Synphilin-1A Inhibits Seven in Absentia Homolog (SIAH) and Modulates α-Synuclein Monoubiquitylation and Inclusion Formation*

Parkinson disease (PD) is characterized by the presence of ubiquitylated inclusions and the death of dopaminergic neurons. Seven in absentia homolog (SIAH) is a ubiquitin-ligase that ubiquitylates α-synuclein and synphilin-1 and is present in Lewy bodies of PD patients. Understanding the mechanisms that regulate the ubiquitylation of PD-related proteins might shed light on the events involved in the formation of Lewy bodies and death of neurons. We show in this study that the recently described synphilin-1 isoform, synphilin-1A, interacts in vitro and in vivo with the ubiquitin-protein isopeptide ligase SIAH and regulates its activity toward α-synuclein and synphilin-1. SIAH promotes limited ubiquitylation of synphilin-1A that does not lead to its degradation by the proteasome. SIAH also increases the formation of synphilin-1A inclusions in the presence of proteasome inhibitors, supporting the participation of ubiquitylated synphilin-1A in the formation of Lewy body-like inclusions. Synphilin-1A/SIAH inclusions recruit PD-related proteins, such as α-synuclein, synphilin-1, Parkin, PINK1, and UCH-L1. We found that synphilin-1A robustly increases the steady-state levels of SIAH by decreasing its auto-ubiquitylation and degradation. In addition, synphilin-1A blocks the ubiquitylation and degradation of the SIAH substrates synphilin-1 and deleted in colon cancer protein. Furthermore, synphilin-1A strongly decreases the monoubiquitylation of α-synuclein by SIAH and the formation of α-synuclein inclusions, supporting a role for monoubiquitylation in α-synuclein inclusion formation. Our results suggest a novel function for synphilin-1A as a regulator of SIAH activity and formation of Lewy body-like inclusions.

Parkinson disease (PD) 3 is characterized by progressive degeneration of dopaminergic neurons in the substantia nigra and the presence of Lewy bodies in surviving neurons (1). The majority of PD cases are sporadic, but mutations in different genes have been found responsible for familial PD (1,2). ␣-Synuclein plays a crucial role in the disease. It is mutated in some familial forms of PD and represents a major component of Lewy bodies in sporadic PD as well (3,4).
Dysfunction of the ubiquitin-proteasome system has been proposed to play a role in PD as the proteasome activity is decreased in the substantia nigra of PD patients (5). Highlighting the role of the ubiquitin-proteasome system in PD is the finding that proteins mutated in the disease, such as Parkin and UCH-L1, are also components of this system (6,7). Furthermore, different PD-related proteins, including ␣-synuclein, synphilin isoforms, UCH-L1 and PINK1, were shown to be ubiquitylated and to accumulate into ubiquitylated inclusions (8 -14).
In addition to ubiquitylated synphilin-1, SIAH interacts with and monoubiquitylates ␣-synuclein at lysines found to be ubiquitylated in ␣-synuclein purified from Lewy bodies (8,14). Monoubiquitylation of ␣-synuclein leads to its aggregation and inclusion formation, which are toxic to dopaminergic cells (14). The presence of SIAH in Lewy bodies of PD patients (8) indicates that it might represent an additional component of the ubiquitin-proteasome system involved in the disease.
We recently identified a novel synphilin-1 isoform, synphilin-1A, that is expressed in the brain of different ␣-synucleinopathies and is a neurotoxic and aggregation-prone protein (25). The importance of synphilin-1A in PD is suggested by its ability to interact with ␣-synuclein and synphilin-1, and to accumulate in insoluble brain fractions of patients with Diffuse Lewy Body disease (25). In this study, we sought to investigate the interaction between SIAH and synphilin-1A. We present evidence that synphilin-1A interacts with SIAH. We found that SIAH ubiquitylates synphilin-1A and increases the formation of synphilin-1A inclusions, but it does not promote synphilin-1A degradation. On the other hand, synphilin-1A decreases SIAH E3 ubiquitin-ligase activity as well as toxicity. Synphilin-1A also decreases the monoubiquitylation of ␣-synuclein promoted by SIAH as well as the formation of ␣-synuclein inclusions. Our data indicate that synphilin-1A is a regulator of SIAH activities, with implications for regulation of ␣-synuclein monoubiquitylation and aggregation.
