Huntington's disease (HD) is an incurable neurodegenerative condition characterized by progressive motor and cognitive dysfunction, and depletion of neurons in the striatum. Recently, BACHD transgenic mice expressing the full-length human huntingtin gene have been generated, which recapitulate some of the motor and cognitive deficits seen in HD. In this study, we carried out a series of extensive behavioural and neuropathological tests on BACHD mice, to validate this mouse for preclinical research. Transgenic C57BL/6J BACHD and litter-matched wild-type mice were examined in a battery of motor and cognitive function tests at regular intervals up to 12 months of age. Brains from these mice were also analysed for signs of neurodegeneration and striatal and cortical volume sizes compared using anatomic 16.4T magnetic resonance imaging (MRI) brain scans. BACHD mice showed progressive motor impairments on rotarod and balance beam tests starting from 3 months of age, were hypoactive in the open field tests starting from 6 months of age, however, showed no alterations in gait and grip strength at any age. Surprisingly, despite these distinct motor deficits, no signs of neuronal loss, gliosis or blood–brain barrier degeneration were observed in the striatum of 12-month-old mice. MRI brain scans confirmed no reduction in striatal or cortical volumes at 12 months of age, and BACHD mice had a normal lifespan. These results demonstrate that classical Huntington's-like motor impairments seen in this transgenic model, do not occur due to degeneration of the striatum, and thus caution against the use of this model for preclinical studies into HD.
Huntington's disease (HD) is a terminal condition, characterized by progressive loss of voluntary motor control, cognitive decline and psychiatric disturbance. Marked neurodegeneration and gliosis are primarily observed in the striatum and in the cortex of HD patients (1), and correlate with motor impairments and cognitive decline. HD is a dominant monogenic disorder caused by an expansion of CAG repeats on the huntingtin (HTT) gene, which encodes for a pathological polyglutamine repeat expansion in the huntingtin protein (htt). Mutated htt is misfolded, can aggregate, and has been in implicated in altered vescicular trafficking, cellular metabolism and gene transcription (2). No disease modifying treatment is available for HD, and death generally occurs ∼15–20 years after disease clinical onset.
The development of effective disease modifying therapies has been hampered by a lack of suitable preclinical experimental models that recapitulate the major pathaological features of the disease. Since the discovery of the HTT gene (3), a number of transgenic animal models expressing the full-length or a fragment of mutated HTT have been generated (4,5). These include mice harboring a fragment of the HTT gene coding for a N-terminal truncated mutant protein (such as: R6/1 and R6/2 mice; HD150QG mice and N586-82Q mice), and full-length HD models (such as: HD48 and 89 and YAC mice, and knock-in mice such as: CAG140, HdhQ92; HdhQ150 and 200), which have varying degrees of suitability for mechanistic and preclinical studies in HD. Most recently, a bacterial artificial chromosome (BAC)-mediated transgenic mouse model expressing a mutated version of the full-length human HTT was developed, which has generated much interest (6). This model, called the BACHD mouse, carries a stable number of 97 CAG repeats (6–8), and was shown to display some motor, cognitive, pathological and physiological abnormalities reminiscent of human HD (6). Interestingly, BACHD transgenic mice also have increased body weight compared with wild-type (WT) littermates (6,9–11), which may confound the results from some of the motor and cognitive tests used to assess Huntington's-like pathology.
There have been numerous reports utilizing the BACHD transgenic mouse in behavioural, mechanistic and therapeutic studies (6,8,9,12–17), however, a comprehensive behavioural analysis conducted to directly assess striatal pathology and degeneration has not yet been reported. We, therefore, aimed to phenotype motor deficits over 12 months in BACHD mice on a C57BL/6J background, and to correlate this with neurodegenerative alterations in the striatum. Surprisingly, despite observing distinct early and late motor deficits in these mice, we could not detect signs of striatal volume reduction, neuronal loss, gliosis or blood–brain barrier (BBB) damage. We, therefore, conclude that the striatum does not undergo degeneration in C57BL/6J BACHD mice, and is not the major contributor to the behavioural deficits observed in this model.
Mutant HTT is abundantly expressed in the brains of BACHD mice
Initially, we confirmed the expression of mutated HTT in the brains of BACHD transgenic mice by performing a western blot using the MW1 antibody which marks polyQ repeats, as found in the expanded repeats of mutant HTT (18). Mutant HTT was observed in both striatal and cortical areas of the brain in transgenic BACHD mice, but was absent in these same regions, in a non-transgenic WT littermate (Supplementary Material, Fig. S1).
