α-Synuclein Binds to Tau and Stimulates the Protein Kinase A-catalyzed Tau Phosphorylation of Serine Residues 262 and 356*

α-Synuclein has been implicated in the pathogenesis of several neurodegenerative disorders based on the direct linking of missense mutations in α-synuclein to autosomal dominant Parkinson’s disease and its presence in Lewy-like lesions. To gain insight into α-synuclein functions, we have investigated whether it binds neuronal proteins and modulates their functional state. The microtubule-associated protein tau was identified as a ligand by α-synuclein affinity chromatography of human brain cytosol. Direct binding assays using 125I-labeled human tau40 demonstrated a reversible binding with a IC50 about 50 pm. The interacting domains were localized to the C terminus of α-synuclein and the microtubule binding region of tau as determined by protein fragmentation and the use of recombinant peptides. High concentrations of tubulin inhibited the binding between tau and α-synuclein. Functionally, α-synuclein stimulated the protein kinase A-catalyzed phosphorylation of tau serine residues 262 and 356 as determined using a phospho-epitope-specific antibody. We propose that α-synuclein modulates the phosphorylation of soluble axonal tau and thereby indirectly affects the stability of axonal microtubules.

␣-Synuclein has been implicated in the pathogenesis of several neurodegenerative disorders based on the direct linking of missense mutations in ␣-synuclein to autosomal dominant Parkinson's disease and its presence in Lewy-like lesions. To gain insight into ␣-synuclein functions, we have investigated whether it binds neuronal proteins and modulates their functional state. The microtubule-associated protein tau was identified as a ligand by ␣-synuclein affinity chromatography of human brain cytosol. Direct binding assays using 125 I-labeled human tau40 demonstrated a reversible binding with a IC 50 about 50 pM. The interacting domains were localized to the C terminus of ␣-synuclein and the microtubule binding region of tau as determined by protein fragmentation and the use of recombinant peptides. High concentrations of tubulin inhibited the binding between tau and ␣-synuclein. Functionally, ␣-synuclein stimulated the protein kinase A-catalyzed phosphorylation of tau serine residues 262 and 356 as determined using a phospho-epitope-specific antibody. We propose that ␣-synuclein modulates the phosphorylation of soluble axonal tau and thereby indirectly affects the stability of axonal microtubules.
Filamentous nerve cell inclusions are shared characteristics of the common neurodegenerative disorders Alzheimer's disease, Parkinson's disease, and dementia with Lewy bodies. The inclusions of Alzheimer's disease, the neuritic tangles, are localized to the neurites and consist of abnormally phosphorylated protein tau (1). In Parkinson's disease and dementia with Lewy bodies the inclusions, designated Lewy bodies and Lewy neurites, are localized to the cell body and neurites, respectively. The filaments in the Lewy lesions contain ␣-synuclein (2,3). The pathogenesis of idiopathic Parkinson's disease is unknown. However, two independent missense mutations in ␣-synuclein have been shown to cause autosomal heritable early-onset Parkinson's disease. Thus, abnormal ␣-synuclein metabolism is linked both to the development of rare cases of heritable Parkinson's disease and to the common lesions in idiopathic Parkinson's disease (4,5).
␣-Synuclein is a member of the conserved synuclein gene family of which at least three species, ␣-, ␤and ␥-synuclein are expressed in the human nervous system (for recent reviews, see Refs. 6 and 7). Synucleins have been identified in various species, in Torpedo by an antiserum raised against synaptic vesicles (8), in bovine brain by their acidic nature (9), in zebra finches as a gene product of highly regulated expression (10), in rats due to selective expression in dorsal root ganglia (11). In man, ␥-synuclein was recognized as a protein that is up-regulated in various carcinomas (12)(13)(14), and ␤-synuclein was identified as a protein cross-reacting with an antibody against phosphorylated tau (15). Human ␣-synuclein was originally identified as the precursor of a peptide, non-A␤ component of Alzheimer's disease amyloid tightly associated to Alzheimer's disease amyloid (16) and later purified as an inhibitor of phospholipase D2 (17).
␣-Synuclein is a small acidic protein of 140 amino acid residues. The N-terminal part has 7 imperfect repeats containing the consensus core sequence Lys-Thr-Lys-Glu-Gly-Val, whereas the C-terminal part has no recognized structural elements. ␣-Synuclein displays an extended unfolded structure and thus belongs to the group of natively unfolded proteins also comprising protein tau (18). ␣-Synuclein partitions between the soluble cytosolic phase and a vesicle-bound fraction (19,20). The binding to vesicles is mediated through determinants in the N-terminal repeat region (20) and might induce an increased helical content in ␣-synuclein as demonstrated after binding to liposomes (21). The Parkinson's disease causing Ala 30 3 Pro mutation inhibits the binding to brain vesicles and may therefore influence fast anterograde axonal ␣-synuclein transport (20).
