Address correspondence and reprint requests to Eliezer Masliah, Department of Neurosciences, University of California, San Diego, La Jolla, CA 92093-0624, USA. E-mail: firstname.lastname@example.org
The aggregation of α-synuclein (α-syn) is believed to play a critical role in the pathogenesis of disorders such as dementia with Lewy bodies and Parkinson's disease. The function of α-syn remains unclear, although several lines of evidence suggest that α-syn is involved in synaptic vesicle trafficking, probably via lipid binding, and interactions with lipids have been shown to regulate α-syn aggregation. In this context, the main objective of this study was to determine whether methyl-β-cyclodextrin (MβCD), a cholesterol-extracting agent, interfered with α-syn accumulation in models of synucleinopathy. For this purpose, we studied the effects of MβCD on the accumulation of α-syn in a transfected neuronal cell line and in transgenic mice. Immunoblot analysis showed that MβCD reduced the level of α-syn in the membrane fraction and detergent-insoluble fraction of transfected cells. In agreement with the in vitro studies, treatment of mice with MβCD resulted in decreased levels of α-syn in membrane fractions and reduced accumulation of α-syn in the neuronal cell body and synapses. Taken together, these results suggest that changes in cholesterol and lipid composition using cholesterol-lowering agents may be used as a tool for the treatment of synucleinopathies.
Recent studies have shown that α-syn interacts in vivo with PUFAs and that, in brains of patients with PD or DLB, or in neuronal cells over-expressing α-syn or its mutants, the levels of PUFA are elevated (Sharon et al. 2003a). Remarkably, in other neurodegenerative disorders that overlap with LBD, such as Alzheimer's disease (AD), previous studies have suggested a potential role for PUFAs and other lipids, such as cholesterol, in the clearance and metabolism of amyloid-β protein (Aβ) (Simons et al. 1998; Wood et al. 2002; Eckert et al. 2003b). In view of these findings, drugs that may regulate lipid metabolism and lower cholesterol, such as statins (drugs that reduce de novo cholesterol synthesis) and cyclodextrins (CDs), are currently being considered as potential therapeutics for AD (Simons et al. 1998; Hutter-Paier et al. 2004; Cole et al. 2005). Cholesterol depletion with a combination of statins and methyl-β-cyclodextrin (MβCD) significantly lowers Aβ production (Simons et al. 1998), and clinical studies have indicated that there is a decreased prevalence of AD associated with the use of statins alone or in combination with MβCD to treat hypercholesterolemia (Yunomae et al. 2003).
CDs are cyclic oligosaccharides of glucopyranosyl (glycosyl) units linked in a ring formation in the α(1–4) position (Pitha et al. 1988). α-, β-, and γ-CDs are crystalline, water-soluble compounds that differ in having six, seven and eight glucose residues, respectively, forming different sizes of hydrophobic cavity that can accommodate non-polar molecules and transfer lipophilic compounds to aqueous media (Pitha et al. 1988; Loftsson and Masson 2001). CDs selectively extract membrane cholesterol by including it in their non-polar core and, unlike other cholesterol-binding agents that incorporate into membranes, CDs are strictly surface-acting. CDs have been used for many years as carriers of lipophilic drugs in pharmacological research, and recently in membrane studies and cholesterol trafficking. βCDs [including βCD, MβCD and hydroxypropyl-βCD (HPβCD)] have been shown to extract membrane cholesterol selectively from a variety of cell types (Ohtani et al. 1989; Kilsdonk et al. 1995; Klein et al. 1995), and represent unique tools for membrane studies as they neither bind nor insert into the plasma membrane.
In view of some studies suggesting that lipid intake in the diet may be a risk factor for PD (Johnson et al. 1999), and that cholesterol and α-syn may interact in lipid rafts (Fortin et al. 2004), it is possible that cholesterol-reducing agents may have a potential effect in ameliorating pathological accumulation of α-syn. In this context, the main objective of the present report was to investigate the effects of MβCD on α-syn in neuronal cell lines and transgenic (tg) mice. These studies showed that the cholesterol-extracting effects of MβCD were accompanied by reduced α-syn levels in the membrane and detergent-insoluble fractions, and reduced phosphorylated α-syn levels in the membrane fraction. In tg mice, MβCD reduced the neuronal accumulation of α-syn and ameliorated the degenerative alterations, suggesting a possible therapeutic treatment using cholesterol-lowering agents for PD and DLB.