BACHD mice gain significantly more weight than their WT littermates
It has been widely reported that BACHD animals have increased body weight (6,9,13). To determine if this also occurred in our colony of C57BL/6J BACHD transgenic, we recorded the weight of mice from weaning to 12 months of age in females and males. As shown in Figure 2, although WT and BACHD had the same body weight at a young age (1 month), both female and male BACHD mice gained significantly more weight over time compared with WT littermates (Fig. 1). BACHD males showed significantly greater weight from 4 months of age, whereas females were significantly heavier from 3 months of age. Female BACHD mice also had a much greater overall weight gain compared with male BACHD mice, with a 35% increase in weight at 12 months, compared with a 15% increase in BACHD males, when compared with their WT littermates. For this reason, we predominantly focussed on male animals for our subsequent behavioural tests, to avoid any confounding factors that might derive from weight.
BACHD mice have impaired motor balance
We performed a 3-day accelerated rotarod test on BACHD and WT littermates and recorded the average latency to fall on each day. The test was performed at 3, 6, 9 and 12 months of age in male animals (Fig. 2). Notably, a separate cohort of mice was used at each time point, to remove any learning confounder effect. BACHD transgenic mice showed an impaired motor behaviour in this test starting from 3 months of age. The graphs show the results for the 2 trial days (Days 1 and 2) and for the test day (Day 3). Importantly, even though the weight of male WT and BACHD mice at 3 months of age was similar, we could still record a difference in the test performance. A very similar trend was also observed in the female cohort (Supplementary Material, Fig. S2A–D).
To further test motor balance and co-ordination in BACHD and WT littermates we used a narrow balance beam test. As shown in Figure 3A, BACHD animals made a higher number of errors (slips) in this test, starting from 3 months of age, indicating a lack in motor co-ordination in transgenic animals compared with the WT littermates.
BACHD mice are hypoactive, but have normal limb strength and stride length
To measure general activity, an open field test was performed on BACHD animals and WT littermates, at 3, 6, 9 and 12 months of age. As shown in Figure 3B, transgenic animals at 3 months of age did not show any significant impairment in this test, whereas, from 6 months of age onwards, BACHD mice had a clear hypoactive phenotype compared with WT littermates. A similar trend was observed in the female cohort (Supplementary Material, Fig. S2E).
To assess whether there were any impairments in neuromuscular performance (strength and co-ordination), mice were evaluated by assessing hind paw grip strength, as well as stride length as a measure of co-ordination under physiological conditions. Interestingly, despite the deficits observed in rotarod and balance beam tests, both experiments clearly showed no differences in BACHD mice at 12 months of age compared with WT littermates (Fig. 3C and D). These tests were also repeated in mice at younger ages, again showing no differences between male BACHD and WT mice (data not shown).
BACHD mice show no significant depressive phenotype before the onset of major motor symptoms
Patients with HD often suffer from depressive symptoms, especially in the first stages of the pathology (19). Moreover, it is known that the striatal-based circuits play an important role in depressive disorders as well as anhedonia (20). For these reasons, we tested 3-month-old male mice for depression-like behaviour using the forced-swim test (21). As shown in Figure 3E, no significant alterations in immobility were observed in BACHD transgenic animals, indicating a lack of depressive phenotype compared with WT littermates at this age. No females were tested in the forced-swim test due to the significant increase in body weight observed in female BACHD mice even by 3 months of age (Fig. 1B), which would likely confound the interpretation of the final results. Furthermore, mice were not tested at older ages due to the evident hypoactivity at these later ages (Fig. 5A and B), which would have directly confounded any results from this test.
BACHD mice do not have striatal lesions at 12 and 15 months of age
Given the marked motor deficits observed, and prior reports of striatal volume reductions for these mice at 12 months of age (6,15), we next aimed to quantitate the extent of striatal lesions in our colony of BACHD transgenic mice. We obtained a highly detailed anatomical 3D representation of the brain by using a powerful 16.4 T magnetic resonance imaging (MRI) scanner. Using these images, we precisely reconstructed the volume of the striatum of each mouse. Unexpectedly, as shown in Figure 4, no differences between BACHD transgenic animals and WT littermates could be detected in striatal volume at 12 months (Fig. 4A), and up to 15 months (Supplementary Material, Fig. S3A) of age. We also did not detect any significant alterations in other brain regions (cortex and cerebellum) linked to HD pathology (Supplementary Material, Fig. S3B). These data clearly demonstrate that no marked neurodegeneration is occurring in these mice at these ages.