The investigation of ␣-synuclein-interacting nerve cell proteins is required to gain insight into the role of ␣-synuclein in the brain under normal and pathological conditions. As a step in this effort, we here identify human brain tau as an ␣-synuclein ligand by affinity chromatography and direct binding assays, and demonstrate their co-localization in axons. Moreover, we identify the interacting domains in the two proteins and show that ␣-synuclein has the propensity to stimulate the phosphorylation of specific serine residues in tau.

Proteins and Peptides
Human ␣-synuclein was expressed in Escherichia coli and purified as described previously (15). Polymerase chain reaction-based site-directed mutagenesis was used to produce the Ala 30 3 Pro and the Ala 53 3 Thr constructs and the deletion mutants ␣-synuclein-(1-87), ␣-synuclein-(30 -140) and ␣-synuclein-(55-140) were also produced by polymerase chain reaction (20). All constructs were verified by DNA sequencing and subcloned into expression plasmid pET-3d (Stratagene). Expression and purification of the recombinant proteins were done as described for ␣-synuclein (15). The identity of the mutant proteins were verified by matrix-assisted laser desorption ionizationmass spectrometry and N-terminal Edman degradation (22). The following human tau proteins were expressed in E. coli and purified as described previously (23,24); tau40 and tau24, containing 4 microtu-* This work was supported by Danmark's Sundhedsfond, Michaelsen Fonden, Danish Cancer Society, and Dansk Parkinsonforening. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. bule (MT 1 -binding tandem repeats and 2 and 0 N-terminal insertions, respectively; tau39 containing 3 MT-binding tandem repeats and 2 N-terminal insertions; recombinant tau proteins corresponding to amino acid residues 192-383 of tau40, and the microtubulebinding domain (MT-BD) corresponding residues 192-291 in tau23. The MT-BD represents 3 MT-binding tandem repeats. The numbering of the tau isoforms is according to Goedert and Jakes (23). Recombinant human MAP-2C was expressed in E. coli and purified as described previously (25). The 30-mer peptide Asp-Leu-Ser-Lys-Val-Thr-Ser-Lys-Cys-Gly-Ser-Leu-Gly-Asn-Ile-His-His-Lys-Pro-Gly-Gly-Gly-Gln-Val-Glu-Ile-Lys-Tyr-Glu-Lys, corresponding to the consensus sequence for a single MT binding tandem repeat in tau40, was synthesized at Kem-En-Tech, Copenhagen, Denmark. Bovine tubulin was from Cytoskeleton. Bovine serum albumin was from Sigma.

Cell Culture and Immunofluorescence Microscopy
Cultures of hippocampal cells were prepared from the brains of 18-day-old rat embryos (27) and used after 15 days of culture (stage 5 neurons). To label ␣-synuclein, tau, and MAP-2, cells were fixed in 4% paraformaldehyde for 30 min, permeabilized in 0.1% Triton X-100 for 10 min and processed for immunofluorescence microscopy as described (28).

Electrophoresis and Electroblotting
Proteins and peptides were resolved by 8 -16% gradient sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and stained by Coomassie Brilliant Blue (22). Immunoblotting was performed after transfer from the polyacrylamide gel onto a Immobilon polyvinylidene difluoride membrane (29).

Labeling of Synuclein and Tau Peptides
Recombinant ␣-synuclein was biotinylated as described previously (22). Recombinant tau40 was iodinated using chloramine T as oxidizing agent to a specific activity of 250 mCi/mg giving a molar ratio of 125 I/tau40 of approximately 0.5. In brief, 6 g of tau40 was incubated with 100 pmol of 125 I in 0.2 M phosphate, pH 8.0, containing 0.1 mg/ml chloramine T for 3 min at 20°C, followed by filtration through a 2-ml Sephadex G25 column equilibrated in 20 mM Hepes, pH 7.4, 0.01% Tween 20. The tracer was stored at Ϫ20°C. The iodinated tau40 CNBr fragments presented in Fig. 4B were prepared using the same procedure that yielded tracers of equal specific activities.

Affinity Purification of Human Brain Tau by ␣-Synuclein
Affinity Chromatography Recombinant human ␣-synuclein was coupled to CNBr-activated Sepharose 4B (Amersham Pharmacia Biotech) at a concentration of 5 mg of ␣-synuclein/ml of gel according to the recommendations of the manufacturer. An equal concentration of rabbit IgG raised against human plasminogen activator inhibitor-2 was coupled to CNBr-activated Sepharose 4B as control. Human brain cytosol was prepared according to Ref. 30. The brain cytosol supernatant (200 ml) was passed through the ␣-synuclein-Sepharose column (2 ml bed volume) at a flow rate of 0.25 ml/min at 4°C. The column was washed with 30 ml of phosphate-buffered saline and 1 mM EDTA, and the bound proteins were eluted by 50 mM glycine, pH 2.5, into 0.1 volume of 1 M Tris, pH 8.0.

