Tyrosine phosphorylation of tau regulates its interactions with Fyn SH2 domains, but not SH3 domains, altering the cellular localization of tau Alessia Usardi 1 , Amy M. Pooler 1 , Anjan Seereeram 1 , C. Hugh Reynolds 1 , Pascal Derkinderen 1 , Brian Anderton 1 , Diane P. Hanger 1 , Wendy Noble 1, * and Ritchie Williamson 1,2, * 1 Department of Neuroscience, MRC Centre for Neurodegeneration Research, Institute of Psychiatry, King’s College London, UK 2 Biomedical Research Institute, Ninewells Medical School, University of Dundee, UK Keywords Fyn-SH2; Fyn-SH3; phosphorylation; tau; tyrosine Correspondence R. Williamson, Biomedical Research Institute, Ninewells Medical School, University of Dundee, Dundee DD1 9SY, UK Fax: +44 1382 740 359 Tel: +44 1382 740 347 E-mail: R.Williamson@dundee.ac.uk *These authors contributed equally to this work (Received 19 April 2011, revised 20 May 2011, accepted 16 June 2011) doi:10.1111/j.1742-4658.2011.08218.x Recent reports have demonstrated that interactions between the microtu- bule-associated protein tau and the nonreceptor tyrosine kinase Fyn play a critical role in mediating synaptic toxicity and neuronal loss in response to b-amyloid (Ab) in models of Alzheimer’s disease. Disruption of interactions between Fyn and tau may thus have the potential to protect neurons from Ab-induced neurotoxicity. Here, we investigated tau and Fyn interactions and the potential implications for positioning of these proteins in membrane microdomains. Tau is known to bind to Fyn via its Src-homology (SH)3 domain, an association regulated by phosphorylation of PXXP motifs in tau. Here, we show that Pro216 within the PXXP(213–216) motif in tau plays an important role in mediating the interaction of tau with Fyn-SH3. We also show that tau interacts with the SH2 domain of Fyn, and that this associa- tion, unlike that of Fyn-SH3, is influenced by Fyn-mediated tyrosine phos- phorylation of tau. In particular, phosphorylation of tau at Tyr18, a reported target of Fyn, is important for mediating Fyn-SH2–tau interactions. Finally, we show that tyrosine phosphorylation influences the localization of tau to detergent-resistant membrane microdomains in primary cortical neurons, and that this trafficking is Fyn-dependent. These findings may have implica- tions for the development of novel therapeutic strategies aimed at disrupting the tau ⁄ Fyn-mediated synaptic dysfunction that occurs in response to ele- vated Ab levels in neurodegenerative disease. Structured digital abstract l Fyn physically interacts with tau by pull down (View interaction) l Fyn physically interacts with tau by pull down (View interaction) Introduction The microtubule-associated protein tau is a predomi- nantly neuronal soluble phosphoprotein that is mainly cytoplasmic, but is also present in nuclear [1,2] and membrane [3–5] compartments of various cell types. Abnormalities in tau, including its aberrant phosphorylation, truncation and aggregation, are causally associated with neuronal loss in a family of neurodegenerative disorders named the tauopathies, which include Alzheimer’s disease (AD), progressive supranuclear palsy and frontotemporal dementia with Abbreviations AD, Alzheimer’s disease; Ab, b-amyloid; CHO, Chinese hamster ovary; CNS, central nervous system; DRM, detergent-resistant microdomain; GST, glutathione-S-transferase; NMDA, N-methyl- D-aspartate; PSD, postsynaptic density; SEM, standard error of the mean; SH, Src homology. FEBS Journal 278 (2011) 2927–2937 ª 2011 The Authors Journal compilation ª 2011 FEBS 2927 Parkinsonism associated with tau mutations on chro- mosome 17 [6,7]. In AD, tau is believed to act in syn- ergy with b-amyloid (Ab) to mediate neuronal loss [8]. Recently, we and others have highlighted the impor- tance of tau interactions with the membrane-anchored nonreceptor tyrosine kinase Fyn during A b-mediated neurodegeneration in cell and animal models of AD [9–12]. Ab-induced neurotoxicity in primary cultured neu- rons is dependent upon both Fyn and tau [9,13]. This toxicity is associated with the accumulation of Ab on plasma membranes and the recruitment of tau into lipid rafts [9], where tau is phosphorylated on the putative Fyn residue Tyr 18 [14]. In addition, interac- tions between Fyn and tau are important for the den- dritic positioning of these proteins, a localization that has been shown to be a critical factor for Ab-induced toxicity [11]. In a transgenic mouse model of AD over- expressing mutant human amyloid precursor protein, tau-dependent positioning of Fyn in dendrites appears to regulate Fyn activation in response to Ab [11]. Subsequent Fyn-dependent stabilization of N-methyl- D-aspartate (NMDA) receptor interaction with the postsynaptic density (PSD) protein PSD-95 results in Ab-induced excitotoxicity [11]. Furthermore, Fyn sensitizes mice to the toxic effects of Ab [15,16], and tau is required for the Fyn-mediated and Ab-mediated synaptic deficits and network impairments observed in other mouse models of AD [12], further supporting the idea that disruption of Fyn–tau interactions may have therapeutic utility in AD. Fyn can phosphorylate tau directly on Tyr18, one of five tyrosines present on tau [17–19], and tau also interacts with Fyn via its Src homology (SH)3 domain [20,21]. The SH3 domain of Fyn binds to proline-rich motifs within the sequence PXXP in interacting pro- teins. Seven such motifs are present in the proline-rich domain of the longest isoform of tau in the human central nervous system (CNS). Two distinct PXXP motifs in tau, residing within residues 213–219 and 233–236, have been suggested to mediate its associa- tion with Fyn-SH3 [20,21], and this interaction is regu- lated by the serine ⁄ threonine phosphorylation status of tau [21]. Fyn also interacts with proteins through its SH2 domain, which recognizes phosphorylated tyrosines on target proteins [22,23]. Such interactions regulate the induction of several signal transduction pathways [24] that could play a role in Ab-induced neuronal loss. Tau is known to be tyrosine phosphorylated in post- mortem AD brain [17,18,25] as well as in transgenic mouse models of tauopathy, in