No histopathological damage is observed in the striatum of BACHD mice
To verify our MRI lesion volume data, we assessed several pathological hallmarks of neurodegeneration in the striatum, to determine if any of the behavioural deficits observed in these mice could be due to subtle, underlying damage to the striatum. DARPP-32 is expressed by 95% of medium spiny neurons and not by any other striatal cell type (22–24), and has been used to detect striatal degeneration in HD patients and animal models (13,25–30). We, therefore, analysed the striatal mRNA levels of DARPP-32 in 3 and 12-month-old transgenic BACHD mice and WT littermates. Consistent with a lack of striatal volume changes, no significant reduction in DARPP-32 mRNA was seen in BACHD mice compared with WT mice (Fig. 5A).
Next, to assess the contribution of possible non-cell-autonomous mechanisms of neurodegeneration, we measured gliosis in BACHD transgenic and WT mice. Glial fibrillary acidic protein (GFAP) and Iba-1 mRNA levels were used as markers for astrogliosis and microgliosis, respectively. No signs of gliosis were detected in the striatum of BACHD transgenic mice compared with WT littermates at both 3, and 12 months of age (Fig. 5B and C). Similarly, no alteration in RNA levels of DARPP-32, GFAP or Iba-1 were observed in cortices dissected from WT and BACHD animals at 12 months of age (Supplementary Material, Fig. S4).
Immunostaining for Iba-1, GFAP and DARPP-32 (for medium spiny neurons) was also performed on 12-month-old BACHD and WT striata, and cell numbers counted. No differences in the number of Iba-1, GFAP and DARPP-32-positive cells were observed in the striatum of BACHD transgenic mice compared with WT littermates (Fig. 6).
BACHD mice show no signs of BBB damage, and live a normal lifespan
We next assessed the integrity of the BBB in BACHD mice, as this is an important facet of neurodegeneration, and striatal blood vessels are altered in HD (31). We utilized the sodium fluorescein-based method and compared the levels of BBB leakage in BACHD transgenic and WT mice at 12 months of age. As shown in Figure 7, no differences were observed in the striatum, cortex or cerebellum of BACHD transgenic mice compared with WT littermates.
Finally, we examined the lifespan of these mice, as a shortened lifespan is a cardinal feature of several neurodegenerative diseases, and commonly occurs in HD transgenic mice such as R6/1, R6/2 and N171-82Q (32,33). An aging cohort (n = 8) of BACHD and WT littermate mice were kept in the animal facility, and all mice survived for over 800 days (>2.2 years). Importantly, BACHD mice did not show any onset of neurological symptoms, even at the oldest ages. This indicates a lack of overt progressive neurodegeneration in these mice, although it does not exclude that more subtle neurological alterations may appear in very old BACHD mice (34).
HD is a terminal neurodegenerative motor and cognitive disorder for which there is no cure: death usually occurs 15–20 years after the onset of symptoms (2). It is caused by a specific mutation due to the expansion of a CAG triplet in the Exon 1 of the HTT gene, which encodes for an expanded polyglutamine region (>36) in the htt protein. From a neuropathological perspective, HD is characterized by neuronal degeneration, gliosis and atrophy involving mainly the striatum, but also the cerebral cortex and other brain areas with disease progression. Due to the multiple facets of HD pathology, it has been difficult to generate a reliable model recapitulating all features of the disease. Multiple transgenic mouse models of HD have been made available in recent years based on different approaches to express the pathological htt protein. Some transgenic strains harbour a fragment of the HTT gene for a N-terminal truncated mutant htt protein (e.g. R6/1 and R6/2 mice; HD150QG mice; N586-82Q mice), whereas others utilized random transgenesis or knock-in technology to express the full-length HTT gene (e.g. HD48 and 89 and YAC mice, and knock-in mice such as: CAG140, HdhQ92; HdhQ150 and 200) (8,35,36). Most recently, a BAC-mediated transgenic mouse model of HD was developed (6). This model has generated much interest for its potential relevance to human HD, due to the fact that it expresses the full-length human HTT with conditional deletion potential. This model was proposed to have widespread utility in drug discovery and the evaluation of molecular mechanisms underpinning HD pathology (14,15,37–42). BACHD mice have been commonly used on a FvB/NJ or a mixed background (FvB/NJ X C57BL/6J). However, since the C57BL/6J background is by far the most widely used mouse strain in preclinical studies, and a common strain for which genetically modified (e.g. knockout) mice are readily available, we aimed to phenotype in detail motor behavioural deficits, and striatal degeneration in pure C57BL/6J BACHD mice in order provide baseline data for the research community. Surprisingly, we found that despite early and robust motor deficits, there was no evidence of striatal degeneration, gliosis, BBB breakdown or shortened lifespan in C57BL/6J BACHD mice.
