Huntington’s disease (HD) is caused by an expanded CAG repeat in exon 1 of the gene coding for the huntingtin protein. The cellular pathway by which this mutation induces HD remains unknown, although alterations in protein degradation are involved. To study intrinsic cellular mechanisms linked to the mutation, we examined dissociated postnatally derived cultures of striatal neurons from transgenic mice expressing exon 1 of the human HD gene carrying a CAG repeat expansion. While there was no difference in cell death between wild-type and mutant littermate-derived cultures, the mutant striatal neurons exhibited elevated cell death following a single exposure to a neurotoxic concentration of dopamine. The mutant neurons exposed to dopamine also exhibited lysosome-associated responses including induction of autophagic granules and electron-dense lysosomes. The autophagic/lysosomal compartments co-localized with high levels of oxygen radicals in living neurons, and ubiquitin. The results suggest that the combination of mutant huntingtin and a source of oxyradical stress (provided in this case by dopamine) induces autophagy and may underlie the selective cell death characteristic of HD.
Huntington’s disease (HD) is a progressive autosomal dominant disorder caused by an expanded CAG repeat in exon 1 of the huntingtin gene. Although huntingtin is expressed throughout the body (1,2), HD pathology is marked by extensive loss of striatal neurons that receive dopaminergic input and express DARPP-32 (dopamine- and cyclic AMP-regulated phosphoprotein of a molecular weight of 32 kDa), as well as neurons in deep layers of neocortex (3,4).
Several transgenic mouse lines have been produced which express the HD mutation (5–9). The most extensively characterized are the Bates R6 lines (5), which express exon 1 with an expanded CAG trinucleotide repeat. These lines exhibit ubiquitinated nuclear and cytoplasmic inclusion bodies (10,11), dark cell degeneration (12,13), altered neurotransmitter receptor levels (14,15), decreased expression of striatal signaling genes (16) including deficiencies in dopamine signaling (17), decreased striatal and total brain size (5,18) and progressive motor (19) and cognitive deficits (20,21).
It is difficult to explain why striatal projection neurons are primarily affected in HD unless there is an additional factor which promotes death in this particular neuronal population. An obvious candidate is the dopaminergic input from ventral midbrain neurons, as the striatum receives the densest dopaminergic innervation in the brain, and this neurotransmitter is well established to be toxic to striatal neurons in vivo and in vitro (22–27). Dopamine oxidation produces a quinone which reacts covalently with cysteinyl residues on proteins, as well as a variety of other reactive oxidized species (reviewed in 28,29).
While there are numerous suggestions in the literature that dopamine plays a role in HD pathogenesis (reviewed in 30), to our knowledge there has been no direct evidence linking dopamine toxicity and the HD mutation to neurodegeneration. Here, we introduce a postnatal striatal cell culture system with a high yield of DARPP-32-expressing neurons. Consistent with most in vivo reports, we found no difference in neuronal survival between cells derived from R6/2 animals compared to wild-type littermates. However, R6/2 neurons were far more susceptible to dopamine-induced stress, leading to a cascade of cellular events involving production of oxyradicals and induction of neuronal autophagy. The results suggest that dopamine may participate in steps that lead to the selective pathology and cell death of striatal neurons that express the HD mutation.
Expanded CAG repeats do not elevate the rate of death of striatal neurons
Neurodegeneration in HD striatum is specific for DARPP-32-expressing neurons. Previous work has shown that striatal cultures derived from embryonic lateral ganglionic eminence display a very low yield of DARPP-32 positive neurons (<5%) (31–33). However, native huntingtin, which shares the same promoter as the R6/2 gene, increases its level of protein expression in neurons from postnatal day (P) 7–15 (34).
In order to examine the striatal neurons at a time point corresponding to higher huntingtin expression in vivo, and possibly to higher DARPP-32 expression, we established a novel system of postnatally derived striatal cultures from P1–3-day-old animals. In these cultures, we obtained a far higher fraction of DARPP-32 positive neurons in both R6/2 HD mutant and wild-type cultures than in embryonically derived cultures, starting with 20% at 2 days, and increasing to 80% after 1 week (Figs 1 and 2).