which tyrosine phos- phorylated tau is associated with the development of tau pathology and neuronal loss [26]. Thus, it is important to establish whether or not tau interacts with Fyn-SH2, as this would probably reveal that the tyrosine phosphorylation status of tau is important for mediating the association of these two proteins. Here, we used Chinese hamster ovary (CHO) cells to characterize the interaction between exogenously expressed human wild-type tau and Fyn. Previous studies using truncated tau constructs have indicated that Fyn-SH3 interactions are mediated by either Pro216 or Pro233 on tau [20,21]. Using mutant forms of full-length tau, we show that direct interaction with Fyn-SH3 is mediated predominantly by Pro216 in tau. In addition, we show that tau interacts with Fyn-SH2 and that tyrosine phosphorylation of tau, particularly on Tyr18, mediates this binding. Furthermore, we show that Fyn-mediated tyrosine phosphorylation of tau is important for its recruitment to detergent-resis- tant microdomains (DRMs) on primary neurons. These findings suggest that the tyrosine phosphoryla- tion-dependent interactions of tau and Fyn may play an important role in regulating the cellular localization of Fyn and tau. Results Tau interacts with Fyn-SH2 Because tau contains tyrosines that could be targeted by Fyn, and tyrosine phosphorylation influences Fyn- SH2 binding to target proteins, we set out to deter- mine whether or not tau binds to Fyn-SH2. CHO cells were transiently cotransfected with plas- mids expressing V5-tagged human 2N4R tau, the lon- gest isoform of tau present in the adult human CNS, and Fyn. Following transfection, CHO cells were trea- ted either with pervanadate, to prevent the dephos- phorylation of tyrosines, or with catalase as a control. Pervanadate is a cell-permeable inhibitor of protein tyrosine phosphatases that acts by irreversible oxida- tion of the catalytic site [27]. We have previously dem- onstrated that pervanadate increases tyrosine phosphorylation of several proteins, including tau, in cell lines [25]. CHO cell lysates were then incubated with glutathione-S-transferase (GST)–Fyn-SH2 and GST–Fyn-SH3 fusion proteins linked to glutathione Sepharose beads, and bound proteins were collected. An antibody directed against total (phosphorylated and nonphosphorylated) tau revealed a primary band of  64 kDa, corresponding to V5-tagged tau, on western blots of cell lysates (Fig. 1A). Tau was also detected in the GST–Fyn-SH2-bound fraction, but not in the GST-only-bound fraction pulldowns (Fig. 1A). Tyrosine phosphorylation-dependent tau–Fyn binding A. Usardi et al. 2928 FEBS Journal 278 (2011) 2927–2937 ª 2011 The Authors Journal compilation ª 2011 FEBS This indicates that a proportion of tau interacts with Fyn-SH2 and that this binding is specific, as it is not related to the presence of GST. Western blotting with a polyclonal antibody against GST confirmed that an equal amount of GST–Fyn-SH2 beads was used in each pulldown. Pervanadate treatment of CHO cells coexpressing tau and Fyn resulted in an increased amount of tau bound to Fyn-SH2, as compared with cells treated with catalase (Fig. 1A). Furthermore, a tau species of  68 kDa was apparent in SH2 pull- downs from pervanadate-treated cells that had been transfected with tau. This may represent a more highly phosphorylated tau species or might indicate a differ- ent conformation of tau with reduced electrophoretic mobility. Densitometric analysis of GST–Fyn-SH2- bound tau, as a proportion of total tau in each cell lysate, revealed that pervanadate increased Fyn-SH2– tau binding by approximately six-fold as compared with controls (Fig. 1B). These results show that inhibi- tion of tyrosine phosphatases with pervanadate results in significantly increased binding of tau to Fyn-SH2, thus suggesting that the interaction between tau and Fyn-SH2 is enhanced by increased tyrosine phosphory- lation in cells. Tyr18 of tau plays an important role in the interaction of tau with Fyn-SH2 To determine whether the tyrosine phosphorylation status of tau itself is important for its interaction with Fyn-SH2, CHO cells were transiently cotransfected with wild-type tau or a mutant construct in which all five tyrosines in tau (Tyr18, Tyr29, Tyr197, Tyr310 and Tyr394) were replaced with phenylalanine, gener- ating YallF tau. This mutant tau species is therefore unable to be phosphorylated on tyrosines. In addition, to determine whether phosphorylation at individual tyrosines in tau is important for its interaction with Fyn-SH2, CHO cells were transiently cotransfected with Fyn together with one of five mutant tau con- structs in which single tyrosines were mutated to phen- ylalanine (Y18F, Y29F, Y197F, Y310F or Y394F). Transfected CHO cells were treated with pervanadate or catalase, as above, prior to pulldown with GST– Fyn-SH2 and determination of bound proteins on western blots. Immunolabelling with an antibody directed against tau confirmed previous findings that pervanadate treat- ment increased the association of wild-type tau with GST–Fyn-SH2. In addition, although a 64-kDa YallF tau band was apparent in lysates from cells cotrans- fected with the Fyn construct, there were only trace amounts of YallF tau bound to GST–Fyn-SH2 (Fig. 2A). Moreover, the association of YallF tau with Fyn-SH2 was not influenced by pervanadate. These results show that prevention of tau tyrosine phosphor- ylation almost completely ablates the ability of tau to bind to Fyn-SH2, indicating that this interaction is dependent on tyrosine phosphorylation of tau. A small proportion of expressed Y18F, Y29F, Y197F, Y310F and Y394F tau each bound to Fyn-SH2 under control conditions, and this binding was elevated with per- vanadate (Fig. 2A). Notably, Y18F tau appeared to be less able than Y29F, Y197F, Y310F and Y394F tau to bind GST–Fyn-SH2, suggesting that phosphorylation of tau at Tyr18 may be particularly important for its interaction with Fyn-SH2 (Fig. 2B). Western blotting with a polyclonal antibody against GST confirmed that the same amount of GST–Fyn-SH2 beads was used in each pulldown. The amount of tau bound to GST–Fyn-SH2 