One of the hallmarks of HD is a gradual bilateral atrophy of the striatum, coupled with a selective neuronal degeneration of the medium spiny neurons in the same region (43). The striatum is known to be a key brain area involved in motor control, balance and co-ordination. For this reason, we conducted a battery of behavioural motor tests, in order to assess the contribution of any striatal alterations in the behaviour the BACHD mice over time. First, we performed a basic motor evaluation using the accelerated rotarod and the open field test. We demonstrated that both female and male transgenic mice displayed deficits in the accelerated rotarod test starting from 3 months of age. One intriguing finding to come from the original reports examining BACHD mice on a FVB/NJ background is the large alternation in body weight of mice expressing the transgene (6,9). In our study, we confirmed that C57BL/6J BACHD mice gain more body weight than WT mice, with transgenic females gaining more overall weight, and beginning earlier, compared with BACHD males. Importantly, despite this, we identified that in 3-month-old male BACHD C57BL/6J mice, an age at which there is no significant alteration of body weight compared with WT mice, the difference in rotarod performance is still significant, indicating that the poorer rotarod performance is not due to any confounding effects of body weight in agreement with a previous report using FVB/NJ F1 × C57BL/6J mice (11). In addition, the distance travelled in the open field test was not significantly different in BACHD animals at 3 months of age, however at 6, 9 and 12 months of age, transgenic mice were hypoactive compared with their healthy littermates. It should be noted, however, that alterations in hypoactivity in the open field test can be strongly influenced by increased body weight in BACHD animals (11). We then tested balance and motor co-ordination using the narrow balance beam test. We found that BACHD transgenic animals make a higher number of errors (slips) in this test from 3 months of age, regardless of their weight, compared with healthy littermates.
The balance beam and rotarod tests are generic tests for motor co-ordination under conditions of activity and stress. To test for motor co-ordination under non-stress conditions, we used the CatWalk stride length test. Interestingly, in this test, we found no differences between BACHD and WT littermates at any age. This suggests that baseline striatal function is not overtly impaired in these mice. To test for this further, we employed a grip strength test to measure muscle performance. In the presence of nigro-striatal lesions, deficits in grip-strength would be expected (44), however, we noted no differences in the grip strength test between WT and transgenic mice, even at later stages, in line with other studies using FvB/NJ BACHD mice (9). From these results, it is clear that under physiological conditions (e.g. CatWalk test) C57BL/6J BACHD mice do not display any alteration in balance and gait, up to 1 year of age, similar to previous results in FvB/NJ BACHD mice (9). On the contrary, under stress-induced conditions, such as those inflicted by the accelerated rotarod and balance beam (a very narrow beam, located at 50 cm from the ground) tests, BACHD transgenic mice show a clear impairment in performance.
To identify if these motor performance difficulties in the rotarod and balance beam tests could be due to increased anxiety in BACHD mice, we subjected mice to a forced-swim test, as a measure of an anxiety/depressive phenotype. Indeed, depressive symptoms peak during the earliest phases of human HD and then diminish at later stages (19). We tested 3-month-old BACHD and WT mice in the forced-swim test; notably, an age at which there is clear deficits in rotarod performance, not affected by body weight alterations. Interestingly, no differences in the forced-swim time were observed from our data, suggesting that anxiety/depression is not a factor in their reduced rotarod/balance beam performance. In support of this, in 3-month-old BACHD animals, the distance travelled by transgenic and WT mice in the open field is the same, indicating no or very low-anxiety levels at this stage. It should be noted that our data in the forced-swim test is in contrast to what has been observed in other studies using BACHD mice on a FVB/N background (13,17,45). However, importantly, the animals used by Pouladi and Aharony were aged 5–12 months, ages at which there are clear body weight alterations, and severe motor impairments, which would confound the interpretation of data from these mice. In their paper, Lundh et al. showed that FVB/N BACHD male mice, but not females, showed a depressive-like behaviour in the forced-swim test at 2 and 6 months of age, though the number of mice used for the analysis was very low. Interestingly another paper by Baldo et al. (39) showed a depressive-like behaviour in FVB/N BACHD females, but not males, at 4 months of age, which potentially could be explained by body weight differences in BACHD females compared with males.
