Address correspondence and reprint requests to P. Jenner at, Wolfson Centre for Age-Related Diseases, Hodgkin Building, Guy's, King's and St Thomas' School of Biomedical Sciences, King's College London, Guy's Campus, London Bridge, London SE1 1UL, UK. E-mail: email@example.com
Mutations in α-synuclein (A30P and A53T) are involved in some cases of familial Parkinson's disease (FPD), but it is not known how they result in nigral cell death. We examined the effect of α-synuclein overexpression on the response of cells to various insults. Wild-type α-synuclein and α-synuclein mutations associated with FPD were overexpressed in NT-2/D1 and SK-N-MC cells. Overexpression of wild-type α-synuclein delayed cell death induced by serum withdrawal or H2O2, but did not delay cell death induced by 1-methyl-4-phenylpyridinium ion (MPP+). By contrast, wild-type α-synuclein transfectants were sensitive to viability loss induced by staurosporine, lactacystin or 4-hydroxy-2-trans-nonenal (HNE). Decreases in glutathione (GSH) levels were attenuated by wild-type α-synuclein after serum deprivation, but were aggravated following lactacystin or staurosporine treatment. Mutant α-synucleins increased levels of 8-hydroxyguanine, protein carbonyls, lipid peroxidation and 3-nitrotyrosine, and markedly accelerated cell death in response to all the insults examined. The decrease in GSH levels was enhanced in mutant α-synuclein transfectants. The loss of viability induced by toxic insults was by apoptosic mechanism. The presence of abnormal α-synucleins in substantia nigra in PD may increase neuronal vulnerability to a range of toxic agents.
Parkinson's disease (PD) results from degeneration of dopaminergic neurons in the substantia nigra (Jenner and Olanow 1998). Although most cases appear sporadic and of unknown cause, familial Parkinson's disease (FPD) can occur and defects in the gene encoding α-synuclein protein are sometimes involved (Polymeropoulos et al. 1997). α-synuclein is a 15-kDa protein containing 140 amino acids and its major expression site is the presynaptic terminal. The amino terminal region contains five consensus sequence repeats, ‘KTKEGV’ and the carboxyl terminal is rich in acidic amino acids (Jakes et al. 1994). α-Synuclein is present in Lewy bodies, which are a characteristic feature of PD, and of dementia with Lewy body disease (Spillantini et al. 1998a). Abnormal staining for α-synuclein is found in neurofilament inclusions in multisystem atrophy (Spillantini et al. 1998b) and in Alzheimer's disease (Goedert et al. 1998). Two mutated α-synuclein proteins (A30P and A53T) have been found in some cases of FPD (Polymeropoulos et al. 1997; Kruger et al. 1998). The mutant α-synucleins have been suggested to self-aggregate, and to aggregate with wild-type α-synuclein and other proteins, to give insoluble deposits that cannot undergo proteolysis by the proteasome (Mezey et al. 1998). Oxidative stress may also promote the aggregation of α-synuclein (Hashimoto et al. 1999a). The proteasome is important in clearing abnormal proteins, including oxidatively damaged proteins within the cells (Grune et al. 1995; Fenteany and Schreiber 1998).
In the present study, we investigated how cell death induced by a variety of insults was modulated by overexpression of wild-type or FPD-linked mutant α-synuclein proteins. Since oxidative stress is involved in PD (Jenner and Olanow 1998), we tested insults related to this process, namely H2O2 and 4-hydroxy-2-trans-nonenal (HNE). The latter is a toxic unsaturated aldehyde produced during lipid peroxidation and which is present in increased amounts in parkinsonian substantia nigra (Yoritaka et al. 1996). We also examined serum deprivation, the mitochondrial complex I inhibitor, MPP+ (Singer et al. 1987), the protein kinase inhibitor, staurosporine (Fujita-Yamaguchi and Kathuria 1998) and an inhibitor of proteasome function, lactacystin (Fenteany and Schreiber 1998). To ensure that our results were not unique to a particular cell line, we compared two different cell lines, namely NT-2/D1 (NT-2) and SK-N-MC.
