the dendritic shaft avoids dendritic spines (red) until neuronal stimulation drives it in (arrowed yellow dot). [Image courtesy of, and with permission of, Frandemiche et al., The Journal of Neuroscience, 2014.]

What might tau be doing at synapses? Since tau binds and stabilizes actin filaments, the authors looked for a relationship between these two proteins. They treated cultured neurons with a compound that induces actin to polymerize, and found that tau migrated into spines just as it did after neuronal stimulation. On the other hand, treatment with a compound that breaks apart actin filaments abolished synaptic tau. The data suggest that tau responds to changes in actin, Buisson noted. As spines mature to store memories, the cytoskeleton remodels and tau may participate in this process, Buisson said. 

To investigate how Alzheimer’s disease changes the picture, the authors added 100 nM synthetic Aβ oligomers to the cultures. Tau, actin, Fyn, PSD-95, and GluA1 migrated into spines just as they did after synaptic stimulation. However, subsequent neuronal stimulation suppressed synaptic levels of tau and actin, and spines shrank.

What accounts for the difference in tau’s behavior in the presence and absence of Aβ? The authors found that the protein is phosphorylated differently in the two conditions. After neuronal stimulation, residue Thr205 became phosphorylated, but after Aβ exposure, Thr205 phosphorylation dropped while Ser404 phosphorylation surged. The difference may affect tau’s ability to bind actin. When the authors blocked Thr205 phosphorylation, tau moved out of the synapse more easily, suggesting that less of it was bound to the cytoskeleton. Blocking Ser404 phosphorylation, on the other hand, prevented Aβ from driving tau to spines. In ongoing work, the authors are examining in vitro how various tau phosphorylations affect its ability to bind actin.

The data help flesh out the model of how tau mediates Aβ’s synaptotoxicity. They largely agree with previous studies, commentators said. For example, Liao and colleagues recently reported that Aβ’s ability to induce the internalization of glutamatergic receptors depends on the phosphorylation state of dendritic tau (see Miller et al., 2014). However, more research will be needed to determine what function tau performs at the synapse, commentators agreed. Hans Zempel at the German Center for Neurodegenerative Diseases (DZNE) in Bonn pointed out that the GFP-tagged tau in this study was overexpressed, leaving unclear whether endogenous tau would behave in the same way. “It is still a matter of debate whether endogenous tau plays a physiological role in dendrites and in spine plasticity,” he wrote to Alzforum (see full comment below). 

Others noted that the findings support previous work linking dendritic tau to the excessive neuronal excitation induced by Aβ (see Aug 2013 news storyDec 2013 conference story). Erik Roberson at the University of Alabama at Birmingham wrote, “Postsynaptic tau in spines may play a role in controlling susceptibility to hyperexcitation and epileptiform activity and/or in regulating long-term depression (LTD), both of which are altered in tau knockout mice.” Dendritic tau seems less likely to play a role in LTP, since tau knockouts are normal in this regard, he added (see full comment below).—Madolyn Bowman Rogers.


  1. This is a very interesting and elegant study that adds support for the idea that tau is normally present in dendrites and plays a physiological role in normal synaptic functioning. The movement of tau into spines with neuronal activity is demonstrated in both dissociated cultured neurons and hippocampal slices, and by both imaging and biochemical approaches. The implications of this change in tau localization, and its effects on postsynaptic function, will be an important focus for future studies. It seems unlikely that this postsynaptic tau plays a critical role in long-term potentiation (LTP), given that multiple studies have found that tau knockout mice have normal hippocampal LTP (Roberson et al., 2011; Shipton et al., 2011). However, postsynaptic tau in spines may play a role in controlling susceptibility to hyperexcitation and epileptiform activity and/or in regulating long-term depression, both of which are altered in tau knockout mice (Roberson et al., 2007, 2011; Ittner et al., 2011; Holth et al., 2013; DeVos et al., 2013; Kimura et al., 2013).

    Tau in spines may also play an important role in disease, of course.  Another interesting aspect of this study is the proposition that there are differences between the tau translocation into spines induced by Aβ and the tau translocation into spines induced by synaptic activity. This suggests that tau in spines is detrimental only under certain circumstances, and it will be important to further elucidate the differences between physiological and pathological tau in spines.


