Cell culture and plasmid constructs

The in vitro HD model used for this study was a PC12 cell line that expresses N-terminal huntingtin protein in response to induction with the insect steroid hormone tebufenozide (TFZ) (Aiken et al., 2004). The huntingtin construct had been cloned into the TFZ-inducible plasmid pBWN (Suhr et al., 1998) and encodes the first 67 amino acids with a polyglutamine tract of 25 (wild-type) or 97 (mutant) repeats and a C-terminal enhanced green fluorescent protein (EGFP) tag (Kazantsev et al., 1999). This cell line was kindly gifted to us by Dr Eric Schweitzer (Brain Research Institute, UCLA, Los Angeles, CA, USA). Clonal populations of each of these lines (PC12 25Q/97Q) were stably transfected with HA-tagged human CB₁ (PC12 25Q CB₁/97Q CB₁) or dopamine receptor 1 (PC12 25Q D1/97Q D1) constructs, under constitutive promotion in the plasmid pEF4a His V5A (Invitrogen, Carlsbad, CA, USA) [Receptor nomenclature conforms to Alexander et al. (2008)]. Receptor constructs were generated by KpnI/PmeI (CB₁) or KpnI/XbaI (D1) restriction digest of HA-tagged receptor sequences from pcDNA 3.1(+) constructs purchased from the Missouri S & T cDNA Resource Centre (Rolla, MO, USA) (catalogue numbers: #CNR01LTN00 and #DRD010TN00), and ligation into KpnI/PmeI- or KpnI/XbaI-digested pEF4a His V5A.

Cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM, Invitrogen) containing high glucose (4500 mg·L⁻¹), L-glutamine (4 mM) and sodium pyruvate (110 mg·L⁻¹) and supplemented with 10% (v/v) horse serum, 5% (v/v) fetal bovine serum, 100 units·mL⁻¹ penicillin, 100 µg·mL⁻¹ streptomycin, 25 mM HEPES and 250 µg·mL⁻¹ geneticin. Also, 500 µg·mL⁻¹zeocin was added for receptor-transfected lines. Cells were grown at 37°C in a humidified 95% air, 5% CO₂ environment.

Reverse transcriptase PCR

Endogenous GPCR expression in PC12 cells compared with CBB6 (CBAxC57/B6) mouse cerebellum (kindly gifted by Associate Professor Anthony Hannan, Howard Florey Institute, Melbourne, Vic., Australia) was examined using reverse transcriptase PCR. One million cells or <100 mg fresh frozen mouse cerebellum (n= 3 animals) were homogenized in 1 mL TRIzol reagent (Invitrogen) on ice and total RNA extracted as per manufacturer’s directions. Two micrograms of RNA was DNase-treated and cDNA was generated using Superscript II reverse transcriptase (Invitrogen) and according to manufacturer’s instructions. Receptor PCR was performed using 1 µL cDNA generated as above, 1 µM forward and reverse primers, 160 µM dNTPs and 0.5 U Taq DNA polymerase (buffered, Roche, Basel, Switzerland) in a GeneAmp PCR system 9700 (Applied Biosystems Inc., Foster City, CA, USA). Cycling conditions were as follows: 94°C, 2 min; (94°C, 15 s; annealing temp, 30 s; 72°C, extend time) × 35 (CB₁, D2, β-actin) or 40 (D1) cycles; 72°C, 7 min.

Where annealing temp: CB₁= 58°C; D1 = 63°C; D2 = 60°C; β-actin = 60°C and extend time: CB₁= 90 s; D1 = 70 s; D2 = 30 s; β-actin = 30 s. Forward and reverse primer sequences were: CB₁: atgaagtcgatcctagatggccttgcaga, tcacagagcctcggcagacgtg (previously validated to amplify endogenous mouse CB₁ (Grahamet al., 2006), rat CB1, and human CB1 in brain samples and transfected cell lines (data not shown); D1: ggtccaaggtgaccaacttct, accgtctctatggcattattcgt; D2: ctggagaggcagaactggag, ggagatggtgaaggacagga; B-actin: gctcgtcgtcgacaacggctc, caaacatgatctgggtcatcttctc. GPCR primers were designed against regions of high identity between rat and human sequences in order to detect both endogenous and transfected receptor simultaneously. Thirty-five to forty cycles of PCR were used to aid detection of endogenous receptors even if expressed at low levels.

[³H]-SR141716A binding assay

To investigate endogenous CB₁ receptor expression and measure transfected CB₁ receptor expression levels, whole cell binding was performed as described previously (Saidak et al., 2006) but with modifications. Briefly, PC12 cells were plated at 150 000 cells·well⁻¹ in poly-L-lysine-coated 24-well plates and allowed to adhere overnight. Cells were washed in assay buffer [DMEM, 25 mM HEPES pH 7.4, 5 mg·mL⁻¹ bovine serum albumin (BSA)] and incubated at 37°C for 1 h with 5 nM [³H]-SR141716A (two batches used with specific activity of 55 µCi·mol⁻¹ and 50 µCi·mol⁻¹, respectively, Amersham, Chalfont St. Giles, UK). Non-specific binding was defined by inclusion of 100 nM SR141716A in the incubation. Cells were then placed on ice, washed twice with 1 mL assay buffer and lysed in 250 µL 0.1 M NaOH for 1 h at room temperature. A total of 200 µL·well⁻¹ lysate was taken into 4 mL scintillation vials (Perkin-Elmer, Waltham, MA, USA) and 2 mL scintillant (Starscint, Packard Bioscience, Meriden, CT, USA) added. Vials were shaken, allowed to disperse overnight, and counted for 2 min·well⁻¹ on a Wallac 1450 Microbeta Jet Trilux Scintillation Counter (Perkin-Elmer). Total protein per well was determined by Bio-Rad DC Protein Assay (Bio-Rad Laboratories, Hercules, CA, USA) of 5 µL of the lysate. Values are expressed as femtomoles of radioligand bound per milligram of cells.

