R6/2 Huntington’s disease Mice Develop Early and Progressive Abnormal Brain Metabolism and Seizures

Abstract

A hallmark feature of Huntington's disease pathology is the atrophy of brain regions including, but not limited to, the striatum. Though MRI studies have identified structural CNS changes in several HD mouse models, the functional consequences of HD pathology during the progression of the disease have yet to be investigated using in vivo functional magnetic resonance imaging (fMRI). To address this issue, we first established the structural and functional MRI phenotype of juvenile HD mouse model R6/2 at early and advanced stages of disease. Significantly higher fMRI-signals (relative cerebral blood volumes-rCBV) and atrophy were observed in both age groups in specific brain regions. Next, fMRI results were correlated with electrophysiological analysis, which showed abnormal increases in neuronal activity in affected brain regions- thus identifying a mechanism accounting for the abnormal fMRI findings. [14C] deoxyglucose (2DG) maps to investigate patterns of glucose utilization were also generated. An interesting mismatch between increases in rCBV and decreases in glucose uptake was observed. Finally, we evaluated the sensitivity of this mouse line to audiogenic seizures early in the disease course. We found that R6/2 mice had an increased susceptibility to develop seizures.

Together, these findings identified seizure activity in R6/2 mice, and show that neuroimaging measures sensitive to oxygen metabolism can be used as in vivo biomarkers, preceding the onset of an overt behavioral phenotype. Since fMRI-rCBV can also be obtained in patients, we propose that it may serve as a translational tool to evaluate therapeutic responses in humans and HD mouse models.

Introduction

HD (Huntington’s disease) is a fatal inherited neurodegenerative disorder characterized by an increased number of glutamine repeats in the N terminal region of the huntingtin protein (mhtt) (Gusella et al., 1983). Clinically, the disorder is characterized by a progressive decline in cognitive and motor function, and the most vulnerable neuronal population in HD is the medium spiny neurons within the striatum (Vonsattel and DiFiglia, 1998). Grossly, HD brains are characterized by marked atrophy in both the caudate nucleus and putamen, producing large ventricular volumes. Recent high-resolution MRI studies in HD patients and mouse models have identified significant extra-striatal volume losses (Aylward et al., 1998; Rosas et al., 2003; Bohanna et al., 2008; Lerch et al., 2008; Sawiak et al., 2009a). While the use of structural MRI capable of detecting subtle variations in brain volumes has increased substantially in recent years to study animal models of disease (Bohanna et al., 2008; Lerch et al., 2011) functional MRI has not been widely implemented to study mouse models, owing partly to its limited spatial resolution. fMRI however holds the advantage of perhaps identifying metabolic abnormalities that appear before structural changes (Wu and Small, 2006). Considering this, we used in vivo functional and structural brain MRI to study R6/2 mice, a HD model well characterized anatomically and behaviorally and widely used for therapeutic trials (Li et al., 2005; Menalled et al., 2010) at time points corresponding to minor and full symptomatology in the animal model.

Although the molecular and cellular mechanisms underlying neurohemodynamic coupling are not fully known (Logothetis et al., 2001; Logothetis and Pfeuffer, 2004; Lauritzen, 2005) there is a general consensus that neuronal activation correlates with changes in hemodynamic MRI signals in the physiologically intact brain (Huppert et al., 2009; Tian et al., 2010). These signals are a complex function of change in cerebral blood flow (CBF), cerebral blood volume (CBV) and blood oxygenation (BOLD). Specifically, the local cerebral metabolic rate of oxygen consumption, CBF and CBV are all increased during neuronal activity (Mueggler et al., 2003; Lauritzen, 2005). The recent development of MRI techniques that generate rCBV maps with high spatial resolution (Moreno et al., 2006; Moreno et al., 2007) allowed us to observe abnormally high MRI-rCBV values in several brain regions that demonstrated age-dependent regional atrophy in R6/2 mice when examined with structural MRI (sMRI).

Experiments demonstrating the integrity of the blood brain barrier of the model used-- an assumption made when performing rCBV measurements are also presented.

It is important to keep in mind that several disorders might directly target brain vasculature and that rCBV mapping might not reflect underlying neuronal function in these cases. Furthermore since fMRI is an indirect measure of brain metabolism/activity, we also assessed in vivo brain metabolism more directly using neuronal glucose uptake and evaluated activity electrophysiologically. Finally, we studied the behavioral correlates of the hyperexcitability we observed in vitro.

The results allow us to evaluate hemodynamic coupling to neural activity (termed neurovascular coupling) at baseline as well as neurometabolic coupling which matches neuronal energy demand to neuronal metabolism. Understanding these relationships is of particular importance considering recent reports of abnormal metabolism in HD pathology (Mochel and Haller, 2011)

Material and Methods

R6/2 mice

Animal care was in accordance with the United States Public Health Service Guide for the Care and Use of Laboratory Animals. Procedures were approved by the Institutional Animal Care and Use Committee at Psychogenics Inc. Columbia University and SUNY Downstate guidelines. R6/2 transgenic mice expressing the N-terminal region of a mutant human huntingtin gene (Menalled et al., 2010), and wild type (WT) littermates were generated by crossing ovarian-transplanted females (from R6/2 CBAxC57Bl/6 female donors) with CBAxC57Bl/6 F1 WT males. Genotype was determined at 15 days of age by PCR of tail snips (Menalled et al., 2010). Laragen Inc. (USA) measured CAG repeat lengths. CAG number mean=251 +/− 8.12. All CAG repeat numbers reported here are those determined directly by Genemapper software. Mice were group-housed (4–5/cage) in shoebox cages with wood shavings and a filter top. The environment was enriched with a play tunnel, shredded paper, and a plastic bone. Breeder animals also received cotton nestlets and igloos, instead of play tunnels. Food and water were available ad libitum.

A second group of 6 week-old male R6/2 and WT littermates was used for electrophysiology, 2DG maps and DCE-MRI. Mean CAG number R6/2 mice= 239 +/− 7 n=40.

