Clioquinol down-regulates mutant huntingtin expression in vitro and mitigates pathology in a Huntington's disease mouse model

Methods

R6/2 Transgenic and Wild-Type Littermate Mice. A colony of transgenic mice strain R6/2 that is heterozygous for huntingtin exon 1 (containing 145 CAG repeats) was maintained, with founders originating from The Jackson Laboratory. Male transgenic R6/2 mice were bred with background strain (B6CBAF1/J) females. Genotyping was done according to Hockly et al. (38).

CQ in water was orally administered to mice at 30 mg/kg per day starting at 3 weeks of age and continued until the mice were deceased or killed for tissues. Just before administration, CQ (≈6 mg) was mixed with 1 ml of drinking water in a 1.5-ml microcentrifuge tube. The tube was mixed by agitation, 0.1 ml of the dispersed CQ solution taken up into a 1-cc syringe attached to a feeding needle, and the dose delivered by oral gavage (similar to ref. 39). Littermate R6/2 mice were gavaged with water vehicle alone. At the end of CQ treatment, mice were anesthetized and transcardially perfused with PBS followed by 4% paraformaldehyde. Blood glucose levels were determined after at least 8 h of fasting in 11-week-old mice (n = 6 for wild type, n = 4 for vehicle-treated R6/2, n = 6 for CQ-treated R6/2 littermates) using a 2300 STAT Plus glucose analyzer (Yellow Springs Instruments). Fasting blood insulin levels in animals fasted for 8 h were measured by using an Ultra Sensitive Rat Insulin ELISA Kit (Crystal Chem, Downers Grove, IL) according to the manufacturer's instructions. All procedures were performed in accordance with protocols approved by the San Francisco Veterans Administration Medical Center Animal Care Committee.

Cell Culture, Transfection, and CQ Treatment. polyQ-htt-GFP fusion constructs, in which the polyQ region is encoded by CAA- and CAG-containing tandem repeats (CAACAGCAACAACAGCAG)_n, were obtained from the Hereditary Disease Foundation, Santa Monica, CA. PC12 and HEK293 cells were maintained in MEM supplemented with 10% FBS, switched to Opti-MEM without serum for 6 h during transfection using Lipofectamine 2000 (Invitrogen), and returned to serum-containing medium thereafter (up to 48 h). Cells were exposed to CQ (Sigma) at 0.5–8 μM for 30 min before and after (but not during) transfection. Control cultures received equivalent amounts of DMSO-containing vehicle (≤0.1% DMSO, final concentration).

Western Blotting. Cells were lysed in PBS buffer containing 1% Triton X-100 and protease inhibitor mixture (Roche Applied Science, Indianapolis) and passed through a 28-gauge needle to ensure the release of Htt inclusions from the nuclei. Whole mouse brains were homogenized in 50 mM Tris·HCl (pH 7.4)/150 mM NaCl/1% Triton X-100/complete protease inhibitor mixture. Brain homogenate was precleared by centrifuging at 7,500 × g for 90 s at 4°C. The supernatant was passed through a 28-gauge needle and boiled in SDS sample buffer for 10 min. Fifty micrograms of total protein was loaded onto 10% SDS gel containing a 4% stacking gel. Monoclonal antibodies anti-polyQ (1:2,000), anti-GAPDH (1:10,000), anti-GFP (1:4,000), and anti-Htt (EM48) in 1:1,000 dilution were used for immunoblotting. All Western blot experiments were performed at least three times, and band intensities were averaged.

Northern Analysis. TRIzol reagent (Invitrogen) was used to isolate total RNA. Ten micrograms of total RNA was electrophoresed through 1.25% agarose gel and transferred onto Hybond n + membrane. The membrane was baked in an 80°C oven for 15 min and UV-crosslinked. Exon 1 of htt containing 103 CAA/CAG repeats was used as a probe for Northern blot detection according to the AlkPhos Direct Labeling and Detection System (Amersham Pharmacia Biosciences). One hundred nanograms of the probe was hybridized to the membrane in a 60°C rotating hybridization oven overnight. The membrane was scanned by using the Typhoon 9410 digital scanner (Amersham Pharmacia Biosciences) to yield the signal image. The experiment was performed three times, and band intensities were averaged.

