Morris M, Knudsen GM, Maeda S, Trinidad JC, Ioanoviciu A, Burlingame AL, Mucke L.
Tau post-translational modifications in wild-type and human amyloid precursor protein transgenic mice.
Nat Neurosci. 2015 Aug;18(8):1183-9. Epub 2015 Jul 20
PubMed.
This is a comprehensive body of work that has identified new sites of tau post-translational modification in wild-type and APP transgenic mice. The new sites include tau methylation, di-methylation, acetylation, and ubiquitination (inferred from GlyGly-modified lysine), as summarized in Figure 1b. This wide-ranging approach to investigating tau modification brings up the interesting point that ubiquitination could provide a highly sensitive system for regulation of tau function through competition at lysine residues, as suggested previously.
It is notable that O-GlcNAc modification of tau was barely detectable in the mice, even following lectin enrichment. This result is consistent with our own unpublished data, as we were unable to detect this modification in postmortem control or Alzheimer human brain tissue by mass spectrometry.
The observation of differential phosphorylation of tau in the postsynaptic density (PSD) of wild-type mice is interesting, since this could, as the authors suggest, affect the regulation or function of tau in this location. The authors found an increase in the amount of doubly phosphorylated tau 386-404 in the PSD and a decrease in the triply phosphorylated tau peptide, relative to unfractionated brain lysate. The potential sites for phosphorylation in this peptide include residues T386, Y394, S396, S400, T403, and S404. Since there is apparently no difference in tau phosphorylation in the PSD of wild-type and APP transgenic mice, and PSD immunoreactivity with PHF1 antibody (S396/S404) is unchanged, regulated phosphorylation/dephosphorylation of at least one of the serine/threonine/tyrosine residues T386, Y394, S400, T403 in this C-terminal region of tau could be critical for PSD localization.
An important take-home message from this study is that differential post-translational modification of tau is likely to direct its binding to other proteins and to be responsible for regulating tau subcellular localization. However, it remains to be seen which tau modifications, either individually or in combination, might be involved in the pathogenesis of neurodegenerative disease.
This is a heroic effort to characterize post-translational modifications of tau in the J20 human APP mouse. Two things are really notable to me:
1. The presence of the mutated human APP gene does not alter tau modification—whether that be phosphorylation, acetylation, methylation, or other. There are no differences between the J20 mice and the wild-type. This is despite the fact that at the age examined, the J20 "show synaptic, network, and behavioral abnormalities." These abnormalities are evidently not due to post-translational modification of tau, unless this is highly regional or a very minor component of the total tau present. Even this latter "escape clause" appears to be ruled out by the authors' regional and subcellular fraction analyses. And yet, if tau levels are reduced, J20 mice are protected from the synaptic, network and behavioral abnormalities.
2. Although not mentioned in the paper, the authors provide no evidence for changes in the phosphorylation of tyrosine in tau, especially tyrosine 18, and I am sure that they looked long and hard for this. Several reports from Mucke's group and from other labs have implicated the tyrosine kinase fyn in responses to Aβ, and fyn phosphorylates tyrosine 18 of tau. This phosphorylation does not seem to have been detected in either wild type or J20 mice.
Much of the talk about tau modification has been as much wishful thinking as anything. Especially with immunocytochemistry, it is possible to detect changes in tau phosphorylation that might be entirely insignificant in quantitative terms. We really need the kind of rigorous analysis the authors have done here. This is a great paper and will be discussed at much length in tau circles.
This is an extensive study on post-translational modifications of tau by mass spectrometry analysis in wild-type (WT) and hAPP transgenic mouse brains. Most of the phosphorylation sites previously reported in postmortem normal human brains were also found in the present study in WT and hAPP transgenic mice. However, O-GlcNAcylation at only site Ser400 was detected in WT and hAPP mice. This finding led the authors to question an extensive O-GlcNAcylation of tau.
There are no stoichiometric data shown for any of the post-translational modifications found in murine tau isoforms. However, it appears that with their methodology the authors were probably able to detect substoichiometric phosphorylations; the same does not appear to be true for O-GlcNAcylation. That may be why they found only one O-GlcNAcylated residue. Substoichiometric O-GlcNAcylations could have been hydrolyzed by the 1 percent perchloric acid used to homogenize the brains for the isolation of tau.
Of course, the overexpression hAPP transgenic mice are a highly artificial model generated to produce Aβ plaques but do not replicate the mechanism by which similar pathology is produced in AD. Also, hAPP transgenic mice do not produce tau pathology. In fact, the discrepancy in tau O-GlcNAcylation between the previous human studies and the present mouse findings reinforces the limitations of overexpression transgenic mouse models. Hopefully Dr. Mucke and his team will use the mass spectrometry methodology they have developed to analyze post-translational modifications of tau in AD and normal human brains.
The issues surrounding protein tau remain among the toughest nuts to crack in normal brain physiology, and even more in brain pathology of AD and FTDP-17, as well as in the 55 or so known tauopathies. It takes the best hands and minds and most sophisticated tools to dig up solid data on which to base insight in—and corrections of—previous data, as well as provide firm ground and directions for future work.
Lennart Mucke and friends at UCSF and Hopkins have done just that: a comprehensive mass-spec study of mouse tau in wild-type and human APP transgenic mouse brain, compiling an unprecedented number of post-translational modifications at more sites than we knew—or suspected. The study highlights the complex post-translational biochemistry of protein tau in vivo, and albeit “only” in mouse brain, it is the only organ/model that matters. We write 2015, i.e., 40 years after the discovery of a "protein bound to micro-Tubules" (Murray et al., 1975) (a period that spans my entire postdoc/professional career ...).
