|Home | About | Journals | Submit | Contact Us | Français|
Controversies surround the usefulness of Coenzyme Q10 (CoQ10) in Huntington’s disease (HD), an autosomal dominant, fatal, neurodegenerative disease with no cure or disease modifying treatment. CoQ10, an endogenous substrate for electron transport and an anti-oxidant, has been shown in some but not all studies to improve symptoms and survival in mouse models of HD. Previous studies have been conducted in fast progressing models that better mimic the juvenile forms of HD than the much more common middle-age onset form, possibly accounting for mixed results. Establishing the usefulness of CoQ10 to alter HD disease course in a model that better recapitulates the progressive features of the human disorder is important because clinical trials of CoQ10, which is safe and well tolerated, are being planned in patients. The CAG140 knock-in (KI) mouse model of HD in which an expanded (approximately 120) CAG repeat is inserted in the mouse gene provides a model of the mutation in the proper genomic and protein context. These mice display progressive motor, cognitive and emotional anomalies, transcriptional disturbances and late striatal degeneration. Homozygote mutant CAG140 KI mice and wild-type littermates were fed CoQ10 (0.2%, 0.6%) in chow, and behavioral and pathological markers of disease were examined. CoQ10 improved early behavioral deficits and normalized some transcriptional deficits without altering huntingtin aggregates in striatum. The lower dose (0.2%) was more beneficial than 0.6%. Similar to previous studies, this low dose also induced deleterious effects in open field and rotarod in WT mice, however these effects are of unclear clinical significance in view of the excellent safety profile of CoQ10 in humans. These data confirm that CoQ10 may be beneficial in HD but suggest that maximum benefit may be observed when treatment is begun at early stages of the disease and that dosage may be critical.
Much work has been devoted over the last decade in performing clinical trials for potential neuroprotective therapies for Huntington’s disease (HD), a progressive neurological disorder with no cure or effective treatment. The disease is inherited in an autosomal dominant manner and is caused by a single mutation in the N terminus of the huntingtin (htt) gene (Huntington’s Disease Collaborative Research Group, 1993). Many functions have been attributed to htt and it is thought that the mutation causes both a toxic gain in function, and a loss of function of the wild type (WT) protein (Cattaneo et al., 2005, Lee et al., 2007, Miller et al., 2010).
CoenzymeQ10 (CoQ10), also known as ubiquinone, is found in nearly all cell membranes, especially in the inner mitochondrial membrane where it acts as an anti-oxidant, in addition to its role in electron transport (Kwong et al., 2002, Chaturvedi and Beal, 2008). These properties, and evidence for oxidative stress in several neurological disorders, have led to the use of CoQ as a potential treatment for many disorders including amyotrophic lateral sclerosis (Ferrante et al., 2005) and Parkinson’s disease (Shults et al., 1998, Shults et al., 2004). In HD, CoQ10 administration increases patient CoQ10 serum levels significantly when compared to untreated patients, and levels are no longer different to controls (Andrich et al., 2004) and CoQ also decreases cortical lactate (Koroshetz et al., 1997). In a large clinical trial, CoQ10 treatment (300mg twice daily) slowed the decline of total functional capacity in HD patients; however the improvement was non-significant (Huntington Study Group, 2001). Importantly, much higher doses of CoQ10 are safe and well tolerated in HD (Feigin et al., 1996, Huntington Study Group Pre2CARE Investigators, 2010), opening the path for further investigation in humans. Notably, recent studies have highlighted the importance of titrating therapies to disease duration (Okamoto et al., 2009).
Several studies have reported beneficial effects of CoQ10 on behavior and pathology in mouse models of HD (Schilling et al., 2001, Ferrante et al., 2002, Schilling et al., 2004, Smith et al., 2006, Stack et al., 2006, Yang et al., 2009). However, a recent study in the fast progressing murine model of HD, the R6/2 mouse, has failed to confirm these data when testing mice in an enriched environment and using actual end of life as an additional endpoint (Menalled et al., 2010). Together with the minimal effect in the early clinical trials in patients, these data indicate that more studies are required to better define the context in which CoQ might be beneficial in HD. In particular, it is possible that CoQ may improve only some aspects of the disorder, or be more effective at early stages, before multiple pathological processes come into play. This is difficult to assess in fast progressing models of the disease, where extensive pathology and dysfunction are observed early on and animals progress to death in a few months (Hickey et al., 2005). Therefore, we have re-examined CoQ effects, using a regimen similar to that used in the negative study in R6/2 mice, in a slowly progressive model of HD, the CAG140 KI mouse, which expresses the full length protein in the proper genomic context and may better reproduce human pathology (Menalled et al., 2003, Hickey et al., 2008).