follow- ing pervanadate treatment was quantified as a propor- tion of tau in cell lysates (Fig. 2B). These results revealed that tau binding to Fyn-SH2 was almost com- pletely ablated when all tyrosines on tau were substi- tuted (YallF). In contrast, all of the tau mutants with substitutions of individual tyrosines were able to bind GST–Fyn-SH2 to some extent. Y18F tau showed Fig. 1. Tyrosine phosphorylation of tau increases its association with Fyn-SH2. CHO cells were cotransfected with plasmids expressing Fyn and V5-tagged wild-type tau. Cells were treated with 100 l M pervanadate (P) or catalase (C). (A) CHO cell lysates and proteins pulled down by GST or GST–Fyn-SH2 beads on wes- tern blots labelled with antibodies against tau (V5) or GST. Num- bers on the left indicate molecular masses (kDa). (B) Bar chart showing the proportion of total tau bound to Fyn-SH2 in CHO cells treated with pervanadate or catalase as mean ± SEM. N =4. ***P < 0.005. A. Usardi et al. Tyrosine phosphorylation-dependent tau–Fyn binding FEBS Journal 278 (2011) 2927–2937 ª 2011 The Authors Journal compilation ª 2011 FEBS 2929 significantly reduced association with GST–Fyn-SH2, as compared with wild-type tau (P < 0.01), whereas phenylalanine substitutions of Tyr29, Tyr197, Tyr310 or Tyr394 in tau did not significantly influence the interaction with Fyn-SH2. These results suggest that Tyr18, the putative Fyn kinase site on tau, plays a major role in mediating interactions between tau and Fyn-SH2. However, a contribution from other tyro- sines cannot be excluded, as a greater proportion of Y18F than of YallF mutant tau cosedimented with GST–Fyn-SH2. Interestingly, whereas pervanadate induced the appearance of an  68-kDa band in wild-type, Y18F, Y29F, Y197F and Y310F tau, this band was not apparent on western blots of lysates of Y394F tau (Fig. 2A). The absence of this larger species could indi- cate that phosphorylation of tau by Fyn is reduced and therefore, in addition to phosphorylation by c-Abl [25], Fyn might also target Tyr394 in tau. Indeed, Fyn has previously been reported to phosphorylate both Tyr18 and Tyr394 [19], although it is clear that Tyr394 is phosphorylated predominantly by c-Abl. However, impaired phosphorylation of Tyr394 did not appear to influence the binding of tau to Fyn-SH2, as there was no significant difference in the amount of Y394F that cosedimented with GST–Fyn-SH2 when compared with wild-type tau. This may be possible because, although Fyn-SH2 binds directly to phosphotyro- sines, the amino acid sequence context of the phosp- hotyrosine site is also important in SH2 domain recognition, a property that allows SH2 domains to display binding preferences for specific sites on target proteins [28]. Tyrosine phosphorylation does not modulate tau binding to Fyn-SH3 We and others have previously demonstrated that Fyn binds to tau predominantly through its SH3 domain, and this interaction is regulated by serine ⁄ threonine phosphorylation of tau [20,21]. To determine whether the tyrosine phosphorylation status of tau also affects the binding of tau to Fyn-SH3, CHO cells were cotransfected with Fyn together with either wild-type or the YF mutant forms of tau. Cell lysates containing equal amounts of tau were subjected to pulldown assays with GST–Fyn-SH3, and GST-bound proteins were then assessed by immunoblotting. Western blotting of lysates from pervanadate-treated cells with an antibody against total tau revealed decreased electrophoretic mobility of tau, with the appearance of an  68-kDa tau species in wild-type and all of the mutant forms of tau except for YallF and Y394F tau (Fig. 3). In contrast to the results obtained with Fyn-SH2, wild-type tau and all of the YF mutant tau proteins were detected following pulldown with GST–Fyn-SH3 beads. There were no significant differ- ences in the proportion of wild-type or mutant YF tau associated with Fyn-SH3. Furthermore, treatment of Fig. 2. The phosphorylation status of Tyr18 on tau is important for Fyn-SH2–tau interaction. CHO cells were cotransfected with Fyn and V5-tagged wild-type (WT) or mutant YallF, Y18F, Y29F, Y197F, Y310F or Y394F tau. Cells were treated with 100 l M pervanadate (P) or catalase (C). (A) CHO cell lysates and proteins pulled down by GST–Fyn-SH2 beads on western blots labelled with antibodies against tau (V5) or GST. Numbers on the left indicate molecular masses (kDa). (B) Bar chart showing the proportion of total tau bound to Fyn-SH2 in CHO cells expressing wild-type or mutant tau. Mean values ± SEM are shown relative to the amount of wild-type tau bound to Fyn-SH2. N =3.**P < 0.01, ***P < 0.005. Fig. 3. Tau interactions with Fyn-SH3 are not influenced by tyro- sine phosphorylation. CHO cells were transiently cotransfected with Fyn and V5-tagged wild-type (WT) or mutant YallF, Y18F, Y29F, Y197F, Y310F or Y394F tau. Cells were treated with pervana- date (P) or catalase (C). Cell lysates and proteins pulled down by GST–Fyn-SH3 beads were assessed on western blots labelled with antibodies against V5 or GST. Numbers on the left indicate molecu- lar masses (kDa). N =3. Tyrosine phosphorylation-dependent tau–Fyn binding A. Usardi et al. 2930 FEBS Journal 278 (2011) 2927–2937 ª 2011 The Authors Journal compilation ª 2011 FEBS CHO cells with pervanadate did not affect the interaction of wild-type or mutant tau with Fyn-SH3 (Fig. 3). These findings indicate that, unlike the case for Fyn-SH2, tyro- sine phosphorylation does not play a role in mediating interactions between tau and Fyn-SH3, and this further suggests that different, and possibly independent, mech- anisms are involved in these interactions of tau with the same protein. Interactions with Fyn-SH3 are regulated by key PXXP motifs in tau Tau has been shown to bind Fyn-SH3 through spe- cific PXXP motifs, seven of which are present in tau. Six of these occur as three pairs of partially over- lapping tandem sequences (tau residues Pro176– Pro182, Pro200–Pro206 and Pro213–Pro219), and the seventh exists as a separate motif at Pro233–Pro236. However, there is some discrepancy over which of these PXXP motifs is most important for tau binding to Fyn-SH3. In