We have reported beneficial effects of the pharmacological increase of O-Glc-NAc-ylation on brain proteins (several hundreds!) but contradicted a defining role of this modification on protein tau itself and/or on its phosphorylation (Borghgraef et al, 2013). We were unable to confirm even minimal O-Glc-NAc-ylation of protein tau by western blotting with antibodies available commercially or provided by the authors of the original studies (Yuzwa et al., 2008; Yuzwa et al., 2011).
Hence, we maintain what we stated in 2013 (and presented at many a meeting): "We conclude that increasing O-GlcNAc-ylation of brain proteins improved the clinical condition and prolonged the survival of ageing Tau.P301L mice, but not by direct biochemical action on protein tau. The pharmacological effect is proposed to be located downstream in the pathological cascade initiated by protein Tau.P301L, opening novel venues for our understanding, and eventually treating the neurodegeneration mediated by protein tau” (Borghgraef et al., 2013).
References:
Murray JJ, Reed PW, Fay FS.
Contraction of isolated smooth muscle cells by inophore A23187.
Proc Natl Acad Sci U S A. 1975 Nov;72(11):4459-63.
PubMed.
Liu F, Iqbal K, Grundke-Iqbal I, Hart GW, Gong CX.
O-GlcNAcylation regulates phosphorylation of tau: a mechanism involved in Alzheimer's disease.
Proc Natl Acad Sci U S A. 2004 Jul 20;101(29):10804-9.
PubMed.
Yuzwa SA, Macauley MS, Heinonen JE, Shan X, Dennis RJ, He Y, Whitworth GE, Stubbs KA, McEachern EJ, Davies GJ, Vocadlo DJ.
A potent mechanism-inspired O-GlcNAcase inhibitor that blocks phosphorylation of tau in vivo.
Nat Chem Biol. 2008 Aug;4(8):483-90.
PubMed.
Arnold CS, Johnson GV, Cole RN, Dong DL, Lee M, Hart GW.
The microtubule-associated protein tau is extensively modified with O-linked N-acetylglucosamine.
J Biol Chem. 1996 Nov 15;271(46):28741-4.
PubMed.
Yuzwa SA, Shan X, Macauley MS, Clark T, Skorobogatko Y, Vosseller K, Vocadlo DJ.
Increasing O-GlcNAc slows neurodegeneration and stabilizes tau against aggregation.
Nat Chem Biol. 2012 Apr;8(4):393-9.
PubMed.
Borghgraef P, Menuet C, Theunis C, Louis JV, Devijver H, Maurin H, Smet-Nocca C, Lippens G, Hilaire G, Gijsen H, Moechars D, Van Leuven F.
Increasing brain protein O-GlcNAc-ylation mitigates breathing defects and mortality of Tau.P301L mice.
PLoS One. 2013;8(12):e84442. Epub 2013 Dec 23
PubMed.
Liu F, Shi J, Tanimukai H, Gu J, Grundke-Iqbal I, Iqbal K, Gong CX.
Reduced O-GlcNAcylation links lower brain glucose metabolism and tau pathology in Alzheimer's disease.
Brain. 2009 Jul;132(Pt 7):1820-32.
PubMed.
Yuzwa SA, Yadav AK, Skorobogatko Y, Clark T, Vosseller K, Vocadlo DJ.
Mapping O-GlcNAc modification sites on tau and generation of a site-specific O-GlcNAc tau antibody.
Amino Acids. 2011 Mar;40(3):857-68. Epub 2010 Aug 13
PubMed.
This manuscript by Morris and colleagues describes the systematic mapping of the post-translational modifications of endogenous mouse tau isolated from wild-type and hAPP transgenic mice by mass spectrometry. Besides revealing the similarity between these two groups, the identification of a substantial number of post-translational modifications beyond phosphorylation provides a more holistic picture of the tau protein. Although phosphorylation constitutes the main type of modification, lysine acetylation and ubiquitination display a substantial presence and may provide new avenues into modulating tau function.
Concerning O-GlcNAcylation, Morris et al. identify S400 as the only site on endogenous mouse tau that carries this modification above the limit of detection. Despite the high sequence similarity between mouse and human tau, it will still be interesting to see how the two homologs compare. As pointed out by the authors, S400 has previously been exploited to raise O tau-specific antibodies and to demonstrate that this modification is present on human tau in vivo (Yuzwa et al., 2011; Cameron et al., 2013). The steady-state levels of tau O-GlcNAcylation tend to be rather low. This explains why O-tau levels at this site can be increased up to 12-fold upon inhibiting O-GlcNAcase with a specific inhibitor (Aug 2014 news). This would not be the case if tau were already extensively modified prior to the treatment with the O-GlcNAcase inhibitor.
Based on the identification of a single site, the authors conclude that a direct competition of phosphorylation versus O-GlcNAcylation is unlikely the underlying mechanism for the reduction of neurofibrillary tangle (NFT) formation by O-GlcNAcase inhibition. As mentioned in the discussion, this view is consistent with the data from other groups (Yuzwa et al., 2012; Graham et al., 2014) that have not seen a change of phosphorylation patterns in the soluble tau fraction upon treatment with the O-GlcNAcase inhibitor Thiamet G. It is worth bearing in mind, however, that a direct competition can only occur if a larger percentage of a specific serine or threonine residue is either phosphorylated or O-GlcNAcylated. Only in that situation would the post-translational modifications be mutually exclusive. Thus, it seems more likely that a different mechanism is at play. O-GlcNAcylation at S400 and other sites could potentially alter the overall propensity of tau to be recruited into neurofibrillary tangles or affect its prion-like spreading. Although S400 is the only detectable O-GlcNAcylation site in mice, it is worth bearing in mind that an O-GlcNAcase inhibitor could potentially influence further sites which are not detectable under native conditions. Basal levels of O-GlcNAcylation are driven by the equilibrium between the addition of GlcNAc by the corresponding transferase OGT and the removal by O-GlcNAcase. Thus, steady-levels are determined by the predominant enzyme activity in the brain and can be substantially modulated by an O-GlcNAcase inhibitor.