Imaging receptor localization

In order to confirm that transfected CB₁ and D1 receptors were appropriately localized to the cell surface, PC12 cell lines were labelled using a live-cell method described previously (Grimsey et al., 2008). Briefly, for cell surface receptor labelling, culture medium was replaced with serum-free media with 5 mg·mL⁻¹BSA (SFM/BSA) with mouse α-HA.11 antibody (Covance, Princeton, NJ, USA) at 1:500 dilution for 30 min at room temperature with rocking. Unbound antibody was removed by 1× wash in fresh SFM/BSA. Cells were then fixed in 4% paraformaldehyde for 10 min and washed in phosphate-buffered saline (PBS) for 2 × 10 min. Secondary antibody labelling was performed post fixation using Alexa 594 goat α-mouse antibody (Molecular Probes, Eugene, OR, USA) at 1:400 in immunobuffer with Triton® x-100 (IB-T, PBS containing 1% v/v goat serum and 0.2% v/v Triton® x-100) at 4°C overnight then rinsed for 2 × 10 min with PBS containing Triton® x-100 (PBS-T, PBS with 0.2% v/v Triton® x-100). Cell nuclei were stained with Hoechst 33258 for 10 min, then rinsed again in PBS.

Images of cells were acquired on a Discovery-1™ automated fluorescence microscope (Molecular Devices, Sunnyvale, CA, USA), fitted with a monochrome peltier cooled 12-bit Hamamatsu CCD camera, using a 40× plan fluorite objective lens (air, NA 0.6) at room temperature with MetaMorph™ software (v6.2r6) (Molecular Devices). Images were acquired using the DAPI (403Ex/465Em), FITC (470Ex/535Em) and TRED (560Ex/650Em) filter sets. Pseudocolouring and merging of images was performed using ImageJ v1.34s (Rasband, 1997–2009).

Confocal imaging

For confocal imaging of the subcellular distribution of huntingtin EGFP, cells were plated at 300 000 cells·well⁻¹ onto poly-L-lysine-coated coverslips (#1.5 thickness) in 24-well plates and induced the next day with 1 µM TFZ in full-serum media. After 51 h incubation at 37°C, media was removed and cells fixed with 4% paraformaldehyde for 10 min. Cells were rinsed for 2 × 10 min with PBS, nuclei were stained with Hoechst 33258, 1:500 for 10 min in the dark, and cells were rinsed for a final 2 × 10 min in PBS. Coverslips were mounted with CFPVOH/AF100 antifadant (CitiFluor, Leicester, UK) by inversion onto glass slides. Images were acquired using Leica Scanware 4.2a and equivalent voltage, on a Leica TCS 4D confocal laser scanning microscope (Leica, Wetzlar, Germany) using a 63× plan apochromat objective lens (oil, NA 1.4) at room temperature.

Cannabinoids and drugs

HU210, WIN55212-2, BAY59-3074 and SR141716A were purchased from Tocris Bioscience (Bristol, UK). Due to their hydrophobicity, cannabinoids tend to adsorb to surfaces in the absence of a carrier molecule. All dilutions of cannabinoid drugs were therefore performed in plasticware silanized using Coatasil™ (dimethyl dichlorosilane, Ajax Finechem Ltd., Auckland, New Zealand) as per manufacturer’s instructions. The host of carrier proteins in cell culture media animal sera might similarly adsorb cannabinoid ligand, therefore cannabinoid delivery was performed in SFM with 5 mg·mL⁻¹ BSA (SFM/BSA) as described by Hillard et al. (1995). While absolute cell number was lower under serum-free conditions, relative huntingtin toxicity was indistinguishable from that seen in full-serum conditions. All other drugs were purchased from Sigma (St. Louis, MO, USA) unless otherwise indicated.

Cell survival assays

For survival assays, cells were plated at 35 000 well⁻¹ in the central 60 wells of poly-L-lysine (0.2 mg·mL⁻¹)-coated 96-well plates. The following day, cells were induced to express huntingtin in the presence of vehicle, cannabinoids, dopaminergics or cAMP modulators. Briefly, media was replaced with SFM/BSA and 0–1 µM TFZ inducer (each with final 0.01% ethanol) with various co-treatments (or their vehicles). These included cannabinoid agonists HU210, WIN55212-2, BAY59-3074 or inverse agonist SR141716A at 0–1 µM, each with final 0.01% ethanol; cAMP-PKA (protein kinase A) binding inhibitor Rp-cAMPS at 100 µM; or the cAMP stimulator forskolin at 10 µM. Due to their superior hydrophilicity, experiments with dopamine (10 µM), D1 antagonist SCH23390 (10 µM), D1 agonist SKF33893 (0–10 µM) and D2 agonist quinpirole (0–10 µM) were performed in full-serum media.

After 72 h, cell viability was assayed by replacing media with SFM/BSA (or full-serum media for dopamine experiments) and the redox indicator Alamar Blue (AbD SeroTec Ltd., Oxford, UK) at 10% (v/v), and returning to incubator for a further 4 h. The fluorescence generated by conversion of Alamar Blue by viable cells was measured at 37°C (Ex/Em 560/590 nm) on a FLUOstar Optima fluorescence plate reader (BMG Labtechnologies, Durham, NC, USA). Percentage cell number was calculated by normalising test cell fluorescence against control cell fluorescence (100%) and fluorescence of Alamar Blue assay solution alone (0%). Depending upon the assay, control cells were represented by cells treated with either vehicle (0.01% ethanol) or the chosen concentration of test compound (cannabinoid, cAMP modulator or dopaminergic) for the duration of the experiment but not induced with TFZ. This allowed specific quantification of drug effect on huntingtin cell death independent of any basal proliferative/toxic effects of the test drug. Figures depict cell death across the TFZ concentration range or at 1 µM TFZ only where the effect was most pronounced at this concentration.

Quantifying the percentage of cells that expressed huntingtin and formed aggregates

After Alamar Blue readings were taken, cells were fixed in 4% paraformaldehyde and then washed in PBS. Cell nuclei were stained with Hoechst 33258 for 10 min, then cells were washed twice in PBS, and imaged at 10× magnification on the Discovery-1™ automated fluorescence microscope. The images were analysed using the Cell Scoring application within MetaMorph™ image analysis software as described previously (Scotter et al., 2008).