Initially, we compared the survival of neurons derived from R6/2 mice and their wild-type littermates by counting the total number of neurons under differential interference contrast (DIC) optics, and immunolabeled DARPP-32 positive neurons at 2, 8, 16 and 32 days in culture (Fig. 2). We observed a gradual loss of total neurons over time (ANOVA, F = 16.33, P < 0.0001) but no difference in the rate of cell death between mutant and wild-type neurons (F = 0.28, not significant). Moreover, genotype had no influence on the survival or expression of DARPP-32 neurons as a whole (F = 0.88, not significant), or in relation to its expression over time in culture (genotype versus time, F = 0.19, not significant). These findings were confirmed in additional observations up to 64 days after plating (data not shown).
Increased susceptibility to dopamine-mediated toxicity in R6/2 striatal neurons
To examine whether genotype affected the response to dopamine-induced neurodegeneration, we exposed cultures derived from R6/2 and wild-type littermates at 8, 16 and 21 days post-plating to dopamine (1 mM) for 24 h. This level of dopamine exposure was previously used for promoting neurodegeneration in cortical and dopaminergic cultures (35), and was 3-fold higher in concentration than that used in studies in embryonic striatal cultures with few astrocytes (26,27). In pilot experiments, we also used lower concentrations of dopamine (0.1 and 0.5 mM), but they did not produce significant toxicity (data not shown).
Cell loss was assessed by counting the total number of neurons 24 h after addition of dopamine (Fig. 3). At all timepoints examined, there were no differences between the survival of R6/2 and wild-type neurons in untreated cultures. However, dopamine decreased the number of surviving neurons to 68 ± 10% of untreated levels in wild-type cultures, and to 38 ± 5% of untreated levels in R6/2 cultures. The R6/2 genotype was thus more susceptible to dopamine neurotoxicity (Mann–Whitney U-test, P < 0.05).
In addition to a decrease in survival of mutant neurons exposed to dopamine, we suspected that there may be differences in the response of those neurons that survived. As ubiquitinated inclusion bodies occur in striatal neurons in HD, we first compared the distribution of ubiquitin using an antibody directed against ubiquitin. Untreated cultures from R6/2 and their wild-type littermates displayed no ubiquitin-labeled puncta (Fig. 4). Dopamine (1 mM for 24 h) induced the formation of ubiquitinated puncta in the cytoplasm, but not in the nucleus, in both wild-type and mutant neurons. However, R6/2 cultures exposed to dopamine expressed ubiquitinated puncta in nearly all neurons, and were far more highly labeled than dopamine-treated wild-type cultures.
To compare this response, we counted the number of ubiquitinated puncta in the neurons. A significantly higher fraction of R6/2 neurons exposed to dopamine expressed ubiquitinated puncta than did dopamine-treated wild-type neurons (Fig. 5; at 8 days post-plating, P = 0.0002, Fisher’s exact test). Similar responses were found in experiments performed at 16 and 21 days post-plating (data not shown).
We examined whether the puncta were also immunoreactive for huntingtin using the EM48 antibody, which is specific for the N-terminal region of the mutant protein. EM48-labeled puncta were not observed in the cytoplasm under any condition. However, occasional neurons in dopamine-exposed R6/2 cultures developed huntingtin-immunoreactive inclusion bodies in the nucleus (Fig. 6A). Therefore, transport of the N-terminal mutant huntingtin epitope to the nucleus sometimes occurred following dopamine exposure. This occurred in <1% of the neurons, and was not observed in untreated R6/2 cultures or in wild-type neurons under any condition. This impression was confirmed by western blot using the EM48 antibody (Fig. 6B). A single band corresponding to the same molecular weight as that reported in mutant HD-transfected HEK293 cells (36) was observed in dopamine-exposed R6/2 cultures, but was not apparent in untreated R6/2 cultures. As expected, the mutant epitope was not detected by western blot in either untreated or dopamine-exposed wild-type cultures.
In order to investigate whether dopamine exerted its toxic effect via a dopamine receptor-mediated pathway, we exposed cultures to the D1-receptor antagonist, SCH23390 (10 µM), and the D2-receptor antagonist, sulpiride (10 µM), beginning 30 min prior to the addition of dopamine. We observed no effect of these receptor antagonists on dopamine-mediated toxicity (one-way ANOVA, F = 20.45, P < 0.0001 for dopamine-induced loss in cell number, no significant differences between the cultures exposed to dopamine with or without antagonists, Tukey–Kramer post-hoc test).
