Huntington’s disease (HD) is a fatal, inherited neurodegenerative disease with no known cure. Patients suffer from the disease if they have >36 polyglutamine (CAG) repeats in the autosomal dominant HD gene with more repeats associated with earlier onset of symptoms which worsen more rapidly and cause death sooner. The R6/2 mouse is the most widely used model of the disease, usually carrying a fragment of the human HD gene with 100 CAG repeats¹. Previous studies of behaviour and longevity of these mice has shown, unexpectedly, that mice with much longer CAG repeats have later onset of symptoms and a longer lifespan². Here, we compared disease onset and progression between sub-strains of mice carrying 250 CAG repeats (R6/2 250 mice) and 350 CAG repeats (R6/2 350 mice) against wildtype controls using high-resolution magnetic imaging and spectroscopy.

Animals were scanned at 4.7T using a Bruker BioSpec 47/40 system (Bruker Inc., Ettlingen, Germany). Isoflurane (1-2.5% in 1l/min O2) was used for anaesthesia. Structural imaging was achieved with a quadrature surface coil using a rapid acquisition with relaxation enhancement (RARE) sequence (scan parameters: TR/TEeff 3.5s/32ms, echo train length 12, field of view 25.6×19.2×10.0mm³, matrix 256×192×100). The final resolution was 100µm isotropic and images were acquired in 1 hour 33 minutes.

Spectroscopy was performed with the manufacturer PRESS sequence (TR 2.5s/TE 20ms) shimmed with FASTMAP and prepared with VAPOR water suppression using manually-adjusted pulses over a voxel of 2×2×2mm3 for 128 averages with retrospectively frequency locking and outer volume suppression enabled.

Tensor-based morphometry (TBM) was performed in SPM8 (Wellcome Trust Centre for Neuroimaging, UCL) with SPMMouse³. Brain images were segmented into grey/white matter portions and these were registered using DARTEL⁴ following⁵ to produce Jacobian determinant maps of local volume change. The false discovery rate was controlled at q < 0.05 to control for type I errors due to multiple comparisons.

Spectra were analysed using LCModel (v6.3) and results are reported relative to total creatine and phosphocreatine.

Typical images from each group of mice are shown at the relative end-stage of each group in Figure 1. For R6/2 250 mice the oldest mice reach 24 weeks, the oldest 350 R6/2 CAG mice reached 58 weeks.

For comparison, a WT mouse at 58 weeks is also shown. The appearance of atrophy is striking in the R6/2 250 animals, particular in the caudate and frontal cortex. Substantial ventricular expansion is also evident in the R6/2 350 animals concomitant with lost striatal volume.

Figure 2 shows local volume changes between groups of R6/2 animals and WT controls at 12-15 weeks. At this stage, substantial atrophy is detected across the whole cerebral cortex and most subcortical structures in the R6/2 250 mouse with apparent sparing only of the midbrain, brainstem and some thalamic structures. R6/2 350 mice show much more selective atrophy corresponding closely with regions highly associated with human HD pathology, in particular the striatum, frontal cortex and sensorimotor cortex.

Concentration ratios from selected metabolites are shown in Figure 3. Significant decreases were seen in the neuronal marker aspartate (NAA) over time for both groups of R6/2 mice, as well as total glutamate and glutamine. Taurine increased in both groups but choline was seen to decrease significantly only in the R6/2 350 CAG mice.

Both groups of R6/2 mice had widespread atrophy that is visible at 8 weeks and develop broadly similar changes over time but at a much slower rate in the 350 CAG animals. This deviation from the expected worsening of phenotype with more CAG repeats is in line with other findings in the literature from behavioural² and electrophysiological studies⁶ but has not been shown before in terms of neuroanatomical or metabolic pathology as shown here. Despite the slower time course, the CAG 350 mice still die from neurological disease and have similar patterns of cerebral atrophy at end-stage to the 250 mice. This animal model, therefore, may prove to be a useful model of HD with its larger window to study disease mechanisms and evaluate potential therapies.