Binding Assays
Microplate Assay-Recombinant ␣-synuclein (100 l of 15 g/ml) in 200 mM NaHCO 3 , pH 9.6, was immobilized on Maxisorb microtiter plates (Nunc) for 2 h on ice, and residual protein binding sites were blocked by further incubation with 5% bovine serum albumin (Sigma) for 2 h. After rinsing, the wells were incubated with 50 pM 125 I-tau40 in the presence of various concentrations of competitors for 16 h at 4°C in binding buffer (150 mM KCl, 2 mM MgCl 2 , 0.01% bovine serum albumin, 20 mM Hepes, pH 7.4). Unbound ligand was removed by rinsing three times in 200 l of binding buffer, and bound tracer was quantified by ␥-counting (Packard Cobra II) after release by incubation with 200 l of 5% SDS.
Plasmon Surface Resonance Assay-All measurements were performed on a BIAcore 2000 instrument (Biosensor, Uppsala, Sweden) equipped with CM5 sensor chips. The carboxylated dextran matrix of the sensor chip was activated by the injection of 60 l of a solution containing 0.2 M N-ethyl-NЈ-(3-dimethylaminopropyl)carbodiimide and 0.05 M N-hydroxysuccimide in water. Recombinant ␣-synuclein was then immobilized at a concentration of 50 g/ml in 10 mM sodium acetate, pH 3.5. A parallel channel of the sensor chip was derivatized under identical conditions in the absence of protein and was used as a control. The remaining binding sites were blocked with 1 M ethanolamine, pH 8.5. The surface plasmon resonance signal from immobilized ␣-synuclein generated 714 BIAcore response units equivalent to 44.6 fmol of ligand/mm 2 . Screening of peptides for binding to ␣-synuclein was performed by injecting aliquots (120 l) of samples (approximately 100 g of protein/ml) onto the derivatized sensor chip. The samples were in 10 mM Hepes, 150 mM NaCl, 1.5 mM CaCl 2 , 1 mM EGTA, pH 7.4, which was also used as running buffer. Regeneration of the sensor chip after each analysis cycle was performed by injecting 20 l of 1.5 M glycine/HCl buffer, pH 2.5, followed by 20 l of 6 M guanidinium HCl, pH 6.5.

CNBr Cleavage and Peptide Purification
Lyophilized tau40 (100 g) was dissolved in 50 l of 70% formic acid containing 0.2 mg of CNBr (Sigma) and incubated for 24 h at 20°C in the dark, whereafter the cleavage reaction was terminated by lyophilization. The peptides were purified on a RPC C2/C18 PC 3.2/3 reversed phase column (Amersham Pharmacia Biotech) using a SMART chromatography system (Amersham Pharmacia Biotech). The peptides were loaded on the column in 0.1% trifluoroacetic acid and eluted using a gradient of acetonitrile in 0.1% trifluoroacetic acid. The eluate was monitored using absorbances of 212 and 280 nm (Peak Monitor, Amersham Pharmacia Biotech) and collected on a Fraction collector (Amersham Pharmacia Biotech) in the peak fractionation mode. The samples specified in Fig. 4 were lyophilized and dissolved in Binding Buffer, and their purity was analyzed by SDS-PAGE. The protein concentration of the samples was analyzed by laser scanning densitometry of the Coomassie Brilliant Blue-stained bands by comparing the integrated densities of the peptides with a serial dilution from 5 to 0.25 g of bovine serum albumin. The synuclein binding activity of the samples was further analyzed by surface plasmon resonance analysis and, after iodination, by binding to immobilized recombinant synuclein (see above).

Tau Phosphorylation Assay
Prior to phosphorylation, tau isoforms and ␣-synuclein were dialyzed against water. The phosphorylation assay was carried out in 40 mM Tris-HCl, pH 7.4, 20 mM Mg acetate at 37°C. First, tau (1 M) was incubated with different concentrations of ␣-synuclein for different time periods to facilitate their interaction. Subsequently, phosphorylation was initiated by the addition of ATP (2 mM) and the catalytic subunit of protein kinase A (PKA; Promega) (0.5 units/ml) and the incubation was allowed to proceed another 2 h. The reaction was terminated by addition of 1/3 volume of SDS-gel electrophoresis sample buffer followed by heating to 95°C for 3 min. The samples were processed by SDS-PAGE and immunoblotted using the mouse monoclonal antibody 12E8 at a dilution of 1/50,000 for the detection of phosphorylated serines 262 and 356 (31). For the detection of all tau, irrespective of its phosphorylation state, the monoclonal antibody 12E8 was stripped of the membrane by three 10-min washes in 50 mM glycine, pH 2.5, and the membrane was further probed by the N-terminal specific tau antiserum BR133.