This application was used to determine either the proportion of cells that expressed detectable levels of huntingtin, or the proportion of cells that contained at least one mutant huntingtin aggregate. This was achieved by quantifying the percentage of Hoechst-positive cells that also had staining in the EGFP wavelength meeting size and intensity criteria for either total huntingtin, or aggregates, as determined from three sample images. For total huntingtin, the criteria were: staining of cytoplasm size, above the intensity of un-induced cells. For aggregates, the criteria were: staining of 1–3 µm, above the maximal intensity achieved by all aggregates in the sample images. We have previously confirmed (Scotter et al., 2008) by filter retardation assay that aggregates formed by PC12 97Q cells are characteristic ‘insoluble, detergent-resistant’ aggregates as seen in other HD models and in human brain (Kazantsev et al., 1999; Hoffner et al., 2005). Figures depict aggregate formation across the TFZ concentration range, or at 1 µM TFZ only where the effect was most pronounced at this concentration.

cAMP accumulation/PKA binding assay

The cellular accumulation of cAMP was assayed by scintillation counting of tritiated cAMP as described previously (Kearn et al., 2005), with the following modifications: cells were plated at 25 000 cells·well⁻¹ the day before assay. For G-protein toxin experiments, 4 h after plating, media was replaced with fresh full-serum media containing Pertussis toxin (PTX) at 100 ng·mL⁻¹ (or 0.05% v/v vehicle: 50% glycerol, 50 mM Tris, 10 mM glycine, 0.5 M NaCl, pH 7.5), and cells were incubated overnight for 18 h. For all assays, on the day of the assay, basal receptor signalling was reduced by ‘serum starvation’; cells were incubated with SFM/BSA containing 500 µM 3-isobutyl-1-methyl xanthine (phosphodiesterase inhibitor) for 30 min at 37°C. For assay of the stimulation of cAMP by forskolin, forskolin was added to serum starvation media at 2× concentration in the same media to give 0–250 µM final. For assay of the modulation of cAMP by HU210, HU210 was added to serum starvation media at 2× concentration in the same media to give 0–1 µM final. For assay of the inhibition of cAMP-PKA binding, a cell-free modification of this assay was performed where cell lysate was substituted for 25 µL Rp-cAMPS at 0–100 µM in cAMP assay buffer.

Western blotting

Western blotting for phosphorylated ERK was performed as described previously (Graham et al., 2006), with the following modifications: cells were plated at 300 000 cells per well in 24-well plates and were stimulated the next day to mimic conditions in cell survival assays. Plating media was aspirated and cells were stimulated with 1 µM TFZ in SFM/BSA with either vehicle (0.01% ethanol), 1 µM HU210, or 1 µM SR141716A. For G_i/o blockade experiments, cells were pretreated for 18 h in full-serum media with PTX at 100 ng·mL⁻¹ (or 0.05% v/v vehicle). For ERK blockade experiments, cells were pretreated for 45 min in full-serum media with UO126 at 10 µM (or 0.01% v/v DMSO vehicle).

After appropriate incubation time at 37°C, plates were placed on ice and harvested in 100 µL 1× laemmli buffer (62.5 mM Tris-HCl, pH 6.8, 2% SDS, 10% glycerol) (Laemmli, 1970) and then denatured at 95°C for 5 min. Protein concentration was determined using the Bio-Rad DC Protein Assay, and samples were stored at −85°C until required. Samples were diluted to 0.8 µg·µL⁻¹ including final 5% β-mercaptoethanol, and 10 µL of each sample was separated on 10% SDS-polyacrylamide gel at 70 V for 4 h. Proteins were transferred as described previously. Blots were then blocked for 30 min in TBS-T (10 mM Tris Base, 150 mM NaCl, 0.05% Tween 20) containing 1% BSA and then probed overnight at 4°C for pERK (1:1000; catalogue number 9101; Cell Signaling Technology, Beverly, MA, USA), or total ERK (1:2000; catalogue number 9102; Cell Signaling Technology) in TBS-T containing 0.2% BSA. Secondary antibody and ECL detection were performed as described previously, with film exposure times ranging from 20 s to 1 min. Films were scanned using a HP Scanjet 3770 scanner (Hewlett-Packard Co., Palo Alto, CA, USA) at 600 dpi resolution in true colour. Optical density (OD) for pERK was determined by measuring the area under the intensity curve for pERK, as plotted using Gel Analyze in ImageJ, and dividing by the area under the intensity curve for total ERK for that lane. Relative OD was calculated by normalising to the average OD for the vehicle-treated samples.

Immunocytochemistry

For immunocytochemistry for phosphorylated ERK, cells were plated at 100 000 cells·well⁻¹ onto poly-L-lysine-coated eight-well culture slides (BD Biosciences, Franklin Lakes, NJ, USA) and were stimulated the next day to mimic conditions in cell death assays. Plating media was aspirated, and cells were stimulated with 1 µM TFZ in SFM/BSA with vehicle (0.01% ethanol), or 1 µM HU210 for 5 min. After appropriate incubation time at 37°C, slides were placed on ice for 2 min during media removal, removed from ice then fixed with 4% paraformaldehyde for 10 min. Cells were rinsed for 2 × 10 min with PBS-T. Epitope unmasking was performed by adding 100 µL·well⁻¹ ice-cold methanol (90% v/v) and cooling at −20°C for 10 min. Cells were rinsed for a further 2 × 10 min in PBS-T then incubated overnight with a distinct α-pERK antibody from that used on Western blots, which was compatible with immunocytochemistry (1:200 in IB-T, catalogue # 4370, raised in rabbit, Cell Signaling Technology). No primary antibody control wells were incubated with IB-T only. Cells were rinsed for 3 × 10 min in PBS-T, then incubated with goat α-rabbit Alexa 488 secondary (1:500 in IB-T, Molecular Probes, Eugene, OR, USA) for 3 h at room temperature. Cells were rinsed for a further 3 × 10 min in PBS-T, then nuclei were stained with Hoechst 33258, 1:500, for 10 min in the dark, before a final 3 × 10 min rinse in PBS-T. The glass slides were mounted with CFPVOH/AF100 antifadant and coverslipped (#1.5 thickness). Images were acquired using a 60× plan fluorite objective lens (air, NA 0.7) at room temperature with NIS-Elements BR software (v3.0) on an Eclipse Ti microscope fitted with a Nikon Coolpix 4500 camera (all Nikon, Melville, NY, USA). Equivalent exposure times for a particular wavelength were used for each drug treatment. Images were merged using Image J 1.34s (http://rsb.info.nih.gov/ij/, NIH, Bethesda, MD, USA).

Statistical analyses

Representative individual experiments are presented in several figures as stated. All other figures or descriptions in the text are the mean ± SEM for pooled data from all repeats. Statistical analyses were performed on pooled data (for each cell type individually) by one-way repeated measures anova with Bonferroni multiple comparison post testing using GraphPad Prism® 4.02 (GraphPad Software Inc., La Jolla, CA, USA).