    . Amyloid-β/Fyn-induced synaptic, network, and cognitive impairments depend on tau levels in multiple mouse models of Alzheimer's disease. J Neurosci. 2011 Jan 12;31(2):700-11. PubMed.

    . Tau protein is required for amyloid {beta}-induced impairment of hippocampal long-term potentiation. J Neurosci. 2011 Feb 2;31(5):1688-92. PubMed.

    . Reducing endogenous tau ameliorates amyloid beta-induced deficits in an Alzheimer's disease mouse model. Science. 2007 May 4;316(5825):750-4. PubMed.

    . Amyloid-β/Fyn-induced synaptic, network, and cognitive impairments depend on tau levels in multiple mouse models of Alzheimer's disease. J Neurosci. 2011 Jan 12;31(2):700-11. PubMed.

    . Dendritic function of tau mediates amyloid-beta toxicity in Alzheimer's disease mouse models. Cell. 2010 Aug 6;142(3):387-97. Epub 2010 Jul 22 PubMed.

    . Tau loss attenuates neuronal network hyperexcitability in mouse and Drosophila genetic models of epilepsy. J Neurosci. 2013 Jan 23;33(4):1651-9. PubMed.

    . Antisense Reduction of Tau in Adult Mice Protects against Seizures. J Neurosci. 2013 Jul 31;33(31):12887-97. PubMed.

    . Microtubule-associated protein tau is essential for long-term depression in the hippocampus. Philos Trans R Soc Lond B Biol Sci. 2014 Jan 5;369(1633):20130144. Print 2014 Jan 5 PubMed.

    View all comments by Erik Roberson
  2. I enjoyed reading this paper. The study adds to the notion that a prime function of tau is that of a scaffolding protein, not only in the axon but also in the dendrite where tau interacts, as this study shows, also with filamentous actin. Interesting is the role specific phosphorylation reactions have in activity—and Aβ-dependent tau localization. While tau is often perceived as being either normally phosphorylated or hyperphosphorylated, this study underscores the notion that there is a physiological role for distinct phosphorylation sites (see Fig. 9). I am convinced that the future will see many more studies into tau trafficking and how it is regulated by site-specific phosphorylation.

    View all comments by Jürgen Götz
  3. This study from the Buisson lab adds more data arguing that tau may contribute to the regulation of synaptic plasticity. It has long been known that tau can bind to f-actin, particularly when phosphorylated at the KXGS-motifs (Mandelkow et al., 2004; Whiteman et al., 2009), and that it can enter spines when overexpressed (Thies et al., 2007; Zempel et al., 2013). When the authors activate synapses or add jasplakinolide, a substance that increases the f-actin content, more tau invades the spines, presumably because the spine volume and f-actin content increase.

    The interesting part is the potential regulation of this process, which may rely on phosphorylation: The lab of Kwangwook Cho at the University of Bristol, England, has previously shown that tau is essential for regulating long-term depression (LTD), an important form of plasticity, and that LTD comes along with phosphorylation of tau at the PHF-1 site (Kimura et al., 2014). When the authors of this study mutated tau on serine 404 (which is part of the PHF-1 epitope when phosphorylated) to alanine, tau did not enter spines anymore upon synaptic activation (but remained in the dendrite as a result of overexpression), indicating that this phosphorylation site might be important for tau to enter spines. Our studies have shown that tau can enter spines and greatly accumulates there when pseudophosphorylated at the KXGS-motifs (by mutating to KXGE), and that this leads to a disassembly of f-actin in spines and spine decay (Zempel et al., 2013). Thus, sequential or combined phosphorylation at the KXGS motifs (Zempel et al., 2013) and at S404/PHF-1 motif (Kimura et al., 2014; this paper) might enable tau to translocate into spines and cause LTD.

    The majority of the Alzheimer’s research field would probably accept that dendritic tau is a pathological sign. Therefore the caveat of the study is that it is still a matter of debate whether endogenous tau plays a physiological role in dendrites and in spine plasticity. Tau is only present in significant amounts in dendrites when overexpressed (this paper) or when endogenous tau is pathologically missorted. Endogenous tau and in particular tau phosphorylated at the PHF-1 epitope predominantly resides in the axon, even in conditions that acutely induce pathological missorting of tau (Kaech and Banker, 2005; Zempel et al., 2010).


    . MARK/PAR1 kinase is a regulator of microtubule-dependent transport in axons. J Cell Biol. 2004 Oct 11;167(1):99-110. PubMed.