RESULTS
Cytosolic Tau Binds to ␣-Synuclein-Affinity chromatography of human brain cytosol preparations was performed on immobilized recombinant human ␣-synuclein to identify ␣-synuclein binding proteins. Fig. 1 shows that proteinaceous material only eluted from the ␣-synuclein-matrix but not from the control matrix. The proteins in fraction 4 from the ␣-synuclein affinity column comprised a complex mixture of proteins in the molecular range from 40 to 250 kDa as determined by silver staining of SDS-polyacrylamide gels (data not shown). We have previously demonstrated specific binding of biotinylated ␣-synuclein to human hippocampal tissue (22). Initial experiments showed that preincubation of the tissue sections with some antibodies against the neuronal cytoskeletal elements ␣or ␤-tubulin, neurofilament chains, and tau inhibited the binding of biotin-␣-synuclein, whereas antibodies against actin, neuronal enolase and synaptophysin had no effect (data not shown). Accordingly, we searched for candidate ␣-synuclein ligands among proteins associated with the cytoskeleton. Only tau was identified as two BR135 anti-tau immunoreactive species of 51 and 54 kDa ( Fig. 1, inset). Antitau serum BR133 decorated the same two bands (data not shown). No ␣-tubulin, ␤-tubulin, or neurofilament chains were detectable in the eluate even though the cytosolic preparation applied to the column contained these components (data not shown).
Synuclein and Tau Are Coexpressed in Axons of Cultured Fetal Rat Hippocampal Neurons-The isolation of tau by ␣-synuclein affinity chromatography suggested that the two molecules might interact in the nerve cell if present in the same cellular compartment. We investigated their colocalization in stage 5 differentiated fetal rat hippocampal neurons cultured for 18 days as these cells represent a well characterized neuronal model. The differentiated stage is demonstrated by the segregation of the somatodendritic marker MAP-2 from the axons that are tau-positive (Fig. 2, panel B versus C). Using the monoclonal synuclein antibody H3C, we localized synuclein in an intensely stained punctate pattern over the cell membrane of the somatodendritic compartment ( Fig. 2A), which repre-sented presynaptic terminals as shown by colocalization with synaptophysin (data not shown; Ref. 32). In Fig. 2A, these punctae have been placed a little out of focus to demonstrate the presence of ␣-synuclein in slender neurites of uniform diameter on the substratum. Controls incubated without primary antibodies showed no staining of these structures (data not shown). The slender ␣-synuclein-containing neurites of uniform diameter are compatible with axons ( Fig. 2A). An identical pattern of ␣-synuclein expression was obtained when using an affinity-purified rabbit antibody against recombinant human ␣-synuclein (data not shown). Tau was distributed in both the somatodendritic and the axonal compartment of the nerve cells as detected by the anti-tau sera BR134 (Fig. 2C) and BR135 (data not shown). These antisera recognize all tau isoforms irrespective of their state of phosphorylation. Accordingly, ␣-synuclein and tau are coexpressed in the axonal compartment (Fig. 2, panels D and F versus E and G).
Characterization of the Interaction between Tau and ␣-Synuclein-To elucidate if the two proteins interact directly, we next performed direct binding experiments of 125 I-labeled recombinant tau40 to ␣-synuclein immobilized in microtiter wells. The 125 I-tau40 tracer migrated predominantly as an approximately 60-kDa band with some minor impurities migrating as a smear around 30 kDa (Fig. 4, panel B, lane 3). Elution of the ␣-synuclein-bound 125 I-tau40 tracer followed by SDS-PAGE and autoradiography revealed that only the 60-kDa band was bound (data not shown). Fig. 3 demonstrates the binding of 16% of the 125 I-tau40 tracer (50 pM) to immobilized ␣-synuclein after a 16-h incubation at 4°C. Time-course experiments showed the tracer binding reached a plateau within 6 -8 h (data not shown). Half-maximal inhibition (IC 50 ) was obtained with about 50 nM tau40 (Fig. 3), and the IC 50 ranged from 10 to 70 nM in three experiments (data not shown). The binding between tau and ␣-synuclein appeared to involve ionic interactions as it was sensitive to increasing ionic strength. KCl concentrations of 300 and 600 mM, as compared with 150 mM, decreased the binding by 64% and 79%, respectively (data not shown). The tau-binding polyanionic glycosaminoglycan, heparin, also inhibited the tau binding to ␣-synuclein, indicating that charge interactions are important (data not shown).
