|Home | About | Journals | Submit | Contact Us | Français|
Aggregation of α-synuclein (α-syn), a process that generates oligomeric intermediates, is a common pathological feature of several neurodegenerative disorders. Despite the potential importance of the oligomeric α-syn intermediates in neuron function, their biochemical properties and pathobiological functions in vivo remain vastly unknown. Here we employed two-dimensional analytical separation and an array of biochemical and cell based assays to characterize α-syn oligomers that are present in the nervous system of A53T α-syn transgenic mice. The most prominent species identified were 53 Å detergent soluble oligomers, which preceded neurological symptom onset, and were found at equivalent amounts in regions containing α-syn inclusions as well as histologically unaffected regions. These oligomers were resistant to SDS, heat, and urea, but were sensitive to proteinase-K digestion. Even though the oligomers shared similar basic biochemical properties, those obtained from inclusion bearing regions were prominently reactive to antibodies that recognize oxidized α-syn oligomers, significantly accelerated aggregation of α-syn in vitro, and caused primary cortical neuron degeneration. In contrast, oligomers obtained from non-inclusion bearing regions were not toxic and delayed the in vitro formation of α-syn fibrils. These data indicate that specific conformations of α-syn oligomers are present in distinct brain regions of A53T α-syn transgenic mice. The contribution of these oligomers to the development of neuron dysfunction appears to be independent of their absolute quantities and basic biochemical properties, but is dictated by the composition and conformation of the intermediates as well as unrecognized brain-region specific intrinsic factors.
α-Synuclein (α-syn) is a soluble, acidic protein that typically assumes a random coil structure, however it acquires α-helical conformation upon binding to anionic phospholipids (Davidson et al., 1998; Ulmer et al., 2005). Although the exact function(s) of α-syn remain uncertain, the preferential localization to presynaptic nerve terminals and its interaction with vesicular phospholipids and proteins suggests a regulatory function associated with synaptic activity, dopamine production and metabolism, lipid vesicle trafficking, and chaperone-like activity (Maroteaux et al., 1988; Iwai et al., 1995; Souza et al., 2000b; Cabin et al., 2002; Chandra et al., 2005).
Data from post-mortem evaluations of Parkinson’s disease (PD) brains revealed that α-syn aberrantly forms highly organized, linear amyloid fibrils that constitute part of the characteristic inclusions found in neuronal perikarya (Lewy bodies) and dystrophic neurites (Lewy neurites). (Forno, 1996; Goedert, 2001; Norris et al., 2004; Mori et al., 2008). Despite the ubiquitous expression of α-syn throughout the central nervous system, these inclusions are found in certain susceptible neuronal subtypes of specific brain nuclei (Braak et al., 2003). Biochemically, α-syn within inclusions is resistant to extraction with non-ionic detergents. However upon extraction with formic acid or SDS/urea α-syn collapses into monomers and SDS and heat stable oligomers (Baba et al., 1998; Tu et al., 1998; Dickson et al., 1999).
In vitro evidence utilizing purified recombinant α-syn has indicated that the conversion of monomers into amyloid fibrils progresses in a nucleation-dependent manner through an initial polymerization stage involving the formation of oligomeric intermediates (Conway et al., 2000b). The polymerization process is concentration dependent and can be accelerated by the PD-causing mutations A53T, A30P, and E46K (Conway et al., 1998; Giasson et al., 1999; Narhi et al., 1999; Conway et al., 2000b; Greenbaum et al., 2005). Although there is considerable confidence that accumulation and polymerization of α-syn plays an important role in neurodegeneration, the contribution of the different species that emerge through the aggregation process has not been fully delineated. Existing efforts identifying the potential pathogenic formations are based on studies where oligomerization of the protein is forced in vitro (Goldberg and Lansbury Jr, 2000; Volles et al., 2001; Danzer et al., 2007). Characterizing α-syn assemblies that are formed in the brain and in living cells is challenging since unstable conformations may be disrupted during the traditional biochemical extraction processes. Notwithstanding this important concern, there is considerable paucity in the biochemical and biological description of the oligomeric α-syn entities that are formed in mice models and humans, and are stable to standard isolation methodologies with mild non-ionic detergents.
In this study we provide a brain regional-specific biochemical and biological characterization of the relatively stable α-syn oligomeric conformations that are formed in the transgenic mouse line expressing human A53T α-syn driven by the mouse PrP promoter (Giasson et al., 2002). The data indicates that despite similarities in basic biochemical properties, α-syn oligomeric intermediates obtained from different neural regions demonstrated unexpected divergence in promoting α-syn amyloid fibril formation and toxicity.
The mice utilized in this study express human A53T α-syn (line M83) or human WT α-syn (line M20) driven by the murine PrP promoter and have been described previously (Giasson et al., 2002). To generate A53T +/+ and non-transgenic (nTg) control mice used in experiments, A53T +/− females were mated with A53T +/− or +/+ males, since A53T +/+ females were found to produce small litters and exhibit poor motherly behavior. Genotyping was performed by both end-point PCR using GeneAmp PCR system 9700 thermal cycler (PE Applied Biosystems, Sunnyvale, CA) and quantitative PCR using Applied Biosystems 7500 real-time PCR system (PE Applied Biosystems) with the ABI MGB primer-probe set for human SNCA (assay ID Hs00240907_m1). SNCA values were normalized to mouse beta-actin (4352341E). Mice identified as homozygous for the α-syn transgene were verified through backcrosses or validated by quantitative western blot analysis of α-syn protein.