Materials and methods
NT-2 is a human teratocarcinoma cell line (Angulo et al. 1995) and SK-N-MC is a human neuroblastoma cell line, which has cholinergic characteristics (Biedler et al. 1978). These cell lines and their α-synuclein transfectants were maintained in 100-mm tissue culture plates (Greiner, Germany) containing high-glucose Dulbecco's modified Eagle's medium (HG-DMEM; Life Technologies, UK), 1 mm sodium pyruvate (Sigma, UK), 10% fetal bovine serum (Life Technologies) and 100 IU/mL penicillin and 100 µg/mL streptomycin (Life Technologies) under humidified 5% CO2 and 95% air.
Site-directed mutagenesis and cloning of α-synuclein genes
α-Synuclein cDNA was a gift from Prof. M. Goedert (MRC Molecular Biology Laboratory, Cambridge, UK) and was inserted in the bacterial expression vector, pRK172 (McLord et al. 1987). Primers for generating A30P mutant protein, 5′-GGT GTG GCA GAA GCA CCA GGA AAG ACA AAA G-3′ and 5′-C TTT TGT CTT TCC TGG TGC TTC TGC CAC ACC-3′ and for A53T mutant protein, 5′-GTG GTG CAT GGT GTG ACA ACA GTG GCT GAG A-3′ and 5′-T CTC AGC CAC TGT TGT CAC ACC ATG CAC CAC-3′ were used for a PCR-based mutagenesis kit (QuikChange site-directed mutagenesis Kit, Stratagene, USA). The mutant cDNAs (100 ng) were transformed into Escherichia coli XL-1 blue competent cells (Stratagene, USA). The plasmid DNAs were prepared (Sambrook et al. 1989) to identify correct mutagenesis by treating with restriction enzymes, MvaI (Boehringer Mannheim, Germany) for A30P and Tsp45 I (Promega, UK) for A53T. In order to confirm successful mutagenesis, whole α-synuclein cDNAs were sequenced with Sequenase version 2.0 DNA sequencing kit (United States Biochemical, USA).
For cloning of wild-type and mutant α-synuclein cDNAs from bacterial to animal expression vector, the cDNAs (10 µg) were cut with NdeI restriction enzyme (Promega) and gap-filled with DNA polymerase I (Promega). The DNAs were ligated by NheI restriction enzyme site (5′-GCTAGC-3′) with T4 DNA ligase (Promega) and treated with HindIII and NheI enzymes (Promega) to collect the α-synuclein cDNAs. An ecdysone-inducible mammalian expression system (In Vitrogen, the Netherlands) was used for cloning the cDNAs.
pIND(SP1)-hygro, pIND(SP1)-hygro(LacZ), pIND(SP1)-wild-type α-synuclein, pIND(SP1)-A30P mutant α-synuclein, pIND(SP1)-A53T mutant α-synuclein were used. At least 10 µg of each cloned DNA in TE buffer (pH 7.4) was mixed with HEPES buffer (pH 7.4) (final volume 50 µL) and added into the mixture containing 30 µL DOTAP (Boehringer Mannheim) and 70 µL HEPES buffer. This mixture was added to cells of approximately 80% confluence in 100-mm tissue culture plates and incubated at 37°C overnight. Transfectants were selected with 1 mg/mL hygromycin (In Vitrogen) and 1 mg/mL Zeocin (In Vitrogen).
Cells were lysed with a buffer containing 5 µg/mL aprotinin, 5 µg/mL leupeptin and 5 µg/mL pepstastin A (Sigma) and 10% sodium dodecyl sulfate (SDS) and placed on ice for 5 min. The lysates were centrifuged (12 000 g, 10 min) and the supernatants were transferred into new Eppendorf tubes. Protein levels were measured (Lowry et al. 1951), and a total of 10 µg protein was then electrophoresed on 8% SDS polyacrylamide gels for 1.5 h (100 V). Separated proteins were transferred to nitrocellulose membranes (Bio-Rad, USA) at 20 V overnight. The membranes were incubated with anti-α-synuclein monoclonal antibody (1 : 500 dilution, Zymed, USA) for 2 h and then washed with phosphate-buffered saline (PBS) (pH 7.4). The membranes were incubated with alkaline phosphatase-conjugated anti-IgG antibody (1 : 500 dilution, Vector stain, UK) for 1 h. A color reaction was performed with 5-bromo-4-chloro-3-indolyl phosphate/nitroblue tetrazolium (BCIP/NBT) (Sigma). The blotting membranes were washed with PBS to stop the color reaction and then were dried for determining the amount of expressed protein in an image analyzer (Imaging Research Inc., UK).