neuroblastoma cells, truncated tau mutants lacking Pro233–Pro236 were used to demon- strate that this region of tau is critical for the binding of tau to Fyn-SH3 [20]. Conversely, using synthetic peptides, we found that Fyn binds strongly to Pro213– Pro219 of tau, but exhibits little interaction with Pro233–Pro236 [21]. To further investigate which PXXP motifs in tau are responsible for Fyn-SH3 binding, we generated alanine-substituted tau mutant constructs, P216A and P233A, for V5-tagged wild-type human tau. CHO cells were cotransfected with Fyn together with V5-tagged wild-type, P216A mutant or P233A mutant tau. Cell lysates were analysed on western blots probed with an antibody against V5 to confirm the equivalence of tau protein expression in CHO cells (Fig. 4A). GST–Fyn-SH3 beads were used to pull down bound tau in cell lysates. Detection of bound proteins by immunoblotting with an antibody against V5 revealed tau bands of  64 kDa in the GST–Fyn- SH3-bound fraction from cells transfected with each of the tau constructs. Densitometric analysis of Fyn-SH3- bound tau as a proportion of that in corresponding cell lysates showed that significantly less P216A tau was pulled down by GST–Fyn-SH3 than by either wild-type or P233A tau (P < 0.05). There were no sig- nificant differences between the amounts of wild-type or P233A tau bound by the GST–Fyn-SH3 beads (Fig. 4B). These results indicate that tau Pro216 plays an important role in mediating the binding of full- length wild-type tau to Fyn-SH3, in agreement with the findings of our previous study with synthetic tau peptides [21]. Tyrosine phosphorylation influences tau content in neuronal DRMs Interactions between tau and Fyn are important for the cellular distribution of these proteins. For example, we have shown that the trafficking of tau to DRMs [9] and plasma membranes [5] is Fyn-dependent. As we found that the tyrosine phosphorylation of tau influ- ences its interaction with Fyn, we therefore set out to investigate the role of tyrosine phosphorylation in intracellular tau trafficking. Primary cortical neurons cultured from wild-type (Fig. 5A) and Fyn-deficient (Fig. 5B) mice, with matched genetic backgrounds, were treated with per- vanadate or catalase, as above. Cell homogenates were collected, and DRMs were isolated and concentrated. Western blotting with an antibody against the DRM marker flotillin-1 was used to demonstrate that DRMs were successfully isolated from wild-type and Fyn-defi- cient mice (Fig. 5). Similarly, immunoblotting with an antibody against Fyn revealed its enrichment in DRMs prepared from wild-type neurons, but not from neu- rons lacking Fyn. Increased protein tyrosine phosphor- ylation following pervanadate treatment of wild-type Fig. 4. Pro216 in tau is important for its interaction with Fyn-SH3. CHO cells were transiently cotransfected with Fyn and V5-tagged wild-type (WT) or mutant P216A or P233A tau. (A) CHO cell lysates and proteins pulled down by GST–Fyn-SH3 beads were probed with antibodies against tau (V5) and GST. Numbers on the left indi- cate molecular masses (kDa). (B) Bar chart showing the proportion of total tau pulled down by GST–Fyn-SH3 beads. Values shown are mean fold change from control ± SEM. N =6.*P < 0.05. A. Usardi et al. Tyrosine phosphorylation-dependent tau–Fyn binding FEBS Journal 278 (2011) 2927–2937 ª 2011 The Authors Journal compilation ª 2011 FEBS 2931 and Fyn-deficient neuronal cultures was detected with the phosphotyrosine antibody 4G10 (Fig. 5). There was no apparent decrease in 4G10 immunoreactivity in homogenates from Fyn-deficient neurons, and this probably reflects compensation for the loss of Fyn by other Src family kinases, as has been previously dem- onstrated [29]. A tau species of  50–55 kDa, corresponding to endogenous mouse tau, was detected in lysates and DRMs isolated from vehicle-treated wild-type and Fyn-deficient neurons, confirming our previous find- ings [9]. In homogenate from wild-type, but not Fyn- deficient, neurons, pervanadate treatment increased the density of an  55-kDa tau band, suggesting increased Fyn-dependent phosphorylation of tau in response to pervanadate in wild-type neurons (Fig. 5). Interest- ingly, there was also an increase in the amount of DRM-associated tau isolated from wild-type neurons (Fig. 5A). In contrast, pervanadate did not appear to induce any change in the amount of tau in the DRM fraction isolated from Fyn-deficient neurons (Fig. 5B). Quantitation of the amount of DRM-associated tau as a proportion of the total tau in the corresponding cell lysates revealed a significant increase in the amount of tau present in DRMs from wild-type mice (P < 0.001), but not from Fyn-deficient mice, when compared with control (catalase-treated) neurons. This finding suggests that increased tyrosine phosphoryla- tion of cellular proteins leads to enhanced trafficking of tau to DRMs, and that this process is mediated by Fyn. As we also found that the tyrosine phosphoryla- tion status of tau regulates its interactions with Fyn-SH2, it is possible that interactions between tau and Fyn-SH2 play an important role in regulating the intracellular trafficking of tau to membrane compartments. Discussion Interactions between tau and Fyn play a critical role in governing neuronal responses to elevated Ab levels in models of AD [8], suggesting that disruption of the association of these proteins could represent a poten- tial therapeutic strategy for the treatment of AD. Here, we used GST fusion proteins of Fyn-SH2 and Fyn- SH3 to further investigate the mechanisms by which tau and Fyn interact. We determined that tau binds to both Fyn-SH2 and Fyn-SH3, and only the former of these associations is mediated by tyrosine phosphoryla- tion. With the methods employed here, it was not pos- sible to accurately quantify differences in the relative proportions of tau capable of binding to Fyn-SH2 and Fyn-SH3, as there may be variations in the affinity of Fyn-SH2 and Fyn-SH3 for GST beads. However, our data suggest that several-fold more tau binds to Fyn- SH3 than to Fyn-SH2. Tau binds predominantly to Fyn-SH3, which has a specific affinity for PXXP