Analysis was performed on 5 month old A53T+/+ males and age-matched nTg male littermates. Motor activity measurements were performed with an Opto-M3 activity meter (Columbus Instruments, Columbus, OH) by quantifying the number of infrared beam-breaks over a 24 hour period using the Multi Device Interface (MDI) software (Version 1.34 Columbus Instruments). Testing was done in the home holding room and home cage under controlled environmental conditions, and food/water was available ad libitum. The animals were acclimated to the apparatus for 2 days, followed by 3 consecutive days of beam-break data collection. The 3-day average was used as n=1 for each mouse, with 5 mice utilized for each group.
Testing was performed on 4–6 month old A53T+/+ males and age-matched nTg male littermates in the home holding room and home cage under controlled environmental conditions. Singly-housed mice were given 100 g of fresh standard animal chow pellets (Laboratory rodent diet, cat#5001; LabDiet Richmond, IN) placed in a suspended wire cage top feeder at the start of the testing period and food consumption was determined by weighing pellets every 24 hr for 5 consecutive days. Mouse body weight was also determined every day during the testing period and used to normalize food intake measurements. The average value of food consumed over the 5 day testing period was used as n=1 for each mouse. The analysis was performed with 6 different mice from two separate litters.
Brain sections from symptomatic A53T+/+ α-syn transgenic mice and age-matched nTg controls were stained with syn505 mouse monoclonal antibody specific for human α-syn (Duda et al., 2002). Perfusions, tissue processing and staining were performed as previously described (Giasson et al., 2002). Adjacent sections were stained with rabbit polyclonal anti-GFAP (ab7260, AbCam, Cambridge MA). Incubation with secondary antibodies conjugated to biotin was followed by addition of avidin-biotin-peroxidase complex (ABC Elite, Vector Laboratories, Burlingame, CA), visualized with 3,3′-diaminobenzidine as the chromogen and counterstained with Hematoxylin (SH30-500D, Fisher Scientific, Hampton, NH).
4–6 month old asymptomatic A53T+/+ α-syn transgenic mice, 11–12 month old symptomatic A53T+/+ α-syn transgenic mice, and 11–12 month old human WT α-syn transgenic mice were utilized for analysis. A53T+/+ symptomatic mice were sacrificed 1–2 days after initial motor symptom onset. A53T+/+ mice were monitored on a daily basis and symptomatic mice were identified by spine stiffness, weight loss, and hind limb paralysis as described previously (Giasson et al., 2002). After removal of the cerebellum (Crb, including the deep cerebellar nuclei; bregma, −5.46 to −7.70), olfactory bulbs (OB; bregma, +4.74 to +3.14), and spinal cord (SC; cervical region), the remainder of the brain was dissected into 1 mm-thick coronal sections using a vibratome (World Precision Instruments, Sarasota, FL) submerged in ice-cold phosphate-buffered saline, to dissect the substantia nigra (SN), hippocampus (Hipp), visual cortex (Ctx) (bregma, −2.44 to −3.88) and the striatum (+0.66 to −1.34). Dissected brain regions were weighed, rapidly frozen on dry ice, and stored at −80 °C until analysis.
Tissues derived from different regions of individual mice were homogenized with 10 volumes of lysis buffer (20 mM HEPES pH 7.4, 150 mM NaCl, 10 % glycerol, 1 % Triton-X 100, 1 mM EGTA and 1.5 mM MgCl2, 1 μM lactacystin, 50 mM NaF, 10 mM NaPyruvate, 2 mM Na3VO4, 1 mM PMSF and protease inhibitor cocktail (P2714, Sigma, St Luis, MO). The tissues were grinded with the aid of mechanical homogenizer and then subjected to four consecutive freeze and thaw cycles of 2 minutes each (−80 °C ethanol bath/37 °C water bath). The Triton-soluble fraction was obtained after ultracentrifugation at 100,000 g for 30 minutes at 4 °C. The resulting pellets were further extracted in 5 volumes of 2 % SDS, 50 mM Tris pH 7.4 and a protease inhibitor cocktail by sonication and then boiling for 20 minutes. Samples were centrifuged for 10 minutes, 16,000 g at 25 °C to obtain the Triton-insoluble (SDS-soluble) fractions. Protein concentration was determined by the microBCA assay. (Pierce, Rockford, IL)
400 μg of Triton-soluble protein from each tissue in a total volume of 250 μl were loaded onto a Superdex 200 HR10/30 column (GE Healthcare Bio-Sciences, Uppsala, Sweden) connected to an Agilent 1100 series HPLC system (Agilent, Palo Alto, CA) for size exclusion chromatographic (SEC) analysis. The mobile phase used to resolve the protein lysates consisted of 25 mM HEPES, 150 mM NaCl, pH 7.25. The fractions were collected and concentrated with 3,000 nominal molecular weight limit (NMWL) Ultracel Microcon filters (Millipore, Bedford, MA). Fractions 2–5 (99-63 Å), 6–9 (63-48 Å), 18–20 (25-20 Å), and 21–24 (20-11 Å) were combined before loaded on the SDS-PAGE gels to enhance the signals. To isolate the 53 Å-size soluble α-syn oligomers the corresponding fractions were separated and concentrated using 3,000 NMWL Ultracel Microcon filters. The gel filtration column was calibrated using a mixture of globular proteins (HMW standards kit, GE Healthcare).