We expressed and purified the two ␣-synuclein mutants Ala 30 3 Pro and Ala 53 3 Thr since they are associated with rare cases of early-onset Parkinson's disease (4,5). However, the mutations did not perturb the tau binding activity as determined by the direct 125 I-tau40 binding assays (data not shown). Immobilized recombinant human ␤-synuclein and ␥-synuclein bound 125 I-tau40 to about the same level as ␣-synuclein (data not shown).
Identification of the Synuclein Binding Domain in Tau-Our initial strategy for identifying the ␣-synuclein binding domains in tau was to fragment the protein and characterize the ␣-synuclein binding peptides. Recombinant human tau40 was fragmented by CNBr, and the digest was subsequently resolved by reversed phase chromatography (Fig. 4A). The two peaks, A and B, were resolved to near base-line levels upon elution of the C 18 column with a gradient of acetonitrile. The eluates corresponding to the dark bars were collected, lyophilized, and dissolved in equal volumes of binding buffer. Peaks A and B each contained one predominating peptide as judged by a Coomassie Blue stained SDS-polyacrylamide gel (Fig. 4B, lanes 1 and 2). Matrix-assisted laser desorption ionization-mass spectrometry of samples from peaks A and B detected only a single peptide in each peak, their masses being compatible with the recombinant tau40 CNBr cleavage products, amino acid residues 128 -250 and 251-419, respectively (data not shown). Their identity was further confirmed by immunoblotting with the epitope-specific FIG. 1. Purification of tau by ␣-synuclein affinity chromatography. A human brain cytosol preparation was passed through 2-ml columns containing either ␣-synuclein-Sepharose (5 mg/ml) (q) or antiplasminogen activator inhibitor-2 rabbit IgG-Sepharose (5 mg/ml) (E) and eluted by 50 mM glycine, pH 2.5. The ordinate shows the protein concentration as expressed by absorption at 280 nm. The presence of tau in the samples was analyzed by anti-tau immunoblotting (inset). Samples (20 l) from fractions 1-8 were supplemented with sample buffer, resolved by 8 -16% gradient SDS-polyacrylamide gel electrophoresis, electroblotted onto a polyvinylidene difluoride membrane, and probed by the anti-tau rabbit serum BR135 in a dilution of 1/1000. Bound antibody was detected by chemiluminescence. The two tau immunoreactive bands migrated corresponding to molecular masses of 51 and 54 kDa, respectively. Silver staining of a SDS-polyacrylamide gel corresponding to the above blotted gel revealed that bands corresponding to the molecular mass of tau represented a significant, but not major, fraction of the proteins eluted from the affinity column.
anti-tau sera BR133 and BR135, respectively. Only peptide B bound BR135, which is raised against amino acid residues 323-335 in the MT-binding domain of tau40 (Fig. 4B, lane 7), whereas none of the peptides were recognized by the N-terminal specific BR133 (data not shown). The small amounts of Coomassie Blue-stained higher molecular weight peptides in peak B (Fig. 4B, lane 2) also contained the epitope for BR135 (Fig. 4B, lane 7).
Peptides A and B were iodinated by the procedure described for tau40 yielding the tracers displayed in Fig. 4B (lanes 4 and  5). In the synuclein binding assay, 125 I-peptide B bound to ␣-synuclein whereas 125 I-peptide A did not bind (Fig. 4C). We next tested the binding activity of the unlabeled peptides A and B by surface plasmon resonance technique, which measures binding of peptide mass to immobilized ␣-synuclein (Fig. 4D). Application of 500 nM peptide B resulted in the binding corresponding to approximately 600 response units as compared with approximately 1100 response units for the positive control tau40 applied at the same molar concentration, and binding of both ligands was reversible upon removal of the unbound li-gand (Fig. 4D). The application of a similar concentration of peptide A to the ␣-synuclein chip generated a small response of about 20% of that of peptide B. Peptide A did not contain any contaminating peptide B as determined by immunoblotting using the BR135 antiserum or matrix-assisted laser desorption ionization-mass spectrometry (data not shown). The small response might reflect nonspecific adsorption or a low affinity binding not detected by the microtiter plate assay (Fig. 4C). Hence, the major ␣-synuclein binding site in tau resides in the C-terminal segment, which also contains the MT-BD (Fig. 5).