motifs in proteins [20,21]. Tau binding to SH3 domains is regulated by the phos- phorylation of tau on specific serine⁄ threonine residues [21], and we show here that Fyn-SH3–tau interactions are not influenced by the tyrosine phosphorylation sta- tus of tau. Using specific tau constructs in which either Pro216 or Pro233 was mutated to disrupt critical PXXP motifs implicated in Fyn binding, we have shown that Pro216 is especially important for the interaction of tau with Fyn-SH3, in line with our pre- vious work [21]. These findings, however, contrast with those of a previous study in which a deletion mutant of tau lacking residues 169–179 displayed a 90% reduction in its binding to Fyn-SH3 [20]. The reason for this discrepancy is unclear; however, it is possible that the deletion mutants of tau used previously may have altered the tau folding⁄ conformation, thus mask- ing the interaction between P216A of tau and Fyn. Fig. 5. Tyrosine phosphorylation influences tau content in neuronal DRMs. Western blots of homogenates and DRMs isolated from neuronal cultures prepared from (A) wild-type (WT) and (B) Fyn-defi- cient (Fyn ) ⁄ ) ) mice. Neurons were treated with pervanadate (P) or catalase (C). Western blots were probed with antibodies against total (phosphorylated and nonphosphorylated) endogenous tau (Dako), the DRM marker flotillin-1, Fyn, or phosphotyrosine (4G10). The arrow indicates an  55-kDa tau species that increases in den- sity in response to pervanadate treatment of WT neurons. Num- bers on the left indicate molecular masses (kDa). The bar charts below show the amount of tau in DRM fractions as a proportion of total tau in cell lysates, expressed as the fold change from control. Values shown are mean ± SEM. N = 8. ***P < 0.005. Tyrosine phosphorylation-dependent tau–Fyn binding A. Usardi et al. 2932 FEBS Journal 278 (2011) 2927–2937 ª 2011 The Authors Journal compilation ª 2011 FEBS In the work presented here, P216A and P233A mutant tau proteins were correctly synthesized and expressed in CHO cells, and both of these proteins were able to interact, but to differing extents, with Fyn-SH3. This suggests that the results shown here are not artefactual, owing to incorrect folding of the tau mutants, at least not in the regions responsible for SH3 binding. Indeed, previous work by others showed that the association between Fyn and tau was reduced, but still maintained, when a tau mutant lacking an entire PXXP motif was used in similar experiments [20]. We therefore conclude that Pro216 in tau plays an important role in its binding to Fyn-SH3. However, tau possesses seven PXXP motifs, five of which have not been investigated, and we therefore cannot exclude the possibility that other prolines in tau may be impor- tant for its binding to Fyn-SH3. Here, we have demonstrated, for the first time, that tau interacts with Fyn-SH2. Proteins harbouring SH2 domains bind to phosphorylated tyrosines on their tar- get proteins, thereby coupling the activity of tyrosine kinases with intracellular signalling pathways [28]. In our model system, in which exogenous tau and Fyn were expressed in non-neuronal cells, we found that only a small proportion of tau is associated with Fyn- SH2, both under control conditions and following per- vanadate treatment to induce tyrosine phosphoryla- tion. The magnitude of the increased tyrosine phosphorylation that we observed following pervana- date treatment is similar to that observed following treatment of cells with physiological amounts of Ab [30], and thus appears to have physiological relevance. The influence of this relatively small pool of Fyn-SH2- bound tau on potential neuronal responses to neuro- toxic insults such as Ab remains to be determined. The use of mutant tau constructs in which individual tyro- sines were mutated to phenylalanine allowed us to demonstrate that the tyrosine phosphorylation status of tau significantly influences its ability to bind Fyn- SH2. Indeed, we determined that Tyr18, the tau resi- due apparently preferred by Fyn [14,17,18], plays a dominant role in mediating the association of tau with Fyn-SH2. Tau phosphorylated at Tyr18 has been detected in a proportion of tangles in early AD brain [31], and in paired helical filament tau extracted from AD brain [17,26]. Furthermore, Tyr18 phosphorylated tau has been found in DRMs following treatment of neuronal cells with Ab [14], and we have previously shown that tau trafficking to DRMs in response to Ab treatment of primary neurons is Fyn-dependent [9]. Here, using wild-type and Fyn-deficient neurons, we show that Fyn mediates the tyrosine phosphorylation-induced recruitment of tau to neuronal DRMs. These mem- brane microdomains are widely recognized as hubs for intracellular signalling [32,33], and changes in the pro- tein composition of DRMs are associated with the induction of Ab-induced neurotoxicity [9]. Ittner et al. [11] have suggested that interactions between tau and Fyn in dendrites play a critical role in mediating Ab- induced neurotoxicity by influencing the stability of NMDA receptor–PSD-95 complexes. Interestingly, both NMDA receptors and PSD-95 shuttle between DRMs and postsynaptic densities under certain condi- tions, including during learning [34], and PSD-95 plays an important role in positioning NMDA receptors in DRMs [34]. Thus, it is conceivable that tau and Fyn might exist in a complex with NMDA receptors and PSD-95 in neurons. Activation of signalling pathways that lead to increased Fyn activity could therefore affect the tyrosine phosphorylation of tau, which could potentially modulate complex formation, and result in altered trafficking into neuronal membrane compartments. In summary, the results presented here suggest that tyrosine phosphorylation mediates the association of tau with Fyn-SH2, but not with Fyn-SH3. This dem- onstrates that different molecular mechanisms exist for these two distinct interactions of Fyn and tau, with probably disparate downstream consequences. Further- more, these results support the view that non-micro- tubule associations of tau are important for normal physiological function in neurons, and reinforce