53 Å-size Triton soluble α-syn oligomers isolated by SEC from various neuronal regions were treated with 100μg/ml of proteinase K for 30 and 60 minutes at 37°C and another set of samples was treated with 8 M urea for 30 and 60 minutes at room temperature. The protein concentration for each sample was adjusted to 0.5 mg/ml. The reactions were stopped by heating to 100°C for 10 minutes after adding SDS-sample buffer. The samples were resolved on SDS-polyacrylamide gels.
Human recombinant α-syn protein used for the assay was expressed and purified as described previously (Giasson et al. 1999). In vitro α-syn aggregation was performed as described (Mazzulli et al. 2007). Briefly, 2.5 mg/ml of purified protein was incubated for 4 days in conditions promoting aggregation (shaking at 1000 rpm, 37°C) with 0.1μg/μl of fractionated lysate corresponding to the 53 Å-size Triton-soluble α-syn oligomers from the OB and SC regions. Control experiments were conducted in parallel by using samples immunodepleted from α-syn (immunodepletion described below), the fractions corresponding to the monomer of α-syn, and also using the purified protein with no addition of fractionated lysates. The corresponding fractions from lysates of WT α-syn transgenic mice were also assessed as additional controls. The progress of fibril formation was monitored every 12 hours by addition of 10μM Thioflavin T (Sigma) solution in the reaction and measuring fluorescence emission at 490 nm upon excitation at 450 nm.
For Western blot analysis of samples sequentially extracted, the following primary antibodies were used: monoclonal anti-α-syn LB509 (Baba et al., 1998; Jakes et al., 1999) (1:1000), monoclonal anti-tyrosine hydroxylase (TH) (1:4000; cat# 657010; EMD Biosciences), polyclonal antineuronal specific enolase (NSE) (1:2000, cat# 16625; Polysciences, Inc. Warrington PA) and monoclonal anti-vimentin (1:500; cat# 550513; BD Pharmingen, San Jose, CA). For Western blot analysis of fractions from gel filtration profiles the following antibodies were used: monoclonal anti-α-syn syn211 (Giasson et al., 2000) (1:1000) and polyclonal anti-NSE (1:2000, Polysciences Inc.). Fractionated lysates were also blotted with monoclonal anti-α-syn syn208 and syn303 (1:1000) (Giasson et al., 2000; Duda et al., 2002). Dopaminergic markers DOPA decarboxylase (DDC, EC 184.108.40.206) and dopamine transporter (DAT) were detected with mouse monoclonal antibodies DDC-11-M (1:500; Alpha Diagnostic, San Antonio TX) and MAB369 (1:800; Chemicon, Bedford MA) respectively. Primary antibodies were detected with either goat anti-mouse IgG Alexa Fluor 680 (1:5000; Invitrogen, Eugene, OR) or goat anti-rabbit IgG IRDye 800 (1:5000; Rockland, Gilbertsville, PA) conjugated secondary antibodies and scanned at intensity level 2 with an Odyssey Infrared Imaging System (Li-Cor Biosciences, Lincoln, NE). The bands from the western blots were quantified by densitometry using Odyssey infrared imaging system software (version 2.1 Odyssey, Li-cor).
All individual α-syn immunoreactive bands from size-exclusion chromatography (SEC)/SDS-PAGE/syn211 western blot analysis were quantified by densitometry using the Odyssey version 2.1 software (see supplemental Fig. 2) (Li-Cor Biosciences). Our previous SEC analysis of WT and A53T α-syn extracted from transfected neuroblastoma cells has shown that in vitro oligomerization does not occur as a result of the extraction and analysis procedure used here (Mazzulli et al, 2006), thus ruling out the possibility that oligomers detected in A53T α-syn mice form in vitro. α-Syn immunoreactive bands that comprised the 53 Å-size α-syn oligomer peaked at a volume corresponding to a 225 kDa globular protein in the gel filtration column, and were found to be made of a 59, 52, and 31 kDa species by SDS-PAGE. A minor oligomer eluted at a volume corresponding to a 36 Å-size particle and was comprised of 64 and 38 kDa bands as determined by SDS-PAGE. The bands eluting at a volume corresponding to a 34 Å-size particle and migrating at 19 kDa in the SDS-PAGE gels were quantified as the α-syn monomer. Each species was quantified and divided by the total α-syn signal (α-syn 53 Å + minor oligomer + α-syn monomer) to determine the percent of each.
Catechols were resolved on a reverse-phase C18 Luna column (150 × 4.6 mm, 5um; Phenomenex, Torrance, CA) connected to an 1100 series Agilent HPLC system (Agilent, Palo Alto, CA) and detected by a Coularray detector (ESA Biosciences, Chemsford, MA) as previously described (Mazzulli et al., 2006).
