Role of Ser129 phosphorylation of α-synuclein in melanoma cells

α-Synuclein (α-syn) is an important molecule clinically, since mutations and copy number variations in its gene have been linked to familial Parkinson's disease (PD) (Polymeropoulos et al., 1997; Krüger et al., 1998; Singleton et al., 2003; Zarranz et al., 2004). In addition, inclusion bodies containing α-syn, referred to as Lewy bodies, are pathological hallmarks of familial and sporadic PD (Fujiwara et al., 2002; Tanji et al., 2006). Importantly, α-syn is predominantly phosphorylated at Ser129 (S129) in Lewy bodies of the PD brain (Fujiwara et al., 2002; Anderson et al., 2006). Thus, α-syn is a central molecule in understanding PD pathobiology, and its phosphorylation at S129 is thought to be involved in PD pathogenesis.

α-Syn is not only related to PD, but also to melanoma, which is the major cause of skin cancer death worldwide (Matsuo and Kamitani, 2010). We recently showed that α-syn is highly expressed in human melanoma cell lines (Matsuo and Kamitani, 2010; Lee and Kamitani, 2011). Moreover, α-syn is frequently expressed in human melanoma tissues, but it is undetectable in tissues of non-melanocytic cutaneous carcinoma and normal skin (Matsuo and Kamitani, 2010). These observations suggest that α-syn plays a role in the pathobiology of melanoma, as well as that of PD. Indeed, epidemiological studies have revealed co-occurrence of melanoma and PD in patients (Olsen et al., 2005; Olsen et al., 2006; Gao et al., 2009). Specifically, melanoma patients have an increased risk of subsequently developing PD (Olsen et al., 2006), and PD patients have an increased risk of subsequently developing melanoma (Olsen et al., 2005). Therefore, it is possible that α-syn is shared as a common pathogenetic component in these two diseases (Matsuo and Kamitani, 2010).

α-Syn is a small soluble protein containing 140 amino acids (17 kDa), and predominantly expresses in the brain. α-Syn localizes to synaptic vesicles on the presynaptic side and the nucleus – hence the name ‘synuclein’ (Maroteaux et al., 1988; Auluck et al., 2010). Although the exact function of α-syn is still unclear, substantial evidence now exists to suggest that this protein is predominantly natively unfolded in solution but can bind to phospholipid membranes by adopting an amphipathic helical conformation in its N-terminal amino acids (Bennett, 2005; McFarland et al., 2008).

α-Syn has at least three experimentally proven phosphorylation sites, S87, S129 and Y125 (Chau et al., 2009). Among these sites, S129 has been studied most in the neuropathology field as described above. With regard to the physiological role of the S129 phosphorylation, there have been several reports. Pronin et al. reported that the S129 phosphorylation leads to the dissociation of α-syn from lipids (Pronin et al., 2000). Lou et al. reported that the S129 phosphorylation reduces the ability of α-syn to regulate tyrosine hydroxylase (TH) and protein phosphatase 2A (Lou et al., 2010). Recently, McFarland et al. showed that the C-terminal peptide of S129-phosphorylated α-syn interacts preferentially with cytoskeletal proteins and vesicular trafficking proteins (McFarland et al., 2008), suggesting that the S129 phosphorylation likely has a profound effect on protein trafficking. Despite these observations, the physiological relevance of S129 phosphorylation is still unclear.

In this report, we investigated a novel, physiological role of the S129 phosphorylation in human melanoma SK-MEL28 and SK-MEL5 cells. We first identified an antibody specific to S129-unphosphorylated α-syn. Using this as well as an antibody specific to S129-phosphorylated α-syn, we investigated the subcellular localization of α-syn and found that the S129-phosphorylated form localizes to dot-like structures at the cell surface and the extracellular space. On the basis of these results and others, we have proposed that the phosphorylation of S129 plays a role in the translocation to the cell surface and subsequent vesicular release of α-syn in melanoma cells.