References:
Yuzwa SA, Yadav AK, Skorobogatko Y, Clark T, Vosseller K, Vocadlo DJ.
Mapping O-GlcNAc modification sites on tau and generation of a site-specific O-GlcNAc tau antibody.
Amino Acids. 2011 Mar;40(3):857-68. Epub 2010 Aug 13
PubMed.
Cameron A, Giacomozzi B, Joyce J, Gray A, Graham D, Ousson S, Neny M, Beher D, Carlson G, O'Moore J, Shearman M, Hering H.
Generation and characterization of a rabbit monoclonal antibody site-specific for tau O-GlcNAcylated at serine 400.
FEBS Lett. 2013 Nov 15;587(22):3722-8. Epub 2013 Oct 7
PubMed.
Yuzwa SA, Shan X, Macauley MS, Clark T, Skorobogatko Y, Vosseller K, Vocadlo DJ.
Increasing O-GlcNAc slows neurodegeneration and stabilizes tau against aggregation.
Nat Chem Biol. 2012 Apr;8(4):393-9.
PubMed.
Graham DL, Gray AJ, Joyce JA, Yu D, O'Moore J, Carlson GA, Shearman MS, Dellovade TL, Hering H.
Increased O-GlcNAcylation reduces pathological tau without affecting its normal phosphorylation in a mouse model of tauopathy.
Neuropharmacology. 2014 Apr;79:307-13. Epub 2013 Dec 8
PubMed.
This is a very important study and useful resource for the field. This is the first comprehensive, unbiased, quantitative mass-spectrometry analysis of tau modification under normal conditions. The study and analyses are rigorously performed, and the data quality is unquestionable. Researchers like myself will be poring over the data set for different types of modifications on specific sites for mechanistic clues.
One of the most striking discoveries is that "73 percent of lysines targeted by ubiquitination were also targeted by acetylation," and that 78 percent of acetylated lysines could also be ubiquitinated. These results highlight the likelihood that these two modifications compete with each other, and that reducing one could enhance the other. This is exciting for us since we showed that reducing tau acetylation elevates tau clearance in vitro and in vivo (Min et al., 2010; Min et al. in revision).
The tau acetylation data agree very well with published findings overall. Some of the acetylation sites reported in AD brains, such as K174 and K274 (from my lab; Grinberg et al., 2013) and K280 (Cohen et al., 2011; Irwin et al., 2012) did not show up in this analysis. One explanation, as discussed in the paper, is that these sites are of very low abundance in the absence of substantial tau pathology, e.g. soluble or insoluble tau aggregation. Acetylation at sites such as K174, K274, and K280 is upregulated when there is substantial tau pathology. Other sites might be undergoing constitutive acetylation, but not regulated by tau pathology. These "housekeeping" sites might be easier to detect in both conditions.
The other reason for the discrepancy with previous studies could be that mouse and human tau are modified differently even when Aβ accumulates. It is conceivable that murine tau pathology in hAPP mice is not comparable to human tau pathology in AD brains, which could explain the lack of neuronal loss and other pathological features (e.g., neurofibrillary tangles) in these animals.
References:
Min SW, Cho SH, Zhou Y, Schroeder S, Haroutunian V, Seeley WW, Huang EJ, Shen Y, Masliah E, Mukherjee C, Meyers D, Cole PA, Ott M, Gan L.
Acetylation of tau inhibits its degradation and contributes to tauopathy.
Neuron. 2010 Sep 23;67(6):953-66.
PubMed.
Grinberg LT, Wang X, Wang C, Sohn PD, Theofilas P, Sidhu M, Arevalo JB, Heinsen H, Huang EJ, Rosen H, Miller BL, Gan L, Seeley WW.
Argyrophilic grain disease differs from other tauopathies by lacking tau acetylation.
Acta Neuropathol. 2013 Apr;125(4):581-93.
PubMed.
Cohen TJ, Guo JL, Hurtado DE, Kwong LK, Mills IP, Trojanowski JQ, Lee VM.
The acetylation of tau inhibits its function and promotes pathological tau aggregation.
Nat Commun. 2011;2:252.
PubMed.
Irwin DJ, Cohen TJ, Grossman M, Arnold SE, Xie SX, Lee VM, Trojanowski JQ.
Acetylated tau, a novel pathological signature in Alzheimer's disease and other tauopathies.
Brain. 2012 Mar;135(Pt 3):807-18.
PubMed.
This study is very interesting and significant because it identifies complex post-translational modifications of physiological tau, and also reveals no difference between APP mice and wild-type mice, indicating that there is much to learn about tau modifications. As the authors correctly pointed out and addressed, such mass spectrometry analysis might not detect low-abundance tau modifications, especially given that tau is so abundant in the brain. In addition, this analysis cannot detect tau conformational changes that may be important for tau-related pathologies.
I am very impressed with the advances here using mass spec technology. I am convinced there are no significant differences in post-translational modification between model and control mice, as reported by the authors. But to me the most important question here is how faithfully the models the authors used represent AD. The authors' assumption cannot explain why tau-only lesions cause FTDP, FTP, or tangle-only dementia. In this aspect I am concerned less about tau post-translational modification and more about its mislocalization (dendritic shifting), the enigma of which would be the key to understanding the pathogenesis of tauopathy.