Cell Viability and Section Measurement. Plasma membrane integrity was evaluated by using propidium iodide at 20 μg/ml final concentration. Approximately 200 cells transfected with the Htt^exon1-Q₁₀₃-egfp plasmid were counted for each treatment group 48 h after transfection. Cells were analyzed at that time, because expanded protein accumulation had reached a plateau, and 90% of cell death occurred by this time point. The experiment was repeated twice, and counts were averaged. Counts of transfected cells were averaged from three independent randomly selected fields. imagej 1.32j (National Institutes of Health, Bethesda) was used to measure the size of the lateral ventricles (n ≥ 6 for each group).

Pulse–Chase Assay. HEK293 cells at 90% confluence on 100-mm² plates coated with poly(d-lysine) (0.1 μg/μl) were transfected for 4 h in Opti-MEM medium before replacement of the medium with MEM supplemented with 10% FBS. After another 5-h incubation, cells were trypsinized and replated in two wells of a six-well plate, then incubated for another 14 h at 37°C. The medium was replaced with methionine-free MEM medium supplemented with 5% FBS and cells incubated for 1.5 h with 4 μM of CQ or DMSO vehicle added during the last 30 min of incubation. One mCi (1 Ci = 37 GBq) of [³⁵S]methionine (Amersham Pharmacia Biosciences) in 12 ml of methionine-free MEM was prepared as a pulse-stock medium. Cells were pulsed for 10 min by replacing the medium with 1 ml of the pulse-stock medium. At different time points, cells were washed twice with ice-cold PBS and lysed in 1 ml of immunoprecipitation lysis buffer (50 mM Tris·HCl, pH 7.5/150 mM NaCl/1% Nonidet P-40/0.5% sodium deoxycholate/one tablet minicomplete protease inhibitor/10 ml of buffer). Cells were harvested by scraping, placed in a 1.5-ml microcentrifuge tube, and lysed with constant agitation at 4°C for at least 30 min before immunoprecipitation of the Htt^exon1-Q₁₀₃-egfp protein using anti-polyQ antibody and protein G agarose beads (Roche Applied Science). Assays were performed at least three times.

Brain Cryosectioning and Immunohistochemistry. Perfused mouse brains were fixed in 4% paraformaldehyde for 1 h at room temperature and stored in 30% sucrose overnight at 4°C. Brain tissue was frozen at -70°C for at least 30 min before cryosectioning at 40-μm thickness by using a Leica CM1850 cryostat (McBain Instruments, Chatsworth, CA). Eleven-week-old mouse brains (n > 5 for each treatment group) sections were incubated at room temperature for 1 h in blocking buffer containing 2% goat serum, 0.2% Triton X-100, and 0.1% BSA in PBS. Then, EM48 anti-Htt monoclonal antibody (Chemicon) at a 1:100 dilution was incubated with the tissue at 4°C overnight. Colorimetric signal detection was performed by using diaminobenzidine substrate and the ABC kit (Vector Laboratories), according to manufacturer's protocol.

Behavioral Tests. Rotarod performance (n ≥ 6 for each treatment group) was assessed as described by Hockly et al. (38) with the modification that the SDI Rota-Rod (San Diego Instruments, San Diego) was set to linearly increase in speed to 40 rpm in 4-min ramp time. For testing clasping, 10-week-old mice (n = 8 for each treatment group) were suspended by the tail for 30 s, and the foot-clasping time was scored such that a 0- to 5-s clasping duration was given a score of 1, 5–10 s a score of 2, and >10 s a score of 3.

Results

CQ Selectively Down-Regulates polyQ Protein Levels and Relieves Cell Death in Vitro. Transiently transfected PC12 cells expressing Htt exon-1 containing a 103-residue polyQ-encoding region fused to enhanced GFP gene (Htt^exon1-Q₁₀₃-egfp) produced a less fluorescent product when treated with CQ at 4 μM, whereas Htt^exon1-Q₂₅–egfp and egfp alone were unaffected at 48 h posttransfection (Fig. 1). Similar results were obtained with HEK293 cells. Differential expression of Htt^exon1-Q₁₀₃-egfp signal could be detected visually at 6 h posttransfection (not shown) and was demonstrable on Western blotting of HEK293 cell extracts beginning at 9 h posttransfection (Fig. 2A). Concordant with this reduction in mutant protein levels, cell survival was improved by CQ in a dose-dependent fashion with a half-maximal effect at 1–2 μM (Fig. 2B). Cell death began to increase at 8 μM, likely due to dose-dependent toxic effects of CQ. Western blotting of cell extracts at 48 h showed a dose-dependent decrease in Htt^exon1-Q₁₀₃-egfp levels beginning at 0.5 μM CQ and reaching 3.2-fold at 2 μM, which was not observed with egfp alone (Fig. 2C). The lack of effects on Htti^exon1-Q₂₅–egfp and egfp alone suggested that CQ was not causing a generalized decrease in protein synthesis or a reduction in transfection efficiency or affecting GFP maturation.