Materials and methods
B103 neuroblastoma cells transfected with human (h)α-syn or empty vector (pCEP4, Invitrogen, Carlsbad, CA, USA) were grown as described previously (Takenouchi et al. 2001). These cells were routinely cultured in Dulbecco's modified Eagle's medium high glucose containing 10% fetal calf serum (Omega, Tarzana, CA, USA) supplemented with 50 µg/mL hygromycin B (Calbiochem, San Diego, CA, USA), 5% v/v sodium pyruvate (Gibco-BRL, Grand Island, NY, USA) and 1% v/v gentamycin (Gibco-BRL) in a 5% CO2/95% air atmosphere. This neuronal cell line, derived from rat neuroblastoma, was selected because of its ability to express under basal conditions molecules associated with lipid rafts, such as flotillin (Eckert et al. 2003a). Moreover, α-syn over-expression in this line results in the formation of discrete aggregates in the cell body and processes accompanied by reduced neurite outgrowth (Takenouchi et al. 2001), mimicking another important aspect of LBD, namely compromised axonal plasticity.
Cell viability assay
To study the fate of the transfected cells treated with MβCD, and to determine the optimum concentration and time course for treating cells with MβCD, cell viability was measured by the trypan blue exclusion assay (Hashimoto et al. 2002). Cells were treated with various concentrations of MβCD (1–10 mm; Sigma-Aldrich, St. Louis, MO, USA) for different periods of time (1–24 h). Cells were harvested and washed with phosphate-buffered saline (PBS; ScyTek Laboratories, Logan, UT, USA) and, after the addition of 0.4% trypan blue (Sigma-Aldrich), the percentage of viable (unstained) cells was determined by counting the cells under the microscope.
Treatments, preparation of cholesterol–MβCD inclusion complex and cholesterol determination
To analyze the effects of MβCD on α-syn accumulation, vector control and α-syn-transfected B103 cells were treated for various lengths of time (0, 1, 3, 6 and 24 h) with 5 mm MβCD, followed by immunoblot analysis. To determine the specificity of the effects of MβCD, control experiments were performed in which B103 cells were either untreated or treated with cholesterol–MβCD inclusion complexes. This is based on previous studies showing that saturation of the preparation with cholesterol abrogates the effects of MβCD (Simons et al. 1998). For this purpose, complexes were prepared as described previously (Simons et al. 1998). Briefly, 30 mg of cholesterol (Sigma-Aldrich) was added slowly to a 5% (w/v) solution of MβCD in water and stirred on a water bath (80°C) until complete dissolution. The complex was stored at room temperature (21°C). A stock solution of 50 mm cholesterol was prepared in chloroform. Free cholesterol (2 µg/mL) and cholesterol–MβCD complex (0.3 mm) were added together or separately for 6 h to non-treated cells and to cells pre-treated with 5 mm MβCD for 6 h.
Quantification of total cholesterol (measuring both cholesterol and cholesteryl ester) was performed using the Cholesterol/Cholesteryl Ester Quantification Kit (BioVision, Mountain View, CA, USA) colorimetric method. Briefly, for each condition, cells were first extracted with 200 µL of hexane–isopropanol and spun down; the supernatant was collected, vacuumed dry and the lipids were dissolved in 200 µL of 2-propanol containing 10% Triton X-100 and assayed.
Fractionation of neuronal cell lines and brain homogenates
As previous studies have shown that α-syn is more abundant in the cytosolic than in the particulate fraction, and that interactions with lipids may favor the translocation to the membrane (Hashimoto et al. 2003; Ulmer et al. 2005), immunoblot analysis was performed with homogenized tissues that were fractionated by centrifugation as described previously (Hsu et al. 1998). Briefly, cells or frozen hemi-brains (50 mg) were sonicated for 30–60 s in buffer [1 mm HEPES, 5 mm benzamidine, 2 mm 2-mercaptoethanol, 3 mm EDTA, 0.5 mm magnesium sulfate, 0.05% sodium azide (all from Sigma-Aldrich) and 0.01 mg/mL protease inhibitor (Calbiochem)]. Homogenates were centrifuged (274 000 g, 1 h, 4°C), the cytosolic fraction was separated and the particulate fraction was resuspended in buffer and sonicated for 30–60 s. The cytosolic fraction represents the liquid phase and contains the cytosol of the cell and everything that normally dissolves in the normal cell environment, whereas the particulate fraction, or the membrane fraction, contains everything that does not dissolve in the liquid phase. The total protein concentration of each sample was determined using BCA protein assay reagents (Pierce, Rockford, IL, USA). Similarly, cell lysates in TNE buffer (10 mm Tris-HCl, pH 7.4, 150 mm NaCl, 5 mm EDTA; all from Sigma-Aldrich) plus complete protease inhibitor (Calbiochem) were sonicated, the homogenates were centrifuged and the protein concentration was determined.