The post-translation modifications (PTMs) of tau have long been proposed to play an important role in tau aggregation and tau-mediated neurodegeneration. However, the overall PTMs of tau under physiological conditions are not well elucidated. By using an unbiased approach, the authors of this study sought to identify the global PTMs of endogenous mouse tau in wild-type mice and hAPP transgenic mice. They not only confirmed many previously identified PTMs, but also successfully identified some new ones. Notably, this study modifies several views held previously. First of all, several PTMs (acetylation at K281 and K163 and methylation at K163, etc.) that were previously regarded as pathological were found in physiological tau as well. Interestingly, phosphorylation at S202, T205, T231, S262, S396 and S404 which are normally found in brains of AD or other tauopathies were also detected in normal mouse tau. This result indicates that it is the excess of phosphorylation at those sites, not just the phosphorylation as such, that is pathological. Secondly, contrary to previous observations of tau hyperphosphorylation in hAPP transgenic mice compared to wild-type mice (Bellucci et al., 2007; Simon et al., 2009; Sturchler-Pierrat et al., 1997), in this study no difference in phosphorylation status of tau was observed between wild-type mice and hAPP transgenic mice, at an age when hAPP mice already show synaptic network and behavior abnormalities. The cause of the discrepancy between these mouse lines with regard to tau phosphorylation is unknown but may be related to the levels, species, and toxicities of Aβ in these mouse models. It may be of interest to quantify what percentage of tau is modified at sites by certain modifications, as this will indicate what type of modification may play a dominant role in regulating tau functions. Thirdly, this study detected only a single O-GlcNac site (S400) in endogenous mouse tau. This argues against the notion that O-GlcNac modifications block pathologic tau phosphorylation by competitively occupying many potential phosphorylation sites (Arnold et al., 1996). However, since in both wild-type mice and hAPP transgenic mice there was no hyperphosphorylation of tau, it is conceivable that in conditions of elevated tau phosphorylation, the O-GlcNac modifications may be elevated as well; this could be tested experimentally. Finally, ubiquitination was previously only detected in pathologically aggregated tau (Cripps et al., 2006; Morishima-Kawashima et al., 1993), but the present study shows that ubiquitination occurs in endogenous mouse tau under physiological conditions. It is currently not yet clear through what type of ubiquitin linkage tau is modified (K6, K11, K48, K63 etc.). Since different ubiquitination linkages may differentially affect substrate protein functions, it would be of interest to further identify the ubiquitination types on tau.
The identification of more PTMs of tau by this study further highlights the complexity of the regulation of tau metabolism and function. It would not be surprising if tau can be post-translationally modified at many sites by different modifications. Tau is a natively unfolded protein, easily accessible to many modifying enzymes, and exceptionally well soluble at up to several hundred micromolar (Tepper et al., 2014), and is unusually abundant in lysine, serine, and threonine residues, which makes tau subject to many different types of PTMs. However, whether PTMs of tau is a fine-tuned process that plays a critical role in regulating tau function and metabolism needs further investigation. For instance, what types of PTMs and what sites are involved in regulating tau-microtubule interactions or pathological aggregation of tau? Acetylation, methylation and ubiquitination may target the same lysine residues of tau, but how are these PTMs regulated, and what are the different influences of these modifications on tau? In addition, since tau is differentially distributed in neuronal compartments, brain cells and brain regions, where do the PTMs of tau occur? What is the effect of aging on the different PTMs of tau? How do the PTMs affect the spreading of tau in Alzheimer disease? All these questions may deserve further investigation. In summary, this paper is an important advance that sets up a framework for understanding tau regulation.
References:
Bellucci A, Rosi MC, Grossi C, Fiorentini A, Luccarini I, Casamenti F.
Abnormal processing of tau in the brain of aged TgCRND8 mice.
Neurobiol Dis. 2007 Sep;27(3):328-38.
PubMed.
Simón AM, Schiapparelli L, Salazar-Colocho P, Cuadrado-Tejedor M, Escribano L, López de Maturana R, Del Río J, Pérez-Mediavilla A, Frechilla D.
Overexpression of wild-type human APP in mice causes cognitive deficits and pathological features unrelated to Abeta levels.
Neurobiol Dis. 2009 Mar;33(3):369-78.
PubMed.
Sturchler-Pierrat C, Abramowski D, Duke M, Wiederhold KH, Mistl C, Rothacher S, Ledermann B, Bürki K, Frey P, Paganetti PA, Waridel C, Calhoun ME, Jucker M, Probst A, Staufenbiel M, Sommer B.
Two amyloid precursor protein transgenic mouse models with Alzheimer disease-like pathology.
Proc Natl Acad Sci U S A. 1997 Nov 25;94(24):13287-92.
PubMed.
Arnold CS, Johnson GV, Cole RN, Dong DL, Lee M, Hart GW.
The microtubule-associated protein tau is extensively modified with O-linked N-acetylglucosamine.
J Biol Chem. 1996 Nov 15;271(46):28741-4.
PubMed.
Cripps D, Thomas SN, Jeng Y, Yang F, Davies P, Yang AJ.
Alzheimer disease-specific conformation of hyperphosphorylated paired helical filament-Tau is polyubiquitinated through Lys-48, Lys-11, and Lys-6 ubiquitin conjugation.
J Biol Chem. 2006 Apr 21;281(16):10825-38.
PubMed.
Morishima-Kawashima M, Hasegawa M, Takio K, Suzuki M, Titani K, Ihara Y.
Ubiquitin is conjugated with amino-terminally processed tau in paired helical filaments.
Neuron. 1993 Jun;10(6):1151-60.
PubMed.
Tepper K, Biernat J, Kumar S, Wegmann S, Timm T, Hübschmann S, Redecke L, Mandelkow EM, Müller DJ, Mandelkow E.
Oligomer formation of tau protein hyperphosphorylated in cells.
J Biol Chem. 2014 Dec 5;289(49):34389-407. Epub 2014 Oct 22
PubMed.
Comments
King's College London, Institute of Psychiatry, Psychology & Neuroscience
This is a comprehensive body of work that has identified new sites of tau post-translational modification in wild-type and APP transgenic mice. The new sites include tau methylation, di-methylation, acetylation, and ubiquitination (inferred from GlyGly-modified lysine), as summarized in Figure 1b. This wide-ranging approach to investigating tau modification brings up the interesting point that ubiquitination could provide a highly sensitive system for regulation of tau function through competition at lysine residues, as suggested previously.