CAG140 homozygote KI mice present behavioral, neuropathological, electrophysiological and molecular changes that emerge at 1–2 months of age and progress slowly towards loss of striatal neurons, a canonical feature of HD, at approximately 2 years of age (Menalled et al., 2003, Cui et al., 2006, Hickey et al., 2008, Phan et al., 2009, Simmons et al., 2009, Cummings et al., 2010, Hickey et al., in press). Specifically, these mice first show behavioral deficits at 1 month of age, with nuclear staining and microaggregates of htt appearing in striatum and other brain regions by 2 months of age. By 4m, the mice exhibit changes in striatal transcripts for several receptors and neuropeptides and nuclear inclusions form in striatal and other neurons. At 12 months KI mice show striatal atrophy (Lerner et al., submitted), profound loss of DARPP-32 (by optical density) and cortical gliosis, with striatal gliosis and striatal neuronal loss by 2 years (Hickey et al. 2008).
Here, we examined the effects of 0.2% and 0.6% CoQ10 in chow in CAG140 KI mice, at an early stage in disease, with the aim of developing therapeutics for treating early disease and neuronal dysfunction. Thus, we have examined pathological endpoints such as transcription and mutant protein aggregation rather than neurodegeneration. CoQ10 was administered in chow, from conception to 4.5 months and we monitored the effects of the treatment on behavioral deficits at 1–4 months of age as well as mutant htt aggregates and striatal transcripts at 4.5 months. Both doses markedly improved behavioral deficits, with the lower dose (0.2%) providing more extensive benefits. In addition, both doses normalized some of the striatal transcriptional deficits induced by mutated htt. These data support the potential usefulness of CoQ10 in HD but stress the need for early treatment to achieve benefits.
All experiments were performed in accordance with the US Public Health Service Guide for Care and Use of Laboratory Animals and were approved by the Institutional Animal Care and Use Committee at UCLA. CAG140 mice were bred in house from Het(erozygote) × Het pairings (non-sibling pairs). Resultant WT and KI mice were used for behavioral analyses. Breeding mice (5 males, 10 females per group) were fed normal chow (NIH-31 number 7013, Harlan Teklad, Indianapolis, IN) or the same chow containing 0.2% or 0.6% CoQ10 (Health Wright Products Inc, Clackamas, OR) for at least 1 week prior to breeding and throughout breeding. Breeding pairs were checked daily for litters, and the number of pups in each litter was noted at birth and at weaning. In some cases, the number of pups increased from the number at birth and this is due to incomplete parturition at the time that the litter was first counted. Five rounds of breeding were used to generate mice for this study (litter size was not recorded in the first round). Progeny were genotyped and weaned by 21 days and CoQ10 or control chow was continuously fed to the mice until 4.5 months of age. Body weight was recorded weekly. Tail samples were taken at the end of the experiment in order to confirm genotyping and to measure CAG repeat length. The mean CAG repeat length (Laragen Inc, Culver City, CA, using an ABI 3730 sequencer and Genemapper software) of KI and Het mice in this trial were control: 120 ± 2; 0.2%: 120 ± 2; 0.6%: 117 ± 1 (subset of KI and HET mice for each treatment analyzed, control, N=22; 0.2%, N=33; 0.6%, N=30; effect of treatment F(2,82)=0.9, ns). Mice were housed in a temperature and humidity controlled room, on a reverse light-dark cycle, with water and food available ad lib.
All testing and analysis was carried out blinded to genotype and food. WT and KI mice were habituated to the testing rooms for 15–20 min prior to all testing.
For open field rearing activity mice were placed individually in the centre of the open field (Truscan, Coulbourn Instruments, Allentown, PA) and rearing activity during 15 min was video recorded for later analysis in bins of 5 min each. Mice were tested in 2nd through 4th h of the dark phase, under a red light (25W). Data for locomotion (distance covered during the 15 min) were also recorded but were unavailable for 8 mice (rearing was quantified from video analysis for these mice).
For climbing, mice were placed on the floor of a wire cylinder (3¾″ height × 4″ diameter) for 5 min and their behavior recorded for later video analysis. Climbing was recorded when all four paws of the mouse were off the floor of the testing bench and on the wall of the wire cylinder (Hickey et al., 2008). Climbing was tested in the light phase, with testing completed at least 1 h before the start of the dark phase.