the suggestion that tau is itself involved in intracellular sig- nalling pathways ⁄ mechanisms. As its interaction with Fyn is important for tau localization in neurons, regu- lation of the cellular signalling function of this micro- tubule-associated protein could also have significant implications during the progression of neurodegenera- tive diseases, such as AD, in which both tau and Fyn are implicated. Experimental procedures Plasmids and cell transfection A plasmid expressing either the longest isoform of human CNS tau, containing two N-terminal inserts (2N) and four microtubule-binding repeats (4R), 2N4R tau, was a generous gift from M. Goedert (Medical Research Council Laboratory of Molecular Biology, Cambridge, UK) [35]. Generation of 2N4R tau constructs, each with a single tyro- sine replaced by phenylalanine (Y18F, Y29F, Y197F, Y310F and Y394F), or all with five tyrosines replaced by phenylalanines (YallF), has been described previously [25]. To generate tau constructs with individual prolines replaced A. Usardi et al. Tyrosine phosphorylation-dependent tau–Fyn binding FEBS Journal 278 (2011) 2927–2937 ª 2011 The Authors Journal compilation ª 2011 FEBS 2933 by alanines, a QuikChange XL site-directed mutagenesis kit (Stratagene, Amsterdam, the Netherlands) was used. The primers used were as follows: P216A forward primer, 5¢-ACC C CG TCC CT T GCA ACC CCA CCC ACC-3 ¢; P216A reverse primer, 5¢-GGT GGG TGG GGT TGC AA G GGA CGG GGT-3¢; P233A forward primer, 5¢-GCA G TG GTC CGG ACT CCA GCC AAG TCG CCG-3¢ (Primer B); and P233A reverse primer, 5¢-CGG CGA CTT GGC TGG AGT CCG GAC CAC TGC-3¢. P216A and P233A were generated by use of appropriate primers with the plas- mid expressing wild-type tau as template DNA. The cDNA sequence of the full insert was determined for each tau con- struct. For expression in mammalian cells, cDNA encoding tau was subcloned into the pcDNA3.1 ⁄ V5-His-TOPOvector (Invitrogen, Paisley, UK), yielding a construct with C-termi- nal V5 and His tags, as described previously [25]. Human Fyn cDNA in a pSG5 vector was a gift from D. Markby (Sugen, San Francisco, CA, USA). Constructs for the bacte- rial expression of GST have been described previously [21]. Plasmids expressing Fyn-SH3 and Fyn-SH2 were obtained from S. Anderson (University of Colorado Health Sciences Center, Denver, CO, USA). GST–Fyn-SH2 and GST–Fyn- SH3 were generated by subcloning Fyn-SH2 and Fyn-SH3 into the plasmid pGEX-5X-2. CHO cell culture and treatment CHO cells were maintained in DMEM supplemented with 10% (v ⁄ v) fetal bovine serum, 2 m ML-glutamine, and 100 UÆmL )1 penicillin ⁄ 100 lgÆmL )1 streptomycin, and incu- bated at 37 °Cina5%CO 2 atmosphere. Cells were plated into six-well dishes and transfected 24 h later by using Lipofectamine Plus Reagent (Invitrogen), following the manufacturer’s instructions. Pervandate and catalase were prepared as described previously [25]. Briefly, vanadate stock solution was prepared as a 200 m M solution of sodium orthovanadate (pH 10). Pervanadate was prepared as a ·100 stock by adding 50 lL of 200 m M sodium ortho- vanadate and 1.6 lL of 30% (w ⁄ w) hydrogen peroxide to 948.4 lL of water for 5 min at room temperature, giving 10 m M sodium orthovanadate and 16.3 mM hydrogen per- oxide. After 5 min at room temperature, the excess hydro- gen peroxide was removed by addition of 200 lgÆmL )1 catalase (520 UÆmL )1 ) and incubation for an additional 5 min, as described by Huyer et al. [27]. Twenty-four hours after transfection, cells were treated with either 100 l M per- vanadate or 2 lgÆmL )1 catalase (control) for 20 min, prior to harvesting for western blots, as described below. Preparation of GST fusion proteins and GST pulldown Generation of GST, GST–Fyn-SH2 and GST–Fyn-SH3 and subsequent coupling to glutathione beads have been described previously [21]. Following CHO cell transfection and ⁄ or cell treatments, cells were washed once with NaCl ⁄ P i and harvested into ice-cold lysis buffer (25 mM Tris ⁄ HCl, pH 7.5, containing 10% (v ⁄ v) glycerol, 0.5% (w ⁄ v) Triton X-100) 1 m M EDTA, 1 mM EGTA, 1 mM sodium orthovanadate, 150 m M sodium chloride, 10 mM sodium fluoride, and complete protease inhibitor cocktail (minus EDTA; Roche, Burgess Hill, UK), and allowed to solubi- lize for 15 min on ice prior to centrifugation at 15 000 g for 5 min at 4 °C to remove insoluble debris. Following deter- mination of protein concentration by bicinchoninic acid protein assay (Thermo Scientific, Rockford, IL, USA), equal amounts of protein were incubated with GST, GST– Fyn-SH2 or GST–Fyn-SH3 beads for 1 h at 4 °C with rotation. The beads were pelleted by centrifugation at 500 g for 1 min at 4 °C, the supernatant was discarded, and the beads washed three times with lysis buffer. Laemmli sample buffer was added, and the samples were heated to 100 °C for 5 min to release bound proteins. Proteins in CHO cell lysates were separated on 10–12% polyacrylamide gels, transferred to nitrocellulose, and probed with an antibody to tau. The amount of tau in each sample was quantified by densitometry, and samples were adjusted by dilution in lysis buffer to ensure equal tau protein concentrations for subsequent determination of the relative amounts of GST- bound proteins. Animals and culture of primary cortical neurons Wild-type and Fyn-deficient (Fyn ) ⁄ ) ) mice were obtained from The Jackson Laboratory (Bar Habor, Maine, USA). Fyn ) ⁄ ) mice were maintained on a mixed B6129F2 ⁄ J back- ground, and the same strain was used to provide wild-type controls. All animal work was licensed under the Animals (Scientific Procedures) Act 1986, reviewed by the ethical review panel of King’s College London, Institute of Psychia- try and the Home Office inspectorate, and performed in accordance with the European Communities Council Direc- tive of 24 November 1986 (86 ⁄ 609 ⁄ EEC). Primary cortical neurons were prepared from embryonic day 16 wild-type and Fyn-deficient mouse embryos, as described previously [9]. Cells were plated onto six-well dishes coated with poly( D- lysine) (5 lgÆmL )1 ) at a density of 1 · 10 6 cells per well, and cultured in Neurobasal medium (Invitrogen) containing 