Cell growth rate by sulforhodamine B test
Cells (103) were plated on a 96-well plate and incubated at 37°C. Medium was removed and cells washed three times with PBS (pH 7.4), then 400 µL of 20% trichloroacetic acid (TCA) was added. The plates were placed at 4°C for 1 h. The samples were washed five times with deionized water and 200 µL sulforhodamine B (SRB; 0.4% in 0.1% acetic acid) was added. The mixtures were placed at 4°C for 1 h. Then the plates were washed five times with deionized water and allowed to dry at room temperature overnight. Finally, 200 µL of 10 mm Tris was added and the absorbance read at 510 nm.
Wild-type α-synuclein transfectants appeared normal and proliferated at the same rate as non-transfectants and vector-only transfectants under normal growth conditions. By contrast, the growth rates of the two mutant α-synuclein transfectants were decreased by approximately 50% (data not shown). Cells transfected with β-galactosidase (used as a control protein) had identical growth rates to non- or vector-only transfectants (data not shown).
Cell viability assays
When cells reached 80% confluence, normal medium was changed to new medium containing 1 µm ponasterone for 1 day (Ecdysone analogue, In Vitrogen) to allow expression of α-synuclein proteins. Serum-free medium, normal medium containing 20 µm HNE, normal medium containing 100 nm staurosporine, normal medium containing 25 µm lactacystin or normal medium containing 100 µm MPP+ was added for a maximum of 5 days. For H2O2 treatment, cells were washed and placed in Hank's balanced salt solution (HBSS) containing a final concentration of 100 µm H2O2, and incubated for 1 h. The H2O2-HBSS was removed and the cells were washed twice with PBS and then returned to normal medium.
The cells were assayed for viability using trypan blue exclusion, which measures the integrity of the plasma membrane. Cells were trypsinized, washed twice with PBS, trypan blue dye added and the number of dye-excluding cells counted on a hemocytometer on triplicate dishes.
The MTT [3-(4,5-dimethyl-thiazol-2-yl)-2,5-diphenyltetrazolium bromide] assay was also used to assess the metabolic activity of cells. Reduction of MTT is largely dependent on the degree of mitochondrial activity in cells (Liu et al. 1997). Cells (2 × 105) were washed twice with PBS (pH 7.4) and grown in 96-well plates overnight. They were then incubated with MTT (Sigma). In living cells, MTT can be reduced to MTT-formazan. The medium-MTT mixture was then removed and the remaining dark blue crystals were dissolved with 500 µL dimethyl sulfoxide (DMSO). Absorbance (550 nm) was measured in a multiwell plate reader (Perseptive Biosystems, UK).
Propidium iodide and Hoechst 33258 for cell shrinkage and chromatin condensation
Trypsinized cells were washed twice with PBS (pH 7.4) and fixed with 70% (v/v) ethanol for 30 min on ice. The fixed cells were washed twice with PBS and 100 µg/mL propidium iodide (PI; Sigma) and 100 µg/mL Hoechst 33258 (Sigma) were added and incubated for 10 min at room temperature. The cells were examined under the fluorescence microscope at 100 times magnification (> 600 nm for PI and around 430–450 nm for Hoechst 33258) and counted in an image analyzer (Imaging Research Inc.).
Cells were trypsinized and washed twice with PBS (pH 7.4). The cells were then fixed with 4% (w/v) paraformaldehyde and incubated on ice for 10 min, and washed twice with PBS. Permeabilization solution (0.1% (w/v) sodium citrate and 0.1% (w/v) Triton X-100) was added and cells incubated on ice for 2 min. Terminal deoxynucleotidyl transferase-mediated dUTP-biotin nick-end labelling (TUNEL) mixture (terminal deoxynucleotidyl transferase plus fluorescein-labelled dUTP in buffer, Boehringer Mannheim) was added and the solution incubated at 37°C for 1 h. The cells were examined under a fluorescence microscope at 450 nm (1000 times magnification) and counted where necessary in an image-analyzer (Imaging Research Inc.).
Measurement of GSH level
GSH level was assessed as described by Hissin and Hilf (1976) with 105 cells. This assay detects GSH by its reaction with o-phthalaldehyde at pH 8. A standard curve was made with GSH (Sigma).
