Glycogen Synthase Kinase 3β Modulates Synphilin-1 Ubiquitylation and Cellular Inclusion Formation by SIAH

α-Synuclein is known to play a major role in the pathogenesis of Parkinson disease. We previously identified synphilin-1 as an α-synuclein-interacting protein and more recently found that synphilin-1 also interacts with the E3 ubiquitin ligases SIAH-1 and SIAH-2. SIAH proteins ubiquitylate synphilin-1 and promote its degradation through the ubiquitin proteasome system. Inability of the proteasome to degrade synphilin-1 promotes the formation of ubiquitylated inclusion bodies. We now show that synphilin-1 is phosphorylated by GSK3β within amino acids 550–659 and that this phosphorylation is significantly decreased by pharmacological inhibition of GSK3β and suppression of GSK3β expression by small interfering RNA duplex. Mutation analysis showed that Ser556 is a major GSK3β phosphorylation site in synphilin-1. GSK3β co-immunoprecipitated with synphilin-1, and protein 14-3-3, an activator of GSK3β activity, increased synphilin-1 phosphorylation. GSK3β decreased the in vitro and in vivo ubiquitylation of synphilin-1 as well as its degradation promoted by SIAH. Pharmacological inhibition and small interfering RNA suppression of GSK3β greatly increased ubiquitylation and inclusion body formation by SIAH. Additionally, synphilin-1 S556A mutant, which is less phosphorylated by GSK3β, formed more inclusion bodies than wild type synphilin-1. Inhibition of GSK3β in primary neuronal cultures decreased the levels of endogenous synphilin-1, indicating that synphilin-1 is a physiologic substrate of GSK3β. Using GFPu as a reporter to measure proteasome function in vivo, we found that synphilin-1 S556A is more efficient in inhibiting the proteasome than wild type synphilin-1, raising the possibility that the degree of synphilin-1 phosphorylation may regulate the proteasome function. Activation of GSK3β during endoplasmic reticulum stress and the specific phosphorylation of synphilin-1 by GSK3β place synphilin-1 as a possible mediator of endoplasmic reticulum stress and proteasomal dysfunction observed in Parkinson disease.

␣-Synuclein is known to play a major role in the pathogenesis of Parkinson disease. We previously identified synphilin-1 as an ␣-synuclein-interacting protein and more recently found that synphilin-1 also interacts with the E3 ubiquitin ligases SIAH-1 and SIAH-2. SIAH proteins ubiquitylate synphilin-1 and promote its degradation through the ubiquitin proteasome system. Inability of the proteasome to degrade synphilin-1 promotes the formation of ubiquitylated inclusion bodies. We now show that synphilin-1 is phosphorylated by GSK3␤ within amino acids 550 -659 and that this phosphorylation is significantly decreased by pharmacological inhibition of GSK3␤ and suppression of GSK3␤ expression by small interfering RNA duplex. Mutation analysis showed that Ser 556 is a major GSK3␤ phosphorylation site in synphilin-1. GSK3␤ co-immunoprecipitated with synphilin-1, and protein 14-3-3, an activator of GSK3␤ activity, increased synphilin-1 phosphorylation. GSK3␤ decreased the in vitro and in vivo ubiquitylation of synphilin-1 as well as its degradation promoted by SIAH. Pharmacological inhibition and small interfering RNA suppression of GSK3␤ greatly increased ubiquitylation and inclusion body formation by SIAH. Additionally, synphilin-1 S556A mutant, which is less phosphorylated by GSK3␤, formed more inclusion bodies than wild type synphilin-1. Inhibition of GSK3␤ in primary neuronal cultures decreased the levels of endogenous synphilin-1, indicating that synphilin-1 is a physiologic substrate of GSK3␤. Using GFPu as a reporter to measure proteasome function in vivo, we found that synphilin-1 S556A is more efficient in inhibiting the proteasome than wild type synphilin-1, raising the possibility that the degree of synphilin-1 phosphorylation may regulate the proteasome function. Activation of GSK3␤ during endoplasmic reticulum stress and the specific phosphorylation of synphilin-1 by GSK3␤ place synphilin-1 as a possible mediator of endoplasmic reticulum stress and proteasomal dysfunction observed in Parkinson disease.