4D6 is a mouse monoclonal anti-α-syn antibody (Golovko et al., 2005), but its epitope had not been identified. We therefore performed epitope mapping. We expressed several deletion mutants of α-syn as fusion proteins in yeast cells and tested for reactivity with 4D6 using western blotting (Fig. 1A). The results mapped the epitope to within 11 amino acid residues (124–134) at the C-terminal region containing the S129 phosphorylation site (see lane 9). The results are summarized in Fig. 1B,C.

We next determined whether phosphorylation at S129 affects the 4D6 immunoreactivity to α-syn. We performed western blot analysis using both unphosphorylated and phosphorylated forms of α-syn. In brief, α-syn was expressed in bacterial cells and purified. The unphosphorylated recombinant α-syn protein was treated with casein kinase-2 (CK2), which is a serine/threonine-selective protein kinase (Fujiwara et al., 2002), to generate α-syn protein phosphorylated at S129. Both unphosphorylated and phosphorylated forms of α-syn were then analyzed by western blotting using anti-α-syn monoclonal antibodies LB509 [for total α-syn (Jakes et al., 1999)], EP1536Y [for S129-phosphorylated α-syn (Fournier et al., 2009; Qing et al., 2009)], and 4D6. As shown in Fig. 2A, LB509 detected both unphosphorylated and phosphorylated forms of α-syn. EP1536Y only detected α-syn that was phosphorylated by CK2. 4D6 detected α-syn that was not phosphorylated by CK2, but did not detect α-syn that was phosphorylated by CK2, suggesting that 4D6 selectively reacts with the S129-unphosphorylated form of α-syn.

We further confirmed the immunoreactivity of 4D6 to S129-unphosphorylated α-syn using another method. Specifically, α-syn (without tag) was overexpressed in SH-SY5Y cells (Chau et al., 2009) and resolved by two-dimensional gel electrophoresis, followed by western blot analysis using anti-α-syn antibodies LB509 (for total α-syn), EP1536Y (for S129-phosphorylated α-syn), and 4D6. As shown in Fig. 2B, LB509 detected at least five species of α-syn (∼17 kDa) at the pH range between 4.4 and 5.0. Since EP1536Y only detected spot no. 1, we identified spot no. 1 as S129-phosphorylated α-syn and spots no. 2–5 as S129-unphosphorylated species. 4D6 detected spots no. 2–5, but not spot no. 1 (bottom panel, Fig. 2B), indicating that 4D6 selectively recognizes α-syn that is not phosphorylated at S129.

α-Syn is predominantly expressed in neurons and localizes to synaptic vesicles and portions of the nucleus (Maroteaux et al., 1988). Recently, we found that α-syn is expressed more in human melanoma tissues and cell lines (Matsuo and Kamitani, 2010; Lee and Kamitani, 2011). However, its subcellular localization in melanoma cells has not been known. To determine this, we immunostained human melanoma SK-MEL28 cells using 4D6 antibody for α-syn unphosphorylated at S129 (Fig. 3A) and pSyn # 64 antibody (Fujiwara et al., 2002) for α-syn phosphorylated at S129 (Fig. 3B). It should be noted that both 4D6 and pSyn # 64 recognize the same 11 amino acids (residues 124–134) of human α-syn (see Fig. 1C). For this immunostaining, we counterstained endogenous NUB1 (Tanji et al., 2006; Tanji et al., 2007) using rabbit anti-NUB1 antibody, because the counterstaining allowed us to clearly see the cell shape and cell surface staining (see below). As shown in Fig. 3, S129-unphosphorylated α-syn localized to the cytoplasm as small dots, but its nuclear localization was limited (left panel, Fig. 3A). In contrast, S129-phosphorylated α-syn localized to the nucleus and cytoplasm (left panel, Fig. 3B). It should be noted that it also localized to the cell surface, which was clearly observed at the dendritic processes of SK-MEL28 cells (see the magnified image on right, Fig. 3B). It appeared that the small dot-like structures containing S129-phosphorylated α-syn were released from the cell surface to the outside (see below for the detail). Importantly, such release of dot-like structures could not be observed on the cell surface stained with 4D6 antibody (see the magnified image on right, Fig. 3A). These results suggest that the phosphorylation of α-syn at S129 is involved in its cell surface localization as well as its nuclear localization. In Fig. 3C, we show additional cells immunostained with 4D6 or pSyn#64 to support the findings described above.