Functional Imaging

Cerebral blood volume (CBV) maps were acquired using methods previously described (Moreno et al., 2007). Subjects were imaged with a Bruker AVANCE 400WB spectrometer (Bruker NMR, Inc., Billerica, MA) outfitted with an 89 mm bore 9.4 Tesla vertical Bruker magnet (Oxford Instruments Ltd., UK), a 30 mm-i.d. birdcage RF probe, and a shielded gradient system (100 G/cm). Anesthesia was used to immobilize the animals and minimize fear and anxiety induced by the loud environment of the scanner. Like other fMRI studies in rodents, we used isoflurane for anesthesia (induction phase 3 vol. % and maintenance 1.1–1.5 vol. % at 1 l/min air flow, via a nose cone), under conditions that do not significantly affect heart rate, respiratory rate or SaO₂ in C57BL/6J mice (Moreno et al., 2006). Although isoflurane causes a decrease in global glucose uptake, it does not significantly uncouple the hemodynamic response at 1 minimal alveolar concentration (MAC) (Lenz et al., 1998).

We relied on T₂-weighted scans for CBV mapping in mice, as T₂-weighted scans generate images with superior contrast compared to T₁-weighted scans at high-fields. Three scout scans were first acquired so we could position the subsequent T₂-weighted images along the standard anatomical orientations in a reproducible manner. Optimal horizontal images were determined empirically by repositioning the animal to ensure the animal's brain was located at the center of the MRI coils. Shimming was performed in order to optimize B0 field homogeneity. By running the auto shim procedure, the three linear shim channels (x,y,z) were automatically adjusted by means of an iterative software routine that maximized the area of the free inductive decay (FID). T₂-weighted images were obtained with a fast spin echo (FSE) sequence with TR/TE_eff =2000 ms/70 ms, rapid acquisition with relaxation enhancement (RARE) factor=16, FOV=20 mm, acquisition matrix=256 × 256, 16 slices, with a slice thickness=0.6 mm, slice gap=0.1 mm and number of excitations (NEX)=28. Each set of images required 15 minutes, and 5 sets of images were acquired sequentially. The first 2 sets were acquired pre-contrast and the subsequent 3 were acquired post-contrast. Gadodiamide [Gd] (Omniscan, Amersham, Oslo, Norway) was injected (10 mmol/kg) via a 0.5 mm catheter placed intraperitonealy (I.P.) and saline (2 ml) was injected I.P. at the end of imaging to improve the clearance of the contrast agent.

Signal to noise ratio (SNR) was calculated as follows: First mean signal intensity (SI) of an appropriate region of interest (ROI) was measured in a pre contrast baseline T₂ weighted image. The average standard deviation (SD) of an identical sized ROI of the surrounding air was then defined as noise. Subsequently, the SNR value was obtained using the following equation: SNR=mean SI-ROI (brain)/SD (air) ROI. Given the effect of Gd on T₂ signal, contrast to noise ratio (CNR) was calculated as CNR=(Mean ROI-SI pre − Mean ROI-SI post) / SD (air) ROI pre.

ROI-rCBV data had SNR with values between 17–25 and CNR values between 5–7. During imaging, acquisition CNR was also assessed by measuring the contrast-induced percent change in signal in the thalamus (comparing the last post-contrast image to the pre-contrast image). If contrast-induced percentile change was less than 20%, an extra dose of 3 mmol/kg Gd was administered and a new post contrast series was obtained 15 minutes later. If a 20% signal change was still not achieved, the images were rejected prior to analysis. Heart rate, respiratory rate and SaO₂ were continuously monitored using pulse oximetry (Model V33304, Sergivet). The probe was attached to the lower abdomen and normality of physiological parameters was used as criterion to continue with the imaging session. No significant differences were observed in the cardiovascular parameters monitored during MR imaging acquisition between controls and experimental groups in both ages.

An investigator blinded to subject grouping performed all image processing. Although anesthesia and soft head fixation effectively minimized head motion, the AIR program was used to co-register the images. A Gnu plot was generated to assess the quality of the co-registration, and an individual study was rejected if a shift greater than 1 in plane voxel dimension was detected. CBV was mapped as changes of transverse relaxation rate (ΔR₂) induced by the contrast agent (CA). Note that R₂ = 1/T₂. When the contrast agent reaches uniform distribution, then CBV maps can be measured from steady state T₂-weighted images as: CBV α ΔR₂= ln (S_pre/S_post)/TE. Where TE is the effective echo time, S_pre is the T₂-weighted signal before CA administration, and S_post is the T₂-weighted signal after CA reaches steady state. We have previously studied the kinetics of gadolinium (I.P.) generated CBV maps in C57BL/6J mice and identified the 37.5 minute time-point as the optimum post-contrast time interval (Moreno et al., 2006). The derived ΔR₂ maps were then divided by 4 voxels with the highest ΔR₂ measured from the posterior cerebral vein (PCV), yielding relative CBV maps (rCBV). We have published a discussion on the necessity for normalization of CBV maps (Moreno et al., 2006; Moreno et al., 2007).

Additional fMRI Methods

Normative rCBV template

Individual rCBV maps were normalized as described above and subsequently masked to include only the cerebrum before being introduced to the template-defining group-wise registration. Briefly, the template was generated using an instantiation of a bidirectional log-domain diffeomorphic registration procedure optimized by an asymmetric cost function, fully described in (Sabuncu et al., 2009). Accounting for the asymmetry in a source-template coregistration yields more robust registration. The objective function used here accounted for such asymmetry by introducing a simple correction factor to the bidirectional objective function: the Jacobian determinant of the warp. Total number of images included (n=43) (Figure. 6). Note that these mice are not only the WT animals used for this study, but include mice used in other studies developed in our laboratory.

Figure 6. Comparison of the standardized rCBV template against conventional 2DG slice.

Qualitative assessment of the images reveals high concordance of 2-DG slice signals (lower panel) with in vivo rCBV MRI maps (upper panel) at similar anatomical levels, see methods for details.

Regions of Interest Selection

Horizontal slices

Striatum (St)--dorsal striatum

The lateral boundary follows the ventromedially facing concavity of the external capsule, which is inserted between its dorsal, lateral and anterior margins and the neocortex. The medial boundary is the lateral ventricle, and globus pallidus, ventrally overlaps with the ventral striatum and with the extended amygdala caudally. Given these anatomical references a slice corresponding to the stereotaxic coordinates (in mm): interaural=4.36, bregma=−5.64, based on the MBL C57/BlJ6 atlas (www.mbl.org) was selected to identify a dorsal striatum (St) ROI.

Motor Cortex (PMC)

Primary motor cortex (M1) was identified in an axial slice corresponding to the stereotaxic coordinates: interaural=6.26, bregma=−3.24, based on the MBL C57/BlJ6 atlas (www.mbl.org).