Cell Culture and Transfections-HEK293 and SH-SY5Y cells were grown in Dulbecco's modified Eagle's medium containing 10% fetal bovine serum in a 5% CO 2 atmosphere. Cells were transiently transfected with N-terminal tagged pRK5 and pFLAG-CMV-2 plasmids utilizing Lipofectamine 2000 (Invitrogen) and processed after 36 h.
In Vitro Binding Assays-Extracts of transfected HEK293 cells were incubated with 5 g/ml GST fusion proteins for 2 h at 4°C in buffer containing 50 mM Tris-HCl, pH 7.4, 140 mM NaCl, 1% Triton X-100, 0.1% SDS, 30 M MG132 and protease inhibitors (Complete; Roche Applied Science). Beads were washed in the same buffer. Bound proteins were detected by Western blot.
Co-immunoprecipitation Assays-Transfected HEK293 cells were lysed in buffer containing 50 mM Tris, pH 7.4, 140 mM NaCl, 1% Triton X-100, 0.1% SDS, 30 M MG132, and protease inhibitor mixture (Complete; Roche Applied Science). Cell extracts were clarified by centrifugation at 13,000 ϫ g for 5 min, and the supernatant was incubated with anti-HA for 4 h as described (8). Immunoprecipitates were washed with lysis buffer containing 500 mM NaCl and 1% Chaps and detected by Western blot.
In Vitro Ubiquitylation Assays-For the in vitro ubiquitylation assays, synphilin-1A, synphilin-1, and SIAH were translated using TNT wheat germ in vitro translation kit (Promega) using [ 35  In Vivo Ubiquitylation Assays-Transfected HEK293 or SH-SY5Y cells were incubated in the absence or in the presence of 10 M lactacystin for 12 h. Transfected cells were then resuspended in buffer containing 50 mM Tris, pH 7.4, 140 mM NaCl, 1% SDS and boiled for 5 min. Cell extracts were sonicated, diluted 10-fold with buffer containing 50 mM Tris, pH 7.4, 140 mM NaCl, 1% Triton X-100, 30 M MG132, and protease inhibitor mixture (Complete, Roche Applied Science), and then incubated for 16 h with anti-HA beads (Sigma). Immunoprecipitates were washed with lysis buffer containing 500 mM NaCl and detected by Western blot.
Pulse-Chase Experiments-Transfected HEK293 cells were washed, incubated with methionine/cysteine-free medium for 1 h, pulsed with methionine/cysteine-free medium containing 100 Ci of [ 35 S]methionine/cysteine (PerkinElmer Life Sciences) for 3 h, and subsequently chased in normal medium for the times specified. Cells were harvested, and HA-SIAH-1 immunoprecipitation was carried out as described above for the in vivo ubiquitylation assays. Immunoprecipitates were resolved on 10% SDS-polyacrylamide gels and the amount of 35 S-labeled SIAH-1 was quantified by PhosphorImager analysis.
Immunocytochemistry Assays-Transfected HEK293 and SH-SY5Y cells were incubated with or without 10 M lactacystin, 10 mM NH 4 Cl, and 10 mM 3-MA for 12 h, fixed with 4% paraformaldehyde for 15 min, and blocked in phosphate-buffered saline containing 0.2% Triton X-100 and 5% normal goat serum. Cells were labeled with anti-HA (Covance) and anti-Myc antibodies (Santa Cruz Biotechnology) as described (8). Immunolabeling was detected using fluorescein isothiocyanate-and Cy 3 -labeled secondary antibodies (The Jackson Laboratories). Statistics of the number of inclusion-containing cells were analyzed by Student's t test.
Cell Death Analysis-Transfected HEK293 and SH-SY5Y cells were transfected and processed for immunocytochemistry as described above. Before mounting, nuclei were stained with Hoechst 33342. Cell death was determined by counting the number of transfected cells that were positive for nuclear condensation and fragmentation. Statistics of cell death was done by Student's t test.