It is notable that O-GlcNAc modification of tau was barely detectable in the mice, even following lectin enrichment. This result is consistent with our own unpublished data, as we were unable to detect this modification in postmortem control or Alzheimer human brain tissue by mass spectrometry.
The observation of differential phosphorylation of tau in the postsynaptic density (PSD) of wild-type mice is interesting, since this could, as the authors suggest, affect the regulation or function of tau in this location. The authors found an increase in the amount of doubly phosphorylated tau 386-404 in the PSD and a decrease in the triply phosphorylated tau peptide, relative to unfractionated brain lysate. The potential sites for phosphorylation in this peptide include residues T386, Y394, S396, S400, T403, and S404. Since there is apparently no difference in tau phosphorylation in the PSD of wild-type and APP transgenic mice, and PSD immunoreactivity with PHF1 antibody (S396/S404) is unchanged, regulated phosphorylation/dephosphorylation of at least one of the serine/threonine/tyrosine residues T386, Y394, S400, T403 in this C-terminal region of tau could be critical for PSD localization.
An important take-home message from this study is that differential post-translational modification of tau is likely to direct its binding to other proteins and to be responsible for regulating tau subcellular localization. However, it remains to be seen which tau modifications, either individually or in combination, might be involved in the pathogenesis of neurodegenerative disease.
View all comments by Diane HangerDeceased
This is a heroic effort to characterize post-translational modifications of tau in the J20 human APP mouse. Two things are really notable to me:
1. The presence of the mutated human APP gene does not alter tau modification—whether that be phosphorylation, acetylation, methylation, or other. There are no differences between the J20 mice and the wild-type. This is despite the fact that at the age examined, the J20 "show synaptic, network, and behavioral abnormalities." These abnormalities are evidently not due to post-translational modification of tau, unless this is highly regional or a very minor component of the total tau present. Even this latter "escape clause" appears to be ruled out by the authors' regional and subcellular fraction analyses. And yet, if tau levels are reduced, J20 mice are protected from the synaptic, network and behavioral abnormalities.
2. Although not mentioned in the paper, the authors provide no evidence for changes in the phosphorylation of tyrosine in tau, especially tyrosine 18, and I am sure that they looked long and hard for this. Several reports from Mucke's group and from other labs have implicated the tyrosine kinase fyn in responses to Aβ, and fyn phosphorylates tyrosine 18 of tau. This phosphorylation does not seem to have been detected in either wild type or J20 mice.
Much of the talk about tau modification has been as much wishful thinking as anything. Especially with immunocytochemistry, it is possible to detect changes in tau phosphorylation that might be entirely insignificant in quantitative terms. We really need the kind of rigorous analysis the authors have done here. This is a great paper and will be discussed at much length in tau circles.
View all comments by Peter DaviesNew York Institute for Basic Research in Developmental Disabilities
New York State Institute for Basic Research
This is an extensive study on post-translational modifications of tau by mass spectrometry analysis in wild-type (WT) and hAPP transgenic mouse brains. Most of the phosphorylation sites previously reported in postmortem normal human brains were also found in the present study in WT and hAPP transgenic mice. However, O-GlcNAcylation at only site Ser400 was detected in WT and hAPP mice. This finding led the authors to question an extensive O-GlcNAcylation of tau.
There are no stoichiometric data shown for any of the post-translational modifications found in murine tau isoforms. However, it appears that with their methodology the authors were probably able to detect substoichiometric phosphorylations; the same does not appear to be true for O-GlcNAcylation. That may be why they found only one O-GlcNAcylated residue. Substoichiometric O-GlcNAcylations could have been hydrolyzed by the 1 percent perchloric acid used to homogenize the brains for the isolation of tau.
Of course, the overexpression hAPP transgenic mice are a highly artificial model generated to produce Aβ plaques but do not replicate the mechanism by which similar pathology is produced in AD. Also, hAPP transgenic mice do not produce tau pathology. In fact, the discrepancy in tau O-GlcNAcylation between the previous human studies and the present mouse findings reinforces the limitations of overexpression transgenic mouse models. Hopefully Dr. Mucke and his team will use the mass spectrometry methodology they have developed to analyze post-translational modifications of tau in AD and normal human brains.
View all comments by Cheng-Xin GongKULeuven
The issues surrounding protein tau remain among the toughest nuts to crack in normal brain physiology, and even more in brain pathology of AD and FTDP-17, as well as in the 55 or so known tauopathies. It takes the best hands and minds and most sophisticated tools to dig up solid data on which to base insight in—and corrections of—previous data, as well as provide firm ground and directions for future work.
Lennart Mucke and friends at UCSF and Hopkins have done just that: a comprehensive mass-spec study of mouse tau in wild-type and human APP transgenic mouse brain, compiling an unprecedented number of post-translational modifications at more sites than we knew—or suspected. The study highlights the complex post-translational biochemistry of protein tau in vivo, and albeit “only” in mouse brain, it is the only organ/model that matters. We write 2015, i.e., 40 years after the discovery of a "protein bound to micro-Tubules" (Murray et al., 1975) (a period that spans my entire postdoc/professional career ...).
One issue that concerns this old scientist and his former young co-workers more than some others is the claim that O-Glc-NAc-ylation of protein tau could be a major disease-defining derivative—and consequently a major therapeutic target (Liu et al., 2004; Yuzwa et al., 2008; Arnold et al., 1996; Yuzwa et al., 2012; Borghgraef et al., 2013; Liu et al., 2009; Yuzwa et al., 2011).