Parkinson disease (PD) 2 is characterized by loss of dopaminergic neurons in the substantia nigra and the presence of cytoplasmic inclusions called Lewy bodies in surviving neurons (1). Hereditary PD can be caused by mutations in the ␣-synuclein gene (2)(3)(4) and in components of the ubiquitin-proteasome system, such as the E3 ubiquitin ligase parkin and UCH-L1 (5,6).
␣-Synuclein is a major component of Lewy bodies in sporadic PD (7). Overexpression of ␣-synuclein inhibits the proteasomal activity (8 -10) and causes cell death in a variety of cell and animal models (11). In agreement, proteasomal activity is decreased in substantia nigra of PD patients (12).
We have shown that synphilin-1 is a presynaptic protein that interacts with ␣-synuclein in vivo (13,14). Synphilin-1 leads to the formation of inclusion bodies when co-transfected with the non-A␤ component portion of ␣-synuclein in cultured cells and is an intrinsic component of Lewy bodies in PD, suggesting that it may play a role in Lewy body formation (13,15). Synphilin-1 seems to have a dual role in cell survival. Synphilin-1 is toxic to cells and inhibits proteasomal activity, raising the possibility that synphilin-1 might contribute to the death of dopaminergic neurons in PD (16 -18). On the other hand, cells containing synphilin-1 inclusions are more resistant to death, indicating that inclusions might be neuroprotective (19 -21). Additional evidence that synphilin-1 may be involved in PD comes from identification of a missense mutation in its gene in two patients that share a rare haplotype (20).
It has been shown that Lewy bodies are ubiquitylated, and understanding the ubiquitylation mechanism of Lewy body proteins may be relevant for clarifying ubiquitin's role in Lewy body formation. We have recently reported that the E3 ubiquitin ligases SIAH-1 and SIAH-2 ubiquitylate and target synphilin-1 for degradation by the proteasome system (19). The inability of the proteasome to degrade the synphilin-1-SIAH complex leads to a robust formation of ubiquitylated cytosolic inclusions containing synphilin-1, SIAH, and ␣-synuclein (19). Ubiquitylation is required for inclusion body formation, since a catalytically inactive mutant of SIAH-1 that binds to synphilin-1 fails to promote inclusions (19). Additionally, both SIAH and synphilin-1 are present in Lewy bodies of PD patients, implying a role in inclusion formation.
In an attempt to better understand the role of synphilin-1 in PD and Lewy body formation, we now sought to investigate mechanisms that regulate synphilin-1 ubiquitylation and inclusion formation. We present data indicating that synphilin-1 is phosphorylated in vivo by GSK3␤, which regulates ubiquitin-dependent degradation of synphilin-1 and inclusion body formation mediated by SIAH. Selective inhibition of GSK3␤ or mutation of a GSK3␤ phosphorylation site greatly increased synphilin-1 aggregation into cytosolic inclusions, suggesting a role of phosphorylation in modulating synphilin-1 aggregation. GSK3␤ inhibitor also enhanced the degradation of endogenous synphilin-1 in neurons, indicating that synphilin-1 is a physiologic substrate of GSK3␤. We also present data indicating that inhibition of proteasome function by synphilin-1 is modulated by its phosphorylation status. Our results shed light on the mechanism regulating synphilin-1 ubiquitylation and aggregation, with implications for inclusion body formation and possibly cell death in PD.
Cell Culture and Transfections-HEK 293 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-terminally tagged pRK5 and pFLAG-CMV-2 plasmids utilizing Lipofectamine 2000 (Invitrogen) and processed after 36 h.