Additional studies of the effects of MβCD on α-syn aggregation were performed with detergent-soluble and detergent-insoluble fractions prepared as described previously with slight changes (Petrucelli et al. 2004; Ho et al. 2005). Briefly, cells were centrifuged for 10 min at 5000 g, 4°C and lysed in TNE buffer containing 1% Triton X-100 (Sigma-Aldrich). Cells were sonicated for 30 s and ultracentrifuged (274 000 g, 1 h, 4°C), and the detergent-soluble proteins were collected in the supernatant fraction. The pelleted detergent-insoluble proteins were dissolved in TNE buffer containing 1% Triton X-100–1% sodium dodecyl sulfate (SDS).
Preparation of fractions in sucrose gradients
Isolation of sucrose fraction membranes was performed as described previously with some modification (Morishima-Kawashima and Ihara 1998). Briefly, cells (a total of 1 × 107 in two confluent 100-mm dishes) were rinsed in PBS, harvested and homogenized in 800 µL of 2-N-morpholino-ethane-sulfonic acid (MES)-buffered saline (MBS; 25 mm MES, pH 6.5, 150 mm NaCl) containing 1.0% Triton X-100 and protease and phosphatase inhibitors. After treatment with DNaseI (10 U/mL) for 1 h, cell extracts were combined with 800 µL of 80% sucrose in MBS and placed at the bottom of ultracentrifuge tubes, and overlaid with a 5%, 35% discontinuous sucrose gradient (each 1.6 mL) in MBS. The gradients were then centrifuged at 40 000 g for 23 h in a SW55Ti rotor (Beckman, Fullerton, CA, USA) at 4°C. After centrifugation, each fraction (400 µL) was collected from the top of the gradient to yield a total of 12 fractions (fractions were numbered from the top of the gradient). Fractions 4-7 represent the lipid raft-like, insoluble components, and soluble components are located in fractions 10-12. Fractions were stored at −80°C until use. Equal volumes of each fraction were subjected to electrophoresis. Proteins that are not associated with lipid rafts or with the membranes remain at the bottom of the gradient.
Western blot analysis
Samples from cytosolic, membrane, detergent-soluble and detergent-insoluble fractions, and fractions from the sucrose gradients, were separated on 12% or 4–12% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) gels (NuPAGE, Invitrogen) and transferred onto 0.22-µm nitrocellulose membranes (Schleicher & Schuell, Keene, NH, USA) using 1 × 3-(cyclohexylamino)-1-propanesulfonic acid (CAPS) transfer buffer containing 20% methanol. Membranes were blocked with 3% milk in PBS containing 0.1% Tween-20 (Sigma-Aldrich) (PBS-T), followed by incubation in primary antibody (1 : 1000) in PBS-T overnight at 4°C. The primary antibodies used were as follows: anti-flotillin-1 and anti-α-syn were purchased from Transduction Laboratories (Newington, NH, USA); anti-actin was purchased from Chemicon International (Temecula, CA, USA); and anti-phosphorylated α-syn was a gift from Dr Iwatsubo (University of Tokyo, Japan). Membranes were further incubated with goat anti-mouse or anti-rabbit IgG secondary antibodies conjugated to horseradish peroxidase (1 : 5000; American Qualex, San Clemente, CA, USA) and visualized by enhanced chemiluminescence (ECL; NEN Life Sciences, Boston, MA, USA) and exposed to film. For determination of the levels of immunoreactivity, ECL-treated membranes were analyzed in the VersaDoc imaging system (Bio-Rad, Hercules, CA, USA) using Quantity One software (Bio-Rad).
Ribonuclease protection assay (RPA)
To ascertain the levels of α-syn messenger RNA (mRNA) expression in transfected cells and mice treated with MβCD, RPA analysis was performed essentially as described previously (Masliah et al. 2000). Briefly, riboprobe templates were amplified by RT-PCR. The DNA template was linearized and the anti-sense probe was used for RPA analysis. The following 32P-labeled anti-sense riboprobes were used to identify specific mRNAs [protected nucleotides (nt) (GenBank accession number)]: hα-syn [nt 210–475 (No. L08850)], murine (m)α-syn [nt 235–459 (No. S69965)] and a murine actin riboprobe [nt 480–565 (No. X03672)] (used to correct for variations in mRNA content and loading). RPA was carried out and signals were quantified with the PhosphorImager (Amersham Biosciences, Pittsburgh, PA, USA).