Prefrontal Cortex (PFC)

This structure is composed of the anterior cingulated, agranular insular cortex and orbitofrontal. A slice at the same level than motor cortex was selected to identify the PFC ROI. The temporal auditory cortices (Au1–2, Aud, based on mbl.org atlas) were identified in the same slice.

Hippocampus (Hc)

A slice acquired through the mid-body of the hippocampal formation provides optimal visualization of hippocampal morphology and internal architecture. The external morphology of the hippocampus was manually traced, as was the internal architecture that follows the hippocampal sulcus and the internal white matter tracts (Figures. 1A&B). Corresponding to the stereotactic coordinates: interaural=4.36, bregma=−8.64. (www.mbl.org).

Figure 1. Abnormal cerebral blood volumes in HD mice.

Horizontal (A&B) and Sagittal (C) sections of T₂-weighted images from a 6-week-old WT C57BL/6J mouse. Regions of interest were identified within: Hc [green], EC [yellow], St [dark blue], TC [grey], PMC [purple], PFC [brown], Cb [light blue], Th [red], Acc [pink] and Hyp [orange].

Mean normalized relative cerebral blood volumes (rCBVs) and SD were measured from each region of interest in R6/2 and WT mice in 6 week old (D) and 14 week old mice (E). Note that both at 6 and 14 weeks there were significant larger rCBV values in R6/2 mice, compared to WT in specific regions.

Hc= hippocampus, EC= entorhinal cortex, St= striatum,TC=temporal cortex, PMC=primary motor cortex, PFC=prefrontal cortex, Cb=cerebellum, Th=thalamus, Acc= accumbens and Hyp=hypothalamus. * denotes p value ≤0.05, and ** p value ≤0.01

Entorhinal cortex (EC)

At the same levels as the hippocampus, the lateral boundary follows the grey/white junction of the perirhinal cortex; the medial boundary is the medial aspect of the temporal lobe; the superior boundary is the beginning of the collateral sulcus; the inferior boundary is the lateral tip of the brain.

Thalamus

was indentified at the same level as Hc and EC, following the mbl.org atlas

Sagittal slices

Nucleus Accumbens (Acc)

is beneath the rostral caudate-putamen and the boundaries are formed by the external capsule. Corresponding to the stereotactic coordinates: bregma=0.3 to 1.5, lateral= 1.1, VD= 6 to 4.8 (based on Allen brain atlas).

Cerebellum (Cb)

This structure was identified in a slice at the same level as Acc.

Hypothalamus (Hyp)

was identified in the same slice as Acc VD= 6 to 5, bregma= 0 to −2 and lateral=1.1.

Structural Imaging

The MRIs generated for functional analysis were used to obtain anatomical volumes as well. 2D-FSE were used to generate brain volumes utilizing the ITK-SNAP program (Yushkevich et al., 2006), which allows for active contour generation. For the present work manual boundary drawing was used, which involved slice-by-slice boundary drawing on available orthogonal views. This method has a coupling cursor between 2D slices and the 3D lay, which significantly helps to reduce slice-by-slice jitter that occurs in this type of segmentation. Manual segmentation was performed by 3 highly trained raters using a standard anatomical atlas. We had previously compared 2D-FSE and 3D-GE generated structural data, although there was a high correlation between 2D and 3D MRI measurements (r=0.87—reported in (Moreno et al., 2011)), 2D images consistently provided slightly higher volumes. The volumes measured in 2D images were obtained by multiplying the pixel size by the interslice distance (SI=0.7 mm). A slight overestimation may have resulted because the volume from where the signal is obtained reflects a smaller slice thickness (ST=0.6mm). Therefore we calculated the average height in the slice dimension for a given ROI (Hs) and multiplied the 2D volumes by the following correction factor: (1-(1-ST/SI)/Hs)). For instance Hs=1.2: (1-(0.6/0.7)/1.2=0.88; Hs = 2.34: (1-(1-0.6/0.7)/2.34)=0.94. We therefore used this correction factor in all data presented. Both rCBV-ROIs and anatomical ROIs-segmentations were compared between the three highly trained raters using common intraclass correlation statistics. The analysis showed high agreement between the raters for all the ROIs analyzed with values between 0.887 to 0.986.

Blood Brain Barrier Permeability Measurements

Dynamic contrast-enhanced (DCE)-MRI was acquired using the protocol described in Vlachos et al., 2010. In short, a gradient echo pulse sequence was used in the temporal DCE-MR images before and after I.P. administration of Gd. Images were acquired with a T₁-weighted 2D FLASH sequence of a 192 × 128 matrix size (reconstructed to 256×128), a resolution of 130×130 µm² (reconstructed to 98×130 µm²), a flip angle of 70°, TR/TE=230/2.9 ms and a slice thickness of 0.6 mm without any interslice gap. The number of excitations (NEX) was equal to four and the acquisition time was 88 s. The Gd bolus was manually injected at a rate of approximately 10 µL s⁻¹. Mice were imaged in the same 9.4 T MRI system described above. Reference region model (RRM) was used for the quantitative assessment of permeability. RRM requires a reference region concentration profile, similar to the examined region, in order to be fitted optimally and estimate the K_trans values correctly. The epicranial muscle was selected as the reference region because of its intrinsic homogeneity in their anatomic structure and its similar longitudinal relaxation time to the gray matter. All the images were filtered prior to the concentration fitting, using the N-D filtering function of the Image Processing ToolboxTM in Matlab R2008b (MathWorks, Inc., Natick, MA, USA). Major vessels were not taken into account in the fitting wherever possible, by avoiding the corresponding regions where Gd concentration followed an exponential decaying pattern. Permeability maps and K_trans were then calculated using methods described in (Yankeelov et al., 2005; Vlachos et al., 2010).

The general kinetic model is described by:

$$\mathrm{d}{C}_t/\mathrm{d}t=K_{\text{trans}}{C}_p-K_{\text{ep}}{C}_{\mathrm{t}}$$ (1)

where C_t is the tracer concentration in the extra-vascular space (ES), t is the time, C_p is the tracer concentration in the blood plasma and K_trans and K_ep are the transfer rate constants from the intravascular system to the ES and from the ES to the intravascular system, respectively.