When SK-MEL28 cells were stained with pSyn#64 (Fig. 4B), but not with 4D6 antibody (Fig. 4A), many dot-like structures were observed around the cells. Since the staining intensity of each dot-like structure was weak, longer exposure was needed for the demonstration in the immunofluorescence microscopy. However, the dot-like structure with S129-phosphorylated α-syn clearly existed around the cells (which looked like a starry sky), suggesting that the phosphorylation of α-syn at S129 is involved in its cell surface localization and subsequent release.

As described above, we observed differences in 4D6 and pSyn#64 immunostainings. To confirm that the differences do not result from non-specific reaction of antibodies, we examined the immunoreactivity of 4D6 and pSyn#64 to α-syn-knockdown SK-MEL28 cells. For this purpose, we stably knocked down endogenous α-syn in SK-MEL28 cells by using lentivirus-mediated RNAi. In brief, SK-MEL28 cells were infected with lentiviruses to express five different sequence-verified shRNAs. We then chose one SK-MEL28 cell line (out of five cell lines) that was infected with the most efficient lentivirus vector harboring the shRNA sequence for α-syn RNAi, based on western blotting using anti-α-syn antibody 4D6. As shown in the upper panel of Fig. 5A, the expression of endogenous α-syn was interfered in all cell lines expressing α-syn shRNA (no. 1–5). Because endogenous α-syn was strongly knocked down in cell line no. 3, we chose this cell line for further experiments. In addition to the cell line no. 3, we established a cell line expressing control shRNA (first lane).

Next, we immunostained these SK-MEL28 cell lines with 4D6 and pSyn#64 antibodies. As shown in Fig. 5B, cells expressing control shRNA were strongly immunostained with both antibodies (left panels), but cells with knockdown of α-syn were not (right panels). Thus, we confirmed the specificity of 4D6 and pSyn#64 antibodies in α-syn immunostaining.

To further investigate the cell surface of SK-MEL28 cells, we performed electron microscopy. Fig. 6A,D shows two representative images. In both panels, many small structures exist at the cell surface. In the magnified images (Fig. 6B,E), we can clearly observe cross and longitudinal sections of microvilli that contain microfilaments, as reported previously (Rodriguez-Vicente et al., 1996). In addition, there are microvesicles (MVs), whose average diameter is ∼150 nm (arrows, Fig. 6B,E). We also observed MVs that were budding from or fusing to the plasma membrane (arrowhead, Fig. 6C).

Next, we performed immuno-electron microscopy to investigate the ultrastructural localization of S129-phosphorylated α-syn at the cell surface and underneath the plasma membrane. As shown in Fig. 7, immuno-electron microscopy revealed that S129-phosphorylated α-syn localizes to small structures (40 to 60 nm in diameter) underneath the plasma membrane (arrows, Fig. 7A). We speculate that S129-phosphorylated α-syn is either oligomerized into small aggregates or accumulated within the 40–60 nm structures. In addition, S129-phosphorylated α-syn localizes to the membranes of MVs, whose average diameter is ∼150 nm (arrows, Fig. 7B–D). Thus, our immuno-electron microscopy revealed that SK-MEL28 cells release S129-phosphorylated α-syn as MVs.

In human melanoma SK-MEL28 cells, S129-phosphorylated α-syn localized to the cell surface as well as the nucleus. To determine whether this is observed in other human melanoma cell lines, we tested the SK-MEL5, A375, MeWo and WM266-4 melanoma cell lines. First, we examined the expression levels of endogenous α-syn using different antibodies, including LB509 (for total α-syn), 4D6 (for S129-unphosphorylated α-syn), pSyn#64 and EP1536Y (for S129-phosphorylated α-syn). As shown in Fig. 8A, SK-MEL5, MeWo and WM266-4 cells, as well as SK-MEL28 cells, expressed both S129-unphosphorylated and phosphorylated forms. Notably, SK-MEL5 cells expressed them at much higher levels. However, the expression levels were extremely low in A375 cells as we described previously (Lee and Kamitani, 2011).

Next, we immunostained HT1080 (as negative control), SK-MEL5, MeWo and WM266-4 cells using anti-α-syn antibodies 4D6 and pSyn#64. A375 cells were not used because S129-phosphorylated α-syn was almost undetectable by western blotting.

As shown in Fig. 8B, S129-unphosphorylated α-syn localized to the cytoplasm as small dots, but its nuclear localization was limited in three melanoma cell lines (top panels). In contrast, S129-phosphorylated α-syn localized to the nucleus and cytoplasm in these cell lines (middle and bottom panels). It should be noted that it also localized to the cell surface in SK-MEL5 cells, but not in MeWo and WM266-4 cells (see the magnified images at the bottom). In supplementary material Fig. S1, we showed additional SK-MEL5 cells immunostained with 4D6 or pSyn#64 to support these findings. Thus, we found that S129-phosphorylation plays a role in localization of α-syn to the cell surface in two (SK-MEL5 and SK-MEL28) out of five human melanoma cell lines tested in this study.

Moreover, we showed that the dot-like structures with S129-phosphorylated α-syn existed around SK-MEL5 cells (data not shown), suggesting that the phosphorylation of α-syn at S129 is involved in its cell surface localization and subsequent release in SK-MEL5 as well as SK-MEL28 cells (also see Fig. 4).

Previously, tubulin was revealed to be an α-syn-binding protein (Alim et al., 2004). In neurons, α-syn was shown to be transported by both kinesin and dynein motor proteins along microtubules (Utton et al., 2005). These observations suggested that a transport system along microtubules is involved in translocation of α-syn in melanoma cells. To test this possibility, we compared the distributions of dot-like structures of α-syn and the microtubule network in melanoma cells. Specifically, we double-immunostained endogenous α-syn (unphosphorylated or phosphorylated at S129) and endogenous α-tubulin of the microtubule network in flat extended SK-MEL5 cells to determine their distributions by fluorescence microscopy. As shown in the upper panels of Fig. 9, α-tubulin formed a fine network structure of microtubules, while S129-unphosphorylated α-syn was observed as dot-like structures. When the images were merged, the microtubule network was abundantly decorated with dot-like structures of α-syn (Fig. 9A). Importantly, the magnified image clearly showed that the dot-like structures are located along the microtubules (Fig. 9B). In contrast, S129-phosphorylated α-syn localized to the nucleus and cell surface predominantly (Fig. 9C), as described above. The magnified image showed that the dot-like structures were located along the microtubules (Fig. 9D). These results suggest that dot-like structures with unphosphorylated α-syn locate on the microtubule network and, upon phosphorylation at S129, they are transported to the nucleus and cell surface.

α-Syn has at least three experimentally proven phosphorylation sites (Chau et al., 2009), of which the S129 residue is thought to be the major site (Wang et al., 2012). The phosphorylation of S129 plays an important role in the pathobiology of neurodegenerative α-synucleinopathy, because it is selectively and extensively phosphorylated in α-synucleinopathy lesions such as the Lewy bodies and Lewy neurites found in the PD brain (Fujiwara et al., 2002; Anderson et al., 2006). However, little is currently known of the physiological (not pathological) relevance of this phosphorylation, partly due to the technological difficulties of the required experiments. Specifically, to clearly determine the physiological role of S129 phosphorylation, we have to examine S129-unphosphorylated α-syn as well as S129-phosphorylated α-syn. For this purpose, two different antibodies to α-syn are needed: one antibody for α-syn unphosphorylated at S129 and another antibody for α-syn phosphorylated at S129. However, although the latter antibody is commercially available, no antibody for S129-unphosphorylated α-syn had previously been identified, making it difficult to study the physiological role of S129 phosphorylation.