Measurement of oxidative damage
DNA extraction and assessment of purity were carried out as described by Lyras et al. (1997). DNA hydrolysis with formic acid and separation of modified 8-hydroxyguanine (8-OHG) by HPLC were performed as described by Kaur and Halliwell (1996).
Protein carbonyl content was determined by method A of Lyras et al. (1996), except that the final protein pellets were dissolved in 1 mL of 6 m guanidinium hydrochloride. Carbonyl content was calculated as nmol/mg protein (Reznick and Packer 1994).
Measurement of 3-nitrotyrosine content was carried out as described by Khan et al. (1998). Peroxynitrite (ONOO−) was prepared in a quenched flow reaction system (Yermilov et al. 1995). Nitrated BSA was prepared by using a method described by Whiteman and Halliwell (1999). The conjugation of HRP with anti-3-nitrotyrosine antibody was performed with the periodate method as described by Harlow and Lane (1988). A standard curve for 3-nitrotyrosine (Sigma) and for nitrated bovine serum albumin (BSA) was made using serially diluted nitrated BSA and 3-nitrotyrosine solution.
For measuring lipid damage, thiobarbituric acid (TBA) reactive material was measured by HPLC as described by Lyras et al. (1998), employing 3 × 106 cells.
Statistical differences were analyzed by one- and two-way anova tests. Trypan blue exclusion and MTT data were analyzed by a two-way anova test. Multiple comparisons were performed with post hoc Bonferroni t-test.
We selected more than 15 clones, stably expressing wild-type or mutant α-synucleins in both cell lines. After control experiments measuring growth rate and protein expression, we chose 10 clones. All showed consistent traits; (1) protein expression which was 3 times higher than vector-only transfectants and (2) slower growth rate than vector-only transfectants. For viability test and oxidative stress (levels of GSH and oxidative damage), 4 and 6 clones were chosen respectively. Three independent experiments were done with each clone.
Overexpression of wild-type and mutant α-synuclein proteins
Both NT-2 and SK-N-MC cell lines contained detectable levels of α-synuclein (Fig. 1). Addition of 1 µm ponasterone to all transfectants for 1 day caused α-synuclein protein over-expression by about 3.5-fold for both wild-type and mutant α-synuclein gene transfectants (3.5 ± 0.3-fold, n = 4, a representative blot is shown in Fig. 1).
Effects of serum withdrawal
Withdrawal of serum from either NT-2 or SK-N-MC cells led to a loss of viability as measured by trypan blue exclusion or MTT reduction. Loss of metabolic activity (as assessed by MTT reduction) occurred faster than loss of dye exclusion, being almost complete at 36–48 h whereas loss of viability as measured by trypan blue was complete at 3–5 days (Fig. 2). Light microscope examination of the cells at 24 h showed morphological features characteristic of apoptosis, which was confirmed by cell morphology change (PI), chromatin condensation (Hoechst 33258), DNA fragmentation (TUNEL assay) (Fig. 3) and by DNA laddering (data not shown).
Overexpression of wild-type α-synuclein delayed cell death induced by serum withdrawal, as measured by three different parameters, trypan blue exclusion, MTT reduction and features of apoptosis. By contrast, both mutant α-synucleins rendered cells more sensitive to serum withdrawal (Figs 2a and b), such that by any of the three assays used the cells lost viability faster. Almost all of the non- or vector-only transfectants failed to exclude trypan blue dye within 5 days after serum deprivation and showed evidence of DNA laddering at 1 day (data not shown). However, almost 40% of wild-type α-synuclein transfectants still excluded trypan blue at 5 days and the appearance of laddering was delayed until 2 days after serum withdrawal. About 35% of cells expressing wild-type α-synuclein showed cell morphology change, chromatin condensation and DNA fragmentation as measured by the PI, Hoechst 33258, and TUNEL assays whereas 50% of non- or vector-only transfectants did so (p < 0.01, Fig. 3). By contrast, cells over-expressing either mutant α-synuclein failed to exclude trypan blue within 3 (NT-2 cell line) or 3–4 days (SK-N-MC cell line). Similarly, metabolic activity as assessed by MTT reduction was largely lost by 24–48 h in non- and vector-only transfectants, but by 12 h in the mutant α-synuclein transfectants as compared with over 72 h when wild-type α-synuclein transfectants were examined (Figs 2c and d). Approximately, 90% of cells expressing either mutant α-synuclein were observed as apoptosis-positive after 1 day of serum deprivation (p < 0.01, Fig. 3).