Different tau isoforms are expressed in the human central nervous system during development. They differ with regard to the presence of zero, one or two N-terminal insertions and the presence of 3 or 4 repeats in the MT-BD. The longest isoform of 440 amino acid residues, tau40 (Fig. 5) contains 2 N-terminal insertions and 4 MT binding repeats. As a further approach to identify the ligand binding segment in tau, we determined the inhibitory effects of truncated recombinant tau peptides on the binding of 125 I-tau40 to immobilized ␣-synuclein. Fig. 6 confirms that a binding site resides in the C-terminal part of tau as demonstrated by the inhibition by the tau40-(192-383) peptide as compared with the lack of inhibitory activity of the Nterminal tau40- . The MT-BD peptide, representing 3 repeats of the MT-BD, possessed full inhibitory activity like tau40 (Fig. 6). We next tested the inhibitory activity of MAP-2C, which is an isoform of the MAP-2 gene product expressed in the fetal brain. Its only structural homology to tau is in the 3 repeat MT-BD that is 69% identical to the repeats 1, 2, and 4 in tau40. Fig. 6 shows that MAP-2C inhibited the 125 I-tau40 binding to the same extent as tau40 and MT-BD. This demonstrates that a 3-repeat MT-BD in a non-truncated form, but without any flanking tau regions, possesses full inhibitory activity. Furthermore, 125 I-labeled purified recombinant tau39 and tau40 isoforms that contain 3 and 4 MT-binding repeats, respectively, bound equally well to immobilized ␣-synuclein (data not shown). We next tested the inhibitory efficacy of the synthetic 30-amino acid peptide Asp-Leu-Ser-Lys-Val-Thr-Ser-Lys-Cys-Gly-Ser-Leu-Gly-Asn-Ile-His-His-Lys-Pro-Gly-Gly-Gly-Gln-Val-Glu-Ile-Lys-Tyr-Glu-Lys. This peptide, corresponding to a single consensus repeat flanked by 4 N-terminal and 8 C-terminal linking residues, caused only a 60% inhibition of the 125 I-tau40 binding even when applied in concentrations up Identification of the Tau Binding Segment in ␣-Synuclein-The primary structure of ␣-synuclein is characterized by the presence of 7 imperfect Lys-Thr-Lys-Glu-Gly-Val consensus core repeats and a positive net charge within the N-terminal two thirds as compared with the remaining C-terminal part, which carries a strong negative charge and no known structural or repeated elements (Fig. 5). To determine the tau binding segment, we expressed and purified 3 deletion mutants, ␣-synuclein-(30 -140), ␣-synuclein-(55-140) and ␣-synuclein-(1-87), which lack the N-terminal 2 and 4 Lys-Thr-Lys-Glu-Gly-Val repeats and the acidic C terminus, respectively. Fig. 7 (upper panel) shows that the acknowledged slow migration of wild type ␣-synuclein (lane 1) is accentuated for the more acidic N-terminal deletion mutants. The identity of all peptides were verified by mass spectrometry and N-terminal amino acid se-quencing. When tested for tau binding activity in the microtiter plate assay, it was evident that the N-terminally truncated proteins retained their binding activity, whereas removal of the C-terminal residues 88 -140 reduced the binding by more than 90% (Fig. 7, lower panel). As a positive control for the correct immobilization of the ␣-synuclein-(1-87) peptide, we measured the binding of 125 I-A␤ to the same peptides. The 125 I-A␤ binding to ␣-synuclein-(1-87) corresponded to 70% of the binding to the wild type ␣-synuclein. This is compatible with the known presence of several A␤-binding sites, some of which are located within residues 1-87 (22,33).
