Progress in genetic analysis (notably in genome-wide association studies) has led to the identification of 24 validated risk loci for AD [
13]. BIN1 is the first of these risk factors to have been linked to Tau pathology [
3]. In the present study, we used biochemical and cellular approaches to further characterize the relationship between Tau and BIN1. Firstly, we found that BIN1’s SH3 domain interacts directly with the Tau PRD. Secondly, we determined that the amino acid sequence [212–231] within the PRD is involved in this interaction. Thirdly, we showed that Tau phosphorylation weakens the Tau-BIN1 interaction both
in vitro and in primary neurons. Fourthly, we found that BIN1-Tau complexes exhibit a co-localization with the actin cytoskeleton in primary neurons.
BIN1 belongs to the amphiphysin protein family, the best-known function of which is to sense and generate membrane curvature through its BAR domain. This protein is also a key regulator of biological functions such as endocytosis, membrane recycling and cytoskeleton regulation [
14]. Accordingly, BIN1 recruits some partners (such as AP-2 and clathrin) through its CLAP domain and others (such as dynamin and synaptojanin) through its SH3 domain [
14]. Tau proteins are microtubule-associated proteins whose main function is to promote the polymerization and stabilization of the cytoskeleton microtubule network [
15]. Tau is also involved in cell signaling pathways via the recruitment of kinases (such as Lck, Fyn, Src [
16,
17] and phosphatidylinositol 3-kinase) and signaling adaptors (such as Grb2 [
11,
18]). These interactions are mediated by Tau’s PRD. In the present work, we extended these results by showing that BIN1-Tau binding occurs through the SH3 domain and the PRD. Given that Tau phosphorylation (strong regulator of Tau function) is deregulated in AD, we sought to determine whether Tau phosphorylation was involved in regulation of the BIN1-Tau interaction. Our data indicate that Tau phosphorylation can indeed modulate the interaction between BIN1 and Tau
in vitro. These results are consistent with previous reports of phosphorylation-dependent binding between Tau PRD and SH3 domains in proteins such as phosphatidylinositol 3-kinase, phospholipase C γ1, Grb2, and Src family kinases [
11,
18]. Importantly, the results of our NMR and cell biology experiments suggest that phosphorylation at Thr231 (which is within the Tau PRD sequence that interacts with the BIN1 SH3 domain) weakens this interaction. These data are coherent with previous studies showing that Tau mutants mimicking phosphorylation at Thr231/Ser235 (T231D/S235D) bind 8 times less avidly (compared with native Tau) to the SH3 domain of Fyn [
18], whereas a single mutation at Ser235 (S235E) does not affect the binding [
11]. In contrast, mimicking phosphorylation at Ser396 and Ser404 (S396D/S404D) did not modify the affinity of 3-repeat Tau isoforms for Fyn/SH3. The latter literature data are consistent with (i) our PLA staining results for pS396-404 Tau and BIN1 and (ii) the fact that 3-repeat Tau isoforms are expressed in primary neuron cultures. Indeed, it has been well established that both 4-repeat and 3-repeat Tau isoforms are expressed in primary neurons after 14 days of in vitro culture [
19]. Moreover, Tau proteins with different phosphorylation patterns (generated with various kinases) failed to interact with BIN1
in vitro (Fig.
3 and Additional file
8 and Additional file
9) but all were phosphorylated at Thr231. This strongly suggests that pThr231 is an important regulator of the BIN1/Tau interaction. Indeed, these findings were confirmed by our PLAs, since no interaction was observed with Thr231-phosphorylated Tau species in primary neurons. However, since Tau phosphorylation patterns are known to be associated with different subcellular locations (e.g. low Thr231 phosphorylation at the neuron’s cortical membrane [
20]), one can argue that cell-sorting mechanisms might be responsible for the lack of interaction of BIN1 with pThr231 Tau in neurons. However, given that other sites in pThr231 Tau may also be phosphorylated, we cannot rule out the possibility that the latter also modify the BIN1/Tau interaction. Further experiments will be required to specify the possible contributions of other Tau phosphorylation sites and identify the mechanisms that underlie the regulation of the BIN1-Tau interaction by phosphorylation. Nevertheless, our finding that phosphorylation of Thr231 strongly influences the BIN1-Tau interaction highlights a putative link to the pathogenesis of AD. Indeed, Thr231 hyperphosphorylation occurs in the very early stages of neurofibrillary degeneration [
21]. The enzymes that control this phosphorylation (such as Cdk5 [
22], Gsk3-β [
23], PP2A [
24] and PIN1 [
25]) are markedly deregulated during the AD process. These findings suggest that Tau hyperphosphorylation during AD (or at least phosphorylation at Thr231) can interfere with the physiological stoichiometry of BIN1-Tau complexes and may have a particular impact in the early stages of AD. We and others have shown that BIN1 expression is elevated in AD brains [
3] and that BIN1 levels are correlated with the presence of neurofibrillary tangles [
26]. Moreover, our previous work demonstrated that Tau toxicity in
Drosophila models was suppressed by downregulating the
Drosophila BIN1 ortholog [
3]. Although our earlier results suggest that BIN1-Tau complexes have toxic effects, we cannot yet clearly define the roles of these complexes in AD or how Tau Thr231 phosphorylation contributes to this pathway during AD pathogenesis. In order to better understand the role of BIN1-Tau complexes, we performed experiments with markers of the various cell compartments involved in fundamental neuronal functions (such as synapses and endocytosis). Our results showed that BIN1-Tau complexes did not co-localize with either pre-synaptic or post-synaptic terminal markers. Moreover, clathrin (a marker of the endocytosis pathway) did not particularly co-localize with BIN1-Tau complexes. These results suggest that BIN1-Tau complexes exert their functions through cellular pathways other than the currently known pathogenic pathways involving Tau at the synapse [
27] or involving BIN1 in endocytosis [
14]. Nonetheless, we demonstrated substantial co-localization of BIN1-Tau complexes and the actin cytoskeleton. Therefore, one can postulate that BIN1 and microtubule-associated Tau have a role at the interface between actin and the microtubule network in neurons. Indeed, previous research has found that BIN1 interacts with the actin cytoskeleton [
14] and serves as a membrane anchoring point for microtubules [
28]. Further studies will be needed to fully understand the roles of BIN1-Tau complexes in neurons.