In vivo treatment of mice with MβCD, tissue processing and determination of cholesterol levels
The in vivo effects of MβCD were tested in non-tg and hα-syn tg mice from line D (Masliah et al. 2000). These tg mice express hα-syn under the regulatory control of the platelet-derived growth factor-β (PDGFβ) promoter, and were selected because they display neuropathology and performance alterations that mimic some aspects of LBD. For these experiments, 16 non-tg and 16 hα-syn tg mice (12 months of age) were used. Mice from each group received a daily intraperitoneal injection of either MβCD (10 mm) (n = 8 non-tg and n = 8 hα-syn tg) or saline (n = 8 non-tg and n = 8 hα-syn tg) for a total of 7 days. At the end of the experiment, mice were killed following National Institutes of Health (NIH) guidelines for the humane treatment of animals. From each mouse, blood was obtained by intracardiac puncture using a heparinized needle. To obtain plasma, blood samples from each mouse were spun at 600 g for 10 min. Plasma was then used for the determination of cholesterol levels.
The quantification of total cholesterol was performed using the Cholesterol/Cholesteryl Ester Quantification Kit (BioVision) colorimetric method. Briefly, serum samples (0.5 µL/assay) and cholesterol standard samples were prepared at 50 µL/well with Cholesterol Reaction Buffer in a 96-well plate, and a total of 50 µL of Reaction Mix was added to each well. The plate was incubated for 60 min at 37°C and the optical density (OD) was measured at 570 nm. The cholesterol concentration in each well was calculated from the standard curve.
After the animals had been killed, brain hemi-sections were obtained from each animal by brain removal and knife bisection along the superior sagittal sulcus from the cortical surface to the extreme ventral surface. The right hemi-brain was snap frozen in liquid nitrogen and divided for subsequent RNA analysis by RPA and for fractionation and western blot analysis. The left hemi-brain was fixed in 4% paraformaldehyde for vibratome sectioning and subsequent immunocytochemical and laser scanning confocal microscopy (LSCM) analysis.
Immunocytochemical analysis and confocal microscopy
Immunohistochemistry of B103 cells was performed as described previously (Takenouchi et al. 2001). Cells were seeded onto poly-l-lysine-coated glass coverslips, grown to 60% confluence, fixed in 4% paraformaldehyde (30 min) and pre-treated for 20 min with 0.1% Triton X-100 in Tris-buffered saline (TBS; ScyTek Laboratories). The coverslips were first incubated overnight at 4°C either with the antibody against hα-syn 72-10 (1 : 5000) (Masliah et al. 2000) or anti-flotillin (1 : 1000). The next day, antibodies were detected with the Tyramide Signal Amplification-Direct (Red) System (NEN Life Sciences), followed by an overnight incubation with the mouse monoclonal anti-microtubule-associated protein-2 (MAP2, 1 : 50; Roche Molecular Biochemicals, Indianapolis, IN, USA) primary antibody. This antibody was then detected with FITC-conjugated anti-mouse secondary antibody (Vector Laboratories, Burlingame, CA, USA). Vector-transfected B103 cells were run in parallel as controls. Other control experiments included immunolabeling in the absence of primary antibodies or with antibodies adsorbed for 48 h with a 20-fold excess of peptide. Coverslips were air-dried overnight, mounted on slides with anti-fading media (Vectashield, Vector Laboratories) and imaged by LSCM (MRC1024, Bio-Rad).
Briefly, as described previously (Masliah et al. 2000), to investigate the effects of MβCD on mα-syn and hα-syn in vivo in the brains of non-tg and tg mice, serially sectioned, free-floating, blind-coded vibratome sections were incubated overnight at 4°C with either the mouse monoclonal antibody against α-syn (syn-1, 1 : 500; Transduction Laboratories) or the rabbit polyclonal anti-hα-syn-specific antibody (72-10, 1 : 500). The syn-1 antibody was prepared by immunizing animals with purified rat α-syn and recognizes both endogenous and hα-syn. The 72-10 antibody (anti-hα-syn) was prepared as described previously (Masliah et al. 2000) by immunizing rabbits with synthetic hα-syn peptides consisting of amino acids 101–124. Incubation with the primary antibody was followed by FITC-tagged horse anti-mouse IgG for the syn-1 antibody or FITC-tagged goat anti-rabbit IgG (1 : 100; Vector Laboratories) for the 72-10 antibody. Sections were analyzed by LSCM (Bio-Rad) in order to determine the levels of endogenous and hα-syn immunoreactivity in the neocortex. For each case, three sections were analyzed and the results were averaged and expressed as the mean pixel intensity.