As the neurotoxic effects of dopamine did not appear to be due to D1- or D2-receptor-mediated responses, we speculated that dopamine toxicity was mediated via its oxidation and production of free radicals, leading to oxidative stress. We monitored oxyradical stress induced by dopamine in living neurons using the fluorogenic oxyradical label, DCF (2,7-dichlorofluorescein diacetate; Fig. 7). Dopamine induced DCF-labeled puncta in the extranuclear cell body and neurites. While dopamine produced DCF labeling in both genotypes, as with ubiquitin, the mutant cultures appeared to be far more sensitive to dopamine and displayed far more label. This impression was confirmed by rating the number of DCF-labeled sites in the fluorescent images of the neurons (Fig. 8), which showed that dopamine-treated R6/2 neurons displayed substantially more DCF puncta than the wild-type neurons (dopamine treatment versus genotype interaction, F = 13.59, P < 0.0005).
To examine whether the DCF puncta co-localized with the ubiquitinated puncta, we photographed dopamine-treated neurons under fluorescent optics, fixed the cells, performed ubiquitin immunocytochemistry, and then re-photographed the neurons. There was some unavoidable alteration of structure during the fixation and subsequent processing (detergent permeabilization), as the puncta often appeared to have shifted from their position in the living preparation. Nevertheless, regions of the neurons that displayed DCF label and the ubiquitinated puncta often appeared to correspond, as can be seen by overlaying arrows indicating DCF puncta onto images processed for ubiquitin immunogenicity (Fig. 9).
To examine the subcellular identity of the DCF-labeled puncta further in a more stable preparation, we co-localized the puncta in living cells with additional fluorescent intracellular markers. No DCF-labeled puncta overlapped with the mitochondrial marker Mitotracker Red (none of 53 DCF-labeled puncta in 23 neurons examined overlapped with mitochondrial label; Fig. 0A). We observed negligible overlap of DCF-labeled puncta with orange/red fluorescently tagged brefeldin A, a marker for endoplasmic reticulum (1 of 81 DCF-labeled puncta in 19 neurons examined overlapped with brefeldin A; Fig. 0B).
To examine autophagic granules, we used monodansylcadaverine (MDC), a highly lipophilic weak base vital dye thought to be specific for these organelles (37,38). We found that MDC distribution was not different in untreated wild-type or R6/2 cultures, nor in dopamine-treated wild-type cultures. However, in R6/2 cultures exposed to dopamine, the average pixel intensity of the MDC fluorescence per neuron was increased by >100% (8–10 cells measured per experimental group, F = 5.9272, P < 0.005), and the number of MDC-labeled puncta per neuron was increased to 160–220% of the level in the other groups (puncta within 8–10 cells measured per experimental group, F = 12.933, P < 0.0001).
MDC label in dopamine-treated R6/2 neurons was present in intracellular puncta as large as 2 µm in diameter (Fig. 0C). DCF and MDC labels were generally exclusive, although we observed a single clear case of overlap in a smaller structure (1 of 56 DCF-labeled puncta in six neurons assayed for colocalization), and five additional cases of partial overlap. As the depth of field under these conditions is ∼0.25 µm (see Materials and Methods), it may be that substructures within autophagic granules are labeled, that DCF and MDC-labeled organelles are occasionally associated (e.g. during fusion of autophagic granules and lysosomes) or that they are separate but occasionally exist in very close proximity.
We obtained the clearest instances of co-localization using Lysotracker Red, a weak base vital dye used to label lysosomes (Fig. 0D). We noted 9 of 34 DCF-labeled puncta in 10 neurons examined that overlapped with the lysosomal marker. The DCF/lysotracker double-labeled puncta were occasionally as large as 2 µm in diameter, consistent with the size of large acidic autophagic granules in the stage prior to fusion with mature lysosomes (see electron micrograph below). As autophagic granules are thought to acquire acidic pH levels after their initial formation, they would not be expected to be labeled by lysosomal indicators at early stages of formation (see Discussion). These results indicate that at least some, if not all, of the puncta displaying high levels of oxyradicals are membrane-delimited autophagic/lysosomal organelles.