CQ Does Not Affect polyQ mRNA Abundance or Protein Degradation Rate. In examining potential mechanisms for the effects of CQ on Htt accumulation, we considered the possibilities that CQ might act to inhibit Htt aggregation or directly influence Htt synthesis or degradation. Because the early accumulation of Htt was inhibited by CQ, we focused on the aspects of Htt synthesis and turnover. To determine whether CQ might down-regulate polyQ-Htt expression by reducing mRNA levels, we examined the levels of Htt^exon1-Q₁₀₃-egfp mRNA present 48 h posttransfection by Northern analysis (Fig. 3A). CQ had no significant effect on mRNA levels at concentrations that markedly diminished protein levels (Fig. 2B). In addition, no differences in mRNA content were observed by using RT-PCR at earlier time points (6, 9, 12, and 18 h; data not shown). Because Htt-exon 1 is subject to ubiquitination and proteasomal degradation (40, 41), we examined the possibility that CQ might work by increasing proteasome-mediated turnover. If this were so, we would expect that the effects of CQ on Htt^exon1-Q₁₀₃-egfp expression might be reversed by a proteasome inhibitor. Htt^exon1-Q₁₀₃-egfp-transfected HEK293 cells were treated with the panproteasome inhibitor MG132 in the presence or absence of CQ, visualized (compare Figs. 1 and 3B) and blotted (Fig. 3C). Suppression of Htt^exon1-Q₁₀₃-egfp protein levels by CQ was unaffected by MG132 treatment. To further support this result and assess the possibility degradation through alternative paths, pulse–chase experiments were performed. No significant differences in the rates of Htt^exon1-Q₁₀₃-egfp degradation were observed in the presence of CQ (Fig. 3D), although there may have been a small trend toward decreased degradation at longer time points in CQ-treated cells.

CQ Inhibits Accumulation of Aggregated Htt in Vivo. We next asked whether CQ would also reduce mutant Htt accumulation in vivo by using the HD transgenic mouse model R6/2. After 8 weeks of daily oral CQ treatment, R6/2 mice had a 3.9-fold reduction in the quantity of aggregated Htt, as indicated by Western blotting of whole-brain extracts (Fig. 4A). To examine the anatomical distribution of the changes in expression, immunohistochemistry with the anti-Htt aggregate antibody EM48 was performed, revealing that CQ treatment inhibited the accumulation of aggregated Htt in both the striatum and motor cortex (Fig. 4B).

CQ Treatment Improves HD Pathology and Enhances Motor Function and Survival in R6/2 Mice. The major sites of cell loss in HD brain are the striatum and the cortex (42, 43) and in R6/2 mice, striatal atrophy is reflected by expansion of the size of the lateral ventricles. Compared with the sham-treatment group, the treated group had a 55% reduction of lateral ventricle area in standardized sections (Fig. 5). Foot clasping, a dystonic posturing of the hind limbs on suspension by the tail (Fig. 6A), is a characteristic behavior of the later stages of the disease in R6/2. CQ treatment decreased the average clasping score at 10 weeks from 2.5 (of 3 maximum) to 1.3, a 48% reduction (Fig. 6B). The treated group also scored significantly better on rotarod performance, a test of agility and muscle coordination, between 9 and 11 weeks of age (Fig. 6C). The initial trend to a therapeutic effect of CQ on rotarod performance in R6/2 mice began when symptoms such as tremors and weight loss developed. With respect to weight changes, without CQ treatment, weight gain in R6/2 diverged from the wild-type curve and reached a plateau between the fifth and seventh weeks, which was followed by a steady weight loss (Fig. 6D). With CQ treatment, weight gain in R6/2 was comparable to that of the vehicle-treated group up to the seventh week of age but trended toward a gain thereafter, with weights significantly greater than untreated animals observed in the tenth and eleventh weeks (Fig. 6D). The average lifespan of untreated animals in this series was ≈76 d, within the originally described range (10–13 weeks) of R6/2 lifespan (44). CQ treatment extended the lifespan to an average of 92 d, a 20% increase in longevity over the littermate control group (Fig. 6E). This falls within the range of survival improvement (7–31%) seen with several therapeutic agents tested in this mouse model [reviewed in ref. 24).