To evaluate the integrity of the dendritic system, blind-coded 40-µm-thick vibratome sections from mouse brains fixed in 4% paraformaldehyde were immunolabeled with the mouse monoclonal antibody against MAP2 (dendritic marker, 1 : 40; Chemicon International) and drebrin (mouse monoclonal, 1 : 500; Medical and Biological Laboratories Co., Nagoya, Japan), as described previously (Mucke et al. 1995). After an overnight incubation with the primary antibodies, sections were incubated with FITC-conjugated horse anti-mouse IgG secondary antibody (1 : 75; Vector Laboratories), transferred to SuperFrost slides (Fisher Scientific, Tustin, CA, USA) and mounted under glass coverslips with anti-fading media (Vector Laboratories). All sections were processed under the same standardized conditions. The immunolabeled blind-coded sections were imaged by LSCM (MRC1024, Bio-Rad) and analyzed with the Image 1.43 program (NIH), as described previously (Toggas et al. 1994; Mucke et al. 1995).
All values in the figures are expressed as the means ± SEM. To determine the statistical significance, the values were compared by two group t-tests using the Statview II statistical package for the Macintosh computer. The differences were considered to be significant if p values were less than 0.05.
MβCD reduces α-syn levels in the membrane fraction of transfected B103 cells
In order to determine the optimum concentration and time course for treatment, cells were exposed to various concentrations of MβCD (1–10 mm) for up to 24 h, and cell viability was determined (Fig. 1). These studies showed that transfected B103 cells were tolerant to MβCD at a 5 mm concentration for up to 24 h (Fig. 1a). On the basis of these results, cells were treated with 5 mm MβCD for 1, 3 and 6 h and western blot analysis was performed. As expected, α-syn immunoreactivity was more abundant in the cytosolic than in the membrane fraction of untreated transfected B103 cells (Fig. 1b). Untransfected cells did not display detectable levels of α-syn immunoreactivity (not shown). Treatment with MβCD resulted in a gradual decrease over time in the levels of α-syn immunoreactivity in the membrane fraction, with the most significant decrease at 6 h of treatment (Figs 1b and c). No significant effects were observed in the levels of α-syn immunoreactivity in the cytosolic fraction (Figs 1b and c). To confirm that MβCD was active in the in vitro model system, levels of cholesterol were determined in cell lysates. This analysis showed that, compared with untreated α-syn-transfected B103 cells (3792 ± 210 ng/well), treatment with MβCD (2114 ± 145 ng/well) resulted in a 44% decrease in the levels of cholesterol. As MβCD decreased α-syn levels in the membrane fraction, and previous studies have shown that α-syn translocation to the membrane fraction may be associated with abnormal aggregation and toxic conversion (Hashimoto et al. 2004), it is possible that MβCD treatment may reduce insoluble α-syn. Consistent with this possibility, immunoblot analysis showed that, although in the untreated transfected B103 cells, higher molecular weight α-syn aggregates were detected in the detergent-insoluble fraction, in the cells treated with MβCD there was a significant decrease in the levels of α-syn oligomers (Figs 1d and e). As the formation of insoluble α-syn aggregates has been associated with phosphorylation at serine 129, and this residue has been shown to be selectively and extensively phosphorylated in synucleinopathy lesions (Okochi et al. 2000; Fujiwara et al. 2002; Hasegawa et al. 2002; Takahashi et al. 2003), we investigated the effects of MβCD on the levels of phosphorylated α-syn (Fig. 2). α-Syn-transfected B103 cells showed the presence of high levels of phosphorylated α-syn by immunoblot. In contrast, treatment with MβCD resulted in a significant decrease in the levels of phosphorylated α-syn in both the membrane (Figs 2a and c) and detergent-insoluble (Figs 2b and c) fractions. The effects on phosphorylated α-syn were comparable with the effects of MβCD on total α-syn, suggesting that the effects of this compound were not directly related to phosphorylation but to α-syn cellular distribution. To further investigate the effects of MβCD on α-syn redistribution and compartmentalization, immunoblot analysis was performed with sucrose gradient fractions (Fig. 3). This analysis showed that, although in the untreated samples most of the monomeric and oligomeric α-syn was found in lipid raft-like insoluble fractions (Figs 3a–c, fractions 4-9), in the cells treated with MβCD there was a shift in the localization of the α-syn aggregates into the soluble fractions (Figs 3b–f, fractions 10-12).