To identify the ultrastructural changes induced by dopamine neurotoxicity, we processed striatal cultures for electron microscopy (Fig. 1). Neurons in untreated R6/2 cultures, untreated wild-type cultures and wild-type cultures exposed to dopamine exhibited normal morphology, with normal lysosomes and no apparent cytoplasmic or nuclear abnormalities. In contrast, the R6/2 cultures exposed to dopamine (1 mM for 24 h) displayed distinctive lysosomes that often contained a lumenal matrix with numerous small electron-dense, vesicle-like bodies of <50 nm in diameter, suggestive of autophagy (Fig. 7D and F). These dopamine-exposed R6/2 neurons also displayed classic double membrane autophagic granules (Fig. 1D and G). The size of the autophagic granules was similar to that of the DCF-, MDC- and lysotracker-labeled puncta. Consistent with a relatively low presence of MDC-labeled structures in wild-type cultures and untreated R6/2 cultures, we observed classic autophagic granules at the electron microscope level only in dopamine-treated R6/2 neurons.
In some vacuolated dopamine-exposed neurons apparently en route to cell death, the electron-dense autophagic/lysosomal structures had apparently lost their surrounding membranes, and appeared as ‘free’ aggregates in the cytoplasm (Fig. 1H). In a number of dopamine-exposed neurons, we observed invaginations of the nuclear membrane, but did not locate any apparent nuclear inclusions under electron microscopy.
In HD, the mutant huntingtin protein is expressed in neuronal and non-neuronal cells throughout the body (1,2), but cell death is relatively specific for DARPP-32-expressing medium spiny neurons of the striatum and some neocortical neurons (4). In contrast to studies in transfected primary neuronal cultures or immortalized cell lines expressing the mutation (39–43), we observed no difference between the survival of untreated wild-type and HD transgenic striatal neurons. However, following neurotoxic dopamine exposure, we found that R6/2 striatal neurons exhibited increased cell death and displayed a variety of structural and functional changes associated with stimulation of autophagy that could represent early steps of a cell death pathway.
Firstly, there was an induction of discrete cytoplasmic puncta which contained high levels of oxyradicals. Dopamine induced oxidative stress in more R6/2 neurons than wild-type neurons, and the R6/2 cultures had a significantly higher number of DCF-labeled free radical-containing puncta per neuron, demonstrating increased dopamine-triggered oxidative stress.
Secondly, we observed that DCF puncta were often colocalized with vital autophagic/lysosomal labels, but not with vital markers for endoplasmic reticulum or mitochondria. Following fixation, we found that the DCF-labeled puncta contained high levels of ubiquitin, which is known to mark oxyradical-damaged proteins for degradation.
Thirdly, electron microscopy confirmed that R6/2 neurons exposed to dopamine displayed classic autophagic granules, as well as a distinctive lysosomal morphology with an electron-dense lumenal matrix indicative of stimulation of autophagic/lysosomal degradative pathways.
The results suggest that in transgenic HD striatal neurons, dopamine triggered free radical-mediated oxidation of macromolecules and stimulated autophagy. Punctate DCF-stained objects have previously been noted in a cell culture model of excitotoxicity (44) and in autophagic granules/lysosomes following neurotoxic methamphetamine regimens (45), and are thought to be due to high levels of oxyradicals (including peroxynitrate) within membrane-bound organelles. In these dopamine-treated mutant neurons, the DCF-labeled puncta are consistent with autophagic granules/lysosomes.
Autophagy involves the bulk sequestration and proteolytic degradation of intracellular components in double-membraned organelles that eventually deliver their contents to lysosomes via membrane fusion. Autophagic granules are suggested to form via enfolding of endoplasmic reticulum (46), and to later develop an acidic interior pH preceding or during lysosomal fusion. Autophagy is promoted by pathological conditions including fasting, metabolic inhibition, hypoxia and ischemia (47,48), and is activated in sympathetic neurons exposed to apoptotic stimuli (38). While there is little literature on autophagic granules in the central nervous system, there is evidence that autophagic granules can be formed in axons, where the huntingtin protein is thought to be expressed normally (49,50), and then transported to the cell body (51), the site where we observed most of these structures.