For pole performance mice were placed head facing upwards on a vertical pole and trained to turn around and descend to the bottom of the pole as described previously (Hickey et al., 2008). Mice were given 2 trials/day for 3 days. On the 4th day (testing day), the time taken to turn and to descend was recorded over 5 successive trials. Testing was carried out half way through the dark cycle, under a dim white light.
For rotarod performance mice were trained to walk on an accelerating rotating axle covered with smooth rubber (smooth axle) (Ugo Basile, Varese, Italy), as described previously, with some modifications (Hickey et al., 2008). Briefly, mice were given 3 trials/day over 4 days (5–40 rpm over 10 min, with approximately 30 min between successive trials) and the latency to fall was measured. On the 5th day the mice received 1 trial at each of 10, 20 and 30 rpm (smooth axle) and on the final (6th) day mice received 1 trial at each of 20, 24, 30, 36 rpm using a grooved axle. Testing was carried out approximately half way through the dark phase, under a red light (25W). Any mice that clung to the axle for 3 successive rotations were removed and the time of removal recorded and used as the latency.
At the end of treatment half of the mice were anesthetised and perfused with 4% paraformaldehyde and 0.5% glutaraldehyde, their brains removed, post fixed for 6–8 h in 4% PFA, cryoprotected in 30% sucrose and frozen for later use. The other mice were quickly decapitated and their brains frozen in powdered dry ice.
Saggital sections from the perfused brains were collected at the levels corresponding to 1.32 mm and 2.28 mm lateral of the midline (Paxinos and Franklin, 2001) for analysis. Tissue sections (35 μm) were stained with EM48 (polyclonal antibody, a generous gift of Dr X. J. Li (Emory University) as described in Menalled et al. (2003). Briefly, sections were washed in 0.01 M PBS and then endogenous peroxidases inactivated by incubating in 1% H2O2 and 0.5% Triton X-100 in PBS, for 20 min. Non-specific binding sites were then blocked by incubating sections for 30 min at room temperature in PBS containing 3% bovine serum albumin (BSA) and 2% normal goat serum (NGS). The primary antibody was diluted (1:300) in PBS containing 3% BSA, 2% NGS, 0.08% sodium azide, and 0.2% Triton X-100 and sections were incubated overnight at room temperature. The following day the sections were washed in PBS and then incubated in biotinylated goat anti-rabbit antibody (1:200; Vector ABC Elite; Vector, Burlingame, CA) for 2 h at room temperature, washed and then reacted with avidin-biotin complex (Vector ABC Elite) in PBS containing 0.2% Triton X-100 for 2 h. Immunoreactivity was visualized by incubation in 0.03% 3–3-diaminobenzidine tetrahydrochloride (Sigma, St. Louis, MO) and 0.0006% H2O2 in 0.05 M Tris buffer, pH 7.6. After rinses in Tris buffer, the sections were dehydrated, defatted, and mounted with Eukitt (Calibrated Instruments, Hawthorne, NY). Control sections, processed in parallel, were incubated in the absence of the primary or secondary antibodies. No staining was noted in control sections (data not shown). The number of 1) stained nuclei, 2) stained nuclei containing microaggregates, 3) stained nuclei containing inclusions, and 4) neuropil aggregates were counted in both sections using Stereo Investigator 5.00 software (Microbrightfield, Colchester, VT). Contours of the striatum were drawn using the software, at 5x magnification. The number of each item was then counted at 100x using a 1.4 NA lens and 1.4 NA oil condenser, with a DVC real-time digital camera and the average number of each item per counting frame (20 μm2) per section per animal was used to generate group means. The counting frames were distributed using a sampling grid of 200 × 200 μm. Densities for each type of aggregated mutant htt were calculated per level and used to generate group means.