Saturation with excess cholesterol decreases the effects of MβCD on α-syn in transfected B103 cells
To demonstrate that the decrease in α-syn levels in the membrane fraction was a result of the cholesterol depletion effect of MβCD, cells were treated with a complex containing excess cholesterol–MβCD or with MβCD for 6 h, and α-syn levels were analyzed by immunoblot (Fig. 4). This experiment showed that treatment with the cholesterol–MβCD complex abrogated the effects of MβCD on α-syn levels in the membrane fraction of transfected cells (Figs 4a and b). In contrast, treatment of cells with the cholesterol saturated complex after the cells had first been treated with MβCD for 6 h resulted in an increase in α-syn levels (Figs 4a and b), suggesting that changes in α-syn were related to cholesterol extraction.
To further verify the extraction function of MβCD, immunoblot analysis was performed with antibodies against flotillin. Flotillin is a caveolae-associated integral membrane protein in neurons and a lipid raft marker (Eckert et al. 2003a; Kokubo et al. 2003; Pandur et al. 2004). Following treatment with MβCD, levels of flotillin immunoreactivity were significantly decreased in the membrane fraction, and the effect was blocked when cells were treated with the cholesterol–MβCD complex (Figs 4a and c).
To determine the effects of MβCD on the cellular distribution of α-syn, double-labeling and confocal analysis were performed. Compared with cells transfected with the vector control, in the hα-syn-transfected cell lines abundant α-syn immunolabeling was detected in the neuronal cell bodies and extended to the neurites (Figs 5a and b). In contrast, in transfected cells treated with MβCD, α-syn immunoreactivity was mainly decreased in the neuritic processes and, to some extent, in the cell bodies of neurons (Fig. 5d). Treatment with the cholesterol–MβCD complex abrogated the effects of MβCD on α-syn immunoreactivity in the neuronal cell bodies and neurites (Fig. 5c). Treatment of cells with the cholesterol-saturated complex after the cells had first been treated with MβCD for 6 h had similar effects on α-syn cellular distribution (Fig. 5e). To investigate whether the effects of MβCD on α-syn were related to transcriptional effects, RPA analysis was performed (Fig. 6). This study showed no significant differences in α-syn mRNA levels under any of the conditions tested (Fig. 6a), supporting the possibility that the effects of MβCD on α-syn levels may be related to the effects on lipid distribution.
Effects of MβCD on α-syn levels in non-tg and hα-syn tg mice brains
Non-tg and hα-syn tg mice were treated with MβCD (10 mm) for 1 week and analyzed by western blot and immunocytochemistry. To confirm that MβCD was active in the mice, cholesterol levels were determined. In plasma samples, compared with vehicle-treated controls (3444 ± 219 ng/well), mice treated with MβCD (2311 ± 141 ng/well) showed a 33% decrease in cholesterol levels. Similarly, the determination of cholesterol levels in brain homogenates showed that treatment with MβCD (98 ± 9 ng/well) resulted in a 20% decrease compared with the saline-treated control group (121 ± 6 ng/well). Consistent with previous studies (Iwai et al. 1994), α-syn immunoreactivity was more abundant in the cytosolic than in the membrane fractions. Compared with saline-treated non-tg mice, in non-tg mice treated with MβCD there was a 40% decrease in the levels of mα-syn and flotillin in the membrane fractions, but no effects were observed in the cytosolic fractions (Figs 7a and b). Similarly, compared with saline-treated hα-syn tg mice, in hα-syn tg mice treated with MβCD there was a 20% decrease in the levels of hα-syn and flotillin in the membrane fractions, but no effects were observed in the cytosolic fractions (Figs 7c and d). By RPA analysis, no significant alterations were detected in the levels of mα-syn or hα-syn mRNA expression in the brains of non-tg and hα-syn tg mice, respectively (Figs 6b and c), supporting the possibility that the effects of MβCD on α-syn are post-transcriptional. Double-immunolabeling analysis in non-tg mice showed that, compared with saline treatment, MβCD treatment resulted in decreased α-syn immunolabeling of presynaptic terminals (Figs 8a, b and e), but no significant effects were observed on neuronal integrity and dendritic complexity, as revealed by levels of MAP2 immunoreactivity (Figs 8c ,d and f). As expected, in saline-treated hα-syn tg mice, there was abundant hα-syn accumulation in neuronal cell bodies and nerve terminals (Fig. 8g), accompanied by neurodegenerative alterations in the dendritic complex as demonstrated by decreased levels of MAP2 (Fig. 8i). In