. Site-specific phosphorylation of tau inhibits amyloid-b toxicity in Alzheimer’s mice. Science. 2016 Nov 18; 354(6314):904-8.


Please login to recommend the paper.


  1. This is an interesting study that suggests the field may have to change the way they think about tau phosphorylation in Alzheimer's disease. Tau phosphorylation has historically been thought of as a contributor to neurofibrillary tangle formation and disease-associated tau toxicity, but in this study Ittner et al. identify that phosphorylation of tau on threonine 205 may actually serve a protective role. This is of note as phosphorylation of tau at this site has often been used as a measure of disease pathology and is the recognition site of the commonly used “AT8” tau antibody as well many commonly used reagents.

    The authors identify p38γ, a previously underappreciated isoform of p38, as a tau kinase that preferentially phosphorylates T205. Surprisingly, loss of p38γ enhanced rather than suppressed tau-based toxicity. This could be explained through the observation that phosphorylation of T205 resulted in dissociation of tau/fyn/PSD-95 complexes, which was identified as a mechanism of tau-based neurotoxicity by the authors in a previous study. From a therapeutic perspective, a wealth of data is presented supporting p38γ as the key kinase regulating tau T205 phosphorylation, though unfortunately the development of compounds that act as kinase activators has proven far less tractable than kinase inhibitors. Therefore, more work is likely needed in order to translate these discoveries into development of new treatments for Alzheimer's disease. 

    Overall the study is comprehensive, well-designed, and contains a number of loss-of-function and gain-of-function experiments with both tau and p38γ that solidify their conclusions. It is a nice addition to our understanding of tau function and will surely provide a starting point for a range of future work.

    View all comments by Joseph Lewcock
  2. This is a well-done study. Although the authors have not pointed it out, one very intriguing possible conclusion that can be drawn is that the  Aβ drug trials, especially immunotherapies, continue to be negative because tau in AD brain is hyperphosphorylated at Thr 205, along with several other sites, which protects AD brain from Aβ-induced neurotoxicity and hence removal of Aβ will not have any beneficial therapeutic effect.

    I believe a serious problem with this present study, and for that matter for many other such studies on the etiopathogenic relationship between Aβ and tau pathologies, is the use of highly artificial overexpression/knockout transgenic mice, which can grossly alter the quantitative effects and hence the outcome.

    View all comments by Khalid Iqbal
  3. The hyperphosphorylation of tau has long been proposed to contribute to the tau pathology in Alzheimer’s disease and other tauopathies. However, owing to the number and heterogeneity of phosphorylation sites on tau, investigating the exact role of phosphorylation of tau in neurodegeneration proves to be a challenge. Although hyperphosphorylation of tau is generally regarded as a culprit of neurodegeneration, hyperphosphorylation of tau occurs in other transient situations (e.g., hibernation, stress) without causing lasting side effects. In their recent paper, the Ittner brothers provide evidence that the phosphorylation of tau at Thr205 by P38γ can even be protective because it can suppress excitotoxicity induced by Aβ or PTZ. This is because the phosphorylation at Thr205 of tau disrupts the formation of NMDA-Receptor/PSD-95/tau/Fyn complexes, which mediate Aβ or PTZ induced excitotoxicity. This interesting result expands on the earlier studies of the Ittner and Götz team (Ittner et al., 2010), which assigned a physiological role to the small fraction of tau found in dendrites (which is otherwise mainly axonal). The Thr205 residue is one of several Ser-Pro or Thr-Pro motifs in tau that are targeted by several proline-directed kinases involved in cellular signaling pathways, including those of the JNK family. This type of phosphorylation has been under intense scrutiny, and therefore the current study can be compared to others, leaving several issues to be clarified in the future:

    (1) Mondragon-Rodriguez and colleagues reported that phosphorylation of tau can suppress excitotoxicity and thus can serve as a regulatory mechanism to prevent NMDA receptor overexcitation (Mondragon-Rodriguez et al., 2012). These authors focused on phosphorylation of tau at the phospho-epitopes AT180, AT8, or AT100, also representing Ser-Pro or Thr-Pro sites, which reduces tau's interaction with PSD-95. This is at variance with the current study demonstrating that only phosphorylation at Thr205 reduces the association of tau with PSD-95 and Fyn.

    (2) The current study showed that the kinase P38γ phosphorylates tau at Thr205 and to a lesser extent at Ser199 (included in the phospho-epitope AT8, a widely used antibody to characterize phosphorylated tau), but not at Ser 202 (included in the phospho-epitope AT8) and hardly any at Ser396 and Ser404 (epitopes of PHF1, another commonly used antibody against phospho-tau). On the other hand, several earlier studies (for instance, Buee-Scherrer and Goedert, 2002; Goedert et al., 1997) revealed that P38γ phosphorylates tau not only at AT8 sites, but also strongly at the PHF1-epitope pSer396/pSer404. It is not clear whether this discrepancy is due to the difference of antibodies used in these studies, since the current Ittner paper used antibodies with epitopes of a single phosphorylation site (pT205, pS202 etc.), while the other studies used antibodies against a combination of multiple phosphorylation sites (e.g. AT8, PHF1).

    The present study also shows that the P38γ level is reduced in old APP23 mice but not in young APP23 mice (Fig.S10), compared to wild-type mice. On the other hand, the APP23 mice display enhanced epileptiform activity at four months, when there is no P38γ reduction. Therefore, there must be mechanisms independent of P38γ underlying the Aβ-induced hyperexcitotoxicity in the young APP mice whose nature deserves attention in future experiments.


    . Phosphorylation of microtubule-associated protein tau by stress-activated protein kinases in intact cells. FEBS Lett. 2002 Mar 27;515(1-3):151-4. PubMed.

    . Phosphorylation of microtubule-associated protein tau by stress-activated protein kinases. FEBS Lett. 1997 Jun 2;409(1):57-62. PubMed.

    . 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.

    . 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.

    View all comments by Yipeng Wang
  4. This is a very interesting paper. I find this study even more intriguing because pT205 is one of the two epitopes recognized by the AT8 antibody, whose staining has been shown to be increased in tauopathy brains in a plethora of studies. I wonder what the authors' opinion on this issue is.

    View all comments by Javier Moron-Oset
  5. This paper is a very important step toward identifying the kinases and site-specific post-translational events that determine tau toxicity.

    These findings are highly suggestive that tau hyperphosphorylation is not the whole story in terms of AD pathogenesis. Furthermore, this study suggests that complex site-specific modifications of tau could either promote or prevent toxicity. This concept could have major implications for how we therapeutically target tau in AD patients.  

    View all comments by Todd Cohen

Make a Comment

To make a comment you must login or register.

This paper appears in the following:


  1. Is Tau Phosphorylation All Bad?