The RRM model is described by;

$$\mathrm{d}{C}_{\text{ref}}/\mathrm{d}t=K_{\mathit{\text{ref-trans}}}{C}_p-K_{\mathit{\text{ref- ep}}}{C}_{\text{ref}}$$ (2)

where C_ref and C_p are the Gd concentrations in the reference region and the blood plasma, respectively. Combining (1) with (2) C_p can be eliminated and the temporal concentration of C_t and C_ref are then fitted, using Levenberg-Marquardt algorithm, yielding relative permeability values. The K_trans value is finally calculated using the product of the relative permeability K_rel-trans and the corresponding reference region permeability.

Electrophysiology

Horizontal hippocampal brain slices, which contain auditory temporal cortex, lateral and medial entorhinal cortex, subicular complex, CA1, CA3, and DG of 400µm thickness were prepared. This slice has an orientation that is similar to the middle hippocampus MRI slice used. Experiments were done in an interface chamber (Fine Scientific Tools, Vancouver Canada). The artificial CSF (aCSF) contained the following in mM: 157 Na⁺, 136 Cl⁻, 5 K⁺, 1.6 Mg²⁺, 2 Ca²⁺, 26 HCO3⁻ and 11_(D)-glucose. Slices were perfused with aCSF continuously bubbled with 95% O₂ / 5% CO₂, to maintain pH near 7.4 and the temperature was set at 34°C. Platinum stimulating electrodes connected to a stimulus isolation-unit (Grass S88) were used to evoke normal excitatory synaptic responses (EPSPs) between superficial layers of EC (stimulated) to Sub (recorded), data not shown. In all experiments the slices were allowed to recover from the isolation procedure for at least 1.5 h before the beginning of the experiments. Borosilicate electrodes of 2–3 MΩ filled with 150 mM NaCl were used to measure local field potentials. Extracellular recordings were performed using an Axoclamp 2B amplifier (Molecular Devices, Palo Alto, CA), filter (0.1 Hz to 10 KHz using −6 dB/octave). Voltage signals were digitized and store in a PC using a digidata 1200 A (Molecular Devices) for off line analysis. Data were analyzed using the Neuromatic routines run in an Igor platform (Wavemetrics).

[¹⁴C] 2-deoxyglucose Procedure

Each animal was injected with 125 µCi/Kg body weight [¹⁴C] 2-deoxyglucose (2DG; specific activity 55 mCi/mM, Perki-Elmer Life Science, Boston MA) via I.P. The traditional 2DG method was carried out as outlined before (Sokoloff et al., 1977). After 45 minutes of 2DG circulation, the animals were sacrificed with a lethal dose of sodium pentobarbital (60 mg) administered I.P. At this time blood samples were obtained. All samples were kept cold on ice until pipetted into scintillation fluid. Plasma ¹⁴C was determined (Beckman LS1800 scintillation counter). The brain was removed and frozen in 2-methylbutane at −20°C for 5 min, coated with matrix and stored at −70°C. Brains were sectioned horizontally at 20 µm and stained with thionin, which binds to acidic proteins and nucleic acids. Sections were exposed to KODAK MinR film along with ¹⁴C standards (Amersham, Arlington Heights, IL) for 15–30 days. Films were analyzed for local cerebral glucose metabolism using the MCID M6 Elite system (Imaging Research Inc, Ont, Canada). Final tissue 2DG concentration was calculated. We determined the cerebral glucose consumption in gray matter of: primary motor cortex (PMC), prefrontal cortex (PFC), auditory temporal cortex (TC), dorsal striatum (St), thalamus (Th), hippocampus (Hc), entorhinal cortex (EC), cerebellum (Cb) and accumbens (Acc) using standard horizontal figure plates atlas (Paxinos George, 2001). Measures were also obtained in corpus callosum. Structures were identified bilaterally in 4 consecutive sections at 20 µm intervals, except for cerebellum and corpus callosum that were drawn as a single ROI in the 4 sections. Reported are the gray matter mean values normalized to the value of corpus callosum. Additionally, cell counts (thionin labeled cells in 1.22 mm²) of 3 consecutive sections in both sides of PMC, PFC and St ROIs were performed.

Additional 2DG analysis methods

Generating a composite 2DG image

Three adjacent 20 µm 2DG maps corresponding to equivalent levels of the rCBV-MRI atlas were used. Each of the three slices representing an equivalent MRI level were transformed to the central image within the level using a 2-dimensional 3 degrees of freedom linear spatial transformation, generating a mean image. 2DG values were normalized to the average 2DG signal in the corpus callosum (Figure 6).

Seizures Analysis

Population audiogenic seizure (AGS) sensitivity was determined by exposing mice to a high intensity siren of 1800–6300 Hz (frequency peak) with an average sound pressure level of 125 dB at 11cm (Personal Alarm, Model 49–417, Tandy Corporation) for up to 10 min in an empty plastic box with a sound-absorbent tile lid under which the siren was mounted (as described in (Yan et al., 2004)). Mice were tested between 03:00 p.m. and 05:00 p.m. for possible circadian variation. They were placed individually in the bell chamber and allowed to explore for 15 s; the bell was then rung for 60 s. None of the mice tested had previous exposure to acoustic stimulation. Mice were videotaped throughout the sound exposure. Video recordings were analyzed in slow motion by the experimenter and independently by a second rater blind to genotype. Animals having seizures in Table 1 were those exhibiting a score ≥ 1. Seizures were counted and rated according to the Racine scale (Racine, 1972) and classified as: grade 1, mouth movements; grade 2, head nodding; grade 3, forelimb clonus; grade 4, rearing with clonus; grade 5, full motor seizures (tonic-clonic). The number of seizures was counted as the number of episodes with the maximum grade of each seizure being used to designate each episode. Thus, if a mouse exhibited a grade 4 seizure which included forelimb clonus and mouth movements, it was counted as a grade 4 seizure. Seizure grades were compared between groups using a non-parametric Mann-Whitney U test. Seizures were also grouped as severe (grades 4–5), mild–moderate (grades 1–3) and then compared using the same non-parametric test. In both cases the result was identical.

Table 1. R6/2 mice have increased susceptibility to audiogenic seizures.

Shown are each mouse that developed seizures out of 8 mice per group (WT and R6/2, n=16). Note that all R6/2 mice had seizures with varying intensities; only two WT mice had abnormal behavior induced by the siren that met criteria for ictal events.