contrast, in tg mice treated with MβCD, levels of α-syn immunolabeling in neuronal cell bodies (Figs 8h and k) and in the neuropil were reduced and levels of MAP2 immunostaining were comparable with those detected in non-tg controls (Figs 8j and l). Additional analysis of the effects of MβCD on neuronal integrity was performed with an antibody against drebrin, a dendritic spine protein associated with the actin cytoskeleton. Consistent with the MAP2 analysis, levels of drebrin immunoreactivity were decreased in saline-treated hα-syn tg mice (mean pixel intensity, 164 ± 34; n = 5; one way anova; post hoc Dunnett's p < 0.05) compared with non-tg controls (mean pixel intensity, 225 ± 25; n = 5). In contrast, MβCD-treated hα-syn tg mice (mean pixel intensity, 213 ± 22; n = 5) displayed levels of drebrin immunolabeling similar to those of non-tg controls.
The present study showed that the cholesterol-depleting agent MβCD decreased the accumulation of α-syn in the membrane and insoluble fractions of an α-syn-transfected neuronal cell line and in the brains of non-tg and hα-syn tg mice. This is of interest because, in PD, misfolded (toxic) forms of α-syn often associate with the membrane, and damage may result from the formation of abnormal pore-like structures (Lashuel et al. 2002; Quist et al. 2005). α-Syn is an abundant cytosolic molecule that concentrates in the presynaptic site (Iwai et al. 1994). Under physiological conditions, α-syn interactions with the plasma membrane are mediated by fatty acids through lipid recognition repeat domains within α-syn (Perrin et al. 2000; Sharon et al. 2001). Previous studies have shown that α-syn associates with a number of membranes, including synaptic vesicles (Maroteaux et al. 1988), lipid droplets (Jensen et al. 1998) and yeast plasma membrane (Outeiro and Lindquist 2003). Moreover, recent studies have shown that lipid rafts mediate the synaptic localization of α-syn, where it may play a role in synaptic plasticity and neurotransmission (Fortin et al. 2004). Lipid rafts are detergent-insoluble, specialized microdomains of the plasma membrane that integrate signaling pathways (Bickel 2002). Lipid rafts are enriched in cholesterol, sphingolipids and glycosylphosphatidylinositol (Simons and Ikonen 1997; Shaikh et al. 2003); however, recent studies have shown that PUFAs may also be an important component (Ma et al. 2004; Wassall et al. 2004). In addition to PUFAs, α-syn may associate with other lipids, such as cholesterol, in the membrane rafts. Consistent with this possibility, previous studies have shown that α-syn association with lipid rafts is sensitive to the cholesterol-depleting effects of MβCD (Fortin et al. 2004) and, in our models, the effects of MβCD on α-syn were paralleled by flotillin, a structural protein enriched in the lipid rafts of neurons (Kokubo et al. 2003).
The mechanisms through which MβCD may decrease α-syn accumulation are not completely clear; however, it is possible that these effects may be directly related to the cholesterol-depleting capabilities of this compound, which, in turn, results in the redistribution of α-syn aggregates from the membrane to the soluble fractions, as demonstrated by the studies with sucrose gradients. However, other indirect mechanisms, such as decreased α-syn expression or increased degradation, may play a role. Alternatively, changes in lipid levels may alter α-syn immunoreactivity in an epitope-dependent manner; however, a previous study has shown that MβCD modifies the distribution of green fluorescent protein-tagged α-syn (Fortin et al. 2004). As MβCD did not affect α-syn mRNA expression or monomeric α-syn levels in the cytosolic fractions, and we did not observe increased generation of C-terminal α-syn degradation products, it is more likely that MβCD effects on α-syn are lipid dependent. In this regard, it is possible that decreased cholesterol in lipid rafts and changes in the membrane composition and fluidity may affect the translocation and transport of α-syn to the membrane. Supporting this possibility, previous studies have shown that MβCD decreases cholesterol levels in the membrane and enhances the membrane fluidity, which, in turn, causes changes in the functional properties of the membrane and in signal transduction (Gniadecki 2004; Larbi et al. 2004). Furthermore, cholesterol extraction from the membrane may interfere with α-syn interactions with PUFAs, which participate in α-syn membrane translocation. Under basal conditions, a critical balance between PUFAs and cholesterol may regulate α-syn location to lipid rafts in synaptic membranes; however, the physiological consequences of this process on synaptic transmission are unclear.