Cultures were prepared from C57Bl/6 mouse embryos at day 17 (stock# 3704732, Charles River, Kingston, NY). After dissection from embryos in Hepes buffered saline (supplemented with 1 mM sodium pyruvate (Invitrogen), 6 mg/ml D-Glucose (Sigma), 100 units/ml penicillin, 100μg/ml streptomycin), the cortex was mechanically dissociated, trypsinized, and seeded at 50,000 cells per well into 96 well plates (# 35–4640, BD biosciences, Bedford MA) freshly coated with 50μg/ml poly-D-Lysine (Sigma). After seeding, neurons were cultured for 2 hours in serum media (10 % heat inactivated fetal bovine serum, 100 units/ml penicillin, 100μg/ml streptomycin, 2 mM glutamine, in neurobasal media (all from Invitrogen)). After 2 hours, the media was changed to B27 media (neurobasal media containing 1X B-27 supplement, 2 mM glutamine, 100 units/ml penicillin, 100μg/ml streptomycin). At 4 days in vitro (DIV), half of the media was replaced with fresh B27 media.
Triton-soluble lysates from the hippocampus, spinal cord, or olfactory bulbs of symptomatic A53T α-syn were subjected to SEC and fractions corresponding to 53 Å-size α-syn oligomers were utilized for experiments. The fractions were concentrated and protein concentration was determined by the BCA protein assay (Pierce). To immunedeplete α-syn oligomers for control samples, 2μg of syn211 antibody (Sigma) was incubated with 100μg of fractionated lysate in a total volume of 200μl for 6 hours at 4°C. Another set of samples was also incubated with mouse immunoglobulin G (IgG, Santa Cruz Biotechnology, Santa Cruz, CA) in parallel. Protein G-conjugated beads (4 % final, Fast Flow, Sigma) were equilibrated in 25 mM Hepes, 150 mM NaCl pH 7.4 then incubated with the immunocomplexes for 1 hour at 4 °C. α-Syn oligomers were then pulled down by centrifugation at 2000 × g for 2 minutes and the protein concentration of the supernatant was determined. 1μg total protein/well of this supernatant was directly added to primary cultures grown in a 96 well plate at 9 DIV, in a total media volume of 200 μl/well. For samples incubated with IgG, the concentration of α-syn oligomers in the supernatant was estimated by quantitative western blot, using purified recombinant α-syn as a standard, and found to be approximately 10 ng oligomers/μg total protein. In addition to fractions immunodepleted of α-syn, total fractionated hippocampal lysate was utilized as a negative control (1μg lysate/well), since this tissue contains a low level of α-syn oligomers (Fig. 1 and and2).2). Cells were harvested 1 and 3 days after addition of the oligomers or immunodepleted supernatants and analyzed for neurodegeneration.
Primary cortical neurons grown on a 96 well plate were washed briefly in warm PBS, followed by immediate fixation in 4 % formaldehyde for 15 minutes. Cells were incubated with PBS containing 0.3 % TritonX-100 for 20 minutes with gentle shaking and then with Odyssey blocking buffer (Li-cor) for 1 hour. Anti-neurofilament antibody (1:1000, mouse IgG 2H3, Developmental Studies Hybridoma Bank, University of Iowa, Iowa City, IA) was then incubated overnight in blocking buffer at 4 °C with gentle shaking, followed by washing in PBS with 0.1 % Tween for 20 minutes. IRdye800-conjugated anti-mouse IgG antibodies (1:1000 dilution, Li-cor) were incubated in blocking buffer for 1 hour. Draq5 (Nuclear stain, 1:10,000 dilution, Li-cor) and Sapphire700 (non-specific cytosolic stain, 1:1000 dilution, Li-cor) were added with the secondary antibody for determination of cell volume and number. Cells were washed four times in PBS 0.1 % Tween and scanned on an odyssey infrared imaging system (Li-cor). Neurofilament intensity was determined using Odyssey software (version 2.1 Li-cor). Toxicity assays were repeated with two separate cell culture preparations with n=6 for each preparation.
One way analysis of variance (ANOVA) with Tukey’s post-hoc test was used to assess neurodegeneration in primary cortical cultures as well as SEC analysis of α-syn oligomers, and p< 0.05 was considered significant. For determination of values that varied significantly from the mean, Grubbs test was used to determine Z values. α-Syn oligomer levels that had Z values over 1.15 (n=3), 1.48 (n=4), and 1.71 (n=6) were excluded from the analysis. For direct tissue-to-tissue comparisons as well as asymptomatic/symptomatic comparisons of oligomers from the SC (shown in supplemental Fig. 2), a paired student’s t-test was used and p values <0.05 were considered significant. Two way ANOVA with a Bonferoni’s post-hoc test was used for the seeding assays shown in Fig. 4, n=4, p values <0.05 were considered significant.