PC12 cells endogenously expressed D2 dopamine but not D1 dopamine or CB₁ cannabinoid receptors

PCR was carried out to examine expression of endogenous dopamine or cannabinoid receptors in PC12 cells. PC12 cells transfected with 25Q or 97Q huntingtin constructs were found to endogenously express the D2 dopamine receptor, at low levels compared with brain, but not the D1 dopamine receptor or the CB₁ cannabinoid receptor (Figure 1). The absence of endogenous CB₁ is consistent with our findings by RT-PCR and binding assay (data not shown), and those of Aiken et al. by immunoblot (Aikenet al., 2004), in the parental PC12 cells. Drug effects on cell lines not transfected with the target receptor could therefore be considered to be independent of activity at these receptors. Whole cell binding assays showed specific binding equivalent to 25.2 ± 4.4 fmol·mg⁻¹ in PC12 97Q CB1 cells and 22.7 ± 1.2 fmol·mg⁻¹ in PC12 25Q CB1 cells. No specific binding was detected in untransfected PC12 cells (data not shown).

Figure 1

(A) RT-PCR analysis of G-protein-coupled receptors expressed by PC12 cells showed that there was a low level of endogenous expression of the D2 dopamine receptor, but no detectable D1 dopamine receptor or CB₁ cannabinoid receptor, unless transfected (−RT, …

Huntingtin cell death was time- and inducer concentration-dependent

Induction of PC12 25Q cells with TFZ leads to diffuse expression of EGFP-tagged huntingtin, throughout the cytoplasm and, to a lesser degree, also the nucleus (Figure 2A). Induction of PC12 97Q cells leads to a similar initial distribution of EGFP-tagged huntingtin; however, diffuse huntingtin formed intensely fluorescent aggregates in an increasing number of cells over time. The majority of aggregate-positive cells contained one large cytoplasmic aggregate; however, there were also instances of more than one large cytoplasmic aggregate or numerous small nuclear/perinuclear aggregates (Figure 2B).

Figure 2

Confocal photomicrographs of (A) PC12 25Q cells and (B) PC12 97Q cells, each induced with 1 µM tebufenozide (TFZ) for 48 h. 25Q huntingtin showed diffuse, predominantly cytoplasmic expression. 97Q huntingtin was also predominantly diffuse in the …

The proportion of cells that expressed detectable levels of huntingtin was well correlated to the concentration of TFZ in this tightly regulated expression system (Figure 2C and D). PC12 cells showed significant and extensive reductions in Alamar Blue conversion in response to huntingtin expression, in a manner that was time-dependent and concentration-dependent with respect to the TFZ inducer (Figure 2E and F). The redox indicator Alamar Blue is converted in live cells by various mitochondrial and cytoplasmic enzymes (Gonzalez and Tarloff, 2001) to give a measure of the number of living cells, rather than a measure of their viability. We found an excellent correlation, across a range of cell viabilities, between Alamar Blue fluorescence and cell counts measured using a Discovery-1™ automated fluorescence microscope and MetaMorph™ image analysis software (data not shown). Alamar Blue was therefore used as an indirect assay for cell number. Lactate dehydrogenase assays confirmed that this TFZ concentration-dependent decrease in cell number was due to cell death rather than an inhibition of proliferation (data not shown).

Interestingly, expression of either 25Q or 97Q exon one huntingtin protein was toxic to PC12 cells, although both the potency and extent of the death response was greater with 97Q Htt expression (Figure 2E, 25Q: EC₅₀= 167.5 ± 47.1 nM, cells remaining with 1 µM TFZ = 56.7 ± 1.6%; 97Q: EC₅₀= 34.0 ± 6.0 nM, cells remaining with 1 µM TFZ = 34.1 ± 1.7%). Hereafter 25Q huntingtin is therefore referred to as ‘wildtype’ to acknowledge that it differs from full-length wild-type huntingtin.

The aggregate formation seen in PC12 97Q cells was dependent upon both the concentration of mutant huntingtin protein, with an EC₅₀ of 30.3 ± 2.7 nM TFZ, and the expression time (Figure 2G and H), consistent with previous reports of amyloid-like, nucleation-dependent aggregation of mutant huntingtin (Scherzinger et al., 1999; Chen et al., 2001).

Huntingtin cell death is modulated by cellular cAMP levels

 

Cell death due to mutant huntingtin expression was alleviated by the cannabinoid agonists HU210 and WIN55212-2 and exacerbated by the inverse agonist SR141716A

As shown in Figure 1, PC12 cells did not endogenously express the CB₁ cannabinoid receptor. Accordingly, CB₁-negative PC12 cells (PC12 25Q/97Q) showed no significant change in the cell death due to expression of ‘wildtype’ or mutant exon one Htt when co-treated with the classical cannabinoid agonist HU210 or inverse agonist SR141716A (Figure 3A and C). In PC12 cells transfected with human CB₁ receptor and expressing ‘wildtype’ exon one Htt (PC12 25Q CB₁), neither ligand significantly altered the cell death profile (Figure 3B). In contrast, in CB₁-transfected cells expressing mutant exon one Htt (PC12 97Q CB₁), HU210 co-treatment caused a small but significant and reproducible reduction in the extent of cell death (Figure 3D, 7.9 ± 2.0% reduction). This alleviation of mutant huntingtin cell death by HU210 was concentration-dependent, with maximal rescue of death seen with 1 µM HU210 (Figure 3E). In accordance with this, the inverse agonist SR141716A caused a small exacerbation of cell death in PC12 97Q CB₁ cells treated with 1 µM TFZ (Figure 3D and F, 3.3 ± 1.3% exacerbation).

Figure 3

Huntingtin-induced death at 72 h in CB₁-negative PC12 cells expressing (A) 25Q or (C) 97Q exon one huntingtin, or in CB₁-transfected cells expressing (B) 25Q or (D) 97Q exon one huntingtin, due to induction with 1 µM tebufenozide (TFZ) and co-treatment …

Two other cannabinoid agonists, the synthetic full agonist WIN55212-2 and the synthetic low-efficacy agonist BAY59-3074, were also investigated as potential cell death modifiers. When applied at 1 µM, there was no effect of either WIN55212-2 or BAY59-3074 on PC12 25Q or PC12 25Q CB₁ cells (Figure 4A and B). However, both compounds showed a small degree of protection in PC12 97Q CB₁ cells induced with 1 µM TFZ (Figure 4D, WIN: 6.4 ± 1.1%, BAY: 4.4 ± 1.7%). Interestingly, BAY59-3074 co-treatment improved HD cell survival similarly in CB₁–negative PC12 97Q cells, indicating a receptor-independent mechanism (Figure 4C, 3.1 ± 1.3%). Indeed, while the alleviation of mutant huntingtin cell death in PC12 97Q CB₁ cells by WIN55212-2 was concentration-dependent, BAY59-3074 conferred protection only at the 1 µM concentration, far higher than its reported K_i at the human CB₁ receptor, of 48.3 nM (De Vry et al., 2004) (Figure 4E).