The identification of the MT-BD of tau as the ␣-synuclein binding site made us investigate whether tubulin inhibits the interaction. Fig. 8 shows that tubulin indeed inhibits the binding of 50 pM 125 I-tau40 to immobilized ␣-synuclein (IC 50 about 500 nM) although less effectively than tau40 itself (Figs. 8 and  3). The inhibition of the binding was not due to tubulin binding to ␣-synuclein (data not shown). The competition between ␣-synuclein and tubulin for the MT-BD in tau was also demonstrated in assays using immobilized tau and radiolabeled tubulin (data not shown). Accordingly, ␣-synuclein is a ligand for soluble tau whereas MT-bound tau is unable to interact with ␣-synuclein. This explains why we were unable to demonstrate ternary complexes between MT-tau and ␣-synuclein in a classical MT spin-down assay (data not shown) and is in agreement with previous studies of MT-associated proteins where ␣-synuclein has not been recognized. Tau-Stimulation of tubulin assembly and stabilization of microtubules are recognized functions of tau caused by its interaction with mono-and polymeric tubulins. The interaction between tubulins and tau is known to be regulated by tau phosphorylation. The interaction between ␣-synuclein and the MT-BD of tau made us investigate whether ␣-synuclein is a candidate neuronal modulator of tau phosphorylation. The phosphorylation of tau is complex with multiple Ser/Thr residues being recognized by several kinases. As a model system for the study of the ␣-synuclein effect on tau phosphorylation we chose to study the phosphorylation of Ser 262/356 in tau by PKA. The reasons were: (i) PKA phosphorylation of tau generates phospho-epitopes recognized by the monoclonal antibody 12E8 in neuritic tangles, (ii) the PKA-catalyzed phosphorylation of Ser 262/356 is specifically recognized by 12E8 (31), and (iii) Ser 262/356 are within the proposed ␣-synuclein binding domain. of monoclonal antibody 12E8 as demonstrated by immunoblotting (Fig. 9A, lane 1). Inclusion of ␣-synuclein in the reaction mixture caused a dose-dependent increase in the 12E8 binding to tau when used at ␣-synuclein/tau ratios of 3, 6, 12, and 24 ( Fig. 9A, lanes 2-5). ␣-Synuclein stimulated phosphorylation of more than one site, as demonstrated by the change in the electrophoretic migration of the phosphorylated tau from a major 66-kDa 12E8 immuno-reactive band to a predominance of a slightly slower migrating species. Laser scanning densitometry of the tau bands in Fig. 9A revealed a 4-and 7-fold increase in the 12E8 immunoreactivity upon coincubation with synuclein/tau ratios of 6 and 24 as compared with tau alone. To check equivalent loading of tau in the lanes in Fig. 9A, we reprobed the membrane with the anti-tau serum BR133 (Fig.  9B, lanes 1-6). Note the presence of BR133 signal, but absence of 12E8 signal, in tau40 incubated with PKA in the absence of ATP (Fig. 9, panel B versus A, lane 6). Recombinant ␣-synuclein proteins containing the Parkinson's disease causing A30P and A53T mutations stimulated the PKA-stimulated tau phosphorylation to the same extent as the wild type peptide (data not shown). The stimulatory effect of ␣-synuclein on the PKAcatalyzed phosphorylation was most pronounced at lower kinase concentrations. Hence, ␣-synuclein in a 24-fold excess to tau increased the phosphorylation by 66% and 20% at kinase concentrations of 1.5 and 2.5 units/ml, respectively (Fig. 9C,  lanes 2 and 3 versus lanes 5 and 6). Fig. 9D shows that ␣-synuclein did not cause the exposure of new 12E8-reactive PKA substrate sites as the mutant tau24(Ser 262 3 Ala/Ser 356 3 Ala) was negative for 12E8 immunoreactivity after phosphorylation in the presence of ␣-synuclein (Fig. 9D, lanes 3 and 4  versus lanes 1 and 2). DISCUSSION We have identified the MT-binding protein tau as a ligand for ␣-synuclein by affinity chromatography of human brain cytosol and confirmed the direct interaction by in vitro binding assays using recombinant human ␣-synuclein and tau. ␣-Synuclein is recognized as a preterminal protein (34) and is, as such, not expected to colocalize with tau in the axon. However, in our study of axonal synuclein transport in the rat optic system, we find ␣-synuclein predominantly is moved by the slow component b of axonal transport and even with approximately 15% moved by the axoskeleton in slow component a. 2 Tau is predominantly moved by slow component a and to a minor extent slow component b (35). Hence, ␣-synuclein and tau have ample opportunities for interactions within the axonal compartment. The ␣-synuclein binding domain in tau was localized to the MT binding repeat region by binding analysis of tau40 fragments and competition analysis using truncated recombinant tau peptides. We observed no difference in the affinity for ␣-synuclein of tau isoforms having 3 or 4 repeats in the MT-BD, whereas a synthetic peptide corresponding to a single repeat displayed a very poor affinity as determined by its competition of tau40 binding to ␣-synuclein. Thus, a cooperative interaction of two or more of the MT-binding repeats seems to occur as also demonstrated for the interaction between tau and microtubules (36, 37). Other ligands binding to the MT binding domain in tau comprise tubulins, heparan sulfate proteoglycans, and presenilin-1 (38, 39), but ␣-synuclein is the first tau ligand whose expression is almost exclusively limited to the nervous system. The tau binding-site in ␣-synuclein resides in the acidic C-terminal 53 residues in analogy to the tau binding acidic C termini of tubulins (40). The importance of charge for the interaction is reflected in the sensitivity to increased ionic strength, the polyanion heparin, and the positive tau binding to ␤and ␥-synuclein, being only 36% and 17% identical in this region, but with a negative net charge. The tau binding site in ␣-synuclein is different from the binding sites for the previously recognized ligands, Alzheimer's disease amyloid peptide A␤ and rat brain vesicles, which both bind within the conserved N-terminal 95 residues that contain the synuclein-specific consensus repeats (20,22,33). This opens the possibility that ␣-synuclein might have bridging functions and bring different classes of ligands together.