The increased presence of autophagic/lysosomal structures we observed, as well as increased autophagy recently reported in a mutant striatal-derived cell line by DiFiglia and colleagues (52), is strikingly similar to findings in humans with HD. Brains of HD patients show an increased presence of endosomal/lysosomal organelles, multivesicular bodies (organelles associated with autophagy), and accumulation of lipofuscin as compared to normal brains (50,53,54). Lipofuscin granules are now known to be autophagic granules with double membranes which are arrested in normal lysosomal fusion/conversion (55,56). It was shown recently that biosynthesis of neuromelanin in the substantia nigra is due to sequestration of cytosolic dopamine–quinone into autophagic granules (57). Such dopamine–quinone may provide the electron-dense appearance of neuromelanin within the lysosomes of dopamine-exposed R6/2 cultures. The lysosomal structures in R6/2 neurons usually maintain membranes and are thus not themselves inclusion bodies. In some cases, however, the autophagic structures appeared to lose membrane integrity, and thus could contribute to inclusion bodies. Alternatively, inclusion bodies could be produced if autophagy is not sufficient to provide a sufficient level of degradation of the cellular material. Furthermore, the finding that stimulation of ubiquitin and autophagic/lysosomal pathways occur in parallel may indicate a role for ubiquitination in the autophagic process.
The EM48 antibody demonstrated rare nuclear inclusions and only in mutant cultures exposed to dopamine. These appear morphologically equivalent to intranuclear inclusions in HD (10,58), and confirm that the mutant huntingtin is produced in cultured neurons and can be translocated to the nucleus. Moreover, the EM48 antibody indicated the presence of mutant HD protein only in dopamine-exposed R6/2 cultures. It is possible that the neurons in culture are normally able to digest the N-terminal huntingtin fragment before sufficient quantities build up to translocate to the nucleus and produce many nuclear inclusions.
An important unanswered question is whether the mutant huntingtin fragment is taken up into autophagic granules and lysosomes. Work by DiFiglia and co-workers (52) demonstrated uptake of huntingtin [using Ab1 (59)] in lysosomes in a clonal striatal cell line overexpressing huntingtin with different lengths of the polyglutamine repeat. In contrast to our results in striatal neurons, the activation of the lysosomal system in the cell line occurred in the absence of an exogenous oxidative insult, and was more marked in cells expressing huntingtin with 46–100 CAG repeats than in cells expressing huntingtin with 18 CAG repeats.
The striatum, the most affected region in the HD brains, has the highest levels of dopamine in the brain. Dopamine levels are highest in the dorsal striatum and decline ventrally (60) showing the same gradient as the neuropathology in HD. Other regions in the brains affected in HD, such as cingulate cortex and hippocampus, also receive dopaminergic terminals (61). Interestingly, deficiencies in dopamine receptors and signaling have been reported in R6/2 mice (14,17). In this study, we show that dopamine-mediated toxicity is not receptor dependent, but rather acts via induction of oxidative stress. It is possible that a lack of dopamine receptor-related signaling in striatal neurons disrupts signaling from the striatum to the substantia nigra pars compacta. This could alter the dynamics of dopamine release in the striatum and change local concentrations of dopamine in HD striatum, possibly contributing to oxidative stress.
In conclusion, this work indicates that expression of the HD mutation increases dopamine-induced damage in striatal neurons, including enhanced oxyradical stress, altered protein degradation, and autophagic/lysosomal abnormalities. As is known to occur following toxin exposure, the stimulation of autophagy in mutant neurons exposed to dopamine may be an attempt by the neurons to protect themselves from damaged cell constituents. Dopamine-induced autophagy may explain the findings of reduced striatal cell size and volume preceding reduction in cell number in mice from R6/2 lines and other transgenic mice (5,18,62–64). Thus, autophagic shrinkage of individual neurons may, in addition to neuronal death, contribute to the severe reduction in striatal volume in humans with HD. More importantly, this in vitro model suggests that increased susceptibility to dopamine toxicity by neurons expressing the HD mutation could act as an upstream trigger for neuronal damage in HD.