We have reported beneficial effects of the pharmacological increase of O-Glc-NAc-ylation on brain proteins (several hundreds!) but contradicted a defining role of this modification on protein tau itself and/or on its phosphorylation (Borghgraef et al, 2013). We were unable to confirm even minimal O-Glc-NAc-ylation of protein tau by western blotting with antibodies available commercially or provided by the authors of the original studies (Yuzwa et al., 2008; Yuzwa et al., 2011).
Hence, we maintain what we stated in 2013 (and presented at many a meeting): "We conclude that increasing O-GlcNAc-ylation of brain proteins improved the clinical condition and prolonged the survival of ageing Tau.P301L mice, but not by direct biochemical action on protein tau. The pharmacological effect is proposed to be located downstream in the pathological cascade initiated by protein Tau.P301L, opening novel venues for our understanding, and eventually treating the neurodegeneration mediated by protein tau” (Borghgraef et al., 2013).
References:
Murray JJ, Reed PW, Fay FS. Contraction of isolated smooth muscle cells by inophore A23187. Proc Natl Acad Sci U S A. 1975 Nov;72(11):4459-63. PubMed.
Liu F, Iqbal K, Grundke-Iqbal I, Hart GW, Gong CX. O-GlcNAcylation regulates phosphorylation of tau: a mechanism involved in Alzheimer's disease. Proc Natl Acad Sci U S A. 2004 Jul 20;101(29):10804-9. PubMed.
Yuzwa SA, Macauley MS, Heinonen JE, Shan X, Dennis RJ, He Y, Whitworth GE, Stubbs KA, McEachern EJ, Davies GJ, Vocadlo DJ. A potent mechanism-inspired O-GlcNAcase inhibitor that blocks phosphorylation of tau in vivo. Nat Chem Biol. 2008 Aug;4(8):483-90. PubMed.
Arnold CS, Johnson GV, Cole RN, Dong DL, Lee M, Hart GW. The microtubule-associated protein tau is extensively modified with O-linked N-acetylglucosamine. J Biol Chem. 1996 Nov 15;271(46):28741-4. PubMed.
Yuzwa SA, Shan X, Macauley MS, Clark T, Skorobogatko Y, Vosseller K, Vocadlo DJ. Increasing O-GlcNAc slows neurodegeneration and stabilizes tau against aggregation. Nat Chem Biol. 2012 Apr;8(4):393-9. PubMed.
Borghgraef P, Menuet C, Theunis C, Louis JV, Devijver H, Maurin H, Smet-Nocca C, Lippens G, Hilaire G, Gijsen H, Moechars D, Van Leuven F. Increasing brain protein O-GlcNAc-ylation mitigates breathing defects and mortality of Tau.P301L mice. PLoS One. 2013;8(12):e84442. Epub 2013 Dec 23 PubMed.
Liu F, Shi J, Tanimukai H, Gu J, Grundke-Iqbal I, Iqbal K, Gong CX. Reduced O-GlcNAcylation links lower brain glucose metabolism and tau pathology in Alzheimer's disease. Brain. 2009 Jul;132(Pt 7):1820-32. PubMed.
Yuzwa SA, Yadav AK, Skorobogatko Y, Clark T, Vosseller K, Vocadlo DJ. Mapping O-GlcNAc modification sites on tau and generation of a site-specific O-GlcNAc tau antibody. Amino Acids. 2011 Mar;40(3):857-68. Epub 2010 Aug 13 PubMed.
View all comments by Fred Van LeuvenAsceneuron SA
This manuscript by Morris and colleagues describes the systematic mapping of the post-translational modifications of endogenous mouse tau isolated from wild-type and hAPP transgenic mice by mass spectrometry. Besides revealing the similarity between these two groups, the identification of a substantial number of post-translational modifications beyond phosphorylation provides a more holistic picture of the tau protein. Although phosphorylation constitutes the main type of modification, lysine acetylation and ubiquitination display a substantial presence and may provide new avenues into modulating tau function.
Concerning O-GlcNAcylation, Morris et al. identify S400 as the only site on endogenous mouse tau that carries this modification above the limit of detection. Despite the high sequence similarity between mouse and human tau, it will still be interesting to see how the two homologs compare. As pointed out by the authors, S400 has previously been exploited to raise O tau-specific antibodies and to demonstrate that this modification is present on human tau in vivo (Yuzwa et al., 2011; Cameron et al., 2013). The steady-state levels of tau O-GlcNAcylation tend to be rather low. This explains why O-tau levels at this site can be increased up to 12-fold upon inhibiting O-GlcNAcase with a specific inhibitor (Aug 2014 news). This would not be the case if tau were already extensively modified prior to the treatment with the O-GlcNAcase inhibitor.
Based on the identification of a single site, the authors conclude that a direct competition of phosphorylation versus O-GlcNAcylation is unlikely the underlying mechanism for the reduction of neurofibrillary tangle (NFT) formation by O-GlcNAcase inhibition. As mentioned in the discussion, this view is consistent with the data from other groups (Yuzwa et al., 2012; Graham et al., 2014) that have not seen a change of phosphorylation patterns in the soluble tau fraction upon treatment with the O-GlcNAcase inhibitor Thiamet G. It is worth bearing in mind, however, that a direct competition can only occur if a larger percentage of a specific serine or threonine residue is either phosphorylated or O-GlcNAcylated. Only in that situation would the post-translational modifications be mutually exclusive. Thus, it seems more likely that a different mechanism is at play. O-GlcNAcylation at S400 and other sites could potentially alter the overall propensity of tau to be recruited into neurofibrillary tangles or affect its prion-like spreading. Although S400 is the only detectable O-GlcNAcylation site in mice, it is worth bearing in mind that an O-GlcNAcase inhibitor could potentially influence further sites which are not detectable under native conditions. Basal levels of O-GlcNAcylation are driven by the equilibrium between the addition of GlcNAc by the corresponding transferase OGT and the removal by O-GlcNAcase. Thus, steady-levels are determined by the predominant enzyme activity in the brain and can be substantially modulated by an O-GlcNAcase inhibitor.