Clinical Seizures
Group	Number of mice with seizures	Racine Scale	Deaths
WT	1	1	0
	1	2
R6/2	2	3	2
	4	4
	2	5

Statistical Methods

Repeated Measures ANOVA was used to identify ROI specific genotype differences. P values were adjusted for multiple comparisons using false discovery rate (FDR) correction. Akaike’s information criterion was used to determine if a heterogeneous model (allowing separate variance term for each ROI) provided better fit. T tests and Mann-Whitney tests were used to compare distributions between groups, Pearson correlation was used to test for associations. The specific selection of variables and methods for each experiment is described in the results section. All repeated measures analyses were implemented using PROC MIXED under SAS, version 9.2 (SAS Institute, Cary, NC).

Results

Early and progressive increase in MRI-rCBV values in R6/2 mice

R6/2 mice and age matched wild type littermates were genotyped for CAG repeat number (see methods) and imaged at 6 and 14 weeks of age (n=10 per genotype and age group, total n=40); sex was equally distributed between the groups. rCBV maps were reconstructed from MR acquisitions before and after intraperitoneal (I.P.) injection of gadodiamide (Gd) contrast as previously described (Moreno et al., 2006; Moreno et al., 2007). Data are presented as normalized ΔR₂ values≈ rCBV (see methods). rCBV mapped from ΔR₂ can be influenced even by relatively small changes in physiological parameters, particularly the final cerebral vascular concentration of the contrast agent. To circumvent such variability when comparing different groups of mice, ΔR₂ measurements were normalized to the maximum (4 pixels) ΔR₂ signal in the posterior cerebral vessels (an area determined to have the highest contrast agent concentration) (Moreno et al., 2006; Moreno et al., 2007).

We first examined whether cerebral rCBV was compromised in 6 week-old R6/2. Region of interest (ROI) analysis of rCBV included but were not restricted to brain structures commonly affected in human and mouse models of HD: the primary motor cortex (PMC), prefrontal Cortex (PFC), temporal cortex (TC), dorsal striatum (St), thalamus (Th), hippocampus (Hc) and entorhinal cortex (EC).In the sagittal views the regions of interest were: cerebellum (Cb), Accumbens (Acc) and Hypothalamus (Hyp) [Figures 1 A–C] (Aylward et al., 1998; Rosas et al., 2003; Bohanna et al., 2008; Lerch et al., 2008; Sawiak et al., 2009a).

Mean ROI rCBV values were compared between R6/2 mice and controls using repeated measures ANOVA, sex was included as covariate. The analysis revealed that six week-old R6/2 mice had significantly higher rCBV values in PFC (t₍₁₆₂₎=3.43; p=0.008), PMC (t₍₁₆₂₎=2.75; p=0.022), TC (t₍₁₆₂₎=2.83; p=0.022) and St (t₍₁₆₂₎ =2.84; p=0.022). Parallel in the 14-week-old group, recapitulated these results- PFC (t₍₁₆₂₎=3.69; p=0.001), PMC (t₍₁₆₂₎=3.11; p=0.006), TC (t₍₁₆₂₎=3.43; p=0.003), St (t₍₁₆₂₎=3.64; p=0.001)- and also demonstrated higher rCBV in Th (t₍₁₆₂₎=2.25; p=0.043) and Hc (t₍₁₆₂₎=3.89; p=0.001) [Figure 1E]. Nota Bene: No sex effects were noted within these comparisons and no age by genotype interaction was observed.

Futhermore, bivariate correlation analysis between rCBV values in R6/2 mice of both age groups and CAG repeat number showed no statistically significant correlation between groups. The results indicate that R6/2 mice develop early focal increases in brain rCBV. The striatum and neocortex, areas known to be affected early in humans with HD (Vonsattel et al., 1985; Rosas et al., 2003) were found to have abnormally high rCBV values in younger R6/2 mice. These differences progressively spread to other brain regions, most notably the thalamus and the hippocampus. Areas such as the entorhinal cortex- the hippocampus’ main input- cerebellum, hypothalamus or nucleus accumbens were unaffected.

In the steady state rCBV method used here, it is assumed that the blood brain barrier (BBB) remains intact (Lin et al., 1999). An in vivo Gd based dynamic contrast-enhanced (DCE)-MRI protocol to evaluate the integrity of BBB (Vlachos et al., 2010) in R6/2 mice and controls (n=6 per group) was performed. In a general kinetic two-compartment model, the tissue is assumed to consist of the blood plasma and extracellular extravascular space, and the tracer exchange (Gd) between these two body compartments is defined by two transfer rate constants, K_trans and K_ep. Permeability maps were generated and the K_trans values were calculated (see methods) for a predefined ROI (neocortex and striatum). Mean ROI K_trans values were compared between R6/2 mice and controls using t-test. The analysis revealed that WT and R6/2 mice had negligible K_trans values and were not significantly different in the regions analyzed: Neocortex (0.00033 +/− 0.00016 vs. 0.0017 +/− 0.0004 min⁻¹, mean +/− SD), striatum (0.00021 +/− 0.0001 vs. 0.00022 +/− 0.0000 min⁻¹). These data indicate that the BBB in both groups has normal permeability values.

Structural abnormalities in R6/2 mice

To investigate whether the fMRI signal changes detected in R6/2 mice accompanied structural changes, we subjected the same horizontal 2D-T₂ images used for the fMRI-analysis to ROI volumetric analysis in the following regions: ventricles (Ven), hippocampus (Hc), neocortex (Neo), striatum (St), cerebellum (Cb) and whole brain (WB) [Figure. 2A].

Figure 2. Volumetric analysis of R6/2 and WT mice.

A. Shown are horizontal sections of T₂-weighted images from ventral to dorsal levels at 0.7 mm intervals, regions measured were: Ven [red], Hc [green], Cb [orange], St [blue], Neo [yellow] and WB. B. The volumes of R6/2 mice were not significantly different from WT controls in any of the brain regions analyzed at 6 weeks of age. C. The volumes of the neocortex, striatum and hippocampus were significantly smaller in 14-week-old R6/2 mice compared to WT. D. A significant inverse linear correlation was found between prefrontal cortex normalized rCBV values and neocortical volumes.

Mean volumes in mm³ and SD for the different areas. * denotes p value < 0.05. Ven=ventricles, Hc=hippocampus, Cb= cerebellum St=striatum, Neo=neocortex and WB=whole brain

Results from different regions were covaried with and separately normalized to whole brain volume. Because the results of the statistical analyses using these normalized values did not vary substantially from those using raw values, only the raw values are reported.