In our present study, we identified 4D6 as an antibody to S129-unphosphorylated α-syn. Importantly, 4D6 recognizes the same C-terminal region as pSyn#64 does. Unlike 4D6, however, pSyn#64 is an antibody to α-syn phosphorylated at S129 and has commonly been used for immunostaining to determine the localization of S129-phosphorylated α-syn (Fujiwara et al., 2002; Tanji et al., 2006; Tanji et al., 2011). Therefore, we expected that immunostaining using both 4D6 and pSyn#64 antibodies would allow us to more clearly and exactly uncover the role of S129 phosphorylation in the subcellular localization of α-syn. Using these antibodies, we immunostained human melanoma SK-MEL28 and SK-MEL5 cells that highly express endogenous α-syn (Matsuo and Kamitani, 2010; Hansen et al., 2011; Lee and Kamitani, 2011). The immunostaining showed that S129-phosphorylated α-syn localizes to the cell surface as small dots, but S129-unphosphorylated α-syn does not. Importantly, we also detected the released dot-like structures containing S129-phosphorylated α-syn by using immunofluorescence microscopy. Furthermore, immuno-electron microscopy revealed that SK-MEL28 melanoma cells release MVs in which S129-phosphorylated α-syn localizes to the vesicular membrane. These results suggest that the S129 phosphorylation plays a role in the cell surface localization and subsequent vesicular release of α-syn in some of the human melanoma cells.

α-Syn release has intensively been studied using neuronal, but not melanoma, cells. These studies revealed that α-syn is released from neuronal cells in a constitutive manner (Lee et al., 2005; Lee et al., 2008). This is supported by a previous clinical finding that α-syn is present in human cerebrospinal fluid at low nanomolar concentrations (Borghi et al., 2000; El-Agnaf et al., 2003). Since both neuronal cells and melanocytes developmentally originate from the neural crest and share several features, the α-syn release may be a common feature in both cell types.

How is α-syn release regulated in melanoma cells? As described above, our immunofluorescence microscopy revealed that the dot-like structures are immunostained by pSyn#64 antibody, but not by 4D6 antibody, at the cell surface and outside of melanoma cells. These results suggest that S129 phosphorylation regulates the α-syn release in melanoma cells. Indeed, McFarland et al. demonstrated that S129 phosphorylation enables α-syn to interact with vesicular trafficking proteins of mouse brain (McFarland et al., 2008). Most recently, Wang et al. detected the selective elevation of S129-phosphorylated α-syn in the cerebrospinal fluid of PD patients (Wang et al., 2012). On the basis of these observations and our results, S129 phosphorylation might regulate the release of α-syn through the vesicular trafficking proteins in melanoma cells as well as neuronal cells. Importantly, our double immunostaining of α-syn and α-tubulin suggested that, upon phosphorylation at S129, α-syn is trafficked along the microtubule network.

What is the biological and clinical relevance of the α-syn release in melanoma? Although not known for melanoma, this has been well investigated in neuronal cells, and an interesting hypothesis has been proposed for α-syn release in the pathogenesis of PD. In this hypothesis, molecules of α-syn with an abnormal structure are released from one neuron and taken up by another, leading to neurodegeneration by prion-like mechanisms (Frost and Diamond, 2010; Olanow and McNaught, 2011; Steiner et al., 2011). In melanoma, epidemiological studies have revealed co-occurrence of melanoma with PD (Olsen et al., 2005; Olsen et al., 2006; Gao et al., 2009). Specifically, melanoma patients have an increased risk of subsequently developing PD (Olsen et al., 2006). Therefore, it is possible that α-syn release from melanoma cells plays a role in the pathogenesis or progression of PD, if it is propagated into neuronal cells. Indeed, this possibility is partially supported by a recent observation that α-syn is propagated from human melanoma SK-MEL5 cells to mouse neuroblastoma N2a cells in co-culture conditions (Hansen et al., 2011).

In conclusion, we have reported that the S129-phosphorylated form, but not the S129-unphosphorylated form, of α-syn localizes to dot-like structures at the cell surface and the extracellular space of melanoma cells. Our studies suggest that the phosphorylation of S129 leads to the cell surface translocation of α-syn along the microtubule network and subsequent vesicular release in melanoma cells. Further studies will be required to determine the clinical relevance of this finding.

The following human cell lines were purchased from American Type Culture Collection (Manassas, VA): SH-SY5Y (human neuroblastoma), SK-MEL28 (human malignant melanoma), SK-MEL5 (human malignant melanoma), A375 (human malignant melanoma), MeWo (human malignant melanoma), WM266-4 (human malignant melanoma), and HT1080 (human lung fibrosarcoma). All cells were maintained in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal calf serum (FCS) and antibiotics.