The genetic background (strain) of a laboratory mouse can have major effects on the phenotype of mice carrying mutations linked to neurodegenerative diseases. This has been clearly shown for Alzheimer's, Parkinson's and amyotrophic lateral sclerosis models (46–48), and it has also been shown to influence the YAC128 and R6/2 models of HD (9). For this reason, it is critical to properly characterize transgenic mouse lines when shifting their genetic background. BACHD mice are commonly used with a FvB/NJ background which is an inbred strain of mice displaying an early onset, severe retinal degeneration due to a mutation on the Pde6brd1 gene (49). Moreover this strain shows higher than average levels of anxiety and activity compared with the more commonly used laboratory strain, C57BL/6J (50,51). Despite these potential strain differences, our motor behavioural data in C57BL6/J BACHD mice shares many parallels with prior studies performed with FvB/NJ or mixed-background BACHD mice. For example, we find a similar a dysfunction in rotarod, open field and balance beam tests. Thus, we conclude that the genetic background confers minimal effects on the motor phenotype in these BACHD transgenic mice.
Our data from the CatWalk stride length motor test suggest that there could be minimal, if any, striatal abnormalities in BACHD mice. To further clarify the contribution of the striatum in the BACHD motor phenotype, we performed a series of neuropathological and biochemical tests on the striatum. To assess any reduction in the striatal volume, we performed a highly detailed anatomical 3D reconstruction of this brain area, using a powerful 16.4 T MRI scanner. We could not detect differences in the volume of the striatum between transgenic and healthy C57BL/6J mice at 12 and up to 15 months of age. This result is in contrast with what was originally described by other groups in FVB/N BACHD mice (6, 13,15). However, more recent studies could not detect any differences in striatal volumes in BACHD animals on the FVB/N background compared with WT littermates using either 7 T MRI or autoradiography techniques (17,42). This suggests our data demonstrating lack of neurodegeneration in C57BL6/J BACHD mice is likely applicable to all BACHD strain backgrounds. Further work clearly identifying if there is any striatal atrophy occurring in the FVB or mixed background BACHD mice is warranted on the basis of our results. We would posit on the basis of available data, however, that any striatal degeneration in any BACHD mouse is unlikely, if proper controlled histopathological and anatomical studies are performed.
Human HD is also characterized by both astrogliosis (43) and progressive microgliosis (52,53), which accompanies, and potentially contributes to, neuronal dysfunction and death. It has been reported that microglial activation is present in subclinical early stages of the pathology (52,54), and its degree correlates with HD severity (55). In addition, gliosis is a hallmark in the majority of other HD models (56). Therefore, we evaluated the activation status of microglia and astrocytes early (at 3 months of age) and at later stages (12 months of age) in BACHD mice compared with their healthy littermates. No differences were observed at any age in striatal mRNA levels of GFAP and Iba1: the two markers used to evaluate astrogliosis and microgliosis respectively. Since gliosis should correlate with HD severity, we analysed glial translational profiles using the same two markers at 12 months of age. No clear patterns of gliosis were observed at this stage comparing transgenic and WT striata. We also investigated the levels of the DARPP-32 neuronal marker, a specific marker for medium spiny neurons (24), the cellular type that should predominantly degenerate in Huntington's. We could not detect differences DARPP-32 mRNA levels comparing transgenic and healthy mice. Our results are in line with what was reported by Pouladi et al. (13) on the FVB/NJ background, but differ from the subtle phenotype presented earlier by Jiang et al. (57) on the same background of BACHD mice. Finally, we performed an analysis of the levels of BBB integrity in different brain areas, since aberrant BBB integrity is found in different neurodegenerative disorders including HD (31). In agreement with our MRI data, we found no apparent BBB leakage in any of the analysed brain regions, including the striatum.