To characterize the regional distribution of aggregated forms of α-syn in A53T α-syn mice, we first sought to biochemically determine the distribution of insoluble α-syn in different neural regions. Tissues from various regions of the nervous system of asymptomatic (4–6 month old) and symptomatic (11–12 month old) mice were sequentially extracted in detergents of increasing strength. Western blot analysis using an antibody specific for human α-syn, LB509, revealed that the spinal cord (SC) and cerebellum (Crb) from symptomatic mice contained more Triton-insoluble α-syn compared to asymptomatic mice (Fig. 1b). Triton-insoluble α-syn from these tissues was present primarily as a 19 kDa monomer, but also contained higher molecular weight bands stable to SDS/heat that likely correspond to dimers, trimers, and multimers. Conversely, in other brain regions such as the substantia nigra (SN), olfactory bulbs (OB), and hippocampus (Hipp), α-syn was mainly extracted in the Triton-soluble fraction with only a modest change in solubility as the mice presented with motor symptoms (Fig. 1a). This is consistent with histological analysis, which demonstrated a paucity of inclusions in these three regions (Giasson et al 2002). Analysis of age-matched transgenic mice expressing human WT α-syn, which do not exhibit α-syn pathology or motor symptoms at any age, revealed undetectable levels of Triton-insoluble α-syn in all brain regions examined (Fig. 1b and supplemental Fig. 1). These biochemical findings were corroborated by histological examination. In accordance with previous reports (Giasson et al 2002) immunohistological analysis revealed abundant α-syn inclusions (both perikaryal and neuritic) in the spinal cord of symptomatic A53T α-syn mice (Fig. 1c, panel H). To assess if indications of neuronal injury are present in areas that contain α-syn inclusions, tissue sections from symptomatic A53T α-syn transgenic mice and nTg mice were histologically examined for gliosis with an anti-GFAP antibody. Robust astrocytic gliosis is observed in the spinal cord of symptomatic A53T α-syn mice (Fig. 1c, panel D), but not in the olfactory bulb (Fig. 1c, panel B) a brain region where inclusions of α-syn are scarcely detected. These data suggest that formation of insoluble α-syn temporally correlates with motor symptom onset and occurs in regions that contain histological α-syn inclusions.
To further investigate α-syn species present in the nervous system of A53T α-syn mice, we initially focused on the SC due to the neurodegenerative changes that occur in this region once the mice develop symptoms (Fig. 1c) (Giasson et al., 2002). Triton-soluble extracts from the SC of symptomatic A53T α-syn transgenic mice were analyzed by size exclusion chromatography (SEC) followed by SDS-PAGE/Western blot analysis. Previous SEC analysis of recombinant purified α-syn had shown that although the monomeric form elutes from a gel filtration column at a volume corresponding to a 34 Å– size particle (or 58 kDa globular protein); in SDS-PAGE it migrates to an apparent molecular weight of nearly 19 kDa (Weinreb et al., 1996). This highly reproducible observation reflects the elongated, unstructured conformation of α-syn in solution. SEC analysis of Triton-soluble SC extracts from A53T α-syn transgenic mice revealed the presence of soluble 53 Å-size oligomers, with apparent molecular weights of 59, 52, and 31 kDa on SDS-PAGE (Fig. 2, and supplemental Fig 2.). These high molecular weight α-syn species were the most abundant oligomeric intermediates detected in the SC of symptomatic mice (Fig. 2). In addition to the 53 Å-size α-syn oligomers, a minor amount of α-syn oligomers eluted at nearly the same volume as the monomeric form, corresponding to a 36 Å-size particle, and migrated at 64 and 38 kDa on SDS-PAGE (Fig. 2, supplemental Fig. 2). Next, SEC profiles of the SC of symptomatic mice were compared to those of asymptomatic mice (supplemental Figs. 2, 3) and revealed a similar SEC profile, indicating that soluble α-syn oligomers accumulate before the formation of insoluble α-syn inclusions and presentation of motor deficits. Additionally, the amounts of oligomers in symptomatic and asymptomatic mice were not significantly different by quantitative densitometry analysis of the SEC/SDS-PAGE blots (Fig. 2). Despite the detection of various oligomeric α-syn species, the SEC analysis did not identify a soluble oligomeric form of α-syn that is specifically correlated with symptom onset (supplemental Fig. 2).
To determine if α-syn oligomers exist in other histologically unaffected regions of symptomatic A53T α-syn transgenic mice, Triton-soluble extracts from the SN, OB, and Hipp were analyzed by SEC. Unexpectedly this analysis revealed a similar elution profile compared to the inclusion bearing regions (SC and Crb), with the 53 Å-size α-syn oligomers as the most prominent species (Fig. 2). Quantification indicated that the relative levels of oligomeric α-syn in the SN, and OB were similar to those in the SC (Fig. 2). Additionally, SEC analysis of hippocampal lysates of A53T α-syn transgenic mice as well as most tissues from control mice expressing WT α-syn showed very low levels of soluble oligomeric α-syn (Fig. 2 and supplemental Fig. 4), indicating that soluble oligomers detected by SEC form as a result of in vivo processes.
A lack of correlation between the relative abundance of soluble oligomers and the formation of insoluble inclusions is unexpected, given the assumption that these oligomeric species are intermediates of the fibril generating process. Therefore, the biochemical nature of the 53 Å-size α-syn Triton-soluble oligomers was further characterized. For this analysis, we focused on the OB and SC because of the clear histological differences between these regions, as well as tissue availability. Following SEC, the fractions corresponding to Stokes radii of 53 Å and higher were separated from the rest of the fractions and concentrated to obtain an enriched solution of oligomeric species. Treatment of 53 Å-size oligomers derived from the SC or OB with urea did not affect the 59, 52, and 31 kDa bands suggesting that these species represent self-associated α-syn molecules that are held together by urea resistant bonds (Fig. 3a). Incubation with proteinase K (PK) was then performed to reveal potential regions of α-syn that might be protected from proteolysis. Oligomers of the SC and OB were equally sensitive to PK treatment, indicating a lack of buried peptide regions protected from digestion (Fig. 3b). Taken together these data indicate that α-syn oligomers in the brain of A53T α-syn transgenic mice are Triton-soluble, SDS and heat stable, resistant to denaturation by urea, and sensitive to PK. Despite the similarity in these biochemical properties, an interesting distinction between oligomers of the OB and SC was evident upon western blot analysis using syn303, a monoclonal antibody generated against oxidized/nitrated recombinant α-syn (Duda et al., 2002). Immunoblotting with syn303 revealed more prominent immunoreactivity with SC oligomers as opposed to the OB in the bands corresponding to dimers and trimers having an apparent molecular weight of 59, 52, and 31 kDa (Fig. 3c). The same samples were analyzed by immunoblot using LB509, a human specific antibody that does not preferentially detect oxidized forms of α-syn, showing roughly equal amounts of oligomers in each region (supplemental Fig. 5). Since syn303 preferentially recognizes oxidized and/or nitrated α-syn in human inclusions (Duda et al., 2002) these data suggest that oxidative modifications to soluble oligomers occur to a greater extent in the SC compared to the OB.