Primary Neuronal Cultures-Embryonic day 18 primary cortical cultures were prepared from Sprague-Dawley rats as described (26). The animals were killed by decapitation according to the protocol approved by the committee for animal experimentation at the Technion-Israel Institute of Technology. Neurons were maintained in neurobasal medium supplemented with B27 (Invitrogen) and 0.5 mM L-glutamine (Biological Industries, Beit-Haemek, Israel). After 2 weeks of culture, neurons were fixed for immunocytochemical studies as described above for cell lines.

Synphilin-1A Interacts in Vitro and in Vivo with SIAH and
Co-localizes in Neurons-We have previously identified a novel synphilin-1 isoform, called synphilin-1A, that is present in Lewy bodies and is toxic to neurons (25). We now investigated if synphilin-1A interacts with SIAH, which was shown to ubiquitylate synphilin-1 (8). To determine whether they interact in vitro, GST pulldown experiments were carried out by incubating GST-SIAH-1 fusion protein with the extract of HEK293 cells transfected with HA-synphilin-1A (Fig. 1A, left panel). HA-Synphilin-1A specifically interacted with GST-SIAH-1 but not with the control protein GST-FKBP12 (Fig. 1A, left panel). The in vitro interaction of synphilin-1A and SIAH-1 was similar to that observed between synphilin-1 and SIAH-1 (Fig. 1A, right panel).
To map the SIAH-binding domain of synphilin-1A, HEK293 cells were co-transfected with different HA-synphilin-1A constructs and myc-SIAH-1. By carrying out co-immunoprecipitation experiments, we identified that SIAH coimmunoprecipitated with the first third as well as the second third of synphilin-1A (Fig. 1C). The last third of synphilin-1A did not interact with SIAH. Thus, the first two-thirds of synphilin-1A seem to be responsible for the interaction with SIAH-1.
Using antibodies that we previously characterized as specific to synphilin-1A (Fig. 1F, panels A and D), SIAH-1 (panel B), and SIAH-1/2 (panel E) (8,14,25), we found in immunocytochemistry experiments of rat primary neuronal cultures that endogenous synphilin-1A co-localizes with endogenous SIAH proteins in neuronal cytosol, including cell bodies and processes, as determined by confocal laser microscopy ( Fig.  1F, panels C and F).

FIGURE 1. SIAH interacts with synphilin-1A in vitro and in vivo.
A, left panel, extract of HEK293 cells transfected with HA-synphilin-1A was incubated with the indicated GST fusion proteins. Binding was analyzed by using an anti-HA antibody. Right panel, extract of HEK293 cells transfected with HA-synphilin-1 was incubated with the indicated GST fusion proteins. Binding was analyzed using anti-HA antibody. The total amount of GST fusion proteins used in the experiments was determined by staining the membranes with Ponceau S (lower panels). B, SIAH-1 co-immunoprecipitates with synphilin-1A from co-transfected HEK293 cells. HA-synphilin-1A was immunoprecipitated (IP) from extracts of HEK293 cells using an anti-HA antibody. Co-immunoprecipitation was determined by Western blot using an anti-Myc antibody. C, mapping the SIAH binding region in synphilin-1A. HA-synphilin-1A fragments were immunoprecipitated from extracts of co-transfected HEK293 cells using an anti-HA antibody. The co-immunoprecipitation of SIAH-1 with synphilin-1A fragments was determined by using the anti-SIAH antibody 990. The detection of immunoprecipitated synphilin-1A fragments was done using an anti-HA antibody. aa, amino acids. D, synphilin-1A (Sph-1A) co-immunoprecipitates with endogenous SIAH-1. SIAH-1 was immunoprecipitated from rat brain homogenate using the anti-SIAH-1 antibody (N-15), and detection of coimmunoprecipitation was carried out using anti-synphilin-1A antibody (25). The detection of immunoprecipitated-SIAH-1 was done using the anti-SIAH-1 antibody (990) (8). E, synphilin-1A also co-immunoprecipitates with endogenous SIAH-2. SIAH-2 was immunoprecipitated from rat brain homogenate using the anti-SIAH-2 antibody (N- 14), and detection of co-immunoprecipitation was carried out using anti-synphilin-1A antibody. The detection of immunoprecipitated-SIAH-2 was done using the anti-SIAH-1/2 antibody (H-18). F, rat cortical neurons grown for 14 days in vitro were immunolabeled against synphilin-1A (panels A and D), SIAH-1 with N-15 antibody (panel B), and SIAH-1/2 with H-18 antibody (panel E). The co-localization of endogenous synphilin-1A and SIAH in the cytoplasm of neurons, including cell bodies and neuronal processes, is observed in the merge pictures (panels C and F). Scale bar, 20 m.
To determine whether SIAH promotes the degradation of synphilin-1A, HEK293 cells were co-transfected with HA-synphilin-1A and increasing amounts of myc-SIAH-1, and the steady-state levels of synphilin-1A were analyzed. No major decrease in the steady-state of synphilin-1A was observed even in the presence of large amounts of SIAH-1 (Fig. 2C, upper panel), suggesting that SIAH-1 does not promote significant degradation of synphilin-1A. By contrast, SIAH-1 promoted a sharp decrease in synphilin-1 steady state levels (Fig. 2C, lower panel) (8). Similar results were obtained when SIAH-2 was used instead SIAH-1 (data not shown). Thus, contrasting with synphilin-1, synphilin-1A seems to be a poor SIAH substrate.
We have shown that overexpression of synphilin-1A into cultured cells promotes its accumulation into inclusions (25). To determine whether SIAH can be recruited to synphilin-1A inclusions, HEK293 cells were co-transfected with HA-synphilin-1A and myc-SIAH-1 or myc-SIAH-1 C55A,C59H,C72S mutant devoid of catalytic activity (8). By carrying out immunocytochemistry experiments, we found that only catalytically active SIAH-1 is recruited to synphilin-1A inclusions (Fig. 3A, left panels). In addition, SIAH-1 co-localized with synphilin-1A inclusions also in the dopaminergic cells SH-SY5Y (Fig. 3A, right panels), indicating that the recruitment of SIAH to synphilin-1A inclusions is not specific to a single cell type. Similar to SIAH-1, SIAH-2 was also recruited to synphilin-1A inclusions (data not shown), further supporting the interaction between SIAH proteins and synphilin-1A. Finally, the co-localization of SIAH-1 and SIAH-2 with synphilin-1A inclusions was observed either in the absence or in the presence of the proteasome inhibitor lactacystin ( Fig. 3A and data not shown). Furthermore, no recruitment of inactive SIAH-1 mutant to synphilin-1A inclusions was observed (Fig. 3A, left and right lower panels).
To confirm the recruitment of SIAH to synphilin-1A inclusions, we investigated the effect of synphilin-1A on the distribution of SIAH-1 into soluble and insoluble fractions by Western blot assays. HEK293 cells were co-transfected with HA-synphilin-1A and myc-SIAH-1 (wild-type or mutant devoid of catalytic activity), and cell homogenates were separated into fractions soluble and insoluble to 1% Triton X-100. In the presence of synphilin-1A, SIAH-1 distributed between soluble and insoluble fractions, whereas SIAH-1 devoid of catalytic activity was present only in the soluble fraction (Fig. 3B). Therefore, we confirmed by biochemical means that insoluble synphilin-1A is able to recruit only active SIAH-1.
We have previously shown that ubiquitylation by SIAH in the presence of proteasome inhibition promotes about 5-fold increase in the formation of synphilin-1 inclusions (8). In the presence of lactacystin, SIAH-1 also increased the formation of synphilin-1A inclusions (Fig. 3C). However, this increase in synphilin-1A inclusions promoted by SIAH-1 was smaller than that previously observed for synphilin-1, as SIAH-1 leads to less than a 2-fold increase in synphilin-1A inclusions (Fig. 3C). This result is compatible with the small ubiquitylation extent of synphilin-1A promoted by SIAH, if compared with that observed with synphilin-1 ( Fig. 2A).
We next characterized synphilin-1A/SIAH inclusions. In immunocytochemistry experiments, synphilin-1A/SIAH inclusions were thioflavine S-and ubiquitin-positive (Fig. 4A). Also, synphilin-1A/SIAH inclusions were able to recruit PD-related proteins, such as ␣synuclein, synphilin-1, Parkin, PINK1, and UCH-L1 (Fig. 4B), indicating that SIAH does not compete with the binding of PD-related proteins to synphilin-1A inclusions. These results were observed for synphilin-1A/SIAH inclusions either in the absence (Fig. 4, A and B) or in the presence of lactacystin (data not shown). No significant recruitment of the control protein FKBP12 to synphilin-1A/ SIAH inclusions was observed, suggesting that, together with SIAH, aggregated synphilin-1A may work as anchor for PD proteins to form Lewy bodies.
Synphilin-1A Decreases SIAH Auto-ubiquitylation and Degradation-Because synphilin-1A strongly binds to SIAH in vivo and in vitro, but is a very poor ubiquitylation substrate, we wondered whether synphilin-1A could in fact affect SIAH functions. To determine whether synphilin-1A could change SIAH ubiquitin-ligase activity, the effect of synphilin-1A on SIAH auto-ubiquitylation was investigated. For this, HEK293 cells were co-transfected with HA-SIAH-1 and increasing amounts of myc-synphilin-1A; SIAH was immunoprecipitated with an anti-HA antibody, and the levels of SIAH auto-ubiquitylation were determined using an anti-ubiquitin antibody. We found that increasing amounts of synphilin-1A significantly decreased SIAH-1 auto-ubiquitylation (Fig. 5A). In vitro ubiquitylation experiments were also carried out by incubating in vitro translated SIAH with purified components of the ubiquitin system in the absence or in the presence of recombinant synphilin-1A. Similar to the in vivo ubiquitylation experiments, addition of synphilin-1A decreased the in vitro auto-ubiquitylation of SIAH (Fig. 5B).

Synphilin-1A Decreases the Catalytic Activity of SIAH
Because auto-ubiquitylation of SIAH is required for its degradation (27), we next examined if synphilin-1A leads to decreased degradation of SIAH. HEK293 cells were co-transfected with HA-SIAH-1 and myc-synphilin-1A, and the steady-state levels of SIAH-1 were analyzed. We found that increasing amounts of synphilin-1A significantly increased the steady-state levels of SIAH-1 (Fig. 6A). In experiments where only synphilin-1A was transfected, increasing amounts of synphilin-1A were able to significantly raise the steady-state levels of endogenous SIAH-1 protein (Fig. 6B). In addition, pulse-chase experiments revealed that SIAH-1 half-life was significantly prolonged in the presence of synphilin-1A (Fig. 6C), indicating that the increase of SIAH-1 steady-state levels promoted by synphilin-1A is because of decreased degradation and not to changes in transcription. Taken together, our data indicate that synphilin-1A decreases the auto-ubiquitylation and degradation of SIAH-1.
Synphilin-1A Decreases SIAH Ligase Activity and Cell Toxicity-To determine whether synphilin-1A also changes SIAH ability to ubiquitylate its substrates, we carried out ubiquitylation of synphilin-1 and DCC promoted by SIAH in the absence and in the presence of synphilin-1A. Increasing amounts of synphilin-1A significantly decreased the ability of SIAH-1 to ubiquitylate synphilin-1 as well as DCC (Fig. 7,  A and C). Also, increasing amounts of synphilin-1A prevented the degradation of both synphilin-1 and DCC promoted by SIAH (Fig. 7, B and D). Therefore, in addition to decreasing the auto-catalytic activity of SIAHs, synphilin-1A decreases the ability of SIAH to ubiquitylate and to degrade its substrates.
To demonstrate the relevance of synphilin-1A in the metabolism of endogenous proteins, we analyzed SIAH activity in the presence of siRNA to synphilin-1A. Because we found low levels of synphilin-1A expression in HEK293 (data not shown), all the siRNA experiments were carried out in SH-SY5Y cells. First, we generated a pair of siRNAs that efficiently reduce synphilin-1A expression (Fig. 8A). We next investigated the ability of synphilin-1A siRNA to interfere with SIAH activity. As steady-state levels of SIAH were shown to correlate with its ligase activity (27), we investigated the effect of synphilin-1A on the steady-state levels of SIAH. We found that siRNA-mediated synphilin-1A knockdown promoted a significant reduction in SIAH steady-state levels (Fig. 8B), suggesting that decreased synphilin-1A expression leads to improvement of SIAH ligase activity with increased auto-ubiquitylation and degradation. Also, siRNA to synphilin-1A decreased the steady state of the SIAH substrates synphilin-1 and DCC (Fig. 8, C and D), indicating that the absence of synphilin-1A increases the activity of SIAH toward its substrates as well. Taken together, our data suggest that the decrease in synphilin-1A expression increases the activity of endogenous SIAH as well as its ability to degrade substrates, confirming that endogenous synphilin-1A negatively modulates the ligase activity of SIAH.
Next, we examined whether the inhibitory activity of synphilin-1A is specific for SIAH-1. For this, we investigated the ability of synphilin-1A to interfere with the ligase activity of two other E3 ubiquitin-ligases, Parkin and CHIP (carboxyl terminus of Hsc70-interacting protein). Because the autoubiquitylation of Parkin and CHIP were shown to reflect their catalytic activity (28 -30), we measured the ability of synphilin-1A to modulate the steady-state levels of Parkin and CHIP. We found that synphilin-1A did not promote the accumulation of Parkin or CHIP as it did for SIAH (Fig. 9, A  and B), indicating that synphilin-1A does not interfere with

Synphilin-1A Decreases the Catalytic Activity of SIAH
the activities of Parkin and CHIP. Also, synphilin-1A did not promote the accumulation of one of the Parkin bona fide substrates p38 (31,32) (Fig. 9C), confirming that synphilin-1A does not inhibit Parkin ligase activity. Therefore, these findings indicate that synphilin-1A specifically modulates SIAH activity.
SIAH has been shown to cause cell death (33). To investigate if synphilin-1A can also decrease the toxicity promoted by SIAH, we transfected HEK293 cells with SIAH, in the absence or in the presence of synphilin-1A, and measured cell toxicity by determining the amount of nuclear condensation/fragmentation by Hoechst 33342. We found that synphilin-1A significantly decreased the cell death promoted by SIAH (Fig. 10). Different from wild-type SIAH, the ligase-inactive mutant of SIAH did not promote cell death when compared with vector alone (Fig. 10). In addition, cell death promoted by synphilin-1A, in the absence or in the presence of ligase-inactive SIAH, was similar to that observed with vector alone, indicating that synphilin-1A, alone or combined with ligase-inactive SIAH, is not toxic to HEK293 cells (Fig. 10). Therefore, synphilin-1A modulates both SIAH ligase activity (Figs. 5-7) and its toxicity (Fig.  10), suggesting that the ligase and the toxic activity of SIAH are interconnected.
Synphilin-1A Decreases the Monoubiquitylation of ␣-Synuclein by SIAH and Its Inclusion Formation-We next investigated the effect of synphilin-1A on the ability of SIAH to ubiquitylate substrates that are not the target for degradation, such as ␣-synuclein (8,14). Addition of increasing amounts of synphilin-1A decreased the monoubiquitylation of ␣-synuclein (Fig.  11A). In addition, no significant change in ␣-synuclein steady-state

Synphilin-1A Decreases the Catalytic Activity of SIAH
levels was observed in the presence of SIAH alone or combined with increasing amounts of synphilin-1A (data not shown).
Because ␣-synuclein monoubiquitylation may affect inclusion formation in dopaminergic cells (14), we investigated the effect of synphilin-1A in the formation of ␣-synuclein inclusions. For this, SH-SY5Y cells were transfected with ␣-synuclein and SIAH-2, in the absence or in the presence of synphilin-1A. To promote the accumulation of monoubiquitylated ␣-synuclein, transfected cells were incubated with proteasome, lysosome, and autophagy inhibitors, and the amounts of monoubiquitylated ␣-synuclein inclusions were determined by immunocytochemistry. In accordance with its ability to inactivate SIAH ligase activity, synphilin-1A significantly decreased the formation of monoubiquitylated ␣-synuclein inclusions promoted by SIAH (Fig. 11B).