As with rCBV, ROI volumetric analysis was carried out in repeated measures ANOVA, sex was included as covariate. Results in the 6 week old group demonstrated no statistically significant ROI volume differences. [Figure. 2B], while 14 week old R6/2 mice showed significant volume loss in the striatum (18.76 +/− 1.8 vs. 23.43 +/− 2.4 mm³, t₍₆₈₎=4.69; p< 0.001), hippocampus (19.19 +/− 1.6 vs. 22.87 +/− 2.5 mm³, t₍₆₈₎=3.79; p= 0.001) and neocortex (101.94 +/− 11.8 vs. 126.40 +/− 11.2, mm³, t₍₆₈₎=5.38; p<0.001) when compared to WT mice [Figure 2C]. A significant age × genotype interaction on the progression of brain atrophy was observed in the hippocampus (F_(3,140)=4.62; p=0.004) neocortex (F_(3,140)=12.3; p< 0.001 and striatum (F_(3,140)=15.01; p< 0.001).

While previous studies demonstrated cerebellar atrophy in R6/2 mice (Sawiak et al., 2009a, b), we only note a trend towards cerebellar volume loss in 14-week-old R6/2 mice (50. 93 +/− 3.7 vs. 53.2 +/−3.0 mm³; t₍₆₈₎= 1.47; p=0.181).

Thus, structural MRI data reveal that R6/2 mice develop progressive, region specific, and age dependent brain atrophy.

Given the demonstration of focal functional and structural abnormalities, we sought a bivariate correlation between rCBV and ROI volumes in R6/2 and WT mice. A significant inverse relationship was observed in R6/2 mice between neocortical volumes and rCBV values in the PMC (Pearson correlation=−0.508, p=0.012) and PFC (Pearson correlation=−0.512, p=0.011) [Figure 2D]. These correlations imply that rCBV values in neocortex are higher in more atrophied brain regions. The influence of volumetric changes on rCBV abnormalities were evaluated in an ANOVA covarying for ROI volume. There was no observed effect in this covariate model, suggesting that rCBV changes in R6/2 are not likely influenced by structural changes.

Increased neuronal synchronized activity in R6/2 mice

Given that increases in fMRI signal components have been linked empirically and theoretically to increases in neuronal activity (Lauritzen, 2005), we sought to correlate local field potentials (LFPs) to rCBV values. Previous experiments have shown that LFPs are correlated to cerebral blood flow (CBF) responses (Lauritzen, 2005; Chaigneau et al., 2007; Piilgaard and Lauritzen, 2009) and that rCBV is related to CBF i.e rCBV=B.rCBF^A, where A and B are obtained by manipulating PCO₂ levels (van Zijl et al., 1998; Rostrup et al., 2005). LFPs were then measured in a separate group of R6/2 and control littermate mice from the same colony (6 week old WT and R6/2 mice, n=8 males per group, with similar CAG repeat number).

Horizontal brain slices including hippocampus, entorhinal and temporal cortices were prepared as described above [Figure 3 A]. Recordings from wild type mouse slices show occasional spontaneous LFPs in several of the regions analyzed in the rCBV and volumetric studies [Figure 3B]. A summary of the characteristics of the events in superficial layers of auditory temporal cortex (TC), lateral entorhinal cortex (LEC), and two subregions of the hippocampus subiculum [Sub] and dentate gyrus [DG] can be seen in figures 3D&E. The frequency and duration of events increased after 40 min of the GABA_A blocker picrotoxin (50 µM) exposure; these changes remained present and grossly stable for the duration of the recording (~ 5 hours).

Figure 3. Characterization of spontaneous local field potentials (LFPs) in hippocampus-entorhinal-temporal cortex slices from wild-type and R6/2 mice.

A. Inset shows a typical slice where the regions recorded are indicated B. Representative traces depicting spontaneous LFPs in the different regions in slices from wild-type mice C. Similar traces as in B from R6/2 mice. D. LFPs durations comparing WT and R6/2 mice in the different regions recorded E. Comparison of LFPs frequencies (mean values and SD) in slices from wild-type mice vs. R6/2 mice. Note that there were significant differences in the frequency of events in all areas analyzed. In contrast, although the durations trend to higher values in mutant mice, only LEC show significant differences..

TC = superficial layers of temporal cortex, LEC= superficial layers of lateral entorhinal cortex, DG= dentate gyrus and Sub=subiculum

Figure 3C shows typical recordings obtained from brain slices prepared from R6/2 mice. The occurrence of higher frequency spontaneous LFPs observed (Figure 3E) was significantly greater than that in WT recordings at: TC (Mann-Whitney U test p=0.0082); Sub (Mann-Whitney U test p=0.0080); DG (Mann-Whitney U test p=0.0043) and LEC (Mann-Whitney U test p=0.016). Low frequency (0.02–0.03 Hz) LFP events lasting ≥ 1.5 secs, characteristic of interictal-like events were evident in the TC of 40% of R6/2 mice; no such discharges were observed in WT mice. Figures 3 D&E summarize population data in the regions described above. Addition of picrotoxin to slice preparations introduced a 5–20 min period of variable activity followed by a statistically significant increase in the frequency of events in all areas analyzed throughout the recording (Figures 4B&D). Further still, the LEC was the only region studied in the preparations that demonstrated prolonged synchronized discharges interspersed among the shorter regular rhythmic discharges induced by GABA_A blockade (Figures 4C&D) (Mann-Whitney U test p=0.0095). Prolonged discharges were considered to be extensions of shorter spontaneous discharges for two reasons: First, they occurred at the time point at which a shorter synchronized discharge would have occurred, as predicted by the ongoing rhythmicity of the waveform, and second, the morphology of the prolonged events resembles that of the shorter synchronized discharges. A summary of the picrotoxin effects in the different regions is presented in figures 4 C&D. Significant increases in picrotoxin-induced LFP discharge frequency (all regions) and duration (LEC only) were events unique to R6/2 mice. Having established LFPs as indicators of spontaneous neuronal activity in R6/2 mice, we next addressed whether such observed hyperexcitablity in vitro results in a propensity towards epileptogenesis. The synchronized discharges seen in R6/2 mice LEC resemble seizure-like events (Figure 4E) as reported previously in vitro in this region (Kovacs et al., 2009) perhaps indicating an increased susceptibility to seizures in this mouse model. Accordingly, population audiogenic seizure (AGS) sensitivity was determined by exposing 6 week old male R6/2 and WT littermate mice to a high intensity siren (Musumeci et al., 2000). When exposed to a 125 dB alarm, R6/2 animals typically layed initial signs of epileptic activity in the form of frenzied running and jumping. (Videos showing the responses are available on request). Seizures were counted and graded according to the Racine scale (Racine, 1972) and results are shown in Table 1. While all 6-week old R6/2 mice developed seizures graded at least level 3 on the Racine scale, including 25% of this group that developed generalized tonic-clonic seizures (GTC-Sz) leading to death, no WT mice developed seizures graded higher than level 2 on the Racine scale (Mann-Whitney U test p=0.01).

Figure 4. Prolonged epileptiform discharges in LEC in slices from R6/2 mice after GABA_A receptor blockade.

A. Shown are spontaneous LFPs after 30 min bath application of 50 µM picrotoxin in the regions indicated in WT mice, to the right of each trace there is a characteristic LFP induced by blocking GABA_A. B. Changes induced by picrotoxin in R6/2 mice differ significantly from WT mice, shown are representative recordings and close up views of single events in the different regions. C. and D. Comparisons of LFPs duration and frequencies (mean values and SD) in the different regions between WT and R6/2 mice. E. Example of an epileptiform discharge induced by 50 µM picrotoxin in LEC in an R6/2 mouse, to the right are plots of the LFP ≥ 1.5 sec in duration in this region.

St= striatum, TC=temporal cortex, PMC=primary motor cortex, PFC=prefrontal cortex

* denotes p value <0.05, and ** p value < 0.01

In vivo measurement of local glucose utilization

As established by numerous studies (Reiman et al., 1996; Costa et al., 1999; Rapoport, 2000; Small et al., 2000a; Small et al., 2000b; Dixon et al., 2002; El Fakhri et al., 2003) rCBV is a correlate of oxygen consumption and therefore typically reflects regional energy metabolism. Glucose utilization is another indicator of energy metabolism. To evaluate glucose utilization, high resolution in vivo [14C] deoxyglucose (2DG) uptake (Torres-Reveron et al., 2006) was measured in 6 week-old R6/2 mice and WT littermates (n=6 males per group).

20µm thick axial whole brain slices were obtained and evaluated for 2DG uptake in all areas analyzed in the rCBV study save for the hypothalamus, which demonstrated sectioning artifact in greater than 20% of the sampled slices (Figure. 5A). Mean local cerebral glucose utilization (LCGU) values were expressed in nCi/gr in each ROI and data were normalized to corpus callosum LGCU to account for the variability of radiotracer concentration available to the brain (see methods).

Figure 5. Decreased brain glucose uptake in R6/2 mice.

A. Representative 2DG autoradiographies, [pseudocolor images] comparing 6 week old WT and R6/2 mice, in basal conditions. ROIs are identified in the horizontal single hemisphere slices at two different ventro-dorsal levels as: PFC, PMC and St. The calibration bar represent higher signals as warmer colors. B. Mean normalized glucose utilization and SD in several ROIs as described above.

St= striatum, PMC=primary motor cortex, PFC=prefrontal cortex * denotes p value <0.05, and ** p value < 0.01

Repeated measures ANOVA demonstrates that R6/2 mice had significantly lower glucose uptake in St (110.14 +/− 6.5 vs. 140.38 +/− 4.72, t₍₁₅₎=6.06; p<0.001), PFC (109.21 +/− 4.2 vs.130.27 +/− 4.6, t₍₁₅₎=4.26; p<0.001) and PMC (100.33 +/− 6.4 vs. 134.25 +/−5.3, t₍₁₅₎=5.93; p<0.001) (Figure 5B). Whether observed reductions in metabolism arise from tissue loss in specific areas (Karow et al., 2010), was addressed by covarying the data analysis with cell number (see methods) of the abnormal ROI. Covariation of LGCU results with this indicator of cell loss did not affect 2DG findings and the average cell number was not found to be a significant predictor (F_(1,15)=0.4; p=0.85). Though regional LGCU findings corroborated observations made in rCBV with respect to the focality of normal and abnormal metabolism (Figures 1D and 5B), evidence for disrupted neurometabolic coupling was found. In fact, LGCU in the PMC, PFC, and St was paradoxically low even though these regions demonstrated higher rCBV.

Discussion

Our main finding is that R6/2 mice develop early, region-specific metabolic abnormalities, particularly in the striatum and neocortex as detected by in vivo MRI and 2DG mapping. Further still, increased regional rCBV correlated well with increased neuronal excitability, manifested in vitro (slices) and in vivo (audiogenic seizures). We propose that this is indirect evidence of preserved neurovascular coupling in R6/2 mice. Brain metabolism requires glucose uptake and oxygen consumption. Neurohemodynamic coupling however is a product of both neurometabolic and neurovascular phenomena (Huppert et al., 2009) and our experiments show a paradoxical decrease in LGCU in R6/2 mice in regions with abnormally high rCBV values. This suggests that R6/2 mice have impaired neurometabolic coupling.

Measuring Brain Structure and Function with MRI in R6/2 mice

Similar to our findings, atrophy in the striatum, cerebral cortex and hippocampus has been previously reported in R6/2 mice using volumetric MRI (Sawiak et al., 2009a, b). The discrepancy from previous reports regarding cerebellar atrophy can be partially attributed to the accelerated disease progression in the R6/2 HD model (our cohort was 6- and 14-week vs. 18 week-old mice).

To validate the physiological underpinnings of our fMRI findings, we performed experiments demonstrating the preserved integrity of the BBB in R6/2 mice and confirmed the validity of our steady-state rCBV findings. Both T₂ and T₂^*-weighted images- each with its own advantages- can be used in rCBV Gd studies. We focused our report on T₂-weighted changes as they provide a more accurate measure of the neuronal metabolism, to which microvascular CBV is most tightly coupled (Weisskoff et al., 1994; Simonsen et al., 2000).

In the present work, MRI-rCBV maps identified an early and progressive change in the HD group. While 6-week old R6/2 mice showed focal increases of rCBV in different areas of the neocortex and striatum, older mice (14 weeks) also demonstrated rCBV elevation in the hippocampus and thalamus. Comparable to our findings in the young R6/2 mice, fMRI-BOLD studies in patients with more than twelve years to estimated onset of manifest HD (Zimbelman et al., 2007) exhibited increased bold signal in several neocortical regions. In contrast to these findings in mice, patients demonstrated decreased activation in the striatum, and cortical hyperactivation that converts to hypoactivation as one nears closer to onset of disease.