To determine if further differences exist between oligomers of the SC compared to the OB, we compared the ability of these oligomers to influence the kinetics of the in vitro polymerization of α-syn. Based on previous structural analysis of in vitro formed oligomers (Conway et al., 2000a; Apetri et al., 2006) the size of the oligomers detected here (53 Å) is consistent with an intermediate species in the pathway of forming α-syn amyloid fibrils. Therefore “on-pathway” oligomers should act as “seeds” in the in vitro polymerization of α-syn and thus accelerate the initial phase of this process (Jarrett and Lansbury, 1993; Wood et al., 1999). Indeed enriched oligomeric fractions from SC shortened the lag phase and increased the rate of recombinant purified α-syn forming amyloid fibrils as indicated by the changes in Thioflavin T fluorescence (Fig. 4a) when compared to identical SC oligomer-enriched fractions that were immunodepleted for α-syn. In contrast oligomeric α-syn extracted from the OB prolonged the lag phase of purified α-syn amyloid fibril formation (Fig. 4b). These effects on the in vitro kinetics of fibril formation were dependent on α-syn since immunodepletion of the oligomers restored the kinetic profile of the purified α-syn aggregation. Moreover, addition of α-syn monomers extracted from SC or OB of A53T α-syn transgenic mice as well as of SEC fractions from SC or OB of mice expressing WT α-syn corresponding to ≥53 Å did not alter the kinetics of α-syn aggregation (Table 1 and supplemental Fig. 6). Collectively the data indicate that oligomers derived from the OB represent either “off-pathway” oligomers, or contain an oligomer-associated factor that kinetically delays the formation of α-syn fibrils. The absence of the detectable α-syn inclusions in the OB of the A53T α-syn transgenic mice supports these findings.
To determine the toxicity of SC and OB oligomers, SEC fractions containing oligomers were added to primary cortical cultures and neuronal toxicity was assessed at 1 and 3 days after treatment by measuring total cell number using a fluorescent non-specific cell stain in combination with a nuclear stain. Neurodegeneration was also assessed by immunostaining for neurofilament (NF) protein, which is a more sensitive assay that can measure degeneration of neurites prior to occurrence of more severe nuclear toxicity (Zala et al., 2005). While toxicity of α-syn is presumed to occur through intracellular events in PD, recent studies have indicated that α-syn aggregates are taken up by neurons in culture through endocytosis and can be disseminated from cell to cell causing widespread neurotoxicity (Desplats et al., 2009). Furthermore, other studies have shown that in vitro formed α-syn oligomers exhibit toxicity when applied externally to cell cultures, likely through disruptions in lipid bilayers (Kayed et al., 2003; Kayed et al., 2004). Therefore, we considered this experimental paradigm as a reliable proxy for the potential toxicity that may occur in vivo. SEC fractions containing the 53 Å-size oligomers were added to cultures and neurodegeneration was assessed. We found that addition of this SEC fraction containing α-syn oligomers derived from the SC resulted in a significant decline in cell viability as determined by analysis of cell volume and total cell number, whereas the same SEC fraction obtained from the OB had no effect (Fig. 5a). We also observed a decrease in neurofilament immunostain intensity, attributed to neurite degeneration, in cultures treated with SC oligomers but not oligomers obtained from the OB (Fig 5b). The toxicity of SC oligomers was observed 1 day after treatment and persisted until 3 days after addition of the oligomers. This effect is presumed to be due to the presence of oligomeric α-syn since depletion with an anti-α-syn antibody reversed the neurodegeneration observed in the SC treated cultures (Fig. 5a, b). Additionally, the observed toxicity differences between the OB and SC did not result from addition of different quantities of oligomers, since we quantified the amount of α-syn oligomers that were added to the primary cultures by immunoprecipitation and western blot (Fig. 5d). These data suggest that despite similar SEC profiles and biochemical properties, α-syn oligomers display different cellular toxicities. The histological analysis shown in Fig. 1 also supports the neuron toxicity findings, since astrogliosis is more prominent in the SC.
Although previous histological analysis of A53T α-syn transgenic mice at 4–5 months indicated normal cell structure without neurodegenerative changes (Giasson et al., 2002), we sought to determine if the presence of detergent-soluble oligomers detected by SEC/SDS-PAGE is reflected on behavioral changes in the mice. Ambulatory activity measurements revealed that A53T α-syn transgenic mice had similar motor activity compared to nTg mice (Table 2A). Food intake and body weight of A53T α-syn transgenic mice were also assessed, since it has been well established that feeding behavior critically depends on the function of the SNpc→ caudate putamen (CPu) circuit (Zigmond and Stricker, 1972; Szczypka et al., 2001). We found that A53T α-syn transgenic mice had similar body weights and consumed slightly more food compared to nTg age-matched controls, suggesting that the SNpc→CPu circuit is functional (Table 2A). Growth curves, by measuring body weight starting at 2 weeks did not reveal any differences in the development of A53T α-syn transgenic mice for up to 11 months of age (data not shown). After 11 months, A53T α-syn transgenic mice experience motor deficits and a rapid decline in body weight at variable time points. These data show that the presence of detergent-soluble α-syn oligomers in the SN is associated with only mild behavioral changes.