2% (v ⁄ v) B-27 serum-free supplement, 0.5 m ML-glutamine, 100 UÆmL )1 penicillin, and 100 lgÆmL )1 streptomycin. Cells were incubated at 37 °C in a 5% CO 2 atmosphere for 7 days prior to use. To obtain total cell lysates from CHO cells and primary cortical cultures, cells were washed in NaCl ⁄ Tris and harvested in lysis buffer, as described above. Isolation of DRMs DRMs were isolated as described previously [36]. Briefly, primary cortical neurons were washed in NaCl ⁄ Tris and lysed in isolation buffer containing 1% (w ⁄ v) Chapso in Tyrosine phosphorylation-dependent tau–Fyn binding A. Usardi et al. 2934 FEBS Journal 278 (2011) 2927–2937 ª 2011 The Authors Journal compilation ª 2011 FEBS MBS buffer (25 mM Mes, 150 mM sodium chloride, pH 6.5) containing 10 m M magnesium chloride, 10 mM sodium fluo- ride, 2 m M sodium orthovanadate, 1 mM EGTA, and 5 mM dithiothreitol. Lysates were homogenized by 20 strokes in a Dounce homogenizer, and incubated on ice for 30 min. One millilitre of homogenate was mixed with 1 mL of 90% (w ⁄ v) sucrose in MBS buffer, and placed in a 12-mL ultracentrifuge tube. A discontinuous 5%–35%–45% sucrose gradient was formed by layering 6 mL of 35% (w ⁄ v) sucrose in MBS buffer on top of the 2-mL homogenate, followed by 4 mL of 5% (w ⁄ v) sucrose in MBS buffer. Samples were centrifuged at 180 000 g for 18 h at 4 °C in a Beckman SW41 rotor (Beckman Instru- ments, Fullerton, CA, USA). Twelve 1-mL fractions were collected from the top of each gradient. To concentrate DRMs in fractions 4–5, 10 mL of MBS buffer was added and the samples were centrifuged at 110 000 g for 1 h at 4 °C in a Beckman SW41 rotor. The resulting pellet was solubilized in 100 lLof20m M Tris ⁄ HCl (pH 7.4) contain- ing 8 M urea, 10 mM NaF, 2 mM Na 3 VO 4 and 5 mM dith- iothreitol. Samples were mixed with Laemlli buffer and heated at 100 °C for 5 min prior to analysis by SDS ⁄ PAGE. SDS ⁄ PAGE and western blotting Denaturing PAGE and western blotting were performed as described previously [37]. Briefly, separated proteins were blotted onto nitrocellulose membranes (Whatman, Maid- stone, UK) and blocked in 5% (w ⁄ v) nonfat milk ⁄ 0.05% (v ⁄ v) Tween-20 in NaCl ⁄ P i for 1 h. After block- ing, membranes were incubated overnight at 4 °C in blocking solution containing primary antibody. The antibodies used were directed against tau (total tau, rabbit polyclonal; Dako, Glostrup, Denmark), V5 (mouse monoclonal; Invitrogen), phosphotyrosine (4G10, mouse monoclonal; Millipore, Bill- erica, MA, USA), Fyn (mouse monoclonal; Santa Cruz Bio- technology, Santa Cruz, CA, USA) and GST (mouse monoclonal; GE Healthcare, Little Chalfont, UK). After three washes in NaCl ⁄ P i containing 0.05% (v ⁄ v) Tween-20, blots were incubated with the appropriate fluorophore-conju- gated goat secondary antibody (IRDye800 goat anti-rabbit; Rockland, Gilbertsville, PA, USA or Alexa Fluor 680 goat anti-mouse; Invitrogen) for 1 h at ambient temperature. Pro- teins were visualized with the Odyssey imaging system (Li-Cor Biosciences, Cambridge, UK). Scanned images were analysed with the public domain IMAGEJ program (http:// www.rsb.info.nih.gov/ij/). Statistics Statistical analyses were performed by t-test or ANOVA with GRAPHPAD PRISM 5.0 (GraphPad Software Inc., La Jolla, CA, USA). Data are presented as mean ± standard error of the mean (SEM). Acknowledgements We thank the following for gifts of materials: M. Goedert (MRC Laboratory of Molecular Biology, Cambridge, UK) for tau 2N4R cDNA; S. Anderson (University of Colorado Health Sciences Center) for pGEX constructs; and D. Markby (Sugen, San Fran- cisco, CA, USA) for human Fyn cDNA. This work was supported by Alzheimer’s Research UK, the Alz- heimer’s Society, and the Medical Research Council. References 1 Brady RM, Zinkowski RP & Binder LI (1995) Presence of tau in isolated nuclei from human brain. Neurobiol Aging 16, 479–486. 2 Sultan A, Nesslany F, Violet M, Begard S, Loyens A, Talahari S, Mansuroglu Z, Marzin D, Sergeant N, Hu- mez S et al. (2011) Nuclear tau, a key player in neuro- nal DNA protection. J Biol Chem 286, 4566–4575. 3 Brandt R, Leger J & Lee G (1995) Interaction of tau with the neural plasma membrane mediated by tau’s amino-terminal projection domain. J Cell Biol 131, 1327–1340. 4 Maas T, Eidenmuller J & Brandt R (2000) Interaction of tau with the neural membrane cortex is regulated by phosphorylation at sites that are modified in paired heli- cal filaments. J Biol Chem 275, 15733–15740. 5 Pooler AM, Usardi A, Evans CJ, Philpott KL, Noble W & Hanger DP (2011) Dynamic association of tau with neuronal membranes is regulated by phosphoryla- tion. Neurobiol Aging, doi: 10.1016/j.neurobiolaging. 2011.01.005 [published online 7 March 2011]. 6 Hanger DP, Seereeram A & Noble W (2009) Mediators of tau phosphorylation in the pathogenesis of Alzhei- mer’s disease. Expert Rev Neurother 9 , 1647–1666. 7 Hanger DP, Anderton BH & Noble W (2009) Tau phosphorylation: the therapeutic challenge for neurode- generative disease. Trends Mol Med 15, 112–119. 8 Ittner LM & Gotz J (2011) Amyloid-beta and tau – a toxic pas de deux in Alzheimer’s disease. Nat Rev Neurosci 12, 65–72. 9 Williamson R, Usardi A, Hanger DP & Anderton BH (2008) Membrane-bound beta-amyloid oligomers are recruited into lipid rafts by a fyn-dependent mechanism. FASEB J 22, 1552–1559. 10 Williamson R, Scales T, Clark BR, Gibb G, Reynolds CH, Kellie S, Bird IN, Varndell IM, Sheppard PW, Everall I et al. (2002) Rapid tyrosine phosphorylation of neuronal proteins including tau and focal adhesion kinase in response to amyloid-beta peptide exposure: involvement of Src family protein kinases. J Neurosci 22, 10–20. 11 Ittner LM, Ke YD, Delerue F, Bi M, Gladbach A, van Eersel J, Wolfing H, Chieng BC, Christie MJ, Napier A. Usardi et al. Tyrosine phosphorylation-dependent tau–Fyn binding FEBS Journal 278 (2011) 2927–2937 ª 2011 The Authors Journal compilation ª 2011 FEBS 2935 IA et al. (2010) Dendritic function of tau mediates amy- loid-beta toxicity in Alzheimer’s disease mouse models. Cell 142, 387–397. 