Correlation of Structure and Function with MRI

It has been proposed that brain tissue can lay spurious metabolic decline, owing mainly to focal tissue loss (Powers et al., 2007). We found that anatomical volumes and normalized rCBV values were inversely correlated and only significantly so in the neocortex of R6/2 mice. Covariation of rCBV group results with tissue loss did not affect the rCBV findings; this exemplifies the view that the relation between macro-structural and functional changes in microvasculature, such as rCBV is not straightforward. Finally, the lack of correlation between CAG number and rCBV values in R6/2 mice may be in part due to the low range of CAG values (243–262). Alternatively, it is possible that rCBV maps cannot detect more subtle differences in brain function related to the length of CAG repeats in R6/2 mice.

Neurovascular Coupling in R6/2 mice

In contrast to several studies providing evidence that CBF and BOLD contrast signals correlate poorly with spiking activity (Logothetis et al., 2001; Kayser et al., 2004) CBF and BOLD signals have been shown to correlate strongly with LFPs, which perhaps indicate the level of neuronal synchronization in a circuit (Lauritzen, 2005; Chaigneau et al., 2007). Given the aforementioned consideration and the relationship between CBV and CBF (Rostrup et al., 2005), we expected that elevated regional rCBV might reveal increased baseline LFPs activity, if neurovascular coupling to neuronal energy demand was indeed preserved in R6/2 mice. Our observation of spontaneous events resembling inter-ictal discharges (Figure 3C) and increased susceptibility to GABA_A blocker induced epileptiform discharges in R6/2 mice in different brain regions (Figures 4B–E) suggests that this was the case.

The relationship of CBF or CBV to neuronal activity is often assessed simultaneously, particularly for evoked responses. Given the steady state nature of rCBV measurements in our analysis, a separate evaluation of these two functions was informative.

Increased neuronal activity in 6 week-old R6/2 mice was also evidenced by increased susceptibility to audiogenic seizures. Note that complex discharges in cortical slices with shorter latencies to in vivo seizure onset induced by GABA_A blockers in R6/2 mice has been reported (Cummings et al., 2009).

Mismatch Between Indictors of Oxygen Consumption and Glucose Utilization

2DG maps from 6-week-old R6/2 mice demonstrated a global decrease in average LCGU in comparison to WTs, contrasting against global increase in rCBV values of R6/2 mice at the same age. LCGU was significantly lower in R6/2 mice with and without accounting for regional atrophy in regions where abnormally high rCBV values were observed (Figures 1D&5B) - mainly St, PMC and PFC. While TC demonstrated elevated rCBV, it revealed no significant difference in 2DG uptake. There was 100% concordance of the canonical relationship of rCBV to LGCU measures in those areas that showed no significant rCBV group differences. The lack of significance in TC LGCU values may be related to the difference in sensitivity of the techniques and/or the lower number of subjects in the 2DG experiments. These results were rather unexpected since previous studies found that basal rCBV values correlate with PET-FDG measures (Gonzalez et al., 1995). Nevertheless our results are consistent with reports that have recognized in HD patients (Antonini et al., 1996) and in mouse models (Tkac et al., 2007; Zhang et al., 2010) decreased brain glucose consumption and abnormal mitochondrial function in different organs (Mochel and Haller, 2011).

The mismatch of a functional brain imaging signal (rCBV) and glucose uptake (2DG) in R6/2 mice could be interpreted at least in three possible scenarios, which are not mutually exclusive: First, it is possible that multiregional neuronal loss and decrease in neuropil observed in R6/2 mice (Li et al., 2005) occurs at a rate faster than the observed microvasculature effects. It has been well documented that there is a mutual influence of nerves and blood vessels on the processes of angiogenesis and axonal guidance, respectively (Weinstein, 2005; Bell et al., 2010). Given this coordinated bidirectional influence, it can be expected that neuronal loss results in vascular reorganization. To this end, it is important to emphasize that the contrast-induced signal changed observed in our MRI sequence (fast spin echo --ΔR₂) is heavily sensitive to changes in microvasculature, and particularly those in the venular system (van Zijl et al., 1998; Wu et al., 2003). Considering the capacitance of the post-capillary venules and the overcompensation of CBF and CBV responses to LFPs, it is feasible to envision relatively large rCBV in a structure despite its metabolic decline. Furthermore, this sustained activity can exist in the setting of impaired glucose utility and perhaps the observed vascular effects serve to re-establish metabolic function in neurons with impaired glucose utilization. Second, it is also possible that decreased glucose utilization represents abnormal activity (silence) in interneuron populations. Though interneurons are believed to be relatively spared in HD (Vonsattel et al., 1985), recent reports have identified a reduction of cortical interneuronal function in R6/2 mice (Cummings et al., 2009). Decreasing interneuron metabolism would produce a robust effect in 2DG signal focally as, at least in the cortex, interneurons have stronger 2DG labeling relative to excitatory neurons (McCasland and Hibbard, 1997). Impaired interneuron function also provides a potential mechanism for the neuronal hyperexcitability observed in R6/2 mice. Finally the abnormal brain glucose metabolism observed may also be related to dysfunction of glial cells. It has been shown that glia in R6/2 mice, accumulate intranuclear htt. R6/2 glial cell dysfunction can contribute, for instance, to the inability to clear extracellular glutamate normally (Shin et al., 2005). It is therefore possible that decreased 2DG uptake represents, at least in part, the putative functional abnormalities observed in glial cells. This possibility does not exclude the other two hypotheses and in fact it provides an additional mechanism of increased neuronal excitability in R6/2 mice. The metabolic interaction between neurons and glia is complex and not free of controversies. Our results cannot conclude which of the cell types is more drastically affected. The current assumption is that glial energy needs range from 5% to 20% (Attwell and Laughlin, 2001; Rothman et al., 2003), whereas 80% to 95% of the energy needs concern the neuronal population. Other reports challenge this assumption by demonstrating equal amounts of 2DG taken up by the two cell populations in the resting brain (Nehlig et al., 2004). From these studies it is clear that brain energy metabolism requires further research at the physiological level and in diseases such as HD.

The possibility that rCBV maps fail to indicate neuronal function appropriately is highly unlikely given the high spatial concordance between rCBV and 2-deoxyglucose maps in wild type mice (Figure 6). Since rCBV mapping may herald the earliest dysfunction seen in HD, it can potentially serve as a marker of disease progression and treatment efficacy in the setting of pharmaceutical interventions.