Numerous studies have reported that tau travels between neurons, but scientists are unsure how the protein exits the cell. In the February 17 Journal of Experimental Medicine, researchers led by David Holtzman at Washington University in St. Louis provide the first in-vivo evidence that neurons in healthy mice release monomeric tau in response to excitatory electrical activity. Whether pathologic forms of tau hijack this physiologic process to spread through the brain remains to be seen. 

“This is an exciting paper. It ties into the idea that tau propagates along neural circuits, which is what we see in human disease and animal models,” said Amy Pooler at King’s College London. She was not involved in the research, but has reported similar activity-dependent tau release from neurons in vitro (see Aug 2013 conference storyPooler et al., 2013). 

Pathogenic forms of tau are now widely believed to propagate themselves through the brain by slipping from one neuron to another along axonal pathways. This  requires the protein to travel through the extracellular space (see, e.g., Feb 2012 news storyAug 2013 conference story). Holtzman and colleagues previously developed a method to measure monomers of tau in the interstitial fluid (ISF) in the brain by sampling it with a microdialysis probe inserted into the hippocampus of wild-type mice (see Sep 2011 news story). 

In the current paper, first author Kaoru Yamada used the same technique to determine whether neuronal activity regulated ISF tau. Yamada tricked neurons to fire in awake, active mice by infusing potassium through the microdialysis probe. In response, ISF tau shot up two- to threefold over several hours, dropping as tau diffused away from the probe site. Other means of stimulating neurons, such as adding the glutamate receptor agonist NMDA, or triggering glutamate release from presynaptic terminals, produced similar results, demonstrating that excitatory presynaptic activity controlled the phenomenon. “It surprised me that a cytoplasmic protein would be regulated by synaptic activity in this way,” Holtzman told Alzforum. “We still don’t know why this happens.”

Surprisingly, blocking neuronal activity did not noticeably lower ISF tau. The authors wondered if that was because the protein just stuck around for long periods. To explore this hypothesis, Yamada and colleagues used transgenic mice developed by co-authors Eva and Eckhard Mandelkow at the German Center for Neurodegenerative Diseases (DZNE) in Bonn (see Feb 2011 news story). The animals express monomeric human tau under the control of the inducible doxycycline promoter. When the authors switched off tau production, it took 11 days for ISF protein levels to drop by half, confirming that tau clears quite slowly. 

Holtzman's group reported previously that neuronal activity also regulates Aβ release in vivo (see Oct 2003 news storyDec 2005 news storySep 2008 news storyAug 2011 news story). He pointed out that activity affects Aβ less than tau, the former inching up by only 30 or 40 percent. The mechanism may also be different, as Aβ release depends on endosome recycling (see Apr 2008 news story), whereas researchers believe tau escapes the cell by other means, Holtzman said. It is not known if there is any connection between Aβ and tau release.

These experiments did not measure tau aggregates, or examine whether ISF tau was modified or truncated in any way. Other work has suggested that aggregated, hyperphosphorylated, fragmented, or pathologically folded forms of tau spread through the brain (see, e.g., Nov 2012 conference storyFeb 2013 news storyNov 2013 news story). In future studies, Holtzman and colleagues will look more closely at what forms of tau populate the ISF and how they behave. Pooler suggested that scientists could use that information to tailor immunotherapies that mop up pathogenic tau while leaving healthy protein alone (see Sep 2013 news story).—Madolyn Bowman Rogers.


  1. The past few years of tau research have provided strong evidence that endogenous tau secretion does occur in the absence of neuronal lysis. One of the next steps is to understand how a cytosolic protein such as tau is secreted. We and others in the field have defined this secretion as noncanonical, i.e., not secreted through the ER/Golgi network. Pooler et al., working in-vitro, and now Yamada et al. with this in-vivo study, are beginning to provide evidence that tau release is driven by the activity of presynaptic glutamatergic hippocampal neurons. It is interesting to note that although interstitial fluid (ISF) tau is elevated after induction of neuronal hyperactivity, the onset is delayed by hours (as compared to minutes-hours for the glucose/lactate changes in response to stimulation, i.e., K+) and persists long after glucose and lactate levels returned to baseline. This may suggest that increased ISF tau levels are a consequence of aberrant neuronal activity. As such, exploring this adaptive process may begin to help define the mechanism by which tau is secreted. Moving forward, it would be interesting to examine the biological consequences of tau secretion. There is accumulating evidence linking tau to epileptic seizures, for example. It would be interesting to see if secreted tau is linked to the increased seizure activity seen in Alzheimer’s disease.