Over-expression of the control protein, β-galactosidase had no effect on cell viability after serum withdrawal using any assay (data not shown). High-level expression of β-galactosidase was confirmed by using its substrate, 5-bromo-4-chloro-3-indolyl-β-d-galactoside to produce a blue chromogen (data not shown).
Cell sensitivity to H2O2
Addition of H2O2 (final concentration, 100 µm) to both cell lines induced loss of cell viability, confirmed as apoptosis by morphological examination, DNA laddering, PI, Hoechst 33258 and TUNEL assays (although higher H2O2 concentrations tended to favor necrotic cell death; data not shown). Over-expression of wild-type α-synuclein protein protected somewhat against the effects of H2O2 in the NT-2 cell line (p < 0.05 from 3 days after the treatment, data not shown) but to a lesser extent in the SK-N-MC cell line as assessed by trypan blue exclusion (p < 0.05 only at day 4 after the treatment, data not shown). By contrast, the mutant α-synuclein transfectants of both cell lines lost viability faster than controls after H2O2 exposure. Measurements of DNA laddering and other features of apoptosis and MTT reduction confirmed these effects, except that the wild-type α-synuclein transfectants were significantly protected against H2O2 as measured by any of these parameters (p < 0.01 from 12 h after the treatment in NT-2 cell line and from 9 h in the SK-N-MC cell line, Fig. 4).
Cell sensitivity to the mitochondrial inhibitor, MPP+
Addition of MPP+ induced a loss of viability in both cell lines, again largely by apoptosis as revealed by morphological examinations and DNA laddering (data not shown). Over-expression of wild-type α-synuclein had no significant effect on susceptibility to MPP+ in either cell line. However, the two mutant α-synucleins caused an accelerated loss of viability as measured by trypan blue exclusion (Fig. 5), MTT reduction, DNA laddering and other chacteristics of apoptosis (Fig. 3). For example, all the mutant α-synuclein transfectants failed to exclude trypan blue dye within 5 days after exposure to MPP+ (p < 0.01 compared with control transfectants from 1 day), and DNA laddering was obvious at 12 h (data not shown). They also lost MTT-reducing activity in 12 h (p < 0.01 compared with control transfectants from 3 h, data not shown). This was confirmed by cell morphology change, chromatin condensation and DNA fragmentation using the PI, Hoechst 33258 and TUNEL assays; 45–50% of the control and wild-type α-synuclein transfectants showed DNA fragmentation at 1 day whereas 90% of both cell types transfected with either mutant α-synuclein were apoptotic (p < 0.01, Fig. 3).
Cell sensitivity to lipid peroxidation end product, HNE
HNE concentrations in the range of 1–100 µm caused a loss of viability in both NT-2 and SK-N-MC cell lines, and 20 µm HNE was selected as producing an approximately 50% loss of viability by 5 days as measured by trypan blue exclusion (data not shown). Again both NT-2 and SK-N-MC cell lines over-expressing mutant α-synucleins showed an enhanced response to HNE, losing viability faster (p < 0.01 compared with control transfectants from 6 h, Figs 6a and b). By contrast to the effects observed on serum withdrawal and exposure to H2O2, cells transfected with wild-type α-synuclein were more sensitive to HNE (p < 0.01 from 6 h). However, in both cell lines, the mutant α-synucleins were more effective in aggravating the effect of HNE than the wild-type α-synuclein (p < 0.05 between 3 and 12 h).
Cell sensitivity to the proteasome inhibitor, lactacystin
Addition of lactacystin caused loss of viability in both cell lines (Figs 6c and d), and cell death occurred largely by apoptosis, as revealed by morphological examination, DNA laddering (data not shown) and other apoptosis characteristics (Fig. 3). Loss of viability was aggravated in cells expressing both wild-type and either mutant α-synucleins, although the mutant α-synucleins induced greater susceptibility to lactacystin than the wild-type α-synuclein. This was also observed in the MTT assay (p < 0.01 between 3 and 12 h in NT-2 cell line and between 6 and 12 h in SK-N-MC cell line). By contrast, overexpression of β-galactosidase in NT-2 or SK-N-MC cells had no effect on the toxicity of lactacystin (data not shown).