In PD, it is possible that this lipid balance in the membrane is altered, resulting in the excessive accumulation of α-syn and the formation of toxic oligomers in the membrane. Supporting this possibility, Sharon et al. (2003a) reported the presence of elevated PUFA levels in PD and DLB brain soluble fractions, and more elevated PUFA levels in membrane fractions, accompanied by increased membrane fluidity in α-syn over-expressing neurons. However, cholesterol levels in the neurons and in PD and DLB brains are unknown. In addition to the effects on membrane cholesterol, MβCD may block α-syn accumulation by interfering with post-transcriptional modifications of α-syn, such as phosphorylation. Amongst other potential kinases, α-syn is phosphorylated at serine 129 by casein kinase II (Lee et al. 2004), and this results in increased aggregation and toxicity (Chen and Feany 2005). Consistent with this possibility, we found that treatment with MβCD also decreased the levels of phosphorylated (serine 129) α-syn in the membrane and insoluble fractions. These effects may be related to the decreased availability of α-syn to kinases because of the effects of MβCD on the membrane cholesterol levels, increased phosphatase activity or decreased kinase activity. As the effects of MβCD on phosphorylated α-syn were comparable with those of total α-syn, it is most probable that further studies will be necessary to investigate which of these potential targets is affected by MβCD and how they are affected.
In the present study, we also showed that MβCD decreased α-syn accumulation in vivo and ameliorated the neurodegenerative alterations in tg mice. These effects on α-syn tg mice are consistent with previous studies showing that CDs can rescue the disease phenotype in other in vivo models of neurological disorders, such as Niemann–Pick type C1 disease (NPC1) (Camargo et al. 2001; Yu et al. 2005). NPC1 is a disorder characterized by a defect in cholesterol trafficking and metabolism resulting in lysosomal pathology and progressive neurodegeneration (German et al. 2002; Gondre-Lewis et al. 2003; Walkley and Suzuki 2004; Chang et al. 2005; Li et al. 2005). Interestingly, recent studies have shown that α-syn accumulates in the brains of patients with Niemann–Pick disease and in NPC1 mutant mice (Mori et al. 2002; Saito et al. 2004). Together, these studies support the concept of a role of cholesterol metabolism in α-syn accumulation and of MβCD as a potential treatment for disorders with parkinsonism and α-syn accumulation (Camargo et al. 2001; Yu et al. 2005).
Notably, alterations in cholesterol levels have been associated with an increased risk for AD, mainly because cholesterol may participate in the regulation of Aβ production (McLaurin et al. 2002; Yanagisawa 2002; Burns et al. 2003; Michikawa 2003; Hartman 2005) and Aβ has been shown to accumulate in lipid rafts (Wood et al. 2002). Consequently, several cholesterol-reducing drugs, including CDs, are currently being evaluated for the treatment of AD (Simons et al. 1998; Cole et al. 2005). Similarly, it is possible that cholesterol may also play a role in LBD. The evidence in this respect is not yet conclusive; for example, some studies have indicated that α-syn associates with cholesterol-related lipid rafts (Fortin et al. 2004), and have suggested that high levels of cholesterol intake may be associated with increased risk for PD (Johnson et al. 1999). In contrast, a recent study has found that a high intake of unsaturated fatty acids may protect against PD, but no significant association has been found with saturated fat, cholesterol or trans-fat intake (de Lau et al. 2005). Furthermore, the direct effects of high cholesterol levels on α-syn accumulation and membrane structures are not entirely clear and require further investigation.
In conclusion, the present study has shown that cholesterol depletion by MβCD treatment may decrease the levels of potentially toxic, detergent-insoluble α-syn which associates with the membrane. Although further studies are necessary to elucidate the mechanism of action of CDs and the interaction between α-syn and cholesterol, cholesterol-lowering agents may be considered as potential treatments for PD.
This work was supported by NIH grants AG18440, AG05131 and AG022074.