MATERIALS AND METHODS
The similar technique of postnatal mouse primary neuronal culture of ventral midbrain has been described in earlier publications (65). Animal protocols were approved by the Institutional Animal Care and Use Committee of Columbia University. Unless noted, reagents are from Sigma.
We used rat (Charles River Laboratories) pups from postnatal days 1–3 for the preparation of astrocyte monolayers. To obtain neurons, we used mouse pups at postnatal days 1–3 resulting from matings between R6/2 males and C57Bl/6 females (Jackson Laboratories). The tissue derived from each mouse pup was processed separately, and a tissue sample from the tail of each animal was immediately genotyped for the transgene (5). Twenty-four hours prior to the plating of neurons, the glial medium was removed. The cultures were then washed twice with serum-free neuronal medium, and serum-free neuronal medium was added once more for long-term culture. Serum-free neuronal medium contained 47% MEM, 40% Dulbecco’s modified Eagle’s medium, 10% Ham’s F-12 nutrient medium, 3.3 mg/ml glucose, 0.25% albumin, 500 µM glutamine, 100 µg/ml transferrin, 15 µM putrescine, 30 nM triiodothyramine, 25 µg/ml insulin, 200 nM progesterone, 115 nM corticosterone, 5 µg/ml superoxide dismutase, 432 U/ml catalase and 500 µM kynurenate.
The striatal tissue was defined by the corpus callosum dorsally and laterally, a vector between the tips of the corpus callosum and lateral ventricle, and the lateral ventricle medially. These neurons thus represent the caudate/putamen and a small portion of the dorsal nucleus accumbens. The striata were divided into 1 mm3 segments and incubated in papain (20 U ml) with kynurenate (500 µM) at 32°C under continuous oxygenation with gentle agitation for 2 h. The tissue segments were rinsed three times in serum-free medium, dissociated by gentle trituration, resuspended in serum-free medium, and plated onto 1 cm2 glass coverslips at a density of 80 000 cells per dish which had been covered with Sprague–Dawley rat-derived cortical astrocytic monolayer established 2 weeks earlier (66,67). Therefore, all astrocyte preparations were identical genetically and the effects reported were due to differences between neuronal genotypes. All cultures were maintained in a total volume of 2 ml, at 37°C in 5% CO2, and were not fed. Neuronal cultures were always maintained in serum-free media and all experiments were conducted in serum-free media. Neurotoxic dopamine exposures were typically 1 mM dopamine for 24 h. Analysis of the cultures was always made directly after the 24 h exposure. SCH23390 (10 µM) and sulpiride (10 µM) were added to some cultures 30 min prior to addition of dopamine.
Cultures were fixed in 4% paraformaldehyde in PBS for 15 min and rinsed three times in PBS. For DARPP-32 immunocytochemistry, cultures were quenched in 3% H2O2 in PBS for 7 min and rinsed three times in PBS followed by a preincubation in 5% normal horse serum (NHS) in PBS with 0.3% Triton X-100 (TPBS). The cultures were incubated overnight at 4°C with a DARPP-32 primary mouse monoclonal antibody (1:20 000; kindly donated by Drs P.Greengard and H.Hemmings, Rockefeller University, New York, NY). The cultures were then incubated with biotinylated horse anti-mouse secondary antibody (Chemicon) for 1 h in TPBS at 1:200, and DARPP-32 immunoreactivity was visualized using an avidin–biotin–peroxidase complex system (Vectastain ABC Elite kit, Vector Laboratories) with 3,3′-diaminobenzidine as a chromogen. Counts of the total number and DARPP-32-positive neurons were performed using Nomarski DIC as described previously (68). Neurons were assessed in 25% of the total culture area.
For ubiquitin immunocytochemistry, a polyclonal rabbit antibody against ubiquitin (1:1000; Dako), and a biotinylated goat anti-rabbit secondary antibody (1:200; Chemicon) were used. The number of ubiquitin-stained puncta were assessed in 50 cells per dish (two to four dishes per condition derived from the same litter, repeated in three independent experiments with similar quantitative results) by a second observer blind to the different experiment conditions.