CAG140 WT and KI mice were quickly decapitated and their brains frozen in powdered dry ice. Total RNA was purified from one striatum of fresh frozen tissue using QiaGen RNeasy mini kit (N = 3 for WT control, N = 4 for others) as described by the manufacturer (QIAGEN Sciences, Maryland, USA). During the RNA extraction procedure DNase I-treatment was performed to remove contaminating genomic DNA (RNase-Free DNase Set, QIAGEN, Hilden, Germany). For cDNA synthesis with oligo dT primers, the Invitrogen ThermoScript RT-PCR System (Invitrogen, Carlsbad, CA, USA) was used. The cDNA was then analyzed by quantitative real time PCR using a Roche LightCycler 480 (UCLA Genotyping and Sequencing Core). PCR was performed using LightCycler FastStart DNA Master plus SYBR Green 1 kit (Roche Diagnostics, Mannheim, Germany). Each assay included: 1) a standard curve of five serial dilution points of control cDNA (mouse EST clone of appropriate fragment of the gene of interest (Invitrogen, CA)), 2) sample cDNA, 3) no template control. All samples were run in triplicate. The PCR cycling parameters were: 95°C –5 min (1cyc); 95°C –10 sec, 65°C –10 sec, 72°C –10 sec (40 cyc). A dissociation protocol was established at the end of each run to verify the presence of a single product. Relative expression of genes of interest was determined using the Pfaffl method (Pfaffl, 2001) normalized to the average expression levels of two housekeeping genes Eif4a and Atp5b (Benn et al., 2008). PCR efficiencies of each primer pair were calculated from standard curve analysis and incorporated into relative quantification calculations. Designed primers yielded a product of about 200 bp for each gene, and sequences are shown in Table 1. The genes of interest were D1, D2, CB1, Enk, DARPP-32 and SubP. Most of these transcripts have been shown to be reduced in striatum in early grade HD patient tissue (Glass et al., 2000). Further, we had previously shown that they were altered (with the exception of SubP, which is preserved in HD) in the CAG140 KI mouse model of HD (Hickey et al., in press). Moreover, DARPP-32 protein is reduced in 1-yr-old CAG140 KI mice (Hickey et al., 2008); thus, protection of the DARPP-32 transcript could prevent this later loss in protein. Finally, CB1 receptors are thought to be important for functional improvement in HD models (Blazquez et al., 2011) and as such provide an important functional outcome measurement. SubP provides an important control, as we have previously shown that it is unchanged despite profound reduction in Enk levels, in both the CAG140 and CAG94 KI mouse models of HD (Hickey et al., in press; Menalled et al., 2000).
GBstat (V8.0) and SAS (V 9.1) were used for statistical analyses. Body weights were analysed using mixed generalized linear model ANOVAs in SAS using Bonferroni’s adjusted Student’s t-tests for post hoc analysis. Other ANOVAs were performed using GBstat, followed by Fisher’s LSD for post hoc analysis. Single timepoint data were compared using Student’s t tests (where data were homoscedastic) or Mann Whitney U-tests (where data were heteroscedastic). The Chi square independence test was used to compare sex of offspring. A critical value of p = 0.05 was used for all analyses.
Mice were fed control or the same dose of CoQ used in a recent study in R6/2 mice (Menalled et al., 2010): 0.2% or 0.6% CoQ10 in chow, however, here 0.6% chow was not supplemented with gamma-cyclodextrin. In the present study CoQ was administered from conception to maximize exposure to the dietary supplement at the earliest stages of the disease. Therefore, we monitored breeding success to ensure normal fecundity. There was no difference between treatment groups in size of litters at birth and weaning (Table 2; no effect of treatment F(2,54)=0.24, ns; no interaction of treatment with age F(2,54)=0.8, ns). The percentage of pups brought through to weaning age was also similar between groups (control, 58.4±14.6%; 0.2%, 75.4±17.2; 0.6%, 82.4±10.7; effect of treatment F(2,30)=0.65, ns), as was the proportion of genders and genotypes represented within each treatment group (males-females group sizes in WT, HET and KI groups: control 7–9; 0.2% 8–14; 0.6% 6–15). Chi square independence tests were used to compare number of WTs and KIs within treatment groups (control ns; 0.2% ns, 0.6%, ns). Pups from each group gained weight, and over the course of the study there was no significant difference in body weight profiles of control-treated mice compared to CoQ10-treated mice (Fig. 1a–d, body weights of WT and KI mice were compared, HETs were also used for analysis of body weights, but were not used for subsequent behavioral testing, males: genotype × treatment × timepoint F(64,1027)=0.84, ns; females: genotype × treatment × timepoint F(64,1020)=0.6, ns). These data are in agreement with the safety profile of CoQ10, which is an over-the-counter dietary supplement (Bhagavan and Chopra, 2006). Thus, treatment with CoQ10 from conception did not have any deleterious effects on the pups or success of breeding.