Cell sensitivity to the protein kinase inhibitor, staurosporine
Staurosporine, at a concentration range of 1 nm to 1 µm, caused apoptosis in both NT-2 and SK-N-MC cell lines, and a concentration of 100 nm was chosen as giving approximately 50% loss of viability during the time course of the experiments (Fig. 6e and f). Staurosporine-induced cell death was accelerated in cells overexpressing either wild-type or either of the two mutant α-synucleins. However, the mutant α-synucleins accelerated loss of cell viability more than the wild-type α-synuclein as shown by trypan blue exclusion (p < 0.01 from 2 days in NT-2 cell line and from 3 days in SK-N-MC cell line), MTT reduction (data not shown), PI, Hoechst 33258, TUNEL (Fig. 3) or DNA laddering (data not shown).
Changes in the level of GSH
GSH levels in the α-synuclein transfectants were examined before, and 3 or 24 h after, exposure to four toxic insults, namely, serum deprivation (Fig. 7a), exposure to H2O2 (data not shown), and the addition of lactacystin (Fig. 7b) or staurosporine (data not shown). GSH levels were somewhat lower in either mutant α-synuclein transfectants in normal medium (p < 0.05). Exposure to all insults produced a decrease in GSH levels, and levels were further decreased in both cell lines expressing either mutant α-synuclein from 3 h after exposure (p < 0.01). The effect of transfection with wild-type α-synuclein paralleled its effects on cell viability in response to the insults, that is, overexpression of wild-type α-synuclein attenuated the decrease in GSH levels under serum deprivation (p < 0.05 compared with non-transfectants from 3 h in NT-2 cell line and between 6 and 12 h in SK-N-MC cell line, Fig. 7a) or addition of H2O2 (p < 0.05 compared with non-transfectants at 3 or 12 h in NT-2 cell line, data not shown), but GSH levels were further decreased (but to a lesser extent than those of either mutant α-synuclein transfectants, p < 0.01) after exposure to lactacystin (p < 0.01 compared with non-transfectants from 12 h in NT-2 cell line and from 9 h in SK-N-MC cell line, Fig. 7b) or staurosporine (p < 0.01 compared with non-transfectants from 6 h in both cell lines, data not shown).
Oxidative DNA damage
The level of 8-OHG in DNA was measured as an index of oxidative DNA damage after serum deprivation, or addition of H2O2, lactacystin or staurosporine. Basal levels were not different from controls in any of the α-synuclein transfectants (Fig. 8a, white bars). All damaging conditions elevated the levels of 8-OHG in both cell lines (p < 0.01). The increase in levels of 8-OHG in cells expressing wild-type α-synuclein after serum deprivation or exposure to H2O2 or staurosporine, was broadly comparable to control transfectants, but was aggravated in the presence of lactacystin (p < 0.01). 8-OHG levels were higher in cells transfected with either mutant α-synuclein in the presence of all injurious conditions (p < 0.05 compared with non-transfectants or vector-only in each injurious condition).
Protein carbonyls were measured as a general index of oxidative damage to proteins (Oliver et al. 1990). Protein carbonyl levels were higher when wild-type (p < 0.05) and mutant α-synucleins were overexpressed (p < 0.01, Fig. 8b). Serum withdrawal or treatment with H2O2, lactacystin or staurosporine raised protein carbonyl levels (p < 0.01). This effect was aggravated after exposure of mutant α-synuclein transfectants in both cell lines to any of the injurious conditions (p < 0.01). Higher protein carbonyl values were also observed in both cell lines over-expressing wild-type α-synuclein after exposure to lactacystin, staurosporine (but to a lesser extent than the mutant α-synuclein transfectants) or H2O2, and, only in the SK-N-MC cell line, after serum deprivation (p < 0.05).
The level of lipid peroxidation in the cells was assessed by using a thiobarbituric acid (TBA)-based test linked to HPLC to remove interfering chromogens (Chirico 1994). Baseline levels of lipid peroxidation in the cells were unaffected in any of the NT-2 cell transfectants, but were elevated in the SK-N-MC cells transfected with mutant (but not wild-type) α-synucleins (p < 0.01, Fig. 8c). All toxic conditions raised the levels of lipid peroxidation in both cell lines (p < 0.01). The elevation in lipid peroxidation levels was greater in both cell lines expressing wild-type or mutant α-synuclein under all injurious conditions (p < 0.01).