12 Roberson ED, Halabisky B, Yoo JW, Yao J, Chin J, Yan F, Wu T, Hamto P, Devidze N, Yu GQ et al. (2011) Amyloid-beta ⁄ Fyn-induced synaptic, network, and cognitive impairments depend on tau levels in mul- tiple mouse models of Alzheimer’s disease. J Neurosci 31, 700–711. 13 Rapoport M, Dawson HN, Binder LI, Vitek MP & Ferreira A (2002) Tau is essential to beta-amyloid- induced neurotoxicity. Proc Natl Acad Sci USA 99, 6364–6369. 14 Hernandez P, Lee G, Sjoberg M & Maccioni RB (2009) Tau phosphorylation by cdk5 and Fyn in response to amyloid peptide Abeta (25–35): involvement of lipid rafts. J Alzheimers Dis 16, 149–156. 15 Chin J, Palop JJ, Yu GQ, Kojima N, Masliah E & Mucke L (2004) Fyn kinase modulates synaptotoxicity, but not aberrant sprouting, in human amyloid precur- sor protein transgenic mice. J Neurosci 24, 4692–4697. 16 Chin J, Palop JJ, Puolivali J, Massaro C, Bien-Ly N, Gerstein H, Scearce-Levie K, Masliah E & Mucke L (2005) Fyn kinase induces synaptic and cognitive impairments in a transgenic mouse model of Alzhei- mer’s disease. J Neurosci 25, 9694–9703. 17 Lee G, Thangavel R, Sharma VM, Litersky JM, Bhas- kar K, Fang SM, Do LH, Andreadis A, Van HG & Ksiezak-Reding H (2004) Phosphorylation of tau by fyn: implications for Alzheimer’s disease. J Neurosci 24, 2304–2312. 18 Lebouvier T, Scales TM, Williamson R, Noble W, Duy- ckaerts C, Hanger DP, Reynolds CH, Anderton BH & Derkinderen P (2009) The microtubule-associated pro- tein tau is also phosphorylated on tyrosine. J Alzhei- mers Dis 18, 1–9. 19 Scales TM, Derkinderen P, Leung KY, Byers HL, Ward MA, Price C, Bird IN, Perera T, Kellie S, Williamson R et al. (2011) Tyrosine phosphorylation of tau by the SRC family kinases lck and fyn. Mol Neurodegener 6, 12. 20 Lee G, Newman ST, Gard DL, Band H & Pancha- moorthy G (1998) Tau interacts with src-family non- receptor tyrosine kinases. J Cell Sci 111 (Pt 21), 3167–3177. 21 Reynolds CH, Garwood CJ, Wray S, Price C, Kellie S, Perera T, Zvelebil M, Yang A, Sheppard PW, Varndell IM et al. (2008) Phosphorylation regulates tau interac- tions with Src homology 3 domains of phosphatidylino- sitol 3-kinase, phospholipase Cgamma1, Grb2, and Src family kinases. J Biol Chem 283, 18177–18186. 22 Anderson D, Koch CA, Grey L, Ellis C, Moran MF & Pawson T (1990) Binding of SH2 domains of phospholi- pase C gamma 1, GAP, and Src to activated growth factor receptors. Science 250, 979–982. 23 Pleiman CM, Clark MR, Gauen LK, Winitz S, Cogge- shall KM, Johnson GL, Shaw AS & Cambier JC (1993) Mapping of sites on the Src family protein tyrosine kin- ases p55blk, p59fyn, and p56lyn which interact with the effector molecules phospholipase C-gamma 2, microtu- bule-associated protein kinase, GTPase-activating pro- tein, and phosphatidylinositol 3-kinase. Mol Cell Biol 13, 5877–5887. 24 Gotoh H, Okumura N, Yagi T, Okumura A, Shima T & Nagai K (2006) Fyn-induced phosphorylation of beta-adducin at tyrosine 489 and its role in their subcel- lular localization. Biochem Biophys Res Commun 346, 600–605. 25 Derkinderen P, Scales TM, Hanger DP, Leung KY, Byers HL, Ward MA, Lenz C, Price C, Bird IN, Perera T et al. (2005) Tyrosine 394 is phosphorylated in Alz- heimer’s paired helical filament tau and in fetal tau with c-Abl as the candidate tyrosine kinase. J Neurosci 25, 6584–6593. 26 Bhaskar K, Hobbs GA, Yen SH & Lee G (2010) Tyro- sine phosphorylation of tau accompanies disease pro- gression in transgenic mouse models of tauopathy. Neuropathol Appl Neurobiol 36, 462–477. 27 Huyer G, Liu S, Kelly J, Moffat J, Payette P, Kennedy B, Tsaprailis G, Gresser MJ & Ramachandran C (1997) Mechanism of inhibition of protein-tyrosine phosphata- ses by vanadate and pervanadate. J Biol Chem 272, 843–851. 28 Pawson T (2004) Specificity in signal transduction: from phosphotyrosine–SH2 domain interactions to complex cellular systems. Cell 116, 191–203. 29 Smith CL, Mittaud P, Prescott ED, Fuhrer C & Burden SJ (2001) Src, Fyn, and Yes are not required for neuro- muscular synapse formation but are necessary for sta- bilization of agrin-induced clusters of acetylcholine receptors. J Neurosci 21, 3151–3160. 30 Luo YQ, Hirashima N, Li YH, Alkon DL, Sunderland T, Etcheberrigaray R & Wolozin B (1995) Physiological levels of beta-amyloid increase tyrosine phosphorylation and cytosolic calcium. Brain Res 681 , 65–74. 31 Horowitz PM, Patterson KR, Guillozet-Bongaarts AL, Reynolds MR, Carroll CA, Weintraub ST, Bennett DA, Cryns VL, Berry RW & Binder LI (2004) Early N-terminal changes and caspase-6 cleavage of tau in Alzheimer’s disease. J Neurosci 24, 7895–7902. 32 Romanelli RJ & Wood TL (2008) Directing traffic in neural cells: determinants of receptor tyrosine kinase localization and cellular responses. J Neurochem 105, 2055–2068. 33 Lingwood D & Simons K (2010) Lipid rafts as a mem- brane-organizing principle. Science 327, 46–50. 34 Delint-Ramirez I, Salcedo-Tello P & Bermudez-Rattoni F (2008) Spatial memory formation induces recruitment of NMDA receptor and PSD-95 to synaptic lipid rafts. J Neurochem 106, 1658–1668. Tyrosine phosphorylation-dependent tau–Fyn binding A. Usardi et al. 2936 FEBS Journal 278 (2011) 2927–2937 ª 2011 The Authors Journal compilation ª 2011 FEBS [...]... et al 35 Smith MJ, Crowther RA & Goedert M (2000) The natural osmolyte trimethylamine N-oxide (TMAO) restores the ability of mutant tau to promote microtubule assembly FEBS Lett 484, 265–270 36 Williamson R, Thompson AJ, Abu M, Hye A, Usardi A, Lynham S, Anderton BH & Hanger DP (2010) Isolation of detergent resistant microdomains from cul- Tyrosine phosphorylation- dependent tau Fyn binding tured neurons:... alterations in protein composition BMC Neurosci 11, 120 37 Kelleher I, Garwood C, Hanger DP, Anderton BH & Noble W (2007) Kinase activities increase during the development of tauopathy in htau mice J Neurochem 103, 2256–2267 FEBS Journal 278 (2011) 2927–2937 ª 2011 The Authors Journal compilation ª 2011 FEBS 2937 . Tyrosine phosphorylation of tau regulates its interactions with Fyn SH2 domains, but not SH3 domains, altering the cellular localization of tau Alessia. of tau with Fyn- SH3. We also show that tau interacts with the SH2 domain of Fyn, and that this associa- tion, unlike that of Fyn- SH3, is influenced by Fyn- mediated