Mice were tested for rearing and locomotor activity in the open field at 1 month of age. Control-fed KI mice reared less than control-fed WT mice (Fig. 2a). Demonstrating a beneficial effect, 0.2% CoQ10 rescued this reduced rearing in KI mice (Fig. 2a, gray stippled bars), while 0.6% (black stippled bars) had no effect. Interestingly, the effect of CoQ10 was opposite in WT mice, with 0.2%-fed WT mice rearing less (gray bars) than control-fed WT mice (Fig. 2a), while 0.6% (black bars) had no effect (genotype × treatment interaction F(2,106)=3.2, p<0.05). The lower dose of CoQ10 also improved initial exploration of the open field, since the reduced locomotion during the first 5 min was rescued in 0.2%-treated KI mice (Fig. 2b, effect of timebin F(2,198)=127.2, p<0.0001). Mice were tested for spontaneous climbing activity two weeks later, at 1.5 months, and as previously shown (Hickey et al., 2008), in control-fed groups, KI mice climbed less than WT mice (Fig. 3; overall effect of genotype F(1,107)=9.7, p<0.01). There was no significant difference in climbing between KI groups (Fig. 3), however CoQ10 0.2% abolished the decreased climbing in KI versus control-treated WT (Fig. 3). At 4 months of age, mice were tested for performance on the pole task and rotarod. Both doses of CoQ10 rescued the impaired pole performance of KI mice (Fig. 4a, interaction of genotype and treatment, F(2,105)=3.2, p<0.05) and neither dose had any effect on WT performance.
For rotarod analysis, mice were trained for 4 days using an accelerating protocol and smooth axel (4–40 rpm over 10 min, 3 trials per day), followed by 1 day of constant speeds using a smooth axel (10, 20, 30 rpm, 1 trial per speed) and a final day of constant speeds with a grooved axel (20, 24, 30 and 36 rpm, 1 trial per speed). No deficit was detected during training in KI performance with either diet or in WT fed CoQ. As demonstrated previously (Hickey et al., 2008), control-fed KI mice showed a subtle deficit in rotarod performance on the 5th day with the smooth axel and constant speeds, and on the 6th day with the grooved axel and constant speeds when compared with control WTs. CoQ10 0.2% rescued rotarod performance of KI mice on the 5th day (Fig. 4b, genotype × treatment × trial interaction F(4,210)=2.9, p<0.05) and on day 6, (eg 24rpm: WT control 327±75s, 0.2% 133±42s, 0.6% 175±63s, KI control, 220 ± 64, 0.2% 296±56, 0.6% 175±49; genotype × treatment effect, F(2,105)=3.6, p<0.03, KI control and KI 0.6% versus WT control p<0.05; 0.2% WT and 0.6% WT versus control WT, p<0.01; 0.2% KI versus WT control, ns). In stark contrast, both doses of CoQ10 impaired performance on the rotarod in WT mice, a finding also shown by Schilling et al. (2004) while Menalled et al. (2010) showed a tendency for impaired performance in CoQ10-treated R6/2 mice. These data suggest that CoQ10 may affect WT mice differently than KI mice. Thus, only at the most difficult part of the rotarod task was the deficit seen in control KI mice, with a beneficial effect noted in the KI mice treated with 0.2% CoQ10, while both doses impaired WT mice.
We observed profound changes in striatal transcripts for receptors and neuropeptides including CB1, D1 and D2 dopaminergic receptors, enkephalin, and DARPP-32 in KI mice fed control chow. At this age, no change in substance P mRNA was observed, supporting our previous results from in situ hybridization studies in a similar line of KI mice (Menalled et al., 2000). CoQ10 did not induce significant differences in KI mice compared to control fed KIs. However, several of the robust genotype differences observed between control fed WT and KI mice were abolished in KI mice treated with CoQ compared to similarly treated WT mice. These include decreases in DARPP-32 mRNA (both doses), D1 and D2 dopamine receptor mRNA (0.6%), and CB1 mRNA (0.2 and 0.6%) (Table 3). Further, CB1 levels in 0.2%-treated KI mice were no longer significantly different to control-treated WT mice.
As previously documented, striatal neurons of 4.5 month old KI mice showed a large number of diffusely stained nuclei, microaggregates, some nuclear inclusions, and abundant neuropil aggregates (Menalled et al., 2003; Hickey et al., in preparation). Despite having beneficial effects on behavior both at 1 month and at 4 months, and some benefit on transcriptional dysregulation, neither treatment dose of CoQ10 affected aggregation of mutant huntingtin in diffusely stained nuclei or in the neuropil at 4.5 months of age (Fig. 5). This further indicates a dissociation between aggregation and these early manifestations of disease in KI mice.
The potential for CoQ10 to improve disease progression in patients with HD has been the object of interest ever since a clinical study showed a non-significant trend to improve total functional capacity in patients (Huntington Study Group, 2001). A beneficial effect has been supported by a number of studies in transgenic mouse models of the disease (Schilling et al., 2001, Ferrante et al., 2002, Schilling et al., 2004, Smith et al., 2006, Stack et al., 2006) but their conclusions have been challenged by more recent work utilising the same model but different experimental conditions (Menalled et al., 2010). Importantly, these models develop quickly, preventing access to early disease stages. Genetic identification of mutation carriers would allow for life long administration of protective therapies providing that they are safe for chronic use. This may be particularly important in HD because recent studies have identified early deficits in gene carriers, years before onset of manifest disease (Aylward et al., 2011, Stout et al., 2011). Therefore, it is meaningful and important to test the effect of life-long treatment administration in models that reproduce the long pre-manifest phase of the disorder.