Rat monoclonal anti-α-tubulin (YL1/2) antibody and mouse monoclonal anti-α-syn antibodies LB509 (specific for amino acid residues 115–122) and 4D6 were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Mouse monoclonal anti-phospho α-syn antibody pSyn#64 (specific for amino acid residues 124–134) was purchased from Wako Chemicals USA (Richmond, VA). Rabbit monoclonal anti-phospho α-syn antibody EP1536Y (specific for amino acid residues 120–140) was purchased from Epitomics (Burlingame, CA). Mouse monoclonal anti-HA antibody 16B12 was purchased from Covance (Richmond, CA). Rabbit polyclonal anti-NUB1 antibody was generated by immunization with a GST fusion protein of NUB1, corresponding to amino acid residues 432–601 (Kito et al., 2001).

For epitope mapping of mouse monoclonal anti-α-syn antibody 4D6, several truncated fragments of α-syn were expressed in yeast cells and detected by western blotting using 4D6. In brief, the cDNAs of truncated α-syn were amplified by polymerase chain reaction (PCR) from pcDNA3/α-syn (see Two-dimensional gel electrophoresis below) using appropriate primers. These cDNAs were inserted into the multi-cloning site of yeast expression vector pGADT7 (Clontech, Mountain View, CA). Since the multi-cloning site was located downstream of the cDNA encoding Gal4-activating domain (AD) and HA tag, the plasmids allow us to express AD-HA-fused α-syn fragments in yeast cells. For the expression, AH109 yeast cells were transformed with the plasmids using the lithium acetate method (Okura et al., 1996). Transformed cells were grown on a Leu⁻ synthetic agar plate for 3 days at 30°C. The yeast cells were then grown in a Leu⁻ synthetic medium. Afterwards, using the boiling-SDS-glass bead method (Matsuo et al., 2004), the total cell lysate was prepared for western blotting.

The cDNA of wild-type α-syn was subcloned into pT7-7 expression plasmid vector (USB, Cleveland, OH) for inducible expression in Escherichia coli. Rosetta (DE3) expression host cells (Novagen, Madison, WI) were then transformed with the plasmid and grown in 1l of LB medium containing ampicillin and chloramphenicol at 37°C with shaking. When the optical density of the culture medium reached 0.6 at 600 nm, α-syn was induced by adding 1 mM isopropyl thiogalactoside, followed by an additional culture for 3 h. Cells were harvested by centrifugation and the pellet was subjected to purification essentially as described previously (Volles and Lansbury, 2007). Briefly, cells were lysed by boiling in the buffer [50 mM Tris-HCl (pH 8.0), 10 mM EDTA, 150 mM NaCl]. Nucleic acids were removed by precipitation with streptomycin sulfate and acetic acid. α-Syn was subsequently purified by ammonium sulfate precipitation. The pellet was resuspended in ammonium acetate, followed by ethanol precipitation three times. The final pellet was resuspended in ammonium acetate, aliquoted into appropriate amounts, subjected to lyophilization, and stored until use for “in vitro phosphorylation assay”.

Recombinant human α-syn (3.5 µg) was incubated in 10 µl of in vitro phosphorylation buffer [20 mM Tris-HCl (pH 7.5), 50 mM KCl, 10 mM MgCl₂, 5 mM ATP] (Fujiwara et al., 2002) with or without 500 units of purified casein kinase-2 (CK2; New England Biolabs, Ipswich, MA) for 18 h at 30°C. After the kinase reaction, protein samples were heated for 1 h in 2% SDS treating solution and analyzed by western blotting.

Protein samples were heated for 30 min in a sample-treating solution containing 2% SDS and 5% β-mercaptoethanol. After SDS-PAGE, proteins were transferred to a polyvinylidene difluoride (PVDF) membrane, Immobilon P (Millipore, Bedford, MA). Western blotting was then performed as described previously (Matsuo and Kamitani, 2010; Lee and Kamitani, 2011). Horseradish peroxidase (HRP)-conjugated antibodies against mouse IgG or rabbit IgG (Santa Cruz Biotechnology) were used as secondary antibodies.