Reactive nitrogen species and 3-nitrotyrosine
Formation of reactive nitrogen species such as peroxynitrite (ONOO−) has been implicated in apoptotic cell death in several cell lines and also in PD (Beal 1997). Using an ELISA assay, basal levels of protein-bound 3-nitrotyrosine, a biomarker of attack of ONOO− upon proteins (Halliwell 1997), were detectable in both cell lines (Fig. 8d). 3-Nitrotyrosine levels were significantly higher in the mutant (but not in the wild-type) α-synuclein transfectants (p < 0.05 in NT-2 cell line and p < 0.01 in SK-N-MC cell line). Serum withdrawal or addition of lactacystin or H2O2 (but not of staurosporine) in culture medium increased 3-nitrotyrosine levels in control cells (p < 0.01). This increase was larger in NT-2/wild-type α-synuclein transfectants after exposure to lactacystin. It was also larger in the mutant α-synuclein transfectants of both cell lines subjected to any of the injurious agents (p < 0.01).
There is considerable evidence that oxidative stress is associated with the progression of sporadic PD and may be an early stage in its development (Sian et al. 1994; Beal 1997; Jenner and Olanow 1998). For example in substantia nigra from PD patients elevated levels of HNE, protein carbonyls and 8-OHG have been reported (Chirico 1994; Alam et al. 1997a, 1997b; Jenner and Olanow 1998). Defects in mitochondria, especially complex I, have also been reported and may themselves lead to further oxidative stress (Schapira 1999). There is growing evidence that apoptosis contributes to neuronal cell death in substantia nigra in PD (Stefani et al. 1997). Considerable excitement has been generated by reports that some cases of FPD are associated with mutations in α-synuclein (Mezey et al. 1998), but the mechanisms by which this leads to cell death are unknown. It is important to determine to what extent α-synuclein mutations affect the response of cells to injurious agents.
Over-expression of wild-type α-synuclein in two different cell lines did not affect growth (data not shown) or viability, and had little or no effect on most of baseline parameters of oxidative damage (8-OHG and lipid peroxidation) or on the levels of 3-nitrotyrosine, although protein carbonyls were significantly elevated. Protein carbonyls are regarded as a general assay of protein damage, generated by direct free radical attack upon, or the binding of certain aldehydes (including HNE) to, amino acid residues in proteins (Oliver et al. 1990).
We examined the effect of various toxic insults, relevant to PD, upon the viability of α-synuclein transfectants, including H2O2, MPP+ and serum withdrawal. H2O2 is generated in increased amounts by monoamine oxidase activity in parkinsonian substantia nigra (Cohen et al. 1997). MPP+ is a mitochondrial electron transport chain complex I inhibitor that models the complex I defect found in PD (Schapira 1999) and HNE is present in increased amounts in parkinsonian substantia nigra (Yoritaka et al. 1996). Serum deprivation is a widely used model for neurotrophic factor deprivation, and such factors are being widely investigated as neuroprotective agents (Bjorklund et al. 1997). Cells transfected with wild-type α-synuclein were less sensitive to cell death induced by H2O2 or serum withdrawal. Death occurred largely by apoptosis, as revealed by morphology examinations, PI, Hoechst 33258, TUNEL and DNA laddering. However, there was no effect of wild-type α-synuclein transfection on death induced by MPP+.
By contrast, both of the mutant α-synucleins examined were deleterious to both cell lines. They slowed growth rate even under normal culture conditions (data not shown), increased the baseline levels of oxidative damage (8-OHG, lipid peroxidation, protein carbonyls and 3-nitrotyrosine), and rendered cells more susceptible to death induced by serum deprivation, H2O2 treatment, exposure to MPP+ or addition of lactacystin or staurosporine, promoting more rapid loss of viability. In general, injurious conditions raised levels of oxidative damage and 3-nitrotyrosine within cells, and this was aggravated by the mutant α-synucleins (Fig. 8). One especially neurotoxic end product of lipid peroxidation is HNE and its increased formation in PD had been described (Yoritaka et al. 1996). HNE caused a loss of viability in both cell lines, which was aggravated by overexpression of mutant α-synucleins. However, the wild-type α-synuclein transfectants were also more sensitive to HNE, although less so than the mutant α-synuclein transfectants. HNE exerts its toxicity in large part by binding to proteins (Oliver et al. 1990), and it is possible that alteration of the protein composition of cells may aggravate its toxicity. Somewhat similar data were obtained with staurosporine, a protein kinase inhibitor that is widely used to induce apoptosis, accompanied by oxidative stress (Ahlemeyer and Krieglstein 2000). The mutant α-synucleins accelerated loss of viability in cells treated with staurosporine. The wild-type α-synuclein did so as well, but to a lesser degree.