Figure 4

Huntingtin-induced death at 72 h in CB₁-negative PC12 cells expressing (A) 25Q or (C) 97Q exon one huntingtin, or in CB₁-transfected cells expressing (B) 25Q or (D) 97Q exon one huntingtin, due to induction with 1 µM tebufenozide and co-treatment …

 

Cell death due to huntingtin expression was exacerbated by forskolin but unchanged by Rp-cAMPS

We examined the potential involvement of cyclic adenosine monophosphate (cAMP) modulation in HU210-mediated protection against mutant huntingtin toxicity. We first validated the ability of the adenylate cyclase activator forskolin to increase cAMP levels, and the ability of Rp-cAMPS to inhibit binding of cAMP to PKA. We found that treatment of cells with 28–250 µM forskolin significantly increased intracellular cAMP concentration (Figure 5A). In a cell-free modification of the same assay, Rp-cAMPS inhibited the binding of [³H]-cAMP to PKA significantly at 100 µM, however only weakly and not at lesser concentrations (Figure 5B, 17.2% inhibition). In cell survival assays, cAMP stimulation by 10 µM forskolin significantly exacerbated cell death due to expression of either ‘wildtype’ or mutant huntingtin (Figure 5C and D, PC12 25Q: 15.0 ± 3.0% increase; PC12 97Q: 9.2 ± 3.2% increase). Rp-cAMPS did not significantly modify cell death due to expression of either ‘wildtype’ or mutant huntingtin; however, this may have been due to its poor efficacy in inhibiting cAMP binding to PKA. Together with the cannabinoid findings, these data suggested that the modulation of cAMP levels may influence cell survival in huntingtin-expressing cells.

Figure 5

(A) Accumulation of cAMP in un-induced (no TFZ) PC12 cells stimulated with forskolin [F(9,96) = 11.26, 28–250 µM Fsk all P < 0.001,n= 2] and (B) inhibition of [³H]-cAMP binding to protein kinase A in the presence of Rp-cAMPs (100 …

 

Cell death due to huntingtin expression was exacerbated by D1 dopaminergic signalling

To further test this hypothesis, we employed D1 dopamine receptor stimulation to investigate the effect of G_s signalling. PC12 cells endogenously expressed the D2 but not D1 subtype, as assessed by RT-PCR (Figure 1). Exogenous dopamine had no significant effect on the viability of D1-negative PC12 cells expressing wild-type or mutant exon one Htt (Figure 6A and B). However, in PC12 cells transfected with human D1 (PC12 97Q D1), exogenous dopamine significantly exacerbated the cell death due to expression of mutant Htt expression at concentrations from 12–333 nM TFZ (Figure 6C). Surprisingly, at low TFZ concentrations, the D1-specific antagonist SCH23390 also significantly exacerbated the extent of cell death in PC12 97Q D1 cells. SCH23390 did not significantly alter the profile of cell death in D1-negative PC12 97Q cells, indicating that the D1 was required for this effect (Figure 6C).

Figure 6

Huntingtin-induced death at 72 h in D1-negative PC12 cells expressing (A) 25Q or (B) 97Q exon one huntingtin, or in D1-transfected cells expressing (C) 97Q exon one huntingtin, due to induction with 0–1 µM tebufenozide (TFZ) and co-treatment …

To further confirm that the detrimental effect of dopamine was D1-mediated, the selective D1 or D2 agonists, SKF38393 and quinpirole, respectively, were tested for their ability to modify mutant huntingtin-induced cell death. Figure 6D shows that in PC12 97Q D1 cells induced with 1 µM TFZ for 72 h, SKF38393, but not quinpirole, mimics dopamine potentiation of huntingtin toxicity. Co-treatment with 1 µM SKF38393 reduced cell number by 8.5 ± 2.6% compared with vehicle co-treatment, whereas 1 µM quinpirole co-treatment did not significantly alter cell death. These data support a detrimental effect of D1 signalling in mutant huntingtin-expressing cells.

HU210-mediated rescue of huntingtin cell death is dependent upon G_i/o and ERK

 

Functional G_i/o was required for HU210-mediated rescue of huntingtin cell death

Results so far indicated that various compounds that elevate cAMP levels exacerbated huntingtin-mediated cell death, and cannabinoid compounds, which decrease cAMP levels (Howlett et al., 1986), alleviated huntingtin-mediated cell death. We therefore tested whether HU210-mediated rescue of huntingtin cell death was in fact dependent upon G_i/o coupling, which links CB₁ to the inhibition of cAMP. PTX ablation of G_i/o was confirmed by cAMP assay, which showed a blockade of the cAMP inhibition caused by HU210, and a switch to G_s coupling, as described previously (Glass and Felder, 1997; Bonhaus et al., 1998) (Figure 7A). It has not been determined whether it is the G_i/o subunits or the βγ they release that are responsible for adenylate cyclase inhibition, therefore the term ‘G_i/o’ is used to encompass all possible effectors. The EC₅₀ values for inhibition or stimulation of cAMP by HU210 were 0.26 ± 0.10 nM and 24.7 ± 14.5 nM, respectively, indicating that at the 1 µM HU210 concentration predominantly used in this study, the CB₁ receptor is capable of activating either G_i/o or G_s.

