Novel BAC Mouse Model of Huntington's Disease with 225 CAG Repeats Exhibits an Early Widespread and Stable Degenerative Phenotype

Abstract

Background: Unusually large CAG repeat expansions (>60) in exon one of Huntingtin (HTT) are invariably associated with a juvenile-onset form of Huntington's disease (HD), characterized by a more extensive and rapidly progressing neuropathology than the more prevalent adult-onset form. However, existing mouse models of HD that express the full-length Htt gene with CAG repeat lengths associated with juvenile HD (ranging between ~75 to ~150 repeats in published models) exhibit selective neurodegenerative phenotypes more consistent with adult-onset HD. Objective: To determine if a very large CAG repeat (>200) in full-length Htt elicits neurodegenerative phenotypes consistent with juvenile HD.

Methods: Using a …bacterial artificial chromosome (BAC) system, we generated mice expressing full-length mouse Htt with ~225 CAG repeats under control of the mouse Htt promoter. Mice were characterized using behavioral, neuropathological, biochemical and brain imaging methods.

Results: BAC-225Q mice exhibit phenotypes consistent with a subset of features seen in juvenile-onset HD: very early motor behavior abnormalities, reduced body weight, widespread and progressive increase in Htt aggregates, gliosis, and neurodegeneration. Early striatal pathology was observed, including reactive gliosis and loss of dopamine receptors, prior to detectable volume loss. HD-related blood markers of impaired energy metabolism and systemic inflammation were also increased. Aside from an age-dependent progression of diffuse nuclear aggregates at 6 months of age to abundant neuropil aggregates at 12 months of age, other pathological and motor phenotypes showed little to no progression.

Conclusions: The HD phenotypes present in animals 3 to 12 months of age make the BAC-225Q mice a unique and stable model of full-length mutant Htt associated phenotypes, including body weight loss, behavioral impairment and HD-like neurodegenerative phenotypes characteristic of juvenile-onset HD and/or late-stage adult-onset HD.

Fig. 1

Transgenic polyQ Htt protein is widely expressed in brain of BAC-225Q mice, and accumulates as large neuropil aggregates. (A) The levels of total Htt protein (MAB2166, n=3) were measured by western blotting for Htt protein. Transgenic polyQ Htt is expressed at about half the level of endogenous Htt in striatum (st) of 3-month old animals and cerebellum (ce), cortex (co), hippocampus (hi) and striatum of 12-month old animals. (B) Full uncut gel western blots of cortex extracts from 3-month-old animals were probed for poly-glutamine proteins (MAB1574, clone 1C2). Aside from the full-length mutant Htt protein band (indicated by arrow) no other abundant BAC-225Q specific protein fragments were detected between WT and BAC animals at either 3-months or 15-months of age. Analysis of the same protein extracts with anti-Htt MAB2166 confirmed the absence of mutant Htt protein fragments in the BAC-225Q cortex (data not shown). (C) Mutant Htt immunostaining was performed in cortex of 6-month (left) and 12-month old (right) BAC-225Q mice and WT mice (not shown) with 4H7H7 antibody. Representative fields from two animals (upper and lower) are shown of cortical layers I, II and III. Animals in upper fields had a relatively lower aggregate abundance compared to the animals in the lower fields. Nuclear accumulation is abundant in 6-month old animals, with sparse neuropil aggregates; in contrast 12-month-old animals exhibit extensive and larger neuropil aggregates with concomitant decrease in nuclear immmunoreactivity. All fields are 225μm x 170μm.

Fig. 2

Decreased body weight in BAC-225Q mice relative to WT. Body weight of BAC-225Q animals and their WT littermates was measured from 1^st to 10^th month of age (n=9–15). Progressively decreased weight is observed in both BAC-225Q males (from 4.5 months, ^*t-test, p<0.05) and females (from 2 months, ^#t-test, p<0.05), when compared to WT mice. Three-way ANOVA reveals effect of genotype (F(1,1862)=737.03, p<0.001), effect of gender (F(1,1862)=1877.94, p<0.001), effect of age (F(38,1862)=63.856, p<0.001 and genotype-age interaction (F(38, 1862)=7.742, p<0.001).

Fig. 3

Behavioral abnormalities in BAC-225Q mice. (A) Motor performance was evaluated using accelerating rotarod in 3- to 15-month-old males (n=9–10). Performance of BAC-225Q animals is impaired relative to WT littermates starting from 3 months of age (^*t-test, p<0.05). Unexpectedly, this impairment is absent in the aged mice, at 12 and 15 months, and at the earlier time points, between 3^rd and 9^th month, the rotarod performance deficiency, though present, is not progressive. These observations are confirmed by ANOVA that identifies effect of genotype (F(1,170)=44.744, p<0.001), but not effect of age or genotype-age interaction. (B) Since the rotarod deficit was identified at the earliest examined time point, 3 months of age, another cohort of animals was assessed from 1.5 to 2.5 months (n=9–13). Of the three examined time points, statistically significant difference between BAC-225Q and WT mice is found only at 2 months of age (^*t-test, p=0.021), however two-way ANOVA reveals strong effect of genotype on rotarod performance between 1.5 and 2.5 months (F(1,66)=8.048, p=0.006). (C–F) Early hyperactive behavior followed by normalization of the symptoms in BAC-225Q mice. Spontaneous locomotor activity of 3- to 12-month-old mice was examined in open field test (n=16–27). At 3 months of age, (C) longer distance travelled (^*t-test, p=0.42) as well as (D) increased vertical counts (^*t-test, p=0.023), (E) jump counts (^*t-test, p=0.023), and (F) average velocity (^*t-test, p=0.022) are observed in BAC-225Q animals compared to WT littermates. These alterations, except (F) average velocity at 6 and 9 months of age (^*t-test, p=0.12, p<0.001, respectively) are not found in the later time points. Two-way multivariate ANOVA reveals effect of age on total distance travelled (F(3,151)=10.015, p<0.001), vertical counts (F(3,151)=5.454, p=0.001), jump counts (F(3,151)=13.939, p<0.001), and average velocity (F(1,151)=20.099, p<0.001). Effect of genotype is found for vertical counts (F(1,151)=8.964, p=0.003) and average velocity (F(3,151)=3.356, p=0.021).