The combined results from our anatomical and pathological studies clearly demonstrate that there are no signs of striatal volume reduction, loss of neurons, gliosis or leakage of the BBB in these BACHD mice up to 12 months of age. Moreover, we show that C57BL/6J BACHD mice show a similar lifespan compared with WT littermates, in line with previous reports in FVB background BACHD mice (8), suggesting a lack of progressive neurodegenerative disease in this model. In contrast, we do highlight distinct motor dysfunction from an early age under forced-motor test conditions (rotarod and balance beam), which are evidently not directly related to any striatal degeneration. Further investigation of other brain regions undergoing degeneration in these mice could clarify the origin of the motor abnormalities in BACHD mice. We do note however, that in our preliminary analysis of the whole BACHD brain, no obvious brain region undergoing degeneration based on brain volume reduction and molecular analyses were identified, indicating that any deficits are likely to be subtle. One possibility for the apparent motor deficits seen early in this model, despite absence of striatal or whole brain degeneration, is an effect of the BAC transgene on the hypothalamus (58), which might act on circadian and sleep cycles in these mice, leading to a modification in their motor behaviour. Interestingly, circadian alterations have been observed in BACHD mice on a C57BL6/J background as early as 3 months of age (59). Moreover, the production of IGF-1 has been shown to be elevated in FVB/N BACHD mice compared with WT littermates (60). This might alter the physiological cycle of exercise and IGF-1 production, potentially altering the motor behaviour of BACHD mice. This suggests that the motor deficits observed in the BACHD model, similar to the body weight alterations, might not be due to any striatal degeneration, further justifying that this model is not ideal to study some of the major pathological features of human HD.
Overall, our results clearly demonstrate that the motor behavioural deficits observed in this model of HD, do not occur due to neurodegeneration of the striatum, as occurs in human HD. Thus, we urge caution when applying experimental results obtained using these mice, to human HD pathophysiology and therapy.
Materials and Methods
BACHD animals on a pure C57BL/6J background were obtained from The Jackson Laboratory, through the CHDI Foundation, and Professor X. William Yang, University of California, Los Angeles. Mice were maintained at the University of Queensland's Biological Resources Animal Facility under specific pathogen-free conditions. For mating, male BACHD animals were crossed with non-transgenic WT C57BL/6J females to generate litter-matched BACHD transgenic and WT mice. Animals were housed in OptiMice® cages (Animal Care Systems, CO) provided with wood shaving and half a cardboard box as an environmental enrichment. Mice had access to ad libitum food and water, and were kept in a regular 12/12 h light/dark circadian cycle. Mice were genotyped using two primers to amplify human HTT (Fw: GAGCCATGATTGTGCTATCG, Rv: AGCTACGCTGCTCACAGAAA), resulting in a 130 bp band for Tg animals and no band for WT animals. All the experimental protocols were approved by the University of Queensland Animal Ethics Committee.
All behavioural tests were performed during the light phase of the light/dark circle, at the Queensland Brian Institute animal facility. Prior to each test, the mice were moved to the testing room for an acclimatization period of at least 30 min. Instruments and tools used for the behavioural tests were cleaned thoroughly with 70% ethanol and rinsed with sterile water between trials.
Mice were weighed daily during test days, or on a weekly basis at other times up to 12 months of age.
The accelerated rotarod test was performed during 3 consecutive days. Three trials per day were performed using a Rotarod (Ugo Basile, Varese, Italy) apparatus with an accelerated speed of 5–40 RMP in 5 min. A resting time of at least 30 min was given between trials. Latency to fall was recorded at each time. Every mouse able to stay on the rotating rod for more than 5 min was removed and its latency recorded as 300 s. The average of the 3 trials for each day are presented. Mice were tested at 3, 6, 9 and 12 months of age and a separate cohort of mice was used at each time point, to remove any learning confounder effect.
For the open field test individual mice were placed in the centre of an infra-red beam-equipped activity chamber (Med Associates, Inc., VT, USA, dimension: 27.3 × 27.3 × 20.3 cm, positioned in standard MDF ventilated cubicles from the same company) on top of which a perforated transparent lid was placed. Mouse activity was recorded for 30 min under 75 lux light conditions using a protocol adapted from (9). The Activity Monitor (Product Number: SOF-811, Med Associates, Inc., VT, USA) software was used to calculate the total distance travelled.
Balance beam test
Mice were tested on a 0.5 cm wide, 1 m long balance beam apparatus. The custom-built balance beam consisted of a transparent Plexiglas® structure that was 50 cm high, and had a dark resting box at the end of the runway. Mice were trained on the beam for three times in the morning, allowing for a resting inter-trial period of at least 15 min. Mice were left in the dark resting box for at least 10 s before being placed back in their home cage. They were then re-tested in the afternoon, at least 2 h after the training session. During test sessions, mice performances were recorded. The test consisted of three trials with a resting inter-trial period of at least 10 min. The number of total paw slips was calculated manually for the last of the three tests. Mice were tested at 3, 6, 9 and 12 months of age and a new battery of mice were used at each time point.