Here, we show that life-long, oral administration of the two doses of CoQ10 used by Menalled et al. (2010) improves several behavioral tests impaired at a young age in a progressive mouse model expressing the full-length mutated protein, more so with the lower (0.2%) dose. In addition, partial normalization of transcriptional dysregulation in striatum further supports a beneficial effect of CoQ10 when administered from very early disease stages in a model that more closely recapitulates the slow development of HD. Oral administration of CoQ10 is highly clinically relevant (Shults et al., 2004, Huntington Study Group Pre2CARE Investigators, 2010) and results in dose-dependent increases in blood levels of CoQ10 in humans (Huntington Study Group Pre2CARE Investigators, 2010) and rodents (Kwong et al., 2002, Kamzalov et al., 2003, Smith et al., 2006). Although regional variability of levels of CoQ10 in brain exist (Fariello et al., 1988), and levels of CoQ10 may be saturated in young animals (1–2 months) (Zhang et al., 1995, Zhang et al., 1996), extended periods of oral administration of CoQ10 increased levels in brain mitochondrial fractions (Matthews et al., 1998, Kwong et al., 2002, Kamzalov et al., 2003) and whole brain (with supplementation, Menalled et al., 2010). In the present study, mice were fed control or 0.2% or 0.6% CoQ10 from conception. The presence of beneficial effects on behavior and brain transcripts in our study further suggests that CoQ reaches functionally significant concentrations in brain, as expected from these previous studies. Nevertheless, it is possible that some effects of CoQ were peripheral in origin, see below.
The CAG140 mice used in our study harbour the HD mutation in its proper genomic context and present a more progressive disease course that allowed us to examine mice at very early stages in disease progression, many months prior to evidence of any striatal atrophy. At the ages examined, pathology is manifest in these mice as transcriptional dysregulation in striatal neurons, aggregated mutant huntingtin, LTP deficits, reduced BDNF and motor, cognitive and emotional behavioral deficits (Menalled et al., 2003, Hickey et al., 2008, Simmons et al., 2009, Hickey et al., in press). The tasks that we used in our trial showed ease of use, and had no inherent effects on disease themselves, and we did not retest animals using the same task repeatedly as we have shown that this may affect WT performance, and reduce sensitivity due to development of habituation (Hickey et al., 2008; Menalled et al., 2003). By 1 year of age, the CAG140 KI mice begin to show reduced weight in comparison to sex-matched controls, neurochemical deficits in striatum, and gliosis (Hickey et al., 2008, Phan et al., 2009). At this later age, electrophysiological alterations are noted in cortex and striatum (Cummings et al., 2009, Cummings et al., 2010) and levels of the transcriptional co-activator PGC-1alpha are profoundly reduced (Cui et al., 2006). PGC-1alpha is involved in energy metabolism and mitochondrial biogenesis and is progressively reduced in HD patient brain and in presymptomatic HD striatum (Cui et al., 2006, Kim et al., 2010). By 2 years of age, the KI mice show 40% loss of striatal neurons, and 38% loss of striatal volume. Importantly, when HD patients show ⅓ to ½ loss of striatal volume, they begin to phenoconvert (Aylward, 2007), and we observed an overt spontaneous deficit in gait and appearance of KI mice at 2 years (Hickey et al., 2008). Thus, these mice provide an excellent model to examine potential therapeutics using behavioral and neuropathological endpoints with high power to detect drug effects (Hickey et al., 2008).