We also examined the effect of lactacystin, a proteasome inhibitor. Proteasomes degrade many cellular proteins, both normal and abnormal, after ubiquitination (Sitte et al. 1998) and there is evidence for impairments of proteasome function in the neurodegenerative diseases (Halliwell and Jenner 1998). Bennett et al. (1999) reported that both wild-type and the A53T mutant α-synuclein are degraded by the proteasome in SH-SY5Y cells, but the mutant is degraded more slowly. Lactacystin addition caused apoptosis, accompanied by increased oxidative damage. A rise in protein carbonyls in lactacystin-treated cells has already been reported (Sitte et al. 1998), consistent with the degradation of oxidatively modified proteins by this mechanism. The mutant α-synucleins again accelerated lactacystin-induced cell death. To a lesser extent, so did wild-type α-synuclein. It may be that overexpression of these proteins, especially the mutants that may be degraded more slowly, cause choking of the proteasome system and disrupt cellular function. However, cells overexpressing β-galactosidase were not more sensitive to lactacystin. It may be, of course, that this protein, which is not normally expressed in NT-2 and SK-N-MC cell lines, is not degraded by the proteasome system. Interestingly, it has been suggested that HNE-modified proteins can also block the proteasome system (Grune et al. 1995), which would be consistent with the results we obtained on addition of this substance.
Our data show that over-expression of wild-type α-synuclein does not in general raise basal levels of oxidative damage or 3-nitrotyrosine in the two cell lines examined. Indeed, it was protective against cell death induced by serum withdrawal and H2O2 treatment, although not against MPP+. However, it accelerated loss of viability caused by HNE, staurosporine or lactacystin (the latter two of which are, of course, non-physiological insults). In all cases, the cell death occurred by apoptosis as evidenced by morphological examination, DNA laddering and other features of apoptosis. It seems that function of wild-type α-synuclein is cell line- and insult-specific, since many previous studies suggest both protective and non-protective properties (Da Costa et al. 2000; Masliah et al. 2000; Zhou et al. 2000).
By contrast, either of the mutant α-synucleins examined accelerated cell death induced by any of the insults examined. They also raised baseline levels of some parameters of oxidative stress. α-Synuclein is one of the abundant proteins in cells (Maroteaux et al. 1988). Data suggest that its expression in transfected cells was 3.5-fold higher than that of untransfected cells. It may result in accumulation and aggregation of normal and abnormal proteins in cytoplasm, as shown in previous studies (El-Agnaf and Irvine 2000; Mattila et al. 2000). These can impair the protein clearing system, and induce oxidative stress, reflected by decreased levels of GSH and increased levels of 8-OHG, protein carbonyls, lipid peroxidation and 3-nitrotyrosine.
Our data argue strongly that the presence of mutant α-synucleins in PD may cause enhanced sensitivity to any type of insult, leading to apoptosis associated with increased oxidative damage. Of course, whether oxidative damage is a consequence of apoptosis or a cause of it remains to be elucidated, and may differ with different insults. For example, cytochrome c released from mitochondria as a result of apoptosis may interact with α-synucleins (Hashimoto et al. 1999b). Consistent with our view, immortalized dopaminergic neurons expressing mutated α-synuclein are hypersensitive to damage by 6-hydroxydopamine (Kholodilov et al. 1999) and other insults (Saha et al. 2000), and this has been observed in studies with primary culture and transgenic mice (Da Costa et al. 2000; Ostrerova-Golts et al. 2000; Van der Putten et al. 2000; Zhou et al. 2000).
We wish to thank Professor M. Goedert (MRC Molecular Biology Laboratory, Cambridge, UK) and Professor A. Schapira (Royal Free Hospital, School of Medicine, London, UK) for providing us with the α-synuclein gene and the cell lines.