A Noldus® Cat Walk apparatus (Wageningen, The Netherlands) was used, which consisted of a narrow corridor with a glass floor equipped with a camera. This system was able to record and precisely analyse gait—a pace pattern that is dysfunctional in HD. Mice were trained once on the first day and then tested on the day after, at the same time of the day. The gait of mice was analysed using the Noldus® Cat Walk software.
We measured hind limb grip strength using a grip-strength meter (IMADA Force Measurement Model DS2-50N, Aichi, Japan). Briefly, after removing the mouse from its cage, it was lowered over the metal bar of the apparatus, keeping its body parallel to the floor and allowing it to grip with the hind paws. The mouse was then gently pulled back by the tail, and the maximum forced applied to the bar recorded. The test was repeated five consecutive times on each animal and the average of the results for each mouse is reported.
For this test, Plexiglas beakers of 20 cm diameter by 30 cm height were used. They were filled with clean water at room temperature (23–25°C, measured by a mercury thermometer), and the water level was kept consistent between animals at 20 cm from the bottom. Three beakers for each session were used. The beakers were placed on a white table with a white background, divided using white plastic dividers to prevent mouse-to-mouse interaction. The room light was kept at 80 lux. To record the test, a video camera was placed 1 m from the test table. After the start of each recording session, mice were gently placed in the water by holding them from the tail. Each mouse was left in the water for exactly 6 min and then carefully removed, dried using blotting paper, allowed to warm, and then placed back in its cage. For the analysis, the last 4 min of the test were examined and the time each mouse spent actively swimming was manually calculated in a blinded fashion. The time spent in immobility was calculated subtracting the mobility time to the total time analysed (4 min).
For a detailed MRI analysis, we performed ex vivo scans of the mouse brains using a 16.4 T MRI Scanner (Burker) housed at The Centre for Advanced Imaging (CAI, the University of Queensland). After deep anaesthesia induction (intraperitoneal injection of Zoletil: tiletamine 40–50 mg/kg, zolazepam 40–50 mg/kg and Ilium Xylazil 15–20 mg/kg), mice were intracardially perfused with phosphate buffered saline (PBS) for 90 s followed by a 4% formaldehyde solution (PFA) in PBS for 3.5 min. We than removed the heads of mice and stored them in 4% PFA overnight at 4°C. The day after, the brains were carefully removed from the skull and transferred in a PBS solution containing 0.2% vol/vol of Magnevist (10 ml of solution contain 4.69 g of dimeglumine gadopentetate, Bayer, Germany). Brains were kept in this solution for 4 days at 4°C, than immersed in a vial containing mineral oil and positioned in the 16.4 T MRI scanner (Ultrashield 700WB Plus; Burker, MA, USA). A 3D T2 anatomic scan of the brains was then performed. Striatal, cortical and cerebellar volumes were calculated using the ITK-SNAP software, version 2.4.0 (61).
Anaesthesia and perfusion steps were performed as described above. The brains were then removed from the skull and kept in 4% PFA overnight at 4°C, then washed in PBS, placed in a 30% sucrose solution in PBS at 4°C, and snap-frozen. 30 μm thick free-floating brain slices were cut with a cryostat, and stained with the appropriate primary and secondary antibodies. Sections were than incubated with DAB-peroxidase substrate (Vector Lab, CA, USA) and mounted on glass slides. Primary antibodies we used were: goat anti-GFAP (C-19, Santa Cruz, TX, USA), goat anti-Iba1 (AB5076, Abcam, Cambridge, UK) and rabbit anti- dopamine and cAMP regulated phosphoprotein of 32 KDa (DARPP-32; AB40801, Abcam, Cambridge, UK). Appropriate secondary horseradish peroxidase-conjugated antibodies (GE Healthcare, UK) and DAB staining kit (SK-4100, Vector Laboratories, CA, USA) were also used. For cell number quantification we set-up an adapted protocol using ImageJ (imagej.nih.gov/ij/).
Brain tissue harvesting
Striata and cortices were isolated on ice, by dissecting fresh brains using a mouse coronal brain matrix (ProSciTech, Australia) and a stereo microscope. Tissues were snap frozen using dry ice.
Western blot analysis
Tissues were homogenized in RIPA buffer [25 mm Tris–HCl pH6.8, 150 mm NaCl, 1% NP-40, 1% sodium deoxycholate, 0.1% sodium dodecyl sulphate (SDS)] using an electric pestle, centrifuged for 30 min at 10000g at 4°C, and supernatants collected. Protein concentration in the samples was measured using the Bradford method (Biorad, CA, USA). Sample concentration was normalized and each sample was diluted 1:1 in two times denaturing Laemmli sample buffer, boiled at 95°C for 10 min and then loaded on a 4–12% acrylamide pre-cast SDS– polyacrylamide gel electrophoresis gel (Bio-Rad). Proteins were electrophoretically transferred on PVDF membranes. Membranes were blocked in a 1 : 1 solution of Odyssey blocking buffer (LI-COR, NE, USA) in PBS for 30 min, then incubated with the appropriate primary and secondary antibody in the same solution also containing 0.01% Tween. The fluorescent signal was revealed using a LI-COR fluorescent scanner. Primary antibodies we used were: mouse MW1 (Developmental Studies Hybridoma Bank, University of Iowa, USA) for mutant HTT, goat anti-GFAP (C-19, Santa Cruz, TX, USA) and goat anti-Iba1 (AB5076, Abcam, Cambridge, UK). Secondary fluorescent IRDye antibodies were also used (LI-COR Millenium Science, NE, USA).
Quantitative real-time polymerase chain reaction analysis
Harvested tissues were homogenized and RNA extracted using the absolutely RNA Miniprep Kit (Agilent, CA, USA), according to manufacturer's instructions. RNA was quantified by nanodrop (Thermo, MA, USA) and sample concentration normalized. cDNA was prepared using the high-capacity cDNA reverse transcription kit (Applied Biosystems, MA, USA). Real-time polymerase chain reaction (RT-PCR) analysis was performed with one time SYBR®Green universal PCR Master mix (Applied Biosystems), and all experimental samples and negative controls were run in duplicate. Specific primers for each gene have been designed using the NHMC primer design tool: for GFAP Fw: AACCAGCCTGGACACCAAAT Rv: TTGTGCTCCTGCTTCGAGTC, for CD11b Fw: ACAGCAATGATGAGGATCTGC Rv: CTCTAGGTGGGTCTTGGGAAC, for DARPP-32 Fw: TGAGCCTGGCACATAAGCTC Rv: GACAGAGTGGGTTTCTGGGG, for Iba1 Fw: ACAGCAATGATGAGGATCTGC, Rv: CTCTAGGTGGGTCTTGGGAAC, for 18S Fw: GATCCATTGGAGGGCAAGTCT Rv: CCAAGATCCAACTACGAGCTT. The 18S gene was run as a stable reference gene for quantification. The mRNA fold increase levels were calculated based on the ddCT method.
BBB integrity experiments
Mice were injected intraperitoneally with sodium fluorescein (NaFl) (i.v., 10 μl of 10% solution in 0.9% sterile saline), and perfused 15 min later with PBS, following the perfusion protocol described above. Brain regions were isolated on ice, using a mouse coronal brain matrix (ProSciTech) and a stereo microscope. NaFl in the tissue was extracted using Tris buffer (pH 7.4) and acetonitrile, and fluorescence measured.
All statistical analyses were performed using the GraphPad software (GraphPad Prism, Version 6c, San Diego, CA, USA). A repeated measures two-way analysis of variance (ANOVA) was used for to analyse rotarod and body weight data, followed by a Sidak post hoc test. A Student’s unpaired t-test was used for all other analysis, as separate cohorts of mice were used for each time point. Results are presented as mean ± standard error of the mean (SEM). P-values <0.05 were considered significantly different.
This work was supported by the Huntington's Queensland Association, the National Health and Medical Research Council of Australia (Project Grant APP1004455) and the Australian Research Council (Future Fellowship FT110100332 to T.M.W.). S.M is supported by the Wesley Medical Research through the Huntington's Disease and Friedreich's Ataxia Thorsen Foundation Fellowship.
We thank Professor X. William Yang and the CHDI Foundation for supplying C57BL6/J BACHD breeders to establish our colony. We thank Dr Nyoman Kurniawan from The Centre for Advanced Imaging at the University of Queensland for technical assistance for the MRI studies. We thank the University of Queensland Biological Resources (UQBR) staff for assisting with mouse breeding and housing. We also acknowledge Ms Jennifer Simpson and Mr Sam Yuan for their technical help with some of the experiments.
Conflict of Interest statement. None declared.