    . Physiological release of endogenous tau is stimulated by neuronal activity. EMBO Rep. 2013 Apr;14(4):389-94. PubMed.

    View all comments by Irene Griswold-Prenner
  2. We agree with the comment by Irene Griswold-Prenner that given the results by both Pooler et al. and Yamada et al., it will be very interesting to determine whether secreted forms of tau are linked in any way to hyperexcitability and possibly seizures.


    . Physiological release of endogenous tau is stimulated by neuronal activity. EMBO Rep. 2013 Apr;14(4):389-94. PubMed.

    View all comments by David Holtzman
  3. This study presents the most clear-cut demonstration of the secretion of tau protein into the interstitial fluid (ISF) by neurons in intact brain, independent of cell death. The authors report converging evidence, which suggests that secretion of tau could be governed by and may occur as a result of "physiological" or close to physiological processes. Similar to Aβ release demonstrated by the same group (Cirito et al., 2008), secretion of tau is revealed to be considerably enhanced upon depolarization of neurons or under abnormally high network activity, e.g., driven by GABAA inhibitors. Surprisingly, the processes mediating release of Aβ and tau appear to rely on canonical SNARE proteins, known to drive the fusion of synaptic vesicles and neurotransmitter release. Indeed, inactivation of v-SNARE VAMP by tetanus toxin strongly suppresses the secretion of tau and Aβ (Pooler et al., 2013; Cirito et al., 2008). Although the authors suggest that the release of these peptides occurs at pre-synaptic terminals of excitatory synapses, no supportive data is provided. To this reader, it is difficult to imagine that peptides of several kD weight released into the cleft of glutamatergic excitatory synapses would escape so rapidly and with such ease into the ISF through layers of constraining barriers designed to contain molecules far smaller, such as aspartate and glutamate. While ectopic (i.e., out of active zones) or extra-synaptic release of tau and Aβ is an attractive possibility (similar to most of neuropeptides) and seems more feasible (based on reported ISF dynamics), it brings up an important question as to why the spread of both tau and Aβ tends to follow the pattern of neuronal connections in the brain. 


    . Endocytosis is required for synaptic activity-dependent release of amyloid-beta in vivo. Neuron. 2008 Apr 10;58(1):42-51. PubMed.

    . Physiological release of endogenous tau is stimulated by neuronal activity. EMBO Rep. 2013 Apr;14(4):389-94. PubMed.

    View all comments by Saak V. Ovsepian

Make a Comment

To make a comment you must login or register.


News Citations

  1. Tales of Traveling Tau: Is Transfer Between Neurons Normal?
  2. Mice Tell Tale of Tau Transmission, Alzheimer’s Progression
  3. Are Protein Strains The Cause of Different Tauopathies?
  4. Brain Microdialysis Reveals Tau, Synuclein Outside of Cells
  5. Making It Stick—Tau Toxicity Linked to Aggregation Propensity
  6. Soluble Aβ: Getting a Grip on Its Fate
  7. Paper Alert: Synaptic Activity Increases Aβ Release
  8. Soluble Aβ—Bane or Boon? Real-time Data in Humans Yield New Insight
  9. Brain Activity and Aβ—The Interstitial Plot Thickens
  10. Link Between Synaptic Activity, Aβ Processing Revealed
  11. SfN: Tau Toxicity in the Limelight
  12. Truncated Tau Triggers Tangles, Transmits Pathology
  13. Tau Biomarkers—More Informative Fragments in Spinal Fluid?
  14. Antibodies Stop Toxic Tau in Its Extracellular Tracks

Paper Citations

  1. . Physiological release of endogenous tau is stimulated by neuronal activity. EMBO Rep. 2013 Apr;14(4):389-94. PubMed.

Further Reading

Primary Papers

  1. . Neuronal activity regulates extracellular tau in vivo. J Exp Med. 2014 Mar 10;211(3):387-93. Epub 2014 Feb 17 PubMed.