Bringing CLARITY to gray matter atrophy
Rory D. Spence a
, Florian Kurth a
, Noriko Itoh b
, Chandler R.L. Mongerson a
, Shannon H. Wailes a
,
Mavis S. Peng b
, Allan J. MacKenzie-Graham a,
⁎
a
Ahmanson-Lovelace Brain Mapping Center, Department of Neurology, University of California, Los Angeles, Los Angeles, CA, USA
b
Multiple Sclerosis Program, Department of Neurology, University of California, Los Angeles, Los Angeles, CA, USA
a b s t r a c ta r t i c l e i n f o
Article history:
Accepted 10 July 2014
Available online 16 July 2014
Gray matter atrophy has been shown to be a strong correlate to clinical disability in multiple sclerosis (MS) and
its most commonly used animal model, experimental autoimmune encephalomyelitis (EAE). However, the relationship
between gray mater atrophy and the spinal cord pathology often observed in EAE has never been
established. Here EAE was induced in Thy1.1-YFP mice and their brains imaged using in vivo magnetic resonance
imaging (MRI). The brains and spinal cords were subsequently optically cleared using Clear Lipid-exchanged
Acrylamide-hybridized Rigid Imaging-compatible Tissue-hYdrogel (CLARITY). Axons were followed 5 mm longitudinally
in three dimensions in intact spinal cords revealing that 61% of the axons exhibited a mean of 22 axonal
ovoids and 8% of the axons terminating in axonal end bulbs. In the cerebral cortex, we observed a decrease in the
mean number of layer V pyramidal neurons and a decrease in the mean length of the apical dendrites of the remaining
neurons, compared to healthy controls. MRI analysis demonstrated decreased cortical volumes in EAE.
Cross-modality correlations revealed a direct relationship between cortical volume loss and axonal end bulb
number in the spinal cord, but not ovoid number. This is the first report of the use of CLARITY in an animal
model of disease and the first report of the use of both CLARITY and MRI.
© 2014 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY NC-SA license
(http://creativecommons.org/licenses/by-nc-sa/3.0/).
Introduction
Multiple sclerosis (MS) is a putative autoimmune demyelinating disease
with a neurodegenerative component whose target organ is the
brain and spinal cord. Progressive gray matter atrophy is a well documented
feature of MS and is widely considered as a marker of irreversible
tissue damage (Chard and Miller, 2009; Charil et al., 2007; Pirko
et al., 2007). Quantitative magnetic resonance imaging (MRI) studies indicate
that gray matter atrophy occurs even in the earliest stages of disease
(Calabrese et al., 2007; Dalton et al., 2004; De Stefano et al., 2003)
and is more closely related to physical disability and cognitive impairment
than either T2 or T1 white matter lesion volume (Amato et al.,
2004; Chard et al., 2002; Sailer et al., 2003; Sanfilipo et al., 2006). Indeed,
it is now considered to be one of the most relevant markers of permanent
disability in MS (Bermel and Bakshi, 2006; Fisher et al., 2002;
van den Elskamp et al., 2010).
More than a decade ago, Ferguson demonstrated axonal ovoids as
the result of axonal dysfunction or injury (Ferguson et al., 1997) and
Trapp further showed axonal end bulbs and axonal transection in MS lesions
in the brain (Trapp et al., 1998), highlighting that axonal injury
and loss are important components of MS pathology. However, it is unknown
if there is a relationship between gray matter atrophy and axonal
injury and transection. These questions cannot be answered in
patients with MS, but they can be addressed in the most widely used
mouse model of MS, experimental autoimmune encephalomyelitis
(EAE).