In Vitro Kinase Assays-HEK 293 cells were transfected with HAsynphilin-1 cDNAs. After 36 h of transfection, cells were lysed as in the co-immunoprecipitation experiments. HA-synphilin-1 was immunoprecipitated with anti-HA antibody and washed in lysis buffer containing 500 mM NaCl. Immunoprecipitated synphilin-1 was incubated with recombinant GSK3␤ (New England Biolabs) at 37°C for 1 h in buffer containing 40 mM Tris, pH 7.6, 2 mM dithiothreitol, 5 mM MgCl 2 , 2 g/ml soybean trypsin inhibitor, 0.5 mM unlabeled ATP, 0.25 mCi/ml [␥-32 P]ATP. Reactions were ended with sample buffer and analyzed by SDS-PAGE using 8% gel. The amount of 32 P-labeled-synphilin-1 was quantified by PhosphorImager analysis. Equal loading of immunoprecipitated HA-synphilin-1 was determined by Western blot or Coomassie Blue staining.
In Vivo Phosphorylation Assays-After overnight starvation in serum-and phosphate-free medium, transfected HEK 293 cells were incubated for 3-6 h at 37°C with serum-free/phosphate-free medium containing 200 -400 Ci/ml [ 32 P]orthophosphate. Cells were harvested and lysed in buffer containing 50 mM Tris-HCl (pH 7.4), 140 mM NaCl, 1% Triton X-100, 0.1% SDS, 20 mM NaF, 2 mM Na 3 VO 4 , 30 M MG132, and protease inhibitor mixture (Complete; Roche Applied Science). Immunoprecipitation of HA-synphilin-1 was carried out with anti-HA antibody for 4 h at 4°C. Beads were washed with lysis buffer supplemented with 500 mM NaCl and analyzed by 8% SDS-PAGE. Densitometric quantification of radiolabeled HA-synphilin-1 was carried out by PhosphorImager analysis. Equal loading of immunoprecipitated HAsynphilin-1 was determined by Western blot or Coomassie Blue staining.
Pulse-Chase Experiments-Transfected HEK 293 cells were washed, incubated with methionine/cysteine-free medium for 1 h, pulsed with methionine/cysteine-free medium containing 200 ml/Ci [ 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-synphilin-1 immunoprecipitation was carried out as described above for the in vivo ubiquitylation assays. Immunoprecipitates were resolved on 8% SDS-polyacrylamide gels, and the amount of 35 S-labeled synphilin-1 was quantified by PhosphorImager analysis.
Immunocytochemistry Assays-Transfected HEK 293 cells were treated for 8 h with 10 M lactacystin, 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 stained with anti-HA (Covance) and anti-Myc (Santa Cruz) as described (19). Immunolabeling was detected using fluorescein isothiocyanate-and Cy 3 -labeled secondary antibodies (Jackson Laboratories). The percentage of cells containing cytosolic inclusions was counted by an investigator unaware of the treatment groups. Statistics of the number of inclusioncontaining cells was analyzed by analysis of variance followed by Tukey's post-test and by paired t test, when appropriate.
Primary Cortical Neuronal Cultures-Primary neuronal cultures were prepared from cerebral cortex of E18 Sprague-Dawley rats as described (22). Cells were cultured in 12-well plates coated with poly-D-lysine in neurobasal medium plus B27 for 2 weeks before use.
In order to further verify the specificity of synphilin-1 phosphorylation by GSK3␤, we carried out in vivo phosphorylation experiments using siRNA to suppress GSK3␤ expression. HEK 293 cells were transfected with full-length HA-synphilin-1 and siRNA to GSK3␤ or negative control siRNA. The siRNA to GSK3␤, but not the control siRNA, was effective in decreasing the expression of GSK3␤ by at least 90% (Fig.  1B). We found that siRNA to GSK3 significantly decreased the phosphorylation of full-length synphilin-1 (Fig. 1B).
Conceivably, the effect of GSK3␤ on synphilin-1 ubiquitylation could be due to interference with the components of the ubiquitin system or with a change of SIAH activity. However, the catalytic activity of SIAH-1 monitored by its autoubiquitylation was not affected by GSK3␤ (Fig.  4B). Moreover, SIAH-1 does not possess consensus sequences for GSK3␤ phosphorylation.
In order to confirm a direct effect of GSK3␤ on synphilin-1 ubiquitylation, we carried out in vitro ubiquitylation experiments using synphilin-1 prephosphorylated by GSK3␤. For this, HEK 293 cells were co-transfected with HA-synphilin-1 and Myc-GSK3␤ and lysed under harsh conditions to avoid co-immunoprecipitation of GSK3␤. Also, to avoid dephosphorylation of synphilin-1, all of the immunoprecipitation steps were carried out in the presence of phosphatase inhibitors. We found that the ubiquitylation of prephosphorylated synphilin-1 by GSK3␤ was significantly smaller than that observed with synphilin-1 immunoprecipitated from cells co-transfected with the control protein FKBP12 (Fig. 4C). This indicates that the effect of GSK3␤ on the ubiquitylation of synphilin-1 is due to a direct effect on synphilin-1 and not on the components of the ubiquitination system. The lack of Myc-GSK3␤ co-immunoprecipitation with HA-synphilin-1 was ascertained by reprobing the immunoprecipitation membrane with an anti-Myc antibody (Fig. 4C).
We next examined if phosphorylation by GSK3␤ interferes with the in vivo ubiquitylation of synphilin-1. For this, ubiquitylation of synphilin-1 by FLAG-ubiquitin was detected by immunoprecipitating synphilin-1 from HEK 293 cells and probing with anti-FLAG (Fig. 4D). The GSK3␤ inhibitor SB415286 promoted a strong increase in synphilin-1 ubiquitylation in HEK 293 cells (Fig. 4D). Conversely, the in vivo ubiquitylation of synphilin-1 was significantly decreased by GSK3␤ overexpression. Cells co-transfected with SIAH-1 had higher levels of ubiquitylated synphilin-1, which were also significantly augmented by SB415286 treatment (Fig. 4D). Overexpression of GSK3␤ decreased synphilin-1 ubiquitylation in SIAH-1-transfected cells, confirming the  effect of GSK3␤ in the ubiquitylation of synphilin-1 (Fig. 4D). To allow a more proper visualization, the effects of SB415286 and GSK3␤ on synphilin-1 ubiquitylation in the presence of SIAH were also shown in a less exposed picture (Fig. 4D).
Most of the observed ubiquitylated synphilin-1 signal in the presence of lactacystin locates above the native molecular mass of synphilin-1 (ϳ120 kDa), indicating the presence of poly-ubiquitylated full-length synphilin-1 (Fig. 4D). However, in conditions where there was a robust increase in synphilin-1 ubiquitylation, such as the presence of SB415286 or SIAH-1 (Fig. 4D), we also observed the presence of ubiquitylated synphilin-1 with faster mobility relative to native synphilin-1. Conceivably, this could be due to accumulation of proteolytic fragments of synphilin-1 (14).
Synphilin-1 Degradation in HEK 293 Cells and Neurons-We wondered whether inhibition of synphilin-1 ubiquitylation promoted by GSK3␤ alters ubiquitin-dependent synphilin-1 degradation. We first carried out Western blot analysis of transfected HEK 293 cells to determine the steady-state levels of synphilin-1 in the presence of GSK3␤.
The effects of GSK3␤ in the ubiquitylation and degradation of synphilin-1 are not due to decreased interaction between synphilin-1 and SIAH. Neither His-GSK3␤ nor GSK3␤ inhibitor SB415286 affected the co-immunoprecipitation of HA-synphilin-1 with Myc-SIAH-1 (data not shown). The casein kinase II inhibitor DRB, previously shown to decrease synphilin-1/␣-synuclein binding (28), had no effect on synphi- lin-1/SIAH-1 interaction, ruling out the participation of phosphorylation by casein kinase II in the effects we observed (data not shown).
In order to determine whether GSK3␤ phosphorylation also modulates endogenous levels of synphilin-1 in neuronal cells, we incubated primary cortical neuronal cultures in the presence of SB415286. Inhibition of GSK3␤ by SB415286 significantly reduced synphilin-1 steadystate levels in cortical neuronal cells (Fig. 6). The proteasomal inhibitor MG132 completely restored synphilin-1 levels, indicating that GSK3␤ is a physiological modulator of endogenous neuronal synphilin-1 degradation by the proteasome system.