The cDNA of human α-syn (wild-type) was amplified using PCR with appropriate primers from a human brain cDNA library (Invitrogen, Carlsbad, CA). The cDNA sequence was confirmed by automated DNA sequencing. The cDNA was inserted into pcDNA3 (Invitrogen) to generate the plasmid pcDNA3/α-syn. The plasmid was then transfected into SH-SY5Y cells using Lipofectamine 2000 (Invitrogen) to express the wild-type α-syn without a tag. Twenty hours after transfection, the cells were collected by centrifugation, lysed in 200 µl of lysis buffer [20 mM Tris-HCl (pH 7.5), 400 mM NaCl, 0.5% NP-40, 10 mM NaF, 1 mM Na₃VO₄, 1 mM EDTA, protease inhibitor cocktail (Roche Applied Science, Indianapolis, IN)] and centrifuged at 18,000 g for 10 min at 4°C. The supernatant was collected and desalted by Sephadex G-25 spin column equilibrated with gel filtration buffer [5 mM Tris-HCl (pH 7.5), 5 mM NaF, 1 mM Na₃VO₄, 1 mM EDTA, protease inhibitor cocktail]. Using the desalted supernatant, two-dimensional gel electrophoresis was performed according to the method reported by Alberio et al. (Alberio et al., 2010), with minor modifications. Briefly, the supernatant (50 µl) was diluted to 320 µl with a sample buffer [8 M urea, 2% CHAPS, 0.5% IPG buffer 3.5–5.0 (GE Healthcare Biosciences), 30 mM dithiothreitol, and traces of bromophenol blue], and loaded on an 18 cm IPG Dry Strip (GE Healthcare Biosciences) with a 3.0–5.6 pH gradient for in-gel rehydration (13 h at 20 V). Isoelectrofocussing was performed at 20°C on Ettan IPGphor3 (GE Healthcare Biosciences) with the schedule: 1 h at 200 V, 1 h linear gradient to 500 V, 2 h linear gradient to 1000 V, 3 h linear gradient to 8000 V, and 3 h at 8000 V. The IPG strip was then equilibrated for 15 min in the SDS treating buffer [50 mM Tris-HCl (pH 8.8), 6 M urea, 30% glycerol, 2% SDS, 1% dithiothreitol, and traces of bromophenol blue]. The IPG strip was placed on a 15%, 1.0 mm-thick separating polyacrylamide gel. SDS-PAGE as a second dimension was carried out at 15 mA (current constant) using Mini-PROTEAN Tetra system (Bio-Rad, Hercules, CA).

Human cell lines were cultured on a coverslip in a 3.5-cm dish. After 24 h, the cells were fixed in a 4% paraformaldehyde solution for 20 min, stained with 0.1 µg/ml of 4′,6-diamidine-2-phenylindole dihydrochloride (DAPI; Roche Diagnostics, Indianapolis, IN) for 10 min, and permeabilized with 0.1% Triton X-100 for 10 min at room temperature. The cells were first labeled with rabbit anti-NUB1 antibody (dilution 1∶5000) and mouse monoclonal anti-α-syn antibody 4D6 (IgG₁ isotype; dilution 1∶2000) or mouse monoclonal anti-phospho α-syn antibody pSyn#64 (IgG₁ isotype; dilution 1∶3000) overnight at 4°C. After washing, the cells were labeled with Alexa Fluor 594-conjugated anti-rabbit IgG (dilution 1∶1000) for NUB1 and Alexa Fluor 488-conjugated anti-mouse IgG (dilution 1∶1000) for α-syn or phospho α-syn for 1 h at room temperature. These secondary antibodies were purchased from Molecular Probes (Eugene, OR). Finally, the cells were analyzed by an Axio Imager M1 microscope (Zeiss, Thornwood, NY). The localization of α-syn or phospho α-syn was shown by the green fluorescence of Alexa Fluor 488. The localization of NUB1 was shown by the red fluorescence of Alexa Fluor 594. Their co-localization was examined by the merging of both fluorescent images.