Our results indicate that CoQ10 supplementation in diet improved early motor phenotypes in the CAG140 KI mice. The behavioral benefits we have observed are reminiscent of previously reported improvements in some transgenic models. In combination with 0.007% remacemide hydrochloride in N171 transgenic mice, 0.2% CoQ10 in chow, for 9 weeks, from a relatively early stage in disease (8 weeks of age), caused transient improvement in rotarod performance in TG mice, but with no effect on WT mice (Schilling et al., 2001). Similar to our results, mutant huntingtin aggregation and accumulation was not affected by this CoQ10 treatment. Furthermore, a later study from the same authors (Schilling et al., 2004) found that a powdered diet of CoQ10 (0.2%), alone, improved rotarod performance in mutants but impaired WT performance (9 week administration, from 8 weeks of age), also similar to our data. Using the more severe R6/2 transgenic mouse model of HD, Ferrante et al. (2002) and Stack et al. (2006) reported beneficial effects of 0.2% CoQ10 in chow, from approximately 3 weeks of age, on survival, body weight, rotarod performance, huntingtin aggregates and striatal neuronal and brain atrophy (Ferrante et al., 2002, Stack et al., 2006). In further studies, Smith et al. (2006) demonstrated optimal beneficial effect of 5000 mg/kg/d CoQ10 (from Chemco Industries Inc) when compared to lower (1000 mg/kg) or higher doses (10,000 mg/kg or 20,000 mg/kg), in R6/2 mice on behavior (rotarod), pathology (reduced mutant htt aggregation, improved striatal and neuronal atrophy) and oxidative stress (reduced levels of 8-hydroxydeoxyguanosine) (Smith et al., 2006). In contrast, testing of 0.2% and 0.6% CoQ in the same R6/2 mouse model from 4 weeks of age under conditions of enriched environment, and with repeated behavioral testing, did not detect any improvement on multiple motor deficits or survival (Menalled et al., 2010). C57Bl/6J mice consume approximately 4g chow per day (Hickey et al., in preparation), which would correlate to daily doses of approximately 400 mg/kg (for 0.2%, for a 20g mouse) in our study. Thus, the doses used in our study and in Menalled et al. (2010) are much lower than those used in Smith et al. (2006), but similar to those used by Ferrante et al. (2002) and Stack et al. (2006).
Importantly, the beneficial effect we have observed in the CAG140 mice in the present study was observed at an age when the KI mice are affected by the mutation in a subtle manner that is only revealed by the use of sensitive tasks. As indicated earlier, KI mice do not show loss in body weight until 1 year of age (Phan et al., 2009), and do not show overt behavioral onset of symptoms until nearly 2 years of age (Hickey et al., 2008). Thus the deficits in control KI mice noted here are indicative of some of the earliest manifestations of the disease. This is different from the deficits examined from 4 to 12 weeks of age by Menalled et al. (2010) in the much more rapid R6/2 model, which begin to lose weight at 10 weeks of age and die with severe motor deficits at approximately 100 days of age even with improved husbandry. In this latter study, treatment was started at an age when extensive behavioral and cellular dysfunction already exists in this model (Klapstein et al., 2001, Hickey et al., 2005). It is possible that the late start of the trial, in addition to differences in model severity, frequency of behavioral testing and the presence of an enriched environment in the home cage, which is beneficial to HD mouse models (Hockly et al., 2002, Schilling et al., 2004), may have obscured the beneficial effects of CoQ10. Thus the different results obtained in Menalled’s and our study may indicate that CoQ10 treatment would be most -or only- beneficial if administered very early in the course of the disease.
A disconcerting aspect of our results is the absence of consistent effects of both doses examined. At 1 month of age, KI mice showed reduced activity and rearing, which were both rescued by 0.2% CoQ10 whereas 0.6% had no effect. In contrast, at 4m of age, both 0.2% and 0.6% CoQ10 improved sensorimotor performance on the vertical pole task, while only 0.2%, improved rotarod performance. Only 0.2% CoQ abolished the deficit in climbing which was observed in KI mice fed a control diet. Furthermore, normalization of striatal transcripts compared to similarly treated WT mice was seen with both doses but on different transcripts, except for DARPP-32 and CB1, which were ameliorated by both doses. The “U-shaped” dose response that we observed not unexpected given the “U-shaped” dose response observed using much higher concentrations in the more severely affected R6/2 model (Smith et al., 2006). Although 0.2%- and 0.6%-treated KI mice performed better than controls in the pole task, higher doses of CoQ are associated with increased risk of GI adverse events in HD patients (Huntington Study Group Pre2CARE Investigators, 2010). It is possible that these events could have obscured a beneficial effect of the higher dose on other behavioral tasks. Since levels of CoQ in striatum were not measured in individual mice in this study, it is also possible that these differences may be related to differences in absorption or accumulation of the dietary supplement. However, they raise questions regarding the optimal dosage of CoQ and the therapeutic window for benefits in patients.