References:
Yuzwa SA, Yadav AK, Skorobogatko Y, Clark T, Vosseller K, Vocadlo DJ. Mapping O-GlcNAc modification sites on tau and generation of a site-specific O-GlcNAc tau antibody. Amino Acids. 2011 Mar;40(3):857-68. Epub 2010 Aug 13 PubMed.
Cameron A, Giacomozzi B, Joyce J, Gray A, Graham D, Ousson S, Neny M, Beher D, Carlson G, O'Moore J, Shearman M, Hering H. Generation and characterization of a rabbit monoclonal antibody site-specific for tau O-GlcNAcylated at serine 400. FEBS Lett. 2013 Nov 15;587(22):3722-8. Epub 2013 Oct 7 PubMed.
Yuzwa SA, Shan X, Macauley MS, Clark T, Skorobogatko Y, Vosseller K, Vocadlo DJ. Increasing O-GlcNAc slows neurodegeneration and stabilizes tau against aggregation. Nat Chem Biol. 2012 Apr;8(4):393-9. PubMed.
Graham DL, Gray AJ, Joyce JA, Yu D, O'Moore J, Carlson GA, Shearman MS, Dellovade TL, Hering H. Increased O-GlcNAcylation reduces pathological tau without affecting its normal phosphorylation in a mouse model of tauopathy. Neuropharmacology. 2014 Apr;79:307-13. Epub 2013 Dec 8 PubMed.
View all comments by Dirk BeherUniversity of California, San Francisco
This is a very important study and useful resource for the field. This is the first comprehensive, unbiased, quantitative mass-spectrometry analysis of tau modification under normal conditions. The study and analyses are rigorously performed, and the data quality is unquestionable. Researchers like myself will be poring over the data set for different types of modifications on specific sites for mechanistic clues.
One of the most striking discoveries is that "73 percent of lysines targeted by ubiquitination were also targeted by acetylation," and that 78 percent of acetylated lysines could also be ubiquitinated. These results highlight the likelihood that these two modifications compete with each other, and that reducing one could enhance the other. This is exciting for us since we showed that reducing tau acetylation elevates tau clearance in vitro and in vivo (Min et al., 2010; Min et al. in revision).
The tau acetylation data agree very well with published findings overall. Some of the acetylation sites reported in AD brains, such as K174 and K274 (from my lab; Grinberg et al., 2013) and K280 (Cohen et al., 2011; Irwin et al., 2012) did not show up in this analysis. One explanation, as discussed in the paper, is that these sites are of very low abundance in the absence of substantial tau pathology, e.g. soluble or insoluble tau aggregation. Acetylation at sites such as K174, K274, and K280 is upregulated when there is substantial tau pathology. Other sites might be undergoing constitutive acetylation, but not regulated by tau pathology. These "housekeeping" sites might be easier to detect in both conditions.
The other reason for the discrepancy with previous studies could be that mouse and human tau are modified differently even when Aβ accumulates. It is conceivable that murine tau pathology in hAPP mice is not comparable to human tau pathology in AD brains, which could explain the lack of neuronal loss and other pathological features (e.g., neurofibrillary tangles) in these animals.
References:
Min SW, Cho SH, Zhou Y, Schroeder S, Haroutunian V, Seeley WW, Huang EJ, Shen Y, Masliah E, Mukherjee C, Meyers D, Cole PA, Ott M, Gan L. Acetylation of tau inhibits its degradation and contributes to tauopathy. Neuron. 2010 Sep 23;67(6):953-66. PubMed.
Grinberg LT, Wang X, Wang C, Sohn PD, Theofilas P, Sidhu M, Arevalo JB, Heinsen H, Huang EJ, Rosen H, Miller BL, Gan L, Seeley WW. Argyrophilic grain disease differs from other tauopathies by lacking tau acetylation. Acta Neuropathol. 2013 Apr;125(4):581-93. PubMed.
Cohen TJ, Guo JL, Hurtado DE, Kwong LK, Mills IP, Trojanowski JQ, Lee VM. The acetylation of tau inhibits its function and promotes pathological tau aggregation. Nat Commun. 2011;2:252. PubMed.
Irwin DJ, Cohen TJ, Grossman M, Arnold SE, Xie SX, Lee VM, Trojanowski JQ. Acetylated tau, a novel pathological signature in Alzheimer's disease and other tauopathies. Brain. 2012 Mar;135(Pt 3):807-18. PubMed.
View all comments by Li GanHarvard Medical School
This study is very interesting and significant because it identifies complex post-translational modifications of physiological tau, and also reveals no difference between APP mice and wild-type mice, indicating that there is much to learn about tau modifications. As the authors correctly pointed out and addressed, such mass spectrometry analysis might not detect low-abundance tau modifications, especially given that tau is so abundant in the brain. In addition, this analysis cannot detect tau conformational changes that may be important for tau-related pathologies.
View all comments by Kun Ping LuDoshisha University
I am very impressed with the advances here using mass spec technology. I am convinced there are no significant differences in post-translational modification between model and control mice, as reported by the authors. But to me the most important question here is how faithfully the models the authors used represent AD. The authors' assumption cannot explain why tau-only lesions cause FTDP, FTP, or tangle-only dementia. In this aspect I am concerned less about tau post-translational modification and more about its mislocalization (dendritic shifting), the enigma of which would be the key to understanding the pathogenesis of tauopathy.