Figure 7

(A) HU210-mediated changes in cAMP accumulation in PC12 97Q CB₁ cells pretreated with (□) vehicle or () Pertussis toxin (PTX, 100 ng·mL⁻¹, 18 h) relative to basal (B) or forskolin (Fsk)-induced cAMP levels. While the CB₁ was coupled …

In PC12 97Q CB₁ cells subjected to this PTX ablation before huntingtin expression and HU210 treatment were initiated, the concentration-dependent alleviation of huntingtin cell death by HU210 was abolished [Figure 7B, vehicle (veh) pretreat: 5.8 ± 1.7% rescue by HU210; PTX pretreat: 0.4 ± 1% rescue by HU210]. Functional G_i/o was therefore required for HU210-mediated rescue of huntingtin cell death. In addition, the extent of huntingtin cell death induced by 1 µM TFZ alone (without HU210 co-treatment) was exacerbated by PTX pretreatment (Figure 7B, veh pretreat: 32.5 ± 1.0% cells remain; PTX pretreat: 24.5 ± 0.6% cells remain). This would be consistent with the uncoupling of all tonically active G_i/o-linked GPCRs with protective signalling capacities, which may comprise various GPCR subtypes, or CB₁alone. Activation of the D2, which is also G_i/o-linked, was unable to improve cell survival. While this may be a function of the low level of endogenous D2 expression compared with brain, D2 receptor signalling has previously been shown to potentiate the toxicity of mutant huntingtin (Charvin et al., 2005). The protective effect of cAMP inhibition may therefore be specific to activation by CB₁ or to certain G_i/o-linked pathways.

Huntingtin aggregate formation

 

HU210 and activators of the cAMP pathway enhanced huntingtin aggregate formation

We have described a small but highly reproducible, concentration-dependent protective effect for HU210 against mutant huntingtin cell death in PC12 97Q CB₁ cells. Surprisingly, however, HU210 caused an increase in the percentage of PC12 97Q CB₁ cells that formed mutant huntingtin aggregates over the 72 h induction period with 1 µM TFZ (Figure 8B, 6.1 ± 2.6% increase). This increase in aggregates was specific to CB₁-transfected cells, indicating a receptor-dependent mechanism, with CB₁-negative PC12 97Q cells showing no change in aggregate formation with HU210 (Figure 8A). Also supporting a CB₁receptor-dependent pathway, the increase in aggregates by HU210 was concentration-dependent, with maximal aggregate formation seen with 1 µM HU210 (Figure 8C).

Figure 8

Formation of 97Q huntingtin aggregates (Agg) at 72 h in (A) CB₁-negative or (B) CB₁-transfected PC12 cells expressing 97Q exon one huntingtin, following induction with 1 µM tebufenozide (TFZ) and co-treatment with vehicle (0.01% EtOH), HU210 (1 …

There was no significant modulation of aggregate formation by WIN55212-2 or BAY59-3074 in either PC12 97Q or 97Q CB₁ cells (data not shown). While there was also no detectable modulation of aggregate formation by SR141716A in Figure 8A and B, the modified treatment paradigm using a constant concentration (1 µM) of TFZ and increasing concentrations of SR141716A revealed a concentration-dependent inhibitory effect of SR141716A upon aggregate formation in PC12 97Q CB₁ cells, which was in keeping with its activity as an inverse agonist of CB₁ (Figure 8C).

 

Huntingtin aggregate formation was unchanged by Rp-cAMPS but enhanced by forskolin in PC12 cells

While our findings for HU210 suggested a correlation between cellular rescue and increased aggregate load, the 9.2 ± 1.9% exacerbation of PC12 97Q cell death by forskolin correlated with an almost doubling in the percentage of these cells containing mutant huntingtin aggregates (Figure 9A, 17.9 ± 3.9% increase). There was no reduction in aggregate load seen with Rp-cAMPs. The finding of increased aggregate load associated with forskolin toxicity is surprising in light of the increased aggregate load associated with a beneficial effect of HU210. It appears that either there is a complex relationship between aggregate load and cell number, or that the pathway for modulation of cell survival, via G_i/o, inhibition of cAMP, and phosphorylation of ERK 1/2, is not involved in the modulation of aggregate formation, and this was investigated further.

Figure 9

(A) Formation of 97Q huntingtin aggregates (Agg) at 72 h in PC12 97Q cells treated with 1 µM tebufenozide (TFZ) and co-treated with vehicle (Veh) (0.02% DMSO), forskolin (Fsk) (10 µM) or Rp-cAMPS (100 µM). There was a significant …

HU210-mediated enhancement of aggregate formation proceeds through a different pathway to HU210-mediated alleviation of huntingtin cell death

Neither functional G_i/o nor ERK phosphorylation were required for HU210-mediated enhancement of aggregate formation

We have shown that the rescue of huntingtin-expressing PC12 97Q CB₁ cells by HU210 could be prevented by PTX pretreatment, indicating that G_i/o was necessary. Using the same treatment paradigms we next investigated whether there was a requirement for G_i/o in the potentiation of aggregate formation by HU210. Figure 10A shows that in both vehicle- and PTX-pretreated PC12 97Q CB₁ cells, there was a concentration-dependent enhancement of aggregate formation by HU210. There was no significant difference in the extent of aggregate potentiation by HU210 between the two pretreatments, either in individual experiments or when data was pooled. Interestingly, there was a trend towards PTX pretreatment enhancing the basal aggregate formation response to induction with 1 µM TFZ. These data indicate that G_i/o coupling is not necessary for the CB₁-mediated potentiation of aggregate formation, and that G_i/o coupling of other GPCRs in the cell may basally inhibit aggregate formation.

Figure 10

Formation of 97Q huntingtin aggregates (Agg) at 72 h in PC12 97Q CB₁ cells pretreated with (A) (□) vehicle (Veh) or () Pertussis toxin (PTX, 100 ng·mL⁻¹, 18 h) or (B) (□) vehicle or () UO126 (10 µM, 45 min) then …

Cell survival

Huntingtin aggregate formation

While the mature aggregates detected in our study may not have caused toxicity per se, their formation is dependent upon, and therefore a marker of, high concentrations of misfolded soluble monomers/oligomers, which are increasingly being associated with toxicity in a range of neurological diseases (Lambert et al., 1998; Watase et al., 2002; Caughey and Lansbury, 2003; Walsh and Selkoe, 2004).

This work was supported by a grant from the Marsden Fund of The Royal Society of New Zealand, the Health Research Council of New Zealand and the National Research Centre for Growth and Development. ELS was supported by the Neurological Foundation of New Zealand; ELS and CEG were supported by The University of Auckland. The cell line used in this paper was kindly gifted by Dr Eric Schweitzer (Brain Research Institute, UCLA). Mouse cerebellar tissue was provided by Associate Professor Anthony Hannan (Howard Florey Institute, Melbourne, Vic., Australia). Dow Agrosciences NZ Ltd. provided the insect steroid hormone TFZ for induction of the cell lines. The authors would also like to thank Dr Hector Monzo and Dr Maurice Curtis for their technical assistance.