Further biochemical evaluation was also performed to determine the integrity of the dopaminergic system in asymptomatic and symptomatic mice. Enzyme levels related to dopamine synthesis and metabolism were quantified, such as tyrosine hydroxylase (TH), dopamine transporter (DAT), and DOPA decarboxylase (DDC) in symptomatic mice and showed no statistically significant difference compared to nTg mice (supplemental Fig. 7). Additionally, striatal catechol levels were measured in both asymptomatic and symptomatic A53T α-syn transgenic mice, as well as in non-transgenic control mice, by HPLC-ECD. Analysis of 4-month-old mice showed that striatal catechol levels of A53T α-syn transgenic mice were similar to non-transgenic age-matched controls (femtomoles/μg protein, non-transgenic mice: DA=598±82.5, DOPAC=100.4±27.9; A53T α-syn transgenic mice: DA=703±30, DOPAC=89.9±11.5; n=3–4, mean values ± SEM). Analysis of symptomatic A53T α-syn transgenic mice revealed slightly elevated levels of striatal dopamine while the levels of other analytes tested including DOPAC, norepinephrine, and serotonin, did not significantly change (Table 2B). Since the ratio of DOPAC to dopamine reflects the relative rate of DA turnover, these data suggest that the increased concentration of DA may be due to increased synthesis or storage. Catechol levels in the OB of symptomatic A53T α-syn transgenic mice were also similar to age-matched non-transgenic controls (data not shown). Collectively, these data suggest that the presence of Triton-soluble oligomeric α-syn in dopaminergic regions is not associated with alterations in the levels of enzymes involved in dopamine metabolism and coincides with subtle differences in the production of catechols in vivo as the levels of dopamine increase with aging in A53T α-syn transgenic mice.
The documentation that Lewy bodies (LB) and other neuronal inclusions are composed of aggregated α-syn has positioned the process of α-syn aggregation as a critical biochemical event in the pathogenesis of neurodegeneration (Spillantini et al., 1997; Baba et al., 1998; Goedert, 2001). Considerable advances in cell free systems and cell model systems have indicated that the process of aggregation includes a committed step of α-syn dimer formation that is held together by hydrophobic interactions induced by a conformational transition to β-sheet structure (Conway et al., 2000b; Serpell et al., 2000; Uversky et al., 2001). The dimer represents a thermodynamically stable entity allowing the formation of oligomers which will continue to grow in linear β-sheets, forming polarized protofibrils and eventually the typical amyloid fibrils. Alternatively, the transition of oligomers to protofibrils can be kinetically arrested by interactions with small molecules, other proteins or unidentified cellular constituents allowing these oligomers to assume secondary structures such as annular pores or spheres (Conway et al., 2001; Li et al., 2004; Norris et al., 2005).
The emerging appreciation of conformationally distinct oligomers of not only α-syn, but also other proteins that form amyloid structures in human neurodegenerative disorders, has sparked a considerable interest in the pathobiological role of oligomers and mechanism by which they induce neuron dysfunction and neurodegeneration. However, in part due to methodological challenges there is limited insight regarding the biochemical and biological properties of oligomeric species formed in vivo. Therefore the in vivo regional distribution and temporal formation of α-syn oligomeric intermediates was investigated in the previously characterized transgenic mouse model expressing human A53T α-syn (Giasson et al., 2002) that recapitulates the aggregation of α-syn and inclusion formation. Detailed biochemical and biological analysis of the in vivo formed α-syn oligomers revealed unexpected new insights regarding regional differences in conformation and composition that may govern their ability to promote inclusion formation and induce neuron dysfunction.
Using the previously validated two-dimensional analysis, the presence of detergent-soluble oligomeric forms of α-syn was confirmed in brain regions that contain inclusions and unexpectedly in brain regions devoid of α-syn inclusions such as the SN and OB. The presence of oligomeric intermediates in regions containing insoluble α-syn inclusions is expected assuming that these α-syn oligomers represent transient intermediates of the polymerization process. The presence of these oligomers in regions that do not form inclusions, at least up to the life-span of the symptomatic mouse, indicates that the process of aggregation has been initiated in these regions but has not progressed beyond the oligomeric stages. Moreover, the data indicated that oligomers are formed before the formation of inclusions and onset of symptoms (at 4–6 months), and continue to be present at similar levels during aging and symptom onset.