Neither dose of CoQ changed the density of nuclear microaggregates or inclusions of htt in striatal neurons, or neuropil aggregates in this brain region in 4.5 month old mice. This suggests that the behavioral improvements observed after CoQ treatment in the KI model are independent of htt accumulation into neuronal nuclei. There is ample precedent for such dissociation after CoQ treatment (Schilling et al., 2001) or other treatments that improve behavioral outcome in mouse models of HD (Hockly et al., 2003, Hockly et al., 2006). This is not surprising in view of the uncertain significance of aggregates in the toxicity of mutated htt (Saudou et al., 1998, Arrasate et al., 2004). In contrast, it is generally accepted that transcriptional deficits are central to the cellular toxicity of mutated htt. Although the effects of CoQ on mRNA levels in striatum did not reach significance compared to vehicle treated KI, the normalization compared to WTs is strongly indicative of a beneficial effect of treatment because the differences normally observed between WT and KI are robust and highly reproducible (Hickey et al., in press). For example, for the CB1 receptor neither treatment affected CB1 levels in WT mice. However, both treatments improved KI levels such that they were normalized with respect to similarly-treated WTs and 0.2%-treated KI mice were no longer significantly different when compared to control-fed WTs. Therefore, CoQ had beneficial effects on this functionally important receptor. In line with this, CoQ10 treatment was beneficial to KI mice in both behavioral tasks carried out at the end of the trial.
It is possible that some beneficial effects of CoQ10 are peripheral in origin. Many markers of metabolic dysfunction are observed in HD (Chaturvedi and Beal, 2008), including changes in peripheral tissues important for metabolism such as adipocytes (Weydt et al., 2006, Phan et al., 2009) and muscle. Muscle deficits in mouse models of HD include extensive gene expression, morphological and pathological changes (Sathasivam et al., 1999, Ribchester et al., 2004, Strand et al., 2005), while activities of muscle complex II/III correlate with disease duration and cognitive score activities in HD patients (Turner et al., 2007). Oxygen consumption is increased in R6/2 mice at least near the time of phenotype onset (van der Burg et al., 2008) and in presymptomatic N171-82Q transgenic mice (Weydt et al., 2006). We noted enhanced expression of genes associated with differentiation of white adipose, at relatively early stages of disease in R6/2 mice, while these genes were downregulated at late stages (Phan et al., 2009). Thus, enhanced metabolism may compensate for reduced energy supply which may underlie the protective effects of exogenous CoQ10 in our study. Surprisingly, CoQ10 had the opposite effect on KI and WT mice, causing reduced activity and rearing at 1.5 months in WT mice (0.2%) and both doses causing impaired rotarod performance in 4 month WTs. A deleterious effect of CoQ10 on rotarod performance has been noted previously in WT mice (Schilling et al., 2004), with non-significant trends noted in R6/2 mice (Menalled et al., 2010). Reactive oxygen species can play both a homeostatic and pathological role in the cell, and it is possible that exogenous anti-oxidants may perturb normal functions of redox signalling. In particular, Underwood et al. (2010) have noted that some antioxidants may reduce basal and induced autophagy. Whether this underlies the impaired performance observed here is unknown, and it should be noted that no deleterious effect was observed in the KI mice. Similarly, in humans CoQ is safe to vulnerable patients with neurodegenerative disease (Shults et al., 2004, Ferrante et al., 2005, Huntington Study Group Pre2CARE Investigators, 2010). Therefore the clinical relevance of these effects in WT mice is unclear.
In summary, we report beneficial effects of 0.2% CoQ10 in chow on motor behavior at early stages of disease in the CAG140 knock-in mouse model of HD. These beneficial effects on behavior occur without changes in htt aggregates at the end of the testing period, although some striatal transcripts were normalized in comparison to similarly treated WT mice. While 0.2% CoQ10 in chow provided more beneficial effects than 0.6%, this concentration also had some detrimental effects in WT mice, causing modest reductions in rearing and activity in the open field, and reducing rotarod performance. The significance of these deleterious effects in WT mice remains unclear, however, as our mutant mice showed no deleterious effects and even very high doses of CoQ are well tolerated in humans treated as adults (Huntington Study Group Pre2CARE Investigators, 2010). These data provide added evidence for potential benefit of CoQ10 for HD but, together with studies of similar doses in transgenic mice (Menalled et al., 2010) suggest that this benefit may be limited to very early stages of disease, when neuropathological changes are minimal, and may require early and sustained administration of the compound. This has important implications for further clinical trials of CoQ10 as a neuroprotective therapy for HD in humans.
We would like to thank Ehud Gruen, Gowry Fernando, Jamal Mourad and Zhongliang Zhao for their expert technical assistance with the mouse colony. We thank Dr. X.-J. Li (Emory University) for the generous gift of polyclonal EM48 antibody.
Supported by: Hereditary Disease Foundation Cure HD Initiative, CHDI Inc. and PHS grant R01 (NS41574)
The authors have no conflict of interest.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.