Identification of GSK3␤ Phosphorylation Site-To identify the GSK3␤ phosphorylation sites among the 22 putative sites throughout the protein, we first mapped the region in synphilin-1 that is preferentially phosphorylated. We carried out in vivo phosphorylation experiments using HEK 293 cells transfected with HA-synphilin-1 fragments that together encompass the whole synphilin-1 open reading frame. We found that synphilin-1 is preferentially phosphorylated in the region encoding amino acids 550 -769 (Fig. 7A). Note that this region displayed a more significant binding to GSK3␤ than other synphilin-1 regions (Fig. 2).
Further breakdown of this amino stretch into two fragments (amino acids 550 -659 and 660 -769) showed that the region between amino acids 550 and 659 is responsible for an important portion of endogenous synphilin-1 phosphorylation (Fig. 7B). Quantitative analysis of the amount of incorporated phosphate (Fig. 7B, upper panel) relative to the amount of immunoprecipitated synphilin-1 fragment measured by densitometry (Fig. 7B, lower panel) showed that the phosphorylation of amino acids 550 -659 of synphilin-1 was ϳ3 fold larger than the phosphorylation of synphilin-1 region encoding amino acids 660 -769 (Fig.  7B). Notice that synphilin-1 550 -659 appeared as a phosphorylated doublet, compatible with different degrees of phosphorylation that affect the mobility of the fragment.
We next examined the protein kinases involved in the in vivo phosphorylation of the amino acid stretch 550 -659 of synphilin-1. Among the several protein kinase inhibitors utilized, only the GSK3␤ inhibitors, SB415286 and kenpaullone, were able to robustly inhibit the phosphorylation of the synphilin-1 550 -659 fragment (Fig. 7C). Protein kinase C inhibitor (Bis I) inhibited synphilin-1 550 -659 phosphorylation to a lesser degree. Inhibitors of casein kinase II (DRB), tyrosine kinase (genistein), protein kinase A (H-89), and phosphatidylinositol 3-kinase (wortmannin) had no effect in the phosphorylation of the synphilin-1 550 -659 fragment (Fig. 7C and data not shown).
We next sought to identify the residues within amino acids 549 -679 of synphilin-1 that are phosphorylated by GSK3␤. We generated a series of full-length HA-synphilin-1 constructs with point mutations of the nine putative GSK3␤ sites within this region. We found that substitution of Ser 556 to Ala decreased the phosphorylation level of full-length synphilin-1 as revealed by in vitro assay using immunopurified synphi- lin-1 and purified GSK3␤ (Fig. 8A). No other mutation of putative GSK3␤ sites within this region affected synphilin-1 phosphorylation in vitro (Fig. 8A). To demonstrate whether Ser 556 is also an important GSK3␤ site in vivo, we co-transfected HEK 293 cells with HA-synphilin-1, either wild-type or mutated in putative GSK3␤ sites, and Myc-GSK3␤. We found that the S556A mutant had significant lower phosphorylation in vivo than wild-type synphilin-1 or other mutants (Fig.  8B). However, phosphorylation of full-length synphilin-1 by GSK3␤ was not completely abolished in S556A mutant. This clearly suggests that Ser 556 is a major GSK3␤ phosphorylation site within synphilin-1, but additional phosphorylation by GSK3␤ or other kinases may also take place in other regions (Fig. 8B).