␣-Synuclein Enhances the Protein Kinase A-catalyzed Phosphorylation of Serines 262 and 356 in
␣-Synuclein has not been recognized among the extensively studied group of MT-associated proteins despite its abundance in brain and its affinity for tau (for a recent review, see Ref. 41). Our data show that tubulin at high concentrations, as found locally in microtubules, inhibits the binding of ␣-synuclein to tau (Fig. 8). This points to ␣-synuclein as a ligand for the pool of soluble tau in contrast to the protein phosphatases 1 and 2A (42,43). An increased susceptibility of soluble tau to phosphorylation could be a functional consequence of the interactions as suggested by the ␣-synuclein-stimulated protein kinase Acatalyzed phosphorylation (Fig. 9). The ␣-synuclein-stimulated phosphorylation occurred at ␣-synuclein concentrations below 3 M that is lower than the synuclein concentration of about 0.1% of the total protein in rat brain homogenates (44). We show that at least two serine residues are susceptible to ␣-synuclein modulation of their protein kinase A-catalyzed phosphorylation by the presence of two 12E8 immunoreactive bands. However, other serines are likely to be phosphorylated, as demonstrated by the ␣-synuclein stimulated phosphorylation of the mutant tau24(Ser 262 3 Ala/Ser 356 3 Ala) that lacks the 12E8 epitopes (Fig. 9D, lanes 1 and 2). However, this question has to be addressed by a different experimental approach. Ser 262 is localized within the MT binding repeat region, and its phosphorylation inhibits tau binding to microtubules (45). Ser 262 phosphorylation is observed in neuritic tangles of Alzheimer's disease and the filamentous deposits in the tauopathies, frontotemporal dementia, and parkinsonism linked to chromosome 17, corticobasal degeneration, and progressive supranuclear palsy (1). A direct involvement of ␣-synuclein in the development of Ser 262 phosphorylation is possible in Alzheimer's disease, where the ␣-synuclein expression is increased at early disease stages (46) and ␣-synucleinpositive neuritic tangles have been identified in layers 5 and 6 of hippocampus (47). Moreover, coexisting abnormalities in the metabolism of tau and ␣-synuclein, leading to the formation of neuritic tangles and Lewy bodies, have been demonstrated within the same amygdala neurons in Alzheimer's disease, Parkinson's disease, and Lewy body dementia (48,49). The absent effect of the Parkinson's disease mutations on the ␣-synuclein-stimulated tau phosphorylation suggests abnormal tau interactions are unlikely to be operating in these families.
Molecular mechanisms modulating the phosphorylation state of tau rely on an interplay between kinase and phosphatases and possibly accessory proteins like ␣-synuclein and presenilin 1 (39). The transmembrane endoplasmic reticulum protein presenilin 1 facilitates tau phosphorylation by assembling the kinase, glycogen synthase kinase-3␤ and its substrate tau (39). Like presenilin 1, ␣-synuclein may facilitate tau phosphorylation by altering the presentation of specific serines to protein kinase A and other Ser 262 -reactive kinases, e.g. Ca 2ϩcalmodulin-dependent protein kinase II and microtubule affinity-regulating kinase (50) as well as non-Ser 262 -directed kinases. The structural change that alters the presentation of the specific serines may have other consequences, e.g. the facilitation of binding to hitherto unrecognized tau ligands. Hypothetically, the segregation of the binding sites for tau and brain vesicles to the C-and N-terminal parts of ␣-synuclein, respectively, may enable ␣-synuclein to function as a molecular tether that brings tau in proximity to vesicle surfaces, and other putative ligands specific for the N termini of ␣-synuclein. Such a vesicle-targeting of tau could be dynamically regulated by phospholipases that regulate the phospholipid composition of the vesicles. Phospholipase D2 converts the neutral phosphatidyl choline to the acidic phosphatidic acid, and this changed phospholipid ratio would favor binding of ␣-synuclein and its cargo (21). In the same instance, ␣-synuclein would inhibit the phospholipase D2 and indirectly the generation of its binding sites (17).
In conclusion, we identify soluble tau as a ligand for ␣-synuclein and show that ␣-synuclein can modulate the phosphorylation of tau. This opens the possibility that axonally transported ␣-synuclein, on its way to the synaptic terminal, indirectly can influence MT dynamics by decreasing the concentration of MT-binding-competent tau. The stimulatory effect of ␣-synuclein on tau phosphorylation, and the recently demonstrated phospholipase D2 inhibitory effect of ␣-synuclein (17), further demonstrate that ␣-synuclein has the propensity to affect diverse intracellular signaling pathways.