DISCUSSION
We have previously identified an isoform of synphilin-1 we called synphilin-1A that is prone to aggregation and that accumulates in the brain of Diffuse Lewy Body disease (25). Also, synphilin-1A interacts with both ␣-synuclein and synphilin-1, suggesting a possible role in PD. In this study, we described the interaction between synphilin-1A and the E3 ubiquitin-ligase SIAH. We found that synphilin-1A is ubiquitylated by SIAH but that this ubiquitylation is not enough to promote synphilin-1A degradation by the proteasome. As a poorly degraded substrate, synphilin-1A inactivates SIAH, inhibiting SIAH ligase and toxic activities. Accumulation of ubiquitylated synphilin-1A leads to increased formation of synphilin-1A inclusions, which recruits SIAH. Synphilin-1A decreases the auto-ubiquitylation of SIAH and its ligase activity toward other substrates, including synphilin-1, DCC, and ␣synuclein. In parallel to a decrease in synphilin-1 and DCC ubiquitylation, synphilin-1A reduced their degradation by the proteasome. The ability of synphilin-1A to interfere with SIAH activity was also observed for endogenous proteins as siRNA mediated synphilin-1A knockdown significantly decreased the steady-state of endogenous SIAH and its substrates, synphilin-1 and DCC. Furthermore, the inhibitory activity of synphilin-1A is specific for SIAH because synphilin-1A did not promote the accumulation of other E3 ubiquitin ligases, such as Parkin and CHIP. Moreover, synphilin-1A decreased ␣-synuclein monoubiquitylation and the formation of ␣-synuclein inclusions, supporting a role   APRIL 24, 2009 • VOLUME 284 • NUMBER 17 of ␣-synuclein monoubiquitylation in inclusion body formation.

Synphilin-1A Decreases the Catalytic Activity of SIAH
SIAH has been shown to interact with its substrates by binding to the consensus VXP (34). Synphilin-1A is devoid of VXP consensus, but it interacts with endogenous SIAH, as observed in co-immunoprecipitation experiments with rat brains. Supporting an interaction of SIAH and synphilin-1A in a VXP-independent manner is the finding that some SIAH substrates, such as synaptophysin, do not contain the VXP motif (34). Also, mutations in the VXP consensus of some substrates, such as Kid, do not change the interaction with SIAH (34), suggesting that some substrates of SIAH may interact in a VXP-independent manner.
Further supporting a specific interaction of synphilin-1A and SIAH, we found in immunocytochemistry experiments of primary neuronal cultures that both endogenous proteins co-localize in the cytosol of neurons, including cell bodies and processes. In agreement, both proteins were previously shown to be widely expressed in the brain and to be present at the presynapse (25,35,36). In addition, SIAH interacts with and promotes the degradation of different synaptic proteins, including the group 1 metabotropic glutamate receptor and the synaptic vesicle protein synaptophysin (36,37). Moreover, endogenous SIAH was shown to be present and to co-localize with synaptophysin in neuritic processes of neuronally differentiated PC12 cells (37).
The inability of SIAH to promote the degradation of synphilin-1A is not unique as several substrates of SIAH, such as PHYL and Vav, are not targeted for degradation by the proteasome (38). Nevertheless, the small extent of synphilin-1A ubiquitylation promoted by SIAH increased the formation of synphilin-1A inclusions in the presence of proteasome inhibitors. Therefore, synphilin-1A/SIAH interaction promotes the formation of synphilin-1A inclusions rather than the degradation of synphilin-1A. Because both SIAH and synphilin-1A are present in Lewy bodies, ubiquitylation of synphilin-1A might be important for the formation of these inclusions. SIAH ligase activity has been shown previously to be inhibited by the protein Disabled-1, but it was not clear whether it constitutes SIAH substrate (27). Synphilin-1A works as a "suicide" substrate as it inhibits both the autoubiquitylation and the ability of SIAH to ubiquitylate the substrates ␣-synuclein, synphilin-1, and DCC. The ability of synphilin-1A to inhibit both auto-ubiquitylation and ubiquitylation of SIAH toward substrates indicates that these activities of SIAH are interconnected. Further supporting the correlation between auto-ubiquitylation and ubiquitylation of substrates is the finding that phosphorylation by Cdk5 inhibits Parkin auto-ubiquitylation as well as ubiquitylation of its substrates p38 and synphilin-1 (29). SIAH is a pro-apoptotic protein that seems to promote cell death by activation of glyceraldehyde-3-phosphate dehydrogenase and JNK pathways (33, 39 -41). In an attempt to dissect the connection between the ligase activity and toxicity of SIAH, we also investigated the effect of synphilin-1A on the ability of SIAH to promote cell death. We could not investigate the effect of synphilin-1A on the toxicity of SIAH in primary neuronal cultures because synphilin-1A is per se toxic in neurons (25). To avoid the toxic component of synphilin-1A, we chose to carry out the cell toxicity experiments in HEK293 cells, where we found that synphilin-1A exerts no significant direct toxicity. We found that, in addition to decreasing the ligase activity of SIAH, synphilin-1A decreased the ability of SIAH to promote cell death in HEK293 cells. Therefore, we propose that the toxicity of SIAH correlates with its ability to ubiquitylate substrates involved in triggering cell death. This observation is in accordance with the finding that activation of JNK requires the E3 ligase activity of SIAH (41).