The regional variation is not a result of differences in the levels of human A53T α-syn protein since previous studies utilizing the mouse PrP promoter have shown similar expression levels in various brain regions (Lee et al., 2002), which was confirmed by quantitative western blot analysis for all regions analyzed (Fig. 1 and and2).2). Furthermore, regional variation was not due to generation of oligomers with divergent basic biochemical properties. Detergent-soluble α-syn oligomers isolated from either inclusion bearing or inclusion free regions were composed primarily of 53 Å-size species that are SDS, heat and urea stable, and are sensitive to digestion by proteinase K. However, subtle differences in conformation are evident by the differential reactivity towards oligomer specific antibodies (Fig. 3c). These antibodies were generated against oxidized and tyrosine nitrated α-syn oligomers and were shown to preferentially recognize detergent-soluble α-syn oligomers as well as pathological α-syn inclusions in human brain tissue (Duda et al., 2002). These antibodies also recognize oxidative post-translationally modified α-syn such as dityrosine cross-linked dimers, which has been considered a critical nucleation event for the assembly of α-syn monomers into fibrils (Souza et al., 2000a; Krishnan et al., 2003). It is possible that the oligomers extracted from SC contain a higher proportion of dityrosine cross-linked oligomers that facilitate amyloid fibril formation. This possibility is in part supported by the data documenting that SC-derived oligomers accelerated the aggregation of purified α-syn in a typical fibril formation assay (Fig. 4a and Table 1). In contrast oligomers isolated from the OB prolonged the lag phase of α-syn aggregation in the same assay (Fig. 4b and Table 1). The contrasting biological function of the different oligomeric populations was also evident in primary neuron cultures where oligomers derived from SC but not OB induced cell death and dramatic neurite degeneration as determined by neurofilament stain intensity (Fig. 5).
One candidate pathway that may explain the kinetic deceleration of fiber formation of the OB-extracted oligomers may relate to the previously recognized ability of oxidation products of catechols, such as the ortho-quinones of dopamine and DOPAC, to kinetically arrest the formation of α-syn fibrils (Conway et al., 2001; Li et al., 2005; Norris et al., 2005). This non-covalent interaction has been documented in cell free systems and recently in cellular model systems (Mazzulli et al., 2006; Mazzulli et al., 2007; Mosharov et al., 2009; Outeiro et al., 2009). Interestingly, we found that striatal catechol levels of A53T α-syn transgenic mice increase with age, while they remain stable in non-transgenic mice. This occurred without an apparent increase in the levels of tyrosine hydroxylase, the DA synthesizing rate-limiting enzyme. Tyrosine hydroxylase utilizes tyrosine to catalyze the formation of L-DOPA, a process that is dependent on molecular oxygen, Fe+2, and the cofactor 6(R)-L-erythro-5,6,7,8-tetrahydrobiopterin (BH4) (Fitzpatrick, 1989). It has been established that tyrosine hydroxylase activity is intricately regulated through feedback inhibition by catechols, allosteric regulation by various polyanions, and posttranslational modifications such as phosphorylation in the N-terminal region of the protein (Kumer and Vrana, 1996). Therefore, it is possible that the observed increase in striatal DA may result from altered levels of cofactors or posttranslational modifications that increase the enzymatic activity without affecting protein levels. While an increase in striatal DA may provide a biochemical mechanism for preventing the formation of fibrils in the SNpc of A53T α-syn transgenic mice, genetic and pharmacological manipulations of dopamine levels are required to verify the biological importance of a potential α-syn-catechol interaction in vivo. In relation to PD, it is possible that either a decline in steady-state dopamine levels or cleavage of the DA-interacting C-terminus of α-syn contribute to the formation of Lewy bodies in the SNpc, resulting in degeneration of the nigrostriatal pathway.
Whilst the dopamine interaction may be prominent for dopaminergic neurons, other intrinsic factors present in non-dopaminergic neurons must be also considered. Previous studies using cell models have shown that molecular chaperones have the ability to alter the levels of soluble oligomeric species (Tetzlaff et al., 2008). It is possible that some regions of A53T α-syn transgenic mice possess elevated levels molecular chaperones or other intrinsic factors that may engage the soluble oligomers and possibly accelerate clearance or alter their conformation, thus preventing their conversion into insoluble fibrils within the life span of the A53T α-syn transgenic mice. This ability may be sustained with age in some brain regions (such as the Hipp) but decline or become overwhelmed in other susceptible neuronal regions (such as the Crb and SC) which also contain oxidatively modified α-syn species that have the potential to promote the polymerization process resulting in the formation of insoluble inclusions (Souza et al., 2000a).
Overall the data provide a systematic identification and biochemical characterization of in vivo formed oligomeric intermediates of α-syn in the A53T α-syn transgenic mice that exhibit neurodegeneration and α-syn inclusion formation. Although the analysis does not include transient unstable oligomeric conformations that remain unaffected through the extraction process, these more stable and well-represented α-syn assemblies were shown to vary in conformation and possibly composition, based on the region of origin. These differences in addition to other unrecognized regional specific factors may be responsible for the ex vivo and in vivo toxicity of the oligomers. The insights gained by the present analysis highlight the importance of therapeutic interventions with agents that selectively enhance the clearance of α-syn. Those compounds are likely to prevent the formation of both soluble and insoluble toxic α-syn assemblies, providing therapeutic benefit in PD and related disorders.
This work was supported by the National Institutes of Health grants AG09215 (B.I.G), NS053488 (B.I.G.), R01NS051303 (D.K.), USPHS AG13966 (H.I.), ES013508 NIEHS Center of Excellence in Environmental Toxicology (H.I), and award number F32NS066730 to J.R.M. from the National Institute Of Neurological Disorders And Stroke. H.I. is the Gisela and Dennis Alter Research Professor of Pediatrics.