Cultures from R6/2 mice and wild-type littermates were processed for huntingtin using the EM48 antibody (generously donated by Dr X.-L. Li, Emory University, Atlanta, GA) at a dilution of 1:1000. EM48 is a rabbit polyclonal antibody generated from a GST fusion protein containing the first 256 amino acids of human huntingtin with a deletion of the polyglutamine and polyproline stretches. This antibody is effective for detection of N-terminal huntingtin fragments with normal or expanded polyglutamine repeats, and is particularly selective in human brain for huntingtin aggregates, having only a weak affinity for rodent huntingtin (11,69). We also used the EM48 antibody to detect mutant huntingtin in HD wild-type cultures with or without dopamine according to the protocol for western blot published previously (36). Forty nanograms of protein were loaded per lane, run on a 5% SDS–PAGE gel and transferred to PVDF membrane.
Fluorescent images were acquired using a Zeiss IM35 microscope equipped with a 40× 1.3 numerical aperture (NA) oil immersion objective, immersion oil with refractive index (ri) of 1.518, timed exposures using a shutter under computer control, and a Photometrics Star 1 CCD camera (Roper Scientific). For the examination of compounds with fluorescein emission of λ ≈ 0.52 µm, the depth of field can be estimated as [λ(ri2 – NA2)0.5]/NA2 = 0.24 µm.
Intracellular oxygen radicals were assessed using DCF (Molecular Probes) in two separate sets of experiments. Aliquots of 10 mM DCF were prepared in dimethyl sulfoxide and stored at –85°C. Immediately before use, an aliquot was diluted in physiological saline at a final DCF concentration of 1 µM. Neuronal cultures were incubated for 15 min at 37°C, rinsed twice with oxygenated physiological saline, and monitored as described previously using a Zeiss fluorescein filter set (45). To avoid artifactual amplification of the fluorescence (70), the neurons were first brought into focus under DIC optics and fluorescent images were subsequently acquired during a single 200 ms exposure with a 90% ND filter. Therefore, there was no prior fluorescent excitation. The number of fluorescent puncta per neuron was assessed using NIH image software 1.61 (Wayne Rasband, National Institutes of Health; http://rsb.info.nih.gov/nih-image). Typically, 10–20 neurons per dish were assessed, with two dishes studied from each condition and genotype.
To study co-localization of the DCF stained profiles with organelles, we used Lysotracker Red DND-99 (100 nM, Molecular Probes) for lysosomes and associated acidic organelles, Mitotracker Red CMX ros (10 nM, Molecular Probes) for mitochondria, red-orange-fluorescent BODIPY-brefeldin A (10 nM, Molecular Probes) for endoplasmic reticulum, and MDC (10 nM, Molecular Probes) for autophagic granules.
To assist plastic embedding, some striatal cultures derived from R6/2 and wild-type littermate mice were cultured on astrocyte monolayers on Aclar (Honeywell Inc.) rather than glass coverslips. Following a 24 h exposure to dopamine (1 mM) at 21 days post-plating, the cultures were fixed for 1 h in ice-cold 4% paraformaldehyde and 0.05% glutaraldehyde in 2 mM CaCl2/100 mM sodium cacodylate pH 7.4. The cultures were maintained in 100 mM sodium cacodylate pH 7.4 until processing for electron microscopy as reported previously (71).
Results were compared using Mann–Whitney U-test, χ2 test, Fisher’s test, one- or two-way analysis of variance (ANOVA), followed by Tukey–Kramer Multiple comparison post-hoc test or Student’s t-test.
We are grateful to Drs Paul Greengard and Hugh Hemmings for the donation of the DARPP-32 antibody, and to Dr Xiao-Jiang Li for the EM48 antibody. We thank Drs Marian DiFiglia, Ana-Marie Cuervo and J.Fred Dice for discussion of the manuscript. We thank Mary Schoenebeck for electron microscopy, and Ping Chen for technical assistance. This work was supported by the Huntington’s Disease Society of America, the Parkinson’s Disease Foundation, a Udall Center Award from NINDS, the Hereditary Disease Foundation, the Swedish Association of the Neurologically Disabled, Swedish Society for Medical Research, and Anders Wall Foundation.
To whom correspondence should be addressed. Tel: +1 212 305 3967; Fax: +1 212 305 5450; Email: firstname.lastname@example.org