Fig. 4

Increase in plasma levels of KC and leptin in BAC-225Q mice. The levels of seven HD-associated blood markers were measured in plasma of BAC-225Q mice. IL-4 and MIP2 are below detection limit, for IFNγ, IL-6 and TNFα no differences are found between BAC-225Q and WT mice, but levels of pro-inflammatory chemokine, KC (mouse homologue of human IL-8), and of energy metabolism-regulating hormone, leptin, are altered (n=12-18, n=12–17, respectively, gender balanced). (A) Plasma concentrations of KC are elevated in 10-month-old BAC-225Q animals (^*t-test, p=0.014). (B) Leptin levels are increased in BAC-225Q mice at 6, 8 and 10 months of age (^*t-test, p<0.001, p=0.002, p=0.038, respectively).

Fig. 5

Progressive loss of brain volume in BAC-225Q mice. Brain volumes were measured using MRI volumetric analysis in BAC-225Q mice at 3 and 10 months of age (n=4). (A) Total brain volume (^*t-test, p=0.039) and volumes of 13 out of 27 analyzed brain regions (^*t-test, p<0.05) are decreased in BAC-225Q mice compared to WT littermates at 3 months of age. (B) At 10 months of age a loss of total brain volume is more prominent (^*t-test, p=0.001) and 19 brain regions with decreased volumes (^*t-test, p<0.05) are identified.

Fig. 6

Gliosis in 12-month old BAC-225Q mouse brain versus WT mouse brain. Microgliosis and astrogliosis were analyzed by immunohistochemical detection of Iba1 and GFAP, respectively. S1 cortex is shown, with quantification in arbitrary units (* p<0.01 t-test, n=6, 3 animals per genotype, left and right hemispheres). All fields are 200μm x 160μm.

Fig. 7

Early, widespread up-regulation of TSPO, a neuroinflammation biomarker, in brain of BAC-225Q animals, and late, region-specific up-regulation in YAC128 mice, an established model of adult-onset HD. (A) Binding of [³H]-DPA-713, a ligand of neuroinflammation biomarker, TSPO, was measured in BAC-225Q mice using autoradiography (n=6-8). TSPO is up-regulated in all examined regions except hippocampus. In most of the analyzed regions TSPO pathology starts early – increased ligand binding is observed at 3, 6 and 10 months in periaqueductal gray (^*t-test, p=0.017, p<0.001, p<0.001, respectively), hypothalamus (p=0.013, p=0.021, p=0.015, respectively), motor cortex (p=0.018, p<0.001, p<0.001, respectively) and striatum, (p=0.011, p=0.003, p<0.001, respectively) and at 6 and 9 months in globus pallidus (p<0.001, p=0.017, respectively) and thalamus (p<0.001, p=0.003, respectively). (B) To validate TSPO as a marker of ongoing neurodegeneration in HD models, autoradiography experiments were performed also in the most commonly used transgenic model of adult-onset HD, YAC128 mice (n=7–10). Only three regions with increased binding of [³H]-DPA-713 are found, and TSPO phenotype is observed later than in BAC-225Q animals – at 6 months in hypothalamus (^*t-test, p=0.026), at 6 and 10 months in thalamus (^*t-test, p=0.021, p=0.036, respectively) and at 10 months in striatum (p=0.048). These findings confirm juvenile-like, widespread degeneration in BAC225Q mice and more specific, adult-like degeneration in YAC128 mice.

Fig. 8

Loss of ligand binding to dopamine receptors in HD mice. (A) Autoradiography experiments using [³H]-SCH1337, D1 receptor ligand and [³H]-spiperone, D2 receptor ligand, were performed in striata of BAC-225Q (n=4-12) and YAC128 animals (n=5–12). In BAC-225Q mice loss of ligand binding to D1 receptor is found at 6 months of age (^*t-test, p=0.002), and to D2 receptor - at 6 and 10 months of age (^*t-test, p<0.001; p=0.035, respectively). In YAC128 animals only loss of D2 ligand binding is observed (^*t-test, p=0.011; p=0.03, at 6 and 10 months respectively). (B) Dopamine receptor ligand binding in striatum was assessed also in spatiotemporal manner. In BAC-225Q mice, the impairment of either receptor is age- and subregion-dependent, starting from D1 ligand binding decrease at 3 months in caudal striatum (^*t-test, p=0.039), followed by loss of both receptor types at 6 months in all the three subdivisions (^*t-test, p=0.021; p<0.001; p=0.003 for D1, and p<0.001; p<0.001; p=0.006 for D2 in respectively, caudal, medial and rostral striatum), and finally D2 loss at 10 months in rostral striatum (^*t-test, p=0.014). In YAC128 animals the pattern of D2 receptor pathology appeared to be similar, but delayed compared to BAC-225Q mice. Loss of D2 ligand binding is observed at 6 months of age in caudal and medial striatum (^*t-test, p=0.024, p=0.022, respectively), and at 10 months in medial striatum (^*t-test, p=0.012). No changes in D2 receptor binding in rostral subdivision is found up to 10 months of age and no D1 pathology is observed in any part of striatum.