Although synphilin-1A is toxic to primary cultured neurons from rat brain (25), we now found that it does not promote  toxicity in HEK293 (Fig. 10) or SH-SY5Y cells (data not shown). The lack of toxicity in these cell types allowed us to investigate the role of synphilin-1A in SIAH toxicity and accumulation of monoubiquitylated ␣-synuclein inclusions. By inhibiting SIAH, synphilin-1A led to a significant decrease in the formation of monoubiquitylated ␣-synuclein inclusions in SH-SY5Y cells (Fig. 11B). On the other hand, our attempts to determine changes in ␣-synuclein inclusion formation by synphilin-1A in primary neuronal cultures were not successful because of synphilin-1A neurotoxicity.
Even though SH-SY5Y cells have a neuronal origin, the ability of synphilin-1A to promote cell death seems to be restricted to primary neurons. Differences in protein toxicity depending on the cell type have been previously reported by several studies. Overexpression of wild-type ␣-synuclein was shown to be toxic to neurons (42), but several studies have found that wild-type ␣-synuclein is not toxic to different types of cultured cells (43)(44)(45)(46). Therefore, we speculate that dissimilarities between the cellular machinery of primary neuronal cultures and cultured cell lines could account for the difference in synphilin-1A toxicity. Moreover, this indicates that synphilin-1A toxicity in neurons may be mediated by still unidentified cellular pathways. Alternatively, aggregation of synphilin-1A, as observed in patients with Diffuse Lewy Body disease (25), could lead to less available synphilin-1A to neutralize SIAH toxicity in vivo. Future studies are necessary to determine the cellular pathways involved in the toxicity of synphilin-1A in neurons.
Similar to ␣-synuclein, synphilin-1 has been shown to play a synaptic role. Synphilin-1 is a presynaptic protein (47) and promotes the release of neurotransmitters in a cell model (22). Because synphilin-1A increases the steady-state levels of synphilin-1 (Fig. 7B), it is possible that synphilin-1A may modulate the synaptic function of synphilin-1 by increasing its concentration in the cell. Furthermore, by changing the steady-state levels of DCC (Fig. 7D), synphilin-1A may modulate axonal outgrowth (48). Likewise, synphilin-1A may affect hypoxic response by changing the degradation of HIF1␣, a process controlled by SIAH through PHD proteins (49). Thus, synphilin-1A may be linked to different physiological events, such as axonal growth, neuronal activity, and hypoxic response.
In summary, we report that synphilin-1A interacts in vivo and inhibits the endogenous catalytic activity of SIAH, promoting accumulation of SIAH itself as well as substrates that it normally degrades, such as synphilin-1 and DCC. Our results also shed light on the mechanisms regulating ␣-synuclein monoubiquitylation, which has been implicated in ␣-synuclein aggregation (14). By interfering with the activity of SIAH, synphilin-1A also leads to decreased monoubiquitylation of ␣-synuclein and its aggregation into inclusions. Because monoubiquitylated ␣-synuclein inclusions are toxic to cells (14,50), the ability of synphilin-1A to decrease the formation of ␣-synuclein inclusions indicates that synphilin-1A represents a novel therapeutic target to PD.