    . Activated actin-depolymerizing factor/cofilin sequesters phosphorylated microtubule-associated protein during the assembly of alzheimer-like neuritic cytoskeletal striations. J Neurosci. 2009 Oct 14;29(41):12994-3005. PubMed.

    . Missorting of tau in neurons causes degeneration of synapses that can be rescued by the kinase MARK2/Par-1. J Neurosci. 2007 Mar 14;27(11):2896-907. PubMed.

    . Amyloid-β oligomers induce synaptic damage via Tau-dependent microtubule severing by TTLL6 and spastin. EMBO J. 2013 Nov 13;32(22):2920-37. PubMed.

    . Microtubule-associated protein tau is essential for long-term depression in the hippocampus. Philos Trans R Soc Lond B Biol Sci. 2014 Jan 5;369(1633):20130144. Print 2014 Jan 5 PubMed.

    . Culturing hippocampal neurons. Nat Protoc. 2006;1(5):2406-15. PubMed.

    . Abeta oligomers cause localized Ca(2+) elevation, missorting of endogenous Tau into dendrites, Tau phosphorylation, and destruction of microtubules and spines. J Neurosci. 2010 Sep 8;30(36):11938-50. PubMed.

    View all comments by Hans Zempel
  4. The authors have published a clear and well-performed study on the role of dendritic tau. Several years ago, the dogma was that tau is an intracellular, axonal protein. Now, we know about the multiple faces of tau. Tau could be inside or outside a neuron, or located at the axon or the dendrites. To explain those different features, it has been suggested that different tau isoforms may play different functions (Ittner et al., 2011; Liu and Götz., 2013). Focusing on dendritic tau, the French team report that synaptic activation and Aβ induces translocation of tau to dendritic spines in different ways, and that this difference relates to tau phosphorylation profiles.

    Previously, it was proposed that Aβ-induced tau phosphorylation changes the morphology or even ablates previously assembled spines (Merino-Serrais et al., 2013). This possibility should be further analyzed. 


    . Dendritic function of tau mediates amyloid-beta toxicity in Alzheimer's disease mouse models. Cell. 2010 Aug 6;142(3):387-97. Epub 2010 Jul 22 PubMed.

    . Profiling murine tau with 0N, 1N and 2N isoform-specific antibodies in brain and peripheral organs reveals distinct subcellular localization, with the 1N isoform being enriched in the nucleus. PLoS One. 2013;8(12):e84849. Epub 2013 Dec 30 PubMed.

    . The influence of phospho-τ on dendritic spines of cortical pyramidal neurons in patients with Alzheimer's disease. Brain. 2013 Jun;136(Pt 6):1913-28. PubMed.

    View all comments by Jesus Avila

Make a Comment

To make a comment you must login or register.


News Citations

  1. Tau’s Synaptic Hats: Regulating Activity, Disrupting Communication
  2. Honolulu: The Missing Link? Tau Mediates Aβ Toxicity at Synapse
  3. The Plot Thickens: The Complicated Relationship of Tau and Aβ
  4. In Adult Mice, Reduced Tau Quiets Agitated Neurons
  5. Is Dendritic Tau to Blame for AD-Related Hyperexcitability?

Paper Citations

  1. . Tau protein is involved in morphological plasticity in hippocampal neurons in response to BDNF. Neurochem Int. 2012 Feb;60(3):233-42. PubMed.
  2. . Interaction of Endogenous Tau Protein with Synaptic Proteins Is Regulated by N-Methyl-D-aspartate Receptor-dependent Tau Phosphorylation. J Biol Chem. 2012 Sep 14;287(38):32040-53. PubMed.
  3. . Tau phosphorylation and tau mislocalization mediate soluble Aβ oligomer-induced AMPA glutamate receptor signaling deficits. Eur J Neurosci. 2014 Apr;39(7):1214-24. PubMed.

Further Reading


  1. . The dendritic hypothesis for Alzheimer's disease pathophysiology. Brain Res Bull. 2014 Apr;103:18-28. Epub 2013 Dec 12 PubMed.

Primary Papers

  1. . Activity-dependent tau protein translocation to excitatory synapse is disrupted by exposure to amyloid-beta oligomers. J Neurosci. 2014 Apr 23;34(17):6084-97. PubMed.