Role of GSK3␤ Phosphorylation in Inclusion Formation and Proteasomal Activity-We next examined the role of GSK3␤ and the phosphorylation deficient-mutant S556A in the formation of synphilin-1-SIAH inclusion bodies in cells. These inclusions were previously shown to be ubiquitylated and to recruit ␣-synuclein (19). We co-transfected HEK 293 cells with full-length HA-synphilin-1 and Myc-SIAH-1, and formation of intracellular inclusions were elicited by incubating with proteasome inhibitor lactacystin. We observed synphilin-1-SIAH inclusions in about 30% of the cells after 8 h (Fig. 9, A and B) and as much as 80% after overnight incubation with lactacystin (19) (data not shown). The addition of SB415286 or kenpaullone doubled the number of cells containing synphilin-1-SIAH inclusions, which after 8 h were found in more than 60% of transfected cells (Fig. 9, A and B). The increase in the formation of synphilin-1-SIAH inclusions promoted by the GSK3␤ inhibitors was not due to changes in the amount of SIAH-1, since both SB415286 and kenpaullone did not affect the expression levels of SIAH-1 by immunocytochemistry and Western blot analysis ( Fig. 9A and data not shown). As a control, the addition of casein kinase II inhibitor DRB did not change the formation of synphilin-1-SIAH inclusion bodies (Fig. 9B). This is consistent with the fact that casein kinase II did not affect the ubiquitylation of synphilin-1 promoted by SIAH or the interaction of synphilin-1 and SIAH ( Fig.  4A and data not shown).
To confirm that inhibition of GSK3␤ specifically increases the formation of synphilin-1-SIAH inclusions, we carried out immunocytochemistry experiments using HEK 293 cells co-transfected with HA-synphilin-1, Myc-SIAH-1, and siRNAs (control siRNA and siRNA to GSK3␤). We found that prevention of GSK3␤ expression by siRNA significantly increased the number of synphilin-1-SIAH inclusions (Fig. 9, C and D).
To further determine the importance of GSK3␤ for synphilin-1 degradation and inclusion formation, we carried out experiments with synphilin-1 S556A mutant that display deficient GSK3␤ phosphorylation (Fig. 8). S556A mutant elicited inclusions in a larger number of cells (40% increase) when compared with wild-type synphilin-1 (Fig. 9B). It is noteworthy that the increase in inclusion formation by the S556A mutant is not due to a change in synphilin-1 half-life. Pulse-chase experiments showed that S556A mutant displayed the same half-life as wildtype synphilin-1 (data not shown).
Synphilin-1 was recently shown to inhibit proteasomal function in vivo, but regulation of this process is unknown (18). A GFPu accumulation assay has been shown to reflect proteasomal impairment (30). Using this assay, we sought to determine whether synphilin-1 phosphorylation by GSK3␤ also alters the ability of synphilin to inhibit the proteasome. Since GSK3␤ possesses numerous targets in addition to synphilin-1, experiments testing GSK3␤ inhibitor effect on proteasomal function will obviously not be informative regarding the role of synphilin. Thus, we took advantage of the phosphorylation-deficient S556A mutant that increases synphilin-1-SIAH inclusions in order to study the role of GSK3␤ phosphorylation at this site. We found that S556A mutant was more effective than wild-type synphilin-1 in promoting accumulation of GFPu (Fig. 10). This implies that inhibition of the proteasome by synphilin-1 is also modulated by GSK3␤ phosphorylation.

DISCUSSION
We have previously shown that synphilin-1 is ubiquitylated by SIAH proteins and that ubiquitylated synphilin-1 is prone to aggregate into inclusions (19). The present paper describes a mechanism that regulates synphilin-1 ubiquitylation with implications for inclusion body formation. In a preliminary report, Wakabayashi and co-workers showed phosphorylation of synphilin-1 by GSK3␤ in vitro, but the function and possible occurrence of in vivo phosphorylation was not explored (31). Using pharmacological and siRNA approaches, we have now demonstrated that synphilin-1 is phosphorylated in vivo by GSK3␤. Phosphorylation of synphilin-1 by GSK3␤ decreased ubiquitylation and degradation of synphilin-1 promoted by SIAH. We show that GSK3␤ activity modulates the endogenous levels of synphilin-1 in neurons, indicating that GSK3␤ physiologically regulates synphilin-1 levels. Inhibition of GSK3␤ promoted a robust increase in the ability of synphilin-1 to form inclusions in the presence of SIAH. Additionally, mutation in GSK3␤ phosphorylation site potentiates synphilin-1 ability to inhibit the proteasome.