View all comments by Yasuo IharaDZNE (German Center for Neurodegenerative Diseases)
Shanghai Qiangrui Biotech Co., Ltd
The post-translation modifications (PTMs) of tau have long been proposed to play an important role in tau aggregation and tau-mediated neurodegeneration. However, the overall PTMs of tau under physiological conditions are not well elucidated. By using an unbiased approach, the authors of this study sought to identify the global PTMs of endogenous mouse tau in wild-type mice and hAPP transgenic mice. They not only confirmed many previously identified PTMs, but also successfully identified some new ones. Notably, this study modifies several views held previously. First of all, several PTMs (acetylation at K281 and K163 and methylation at K163, etc.) that were previously regarded as pathological were found in physiological tau as well. Interestingly, phosphorylation at S202, T205, T231, S262, S396 and S404 which are normally found in brains of AD or other tauopathies were also detected in normal mouse tau. This result indicates that it is the excess of phosphorylation at those sites, not just the phosphorylation as such, that is pathological. Secondly, contrary to previous observations of tau hyperphosphorylation in hAPP transgenic mice compared to wild-type mice (Bellucci et al., 2007; Simon et al., 2009; Sturchler-Pierrat et al., 1997), in this study no difference in phosphorylation status of tau was observed between wild-type mice and hAPP transgenic mice, at an age when hAPP mice already show synaptic network and behavior abnormalities. The cause of the discrepancy between these mouse lines with regard to tau phosphorylation is unknown but may be related to the levels, species, and toxicities of Aβ in these mouse models. It may be of interest to quantify what percentage of tau is modified at sites by certain modifications, as this will indicate what type of modification may play a dominant role in regulating tau functions. Thirdly, this study detected only a single O-GlcNac site (S400) in endogenous mouse tau. This argues against the notion that O-GlcNac modifications block pathologic tau phosphorylation by competitively occupying many potential phosphorylation sites (Arnold et al., 1996). However, since in both wild-type mice and hAPP transgenic mice there was no hyperphosphorylation of tau, it is conceivable that in conditions of elevated tau phosphorylation, the O-GlcNac modifications may be elevated as well; this could be tested experimentally. Finally, ubiquitination was previously only detected in pathologically aggregated tau (Cripps et al., 2006; Morishima-Kawashima et al., 1993), but the present study shows that ubiquitination occurs in endogenous mouse tau under physiological conditions. It is currently not yet clear through what type of ubiquitin linkage tau is modified (K6, K11, K48, K63 etc.). Since different ubiquitination linkages may differentially affect substrate protein functions, it would be of interest to further identify the ubiquitination types on tau.
The identification of more PTMs of tau by this study further highlights the complexity of the regulation of tau metabolism and function. It would not be surprising if tau can be post-translationally modified at many sites by different modifications. Tau is a natively unfolded protein, easily accessible to many modifying enzymes, and exceptionally well soluble at up to several hundred micromolar (Tepper et al., 2014), and is unusually abundant in lysine, serine, and threonine residues, which makes tau subject to many different types of PTMs. However, whether PTMs of tau is a fine-tuned process that plays a critical role in regulating tau function and metabolism needs further investigation. For instance, what types of PTMs and what sites are involved in regulating tau-microtubule interactions or pathological aggregation of tau? Acetylation, methylation and ubiquitination may target the same lysine residues of tau, but how are these PTMs regulated, and what are the different influences of these modifications on tau? In addition, since tau is differentially distributed in neuronal compartments, brain cells and brain regions, where do the PTMs of tau occur? What is the effect of aging on the different PTMs of tau? How do the PTMs affect the spreading of tau in Alzheimer disease? All these questions may deserve further investigation. In summary, this paper is an important advance that sets up a framework for understanding tau regulation.
References:
Bellucci A, Rosi MC, Grossi C, Fiorentini A, Luccarini I, Casamenti F. Abnormal processing of tau in the brain of aged TgCRND8 mice. Neurobiol Dis. 2007 Sep;27(3):328-38. PubMed.
Simón AM, Schiapparelli L, Salazar-Colocho P, Cuadrado-Tejedor M, Escribano L, López de Maturana R, Del Río J, Pérez-Mediavilla A, Frechilla D. Overexpression of wild-type human APP in mice causes cognitive deficits and pathological features unrelated to Abeta levels. Neurobiol Dis. 2009 Mar;33(3):369-78. PubMed.
Sturchler-Pierrat C, Abramowski D, Duke M, Wiederhold KH, Mistl C, Rothacher S, Ledermann B, Bürki K, Frey P, Paganetti PA, Waridel C, Calhoun ME, Jucker M, Probst A, Staufenbiel M, Sommer B. Two amyloid precursor protein transgenic mouse models with Alzheimer disease-like pathology. Proc Natl Acad Sci U S A. 1997 Nov 25;94(24):13287-92. PubMed.
Arnold CS, Johnson GV, Cole RN, Dong DL, Lee M, Hart GW. The microtubule-associated protein tau is extensively modified with O-linked N-acetylglucosamine. J Biol Chem. 1996 Nov 15;271(46):28741-4. PubMed.
Cripps D, Thomas SN, Jeng Y, Yang F, Davies P, Yang AJ. Alzheimer disease-specific conformation of hyperphosphorylated paired helical filament-Tau is polyubiquitinated through Lys-48, Lys-11, and Lys-6 ubiquitin conjugation. J Biol Chem. 2006 Apr 21;281(16):10825-38. PubMed.
Morishima-Kawashima M, Hasegawa M, Takio K, Suzuki M, Titani K, Ihara Y. Ubiquitin is conjugated with amino-terminally processed tau in paired helical filaments. Neuron. 1993 Jun;10(6):1151-60. PubMed.
Tepper K, Biernat J, Kumar S, Wegmann S, Timm T, Hübschmann S, Redecke L, Mandelkow EM, Müller DJ, Mandelkow E. Oligomer formation of tau protein hyperphosphorylated in cells. J Biol Chem. 2014 Dec 5;289(49):34389-407. Epub 2014 Oct 22 PubMed.
View all comments by Yipeng WangMake a Comment
To make a comment you must login or register.