Abbreviations:

BSA

bovine serum albumin

CB₁

cannabinoid receptor 1

D1

dopamine receptor 1

D2

dopamine receptor 2

ERK

extracellular signal-regulated kinase

G_i/o

G-protein alpha subtype i/o

G_s

G-protein alpha subtype s

HD

Huntington’s disease

PC12

pheochromocytoma

PKA

protein kinase A

PTX

Pertussis toxin

SFM

serum-free media

SFM/BSA

serum-free media with BSA

TFZ

tebufenozide

• Aiken CT, Tobin AJ, Schweitzer ES. A cell-based screen for drugs to treat Huntington’s disease. Neurobiol Dis.2004;16:546–555. [PubMed]
• Alexander SP, Mathie A, Peters JA. Guide to Receptors and Channels (GRAC), 3rd edition. Br J Pharmacol.2008;153:S1–S209. [PMC free article] [PubMed]
• Andersson M, Usiello A, Borgkvist A, Pozzi L, Dominguez C, Fienberg AA, et al. Cannabinoid action depends on phosphorylation of dopamine- and cAMP-regulated phosphoprotein of 32 kDa at the protein kinase A site in striatal projection neurons. J Neurosci. 2005;25:8432–8438. [PubMed]
• Biegon A, Kerman IA. Autoradiographic study of pre- and postnatal distribution of cannabinoid receptors in human brain. Neuroimage. 2001;14:1463–1468. [PubMed]
• Bonhaus DW, Chang LK, Kwan J, Martin GR. Dual activation and inhibition of adenylyl cyclase by cannabinoid receptor agonists: evidence for agonist-specific trafficking of intracellular responses. J Pharmacol Exp Ther. 1998;287:884–888. [PubMed]
• Caughey B, Lansbury PT. Protofibrils, pores, fibrils, and neurodegeneration: separating the responsible protein aggregates from the innocent bystanders. Annu Rev Neurosci. 2003;26:267–298. [PubMed]
• Caunt CJ, Finch AR, Sedgley KR, McArdle CA. Seven-transmembrane receptor signalling and ERK compartmentalization. Trends Endocrinol Metab.2006;17:276–283. [PubMed]
• Charvin D, Vanhoutte P, Pages C, Borrelli E, Caboche J. Unraveling a role for dopamine in Huntington’s disease: the dual role of reactive oxygen species and D2 receptor stimulation. Proc Natl Acad Sci USA. 2005;102:12218–12223. [PMC free article] [PubMed]
• Chen S, Berthelier V, Yang W, Wetzel R. Polyglutamine aggregation behavior in vitro supports a recruitment mechanism of cytotoxicity. J Mol Biol. 2001;311:173–182.[PubMed]
• Daumas F, Destainville N, Millot C, Lopez A, Dean D, Salome L. Confined diffusion without fences of a G-protein-coupled receptor as revealed by single particle tracking.Biophys J. 2003;84:356–366. [PMC free article][PubMed]
• De Vry J, Denzer D, Reissmueller E, Eijckenboom M, Heil M, Meier H, et al. 3-[2-cyano-3-(trifluoromethyl)phenoxy]phenyl-4,4,4-trifluoro-1-butanesulfo nate (BAY 59-3074): a novel cannabinoid Cb1/Cb2 receptor partial agonist with antihyperalgesic and antiallodynic effects. J Pharmacol Exp Ther.2004;310:620–632. [PubMed]
• van Dellen A, Blakemore C, Deacon R, York D, Hannan AJ. Delaying the onset of Huntington’s in mice. Nature.2000;404:721–722. [PubMed]
• Ebisuya M, Kondoh K, Nishida E. The duration, magnitude and compartmentalization of ERK MAP kinase activity: mechanisms for providing signaling specificity. J Cell Sci.2005;118:2997–3002. [PubMed]
• Ferrante RJ, Kowall NW, Beal MF, Richardson EP, Jr, Bird ED, Martin JB. Selective sparing of a class of striatal neurons in Huntington’s disease. Science. 1985;230:561–563. [PubMed]
• Glass M, Felder CC. Concurrent stimulation of cannabinoid CB1 and dopamine D2 receptors augments cAMP accumulation in striatal neurons: evidence for a Gs linkage to the CB1 receptor. J Neurosci. 1997;17:5327–5333. [PubMed]
• Glass M, Northup JK. Agonist selective regulation of G proteins by cannabinoid CB1 and CB2 receptors. Mol Pharmacol. 1999;56:1362–1369. [PubMed]
• Glass M, Dragunow M, Faull RL. Cannabinoid receptors in the human brain: a detailed anatomical and quantitative autoradiographic study in the fetal, neonatal and adult human brain. Neuroscience. 1997;77:299–318. [PubMed]
• Glass M, Dragunow M, Faull RL. The pattern of neurodegeneration in Huntington’s disease: a comparative study of cannabinoid, dopamine, adenosine and GABA(A) receptor alterations in the human basal ganglia in Huntington’s disease. Neuroscience.2000;97:505–519. [PubMed]
• Glass M, van Dellen A, Blakemore C, Hannan AJ, Faull RLM. Delayed onset of Huntington’s disease in mice in an enriched environment correlates with delayed loss of cannabinoid CB1 receptors. Neuroscience. 2004;123:207–212. [PubMed]
• Gonzalez RJ, Tarloff JB. Evaluation of hepatic subcellular fractions for Alamar blue and MTT reductase activity.Toxicol In Vitro. 2001;15:257–259. [PubMed]
• Graham ES, Ball N, Scotter EL, Narayan P, Dragunow M, Glass M. Induction of Krox-24 by endogenous cannabinoid type 1 receptors in Neuro2A cells is mediated by the MEK-ERK MAPK pathway and is suppressed by the phosphatidylinositol 3-kinase pathway. J Biol Chem.2006;281:29085–29095. [PubMed]
• Graveland GA, Williams RS, DiFiglia M. Evidence for degenerative and regenerative changes in neostriatal spiny neurons in Huntington’s disease. Science.1985;227:770–773. [PubMed]
• Grimsey NL, Narayan PJ, Dragunow M, Glass M. A novel high-throughput assay for the quantitative assessment of receptor trafficking. Clin Exp Pharmacol Physiol.2008;35:1377–1382. [PubMed]
• Hillard CJ, Edgemond WS, Campbell WB. Characterization of ligand binding to the cannabinoid receptor of rat brain membranes using a novel method: application to anandamide. J Neurochem. 1995;64:677–683. [PubMed]
• Hoffner G, Island M-L, Djian P. Purification of neuronal inclusions of patients with Huntington’s disease reveals a broad range of N-terminal fragments of expanded huntingtin and insoluble polymers. J Neurochem.2005;95:125–136. [PubMed]
• Houslay MD. Compartmentalization of cyclic AMP phosphodiesterases, signalling ‘crosstalk’, desensitization and the phosphorylation of Gi-2 add cell specific personalization to the control of the levels of the second messenger cyclic AMP. Adv Enzyme Regul. 1995;35:303–338. [PubMed]
• Howlett AC, Qualy JM, Khachatrian LL. Involvement of Gi in the inhibition of adenylate cyclase by cannabimimetic drugs. Mol Pharmacol. 1986;29:307–313. [PubMed]
• Jin K, LaFevre-Bernt M, Sun Y, Chen S, Gafni J, Crippen D, et al. FGF-2 promotes neurogenesis and neuroprotection and prolongs survival in a transgenic mouse model of Huntington’s disease. Proc Natl Acad Sci USA.2005;102:18189–18194. [PMC free article] [PubMed]
• Kazantsev A, Preisinger E, Dranovsky A, Goldgaber D, Housman D. Insoluble detergent-resistant aggregates form between pathological and nonpathological lengths of polyglutamine in mammalian cells. Proc Natl Acad Sci USA. 1999;96:11404–11409. [PMC free article] [PubMed]
• Kearn CS, Blake-Palmer K, Daniel E, Mackie K, Glass M. Concurrent stimulation of cannabinoid CB1 and dopamine D2 receptors enhances heterodimer formation: a mechanism for receptor cross-talk? Mol Pharmacol.2005;67:1697–1704. [PubMed]
• Laemmli UK. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature.1970;227:680–685. [PubMed]
• de Lago E, Fernandez-Ruiz J, Ortega-Gutierrez S, Cabranes A, Pryce G, Baker D, et al. UCM707, an inhibitor of the anandamide uptake, behaves as a symptom control agent in models of Huntington’s disease and multiple sclerosis, but fails to delay/arrest the progression of different motor-related disorders. Eur Neuropsychopharmacol.2006;16:7–18. [PubMed]
• Lambert MP, Barlow AK, Chromy BA, Edwards C, Freed R, Liosatos M, et al. Diffusible, nonfibrillar ligands derived from Abeta1-42 are potent central nervous system neurotoxins. Proc Natl Acad Sci USA. 1998;95:6448–6453. [PMC free article] [PubMed]
• Lastres-Becker I, Bizat N, Boyer F, Hantraye P, Brouillet E, Fernandez-Ruiz J. Effects of cannabinoids in the rat model of Huntington’s disease generated by an intrastriatal injection of malonate. Neuroreport.2003a;14:813–816. [PubMed]
• Lastres-Becker I, de Miguel R, De Petrocellis L, Makriyannis A, Di Marzo V, Fernandez-Ruiz J. Compounds acting at the endocannabinoid and/or endovanilloid systems reduce hyperkinesia in a rat model of Huntington’s disease. J Neurochem. 2003b;84:1097–1109. [PubMed]
• Lehrach H, Wanker EE. Huntington’s disease: from gene to potential therapy. Dialogues Clin Neurosci. 2001;3:17–23.[PMC free article] [PubMed]
• Rasband WS. ImageJ, 1.34s edn. 1997. –2009. Bethesda, Maryland, USA: U. S. National Institutes of Health [WWW document]. URL http://rsb.info.nih.gov/ij/
• Robinson P, Lebel M, Cyr M. Dopamine D1 receptor-mediated aggregation of N-terminal fragments of mutant huntingtin and cell death in a neuroblastoma cell line.Neuroscience. 2008;153:762–772. [PubMed]
• Ross CA, Margolis RL. Huntington’s disease. Clin Neurosci Res. 2001;1:142–152.
• Saidak Z, Blake-Palmer K, Hay DL, Northup JK, Glass M. Differential activation of G-proteins by mu-opioid receptor agonists. Br J Pharmacol. 2006;147:671–680.[PMC free article] [PubMed]
• Sanchez I, Mahlke C, Yuan J. Pivotal role of oligomerization in expanded polyglutamine neurodegenerative disorders.Nature. 2003;421:373–379. [PubMed]
• Scherzinger E, Sittler A, Schweiger K, Heiser V, Lurz R, Hasenbank R, et al. Self-assembly of polyglutamine-containing huntingtin fragments into amyloid-like fibrils: implications for Huntington’s disease pathology. Proc Natl Acad Sci USA. 1999;96:4604–4609.[PMC free article] [PubMed]
• Scotter EL, Narayan P, Glass M, Dragunow M. High throughput quantification of mutant huntingtin aggregates. J Neurosci Methods. 2008;171:174–179.[PubMed]
• Spires TL, Grote HE, Varshney NK, Cordery PM, van Dellen A, Blakemore C, et al. Environmental enrichment rescues protein deficits in a mouse model of Huntington’s disease, indicating a possible disease mechanism. J Neurosci.2004;24:2270–2276. [PubMed]
• Steffan JS, Bodai L, Pallos J, Poelman M, McCampbell A, Apostol BL, et al. Histone deacetylase inhibitors arrest polyglutamine-dependent neurodegeneration in Drosophila. Nature. 2001;413:739–743. [PubMed]
• Suhr ST, Gil EB, Senut MC, Gage FH. High level transactivation by a modified Bombyx ecdysone receptor in mammalian cells without exogenous retinoid X receptor. Proc Natl Acad Sci USA. 1998;95:7999–8004.[PMC free article] [PubMed]
• Takashima A, Koike T. Relationship between dopamine content and its secretion in PC12 cells as a function of cell growth. Biochim Biophys Acta. 1985;847:101–107.[PubMed]
• Wachtel SR, White FJ. The dopamine D1 receptor antagonist SCH 23390 can exert D1 agonist-like effects on rat nucleus accumbens neurons. Neurosci Lett. 1995;199:13–16. [PubMed]
• Walsh DM, Selkoe DJ. Oligomers on the brain: the emerging role of soluble protein aggregates in neurodegeneration.Protein Pept Lett. 2004;11:213–228. [PubMed]
• Watase K, Weeber EJ, Xu B, Antalffy B, Yuva-Paylor L, Hashimoto K, et al. A long CAG repeat in the mouse Sca1 locus replicates SCA1 features and reveals the impact of protein solubility on selective neurodegeneration. Neuron.2002;34:905–919. [PubMed]