Our strategy to first map the region of phosphorylation to the 550 -659 fragment and then carry out site-directed mutagenesis allowed us to identify Ser 556 as a major site for phosphorylation by GSK3␤ among 22 putative sites. Mutation of Ser 556 significantly inhibited synphilin-1 phosphorylation in vitro and in vivo. Despite the evidence that GSK3␤ inhibitor modulates endogenous synphilin-1 levels and half-life, S556A mutation did not change synphilin-1 degradation promoted by SIAH. This indicates that other phosphorylation sites within synphilin-1 may be required to regulate its degradation rate, since the mutation did not completely abolish phosphorylation by GSK3␤. The mutagenesis of additional phosphorylation sites may be necessary to fully reveal the role of phosphorylation in the ubiquitylation and degradation of synphilin-1. Nevertheless, the S556A mutant disclosed a new role for GSK3␤ phosphorylation in modulating synphilin-1 inclusions. The evidence that the S556A mutant, GSK3␤ inhibitors, and siRNA to GSK3␤ increased inclusion body formation indicates that phosphorylation at Ser 556 modulates intracellular synphilin-1 aggregates. We have previously shown that ubiquitylation of synphilin-1 is required for inclusion body formation elicited by SIAH (19). Our data now suggest that the modulation of inclusion body formation also depends on GSK3␤ phosphorylation. Recently, Mouradian and co-workers found that casein kinase II phosphorylation at a still unidentified region of synphilin-1 is required for synphilin-1/␣-synuclein aggregation (29). Under our experimental conditions, casein kinase II inhibitor had no effect on synphilin-1 ubiquitylation or its ability to form cytosolic aggregates with SIAH-1.
Accumulation of abnormal or unfolded proteins leads to ER stress, which can ultimately lead to cell death (32). Several lines of evidence suggest that ER stress may occur in PD. The parkin substrate pael-r was shown to cause ER stress in dopaminergic neuroblastoma cells (33). 6-Hydroxydopamine and 1-methyl-4-phenylpyridinium were shown to activate ER stress-related proteins, and, more recently, GSK3␤ was shown to mediate 6-hydroxydopamine-induced neuronal death (34,35). In addition, GSK3␤ is known to mediate cell death during ER stress and is associated with neurodegenerative diseases, including Alzheimer disease (36,37). In this context, the decrease we found in neuronal synphilin-1 levels by GSK3␤ inhibition may have important pathological implications. Soluble synphilin-1 was shown to promote cellular toxicity, whereas synphilin-1 inclusions were shown to protect cells from dying (19 -21). Our results suggest that inhibition of phosphorylation by GSK3␤ will decrease synphilin-1 levels and also favor its deposition into inclusions, implying that GSK3␤ may play a role in the ability of synphilin-1 to aggregate into Lewy bodies. On the other hand, activation of GSK3␤ by ER stress and consequent increase of synphilin-1 levels might contribute to the neurotoxicity observed in PD.
The present study shows that 14-3-3 protein potentiates phosphorylation by GSK3␤. Since 14-3-3 was previously shown to associate with ␣-synuclein (26), we raise the possibility that a 14-3-3 and GSK3␤ may be part of a macromolecular complex together with synphilin-1, ␣-synuclein, and SIAH. Further analysis of synphilin-1 phosphorylation status in the substantia nigra of patients with PD will help to determine the role of synphilin-1 in the death of dopaminergic neurons.
Our study now demonstrated that the GSK3␤ phosphorylation-deficient mutant is more effective than wild-type synphilin-1 in inhibiting proteasomal activity. Although the mechanism of proteasomal dysfunction is not known, our results imply that GSK3␤ phosphorylation of synphilin-1 may contribute to proteasomal dysfunction thought to occur in PD (12).
In summary, we showed that synphilin-1 is phosphorylated in vivo by GSK3␤, causing a decrease in ubiquitin-dependent degradation of synphilin-1. Inhibition of GSK3␤ increased the formation of intracellular inclusions induced by proteasome inhibitors, suggesting that GSK3␤ activity may be relevant for Lewy body formation.