SK-MEL5 cells were cultured on a coverslip, fixed in methanol at −20°C, permeabilized in PBS containing 0.1% Triton X-100, blocked with 5% horse serum, and then immunostained with rat monoclonal anti-α-tubulin (YL1/2) antibody and mouse monoclonal anti-α-syn antibody (4D6 or pSyn#64) as we described previously (Tanaka et al., 2010). The localization of endogenous α-tubulin was shown by the green fluorescence of Alexa Fluor 488. The localization of endogenous α-syn was shown by the red fluorescence of Alexa Fluor 555. Nuclear counterstaining is shown by the blue fluorescence of DAPI. Their colocalization was examined by the merging of fluorescent images.

We stably knocked down endogenous α-syn by RNAi and established an α-syn-knockdown cell line. We used lentivirus-mediated RNAi, called Mission® short hairpin RNA (shRNA) system (Sigma-Aldrich). Five sequence-verified shRNA lentiviral plasmids (pLKO.1-puro) for α-syn knockdown were purchased from Sigma-Aldrich. Lentiviruses were generated by triple transfection of HEK293T cells with pLKO.1-puro plasmid (for shRNA), pMDG-2 plasmid (for envelope), and psPAX2 plasmid (for gag-pol) using FuGENE6 (Roche). After 24 h, lentivirus-containing medium was harvested. SK-MEL28 cells were then infected using the lentivirus-containing medium and selected by puromycin. Finally, we chose one SK-MEL28 cell line (out of five cell lines) that was infected with the most efficient lentivirus vector harboring the shRNA sequence for α-syn RNAi, based on western blotting using anti-α-syn antibody 4D6.

SK-MEL28 cells were fixed with 0.1 M sodium cacodylate (NaCac) buffer (pH 7.4) containing 2% glutaraldehyde for 3 min on a culture dish, scraped with a cell scraper, and pelleted by centrifugation. After 1 h, the cells were postfixed in 0.1 M NaCac buffer (pH 7.4) containing 2% osmium tetroxide for 1 h, stained en bloc with 2% uranyl acetate, dehydrated with a graded ethanol series, and embedded in epon-araldite resin. Thin sections were cut with a diamond knife on an EM UC6 ultramicrotome (Leica Microsystems, Inc., Bannockburn, IL), collected on copper grids, and sequentially stained with 2% uranyl acetate for 5 min and Reynolds lead citrate for 3 min. The sections were observed by a JEM 1230 transmission electron microscope (JEOL USA Inc., Peabody, MA) at 110 kV and imaged with UltraScan 4000 CCD Camera and First Light Digital Camera Controller (Gatan Inc., Pleasanton, CA).

SK-MEL28 cells were fixed on a culture dish for 3 min in 0.1 M NaCac buffer (pH 7.4) containing 4% paraformaldehyde and 0.2% glutaraldehyde, scraped with a cell scraper, and pelleted by centrifugation. After 1 h, the cells were dehydrated with a graded ethanol series and embedded in LR White resin (Electron Microscopy Sciences, Hatfield, PA). Thin sections were cut with a diamond knife on an EM UC6 ultramicrotome and collected on nickel grids. Sections were incubated in the blocking buffer [50 mM Tris-HCl (pH 7.4), 118 mM NaCl, 10 mM NaF, 5% BSA, 3% normal donkey serum, 0.05% Tween-20] at room temperature for 2 h, followed by incubation with mouse monoclonal anti-phospho α-syn antibody pSyn#64 (dilution 1∶75) overnight at 4°C. The sections were washed and stained for 2 h at room temperature with donkey anti-mouse IgG (Abcam, Cambridge, MA) that was conjugated with colloidal gold particles (15 nm in diameter) as described previously (Slot and Geuze, 1985). After washing, the sections were stained with 2% uranyl acetate alone for the cytoplasm (Fig. 7A) or with 2% uranyl acetate and 0.04% bismuth subnitrate sequentially for cell surface structures (Ainsworth et al., 1972) (Fig. 7B–D). The stained sections were then observed by a JEM 1230 transmission electron microscope as described above.

We thank Mr Robert Smith (Georgia Regents University) for excellent technical assistance.