CQ Mitigates HD of the Endocrine Pancreas. R6/2 mice develop diabetes mellitus due to dysregulation of insulin gene expression associated with sequestration of transcription factors by mutant Htt, which has suggested a mechanism for the elevated incidence of diabetes in individuals with HD (45–47). We found that after 8 weeks of CQ treatment, 11-week-old R6/2 mice had normal fasting blood glucose levels (4.1 ± 0.3 mM), comparable to that of wild-type littermate controls (5.2 ± 0.8 mM). R6/2 sham-treated animals had significantly higher fasting blood glucose levels (7.6 ± 0.4), characteristic of diabetes (Table 1). Consistent with the blood glucose level was a reciprocal change in blood insulin levels. The sham group had a significantly lower fasting blood insulin level (166 ± 5.13 pg/ml) than the treated group (415 ± 54.60 pg/ml) (Table 1), which was similar to that of wild-type mice in these experiments and in those reported previously (24).

Discussion

We have found that CQ decreases polyQ-expanded protein accumulation and promotes cell survival in an in vitro model of HD and decreases symptoms and increases lifespan in HD model animals. To begin to elucidate the mechanism of action of CQ in these models, we investigated the metabolism of the mutant protein by examining changes in mRNA and protein levels and degradation on exposure to the drug. Northern analysis suggested that mRNA levels were not affected by CQ. Early rates of mutant protein accumulation were reduced, whereas inhibition of proteasomal degradation had little effect; moreover, pulse–chase experiments suggested that overall rates of degradation were unchanged or slowed. These findings are consistent with models in which CQ interacts directly or indirectly with mRNA, RNA-binding proteins, or the translation apparatus, but they do not completely rule out other effects on RNA metabolism or posttranslational events, such as aggregation. Moreover, the relevance of the in vitro findings to the effects of CQ in transgenic mice remains unclear, and other compounds such as creatine and caspase inhibitors (reviewed in ref. 24) decrease aggregate accumulation in such models as well.

As a metal-binding compound, CQ may affect numerous cellular processes. Iron, zinc, copper, manganese, magnesium, and other metals play important roles in the structure and function of DNA, RNA, and some proteins. Of note, CQ is a relatively weak chelator and will not significantly recruit tightly bound metals (48), but under various pathologic conditions, metals may be released, or their complexes loosened, allowing effects of even low-affinity chelators. Chelation of iron has been implicated as a mechanism of CQ inhibition of MPTP toxicity (49) and has been suggested to be broadly neuroprotective (36). Interestingly, Htt may be indirectly transcriptionally regulated by iron or the redox state (50), although no such regulatory element sequences are present in the constructs used in our in vitro experiments. Copper chelation can reduce the production of hydrogen peroxide formed by copper–Aβ peptide and perhaps other complexes (34), and accumulation of copper is a feature of the affected regions in HD (27); however, its form and reactivity have not been characterized. Zinc is a structural component of numerous proteins involved in RNA and protein synthesis, and its binding is regulated by levels of ROS in the cell (51). Thus, through indirect antioxidant effects or direct effects on zinc trafficking, CQ may have significant effects on cellular metabolism. In addition, independent of chelation, quinols similar to CQ have been found to bind to RNA and modulate RNA–protein interactions (52). Thus, CQ might influence mutant protein accumulation and cell survival through chelation of iron, copper, zinc, or other ion or through direct interactions with RNA or protein, by stabilizing secondary structures or interfering with RNA–protein interactions or by other effects.

Conclusion

This work suggests that CQ may have therapeutic value for HD and raises the possibility that the elucidation of its mechanism of action will identify new specific targets for the mitigation of expanded polyQ toxicity.