EAE has been used extensively to understand immune and neurodegenerative
mechanisms in MS. Gray matter atrophy in both the cerebellum
and cerebral cortex has been observed in MS (De Stefano et al.,
2003; Edwards et al., 1999; Iannucci et al., 1999; Sailer et al., 2003)
and has recently been reported in vivo in EAE (MacKenzie-Graham
et al., 2009, 2012). Cerebellar and cerebral cortex atrophy were found
to correlate strongly with Purkinje cell (MacKenzie-Graham et al.,
2009) and cortical neuron number (MacKenzie-Graham et al., 2012) respectively,
suggesting that gray matter atrophy is closely related to neuronal
loss. Furthermore, seminal work has demonstrated inflammation,
demyelination, neuronal, and axonal loss in the spinal cord (Bannerman
et al., 2005; Cross et al., 1993; Pender, 1987; Wujek et al., 2002) and
more recently in the brain (MacKenzie-Graham et al., 2009; Mangiardi
et al., 2011; Zhu et al., 2003; Ziehn et al., 2010), a recapitulation of many
findings in MS (Bo et al., 2003; Peterson et al., 2001; Trapp et al., 1998).
However, all of these observations are based on sparse sampling of the
brain and spinal cord, not a complete evaluation of these processes
throughout the entire CNS. The ability to evaluate these measures in
NeuroImage 101 (2014) 625–632
⁎ Corresponding author at: Department of Neurology, David Geffen School of Medicine
at UCLA, 710 Westwood Plaza, Los Angeles, CA 90095, USA.
E-mail address: amg@ucla.edu (A.J. MacKenzie-Graham).
http://dx.doi.org/10.1016/j.neuroimage.2014.07.017
1053-8119/© 2014 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY NC-SA license (http://creativecommons.org/licenses/by-nc-sa/3.0/).
Contents lists available at ScienceDirect
NeuroImage
journal homepage: www.elsevier.com/locate/ynimg
three dimensions (3D) across an intact CNS would provide unique insights
into relationships between these processes during disease.
Tissues are opaque to conventional light microscopy due primarily
to light scattering by lipids (Chung et al., 2013; Susaki et al., 2014). In
order to permit microscopic imaging of intact preparations, several
techniques for the optical clearing of tissue have recently been developed.
One of the earliest optical clearing methods developed used a
mixture of benzyl alcohol and benzyl benzoate (BABB) to match the refractive
index of fixed tissue (Dodt et al., 2007). This approach yielded
only partial clearing of myelinated tissues and the clearing solutions
led to the rapid loss of fluorescent signals (Erturk et al., 2012a). Nonetheless,
the use of organic solvents for refractive index matching has
produced very fast optical clearing times. 3D Imaging of Solvent Cleared
Organs (3DISCO) can clear a brain in approximately 1 day (Erturk et al.,
2012a), however, the half-life of the fluorescent signal is 1–2 days and
YFP is particularly unstable in the final clearing solution. Scale uses a
water-based clearing solution that preserves fluorescent signals, but requires
long incubation times (weeks to months) and yields incomplete
tissue clearing (Hama et al., 2011). SeeDB makes use of a hydrophilic
optical clearing solution containing fructose and thiols to achieve
rapid tissue clearing (3 days), but, like Scale, does not yield complete
clearing of adult brain tissues (Ke et al., 2013).
Clear Lipid-exchanged Acrylamide-hybridized Rigid Imagingcompatible
Tissue-hYdrogel or CLARITY is a recently developed optical
clearing technology that permits intact imaging of the entire brain
with minimal protein loss while preserving native fluorescence
(Chung et al., 2013; Tomer et al., 2014). The CLARITY protocol calls for
aggressively removing lipids by electrophoresis, leading to a clearing
time of approximately 10 days. However, this requires the use of specialized
electrophoresis devices that must be operated within a narrow
range of parameters to yield meaningful results. The replacement of the
electrophoretic tissue clearing step with passive diffusion permits for
greater scalability and improved reproducibility, albeit at the cost of longer
incubation times (Tomer et al., 2014).
The purpose of this study was to examine the relationship between
gray matter atrophy, as visualized by MRI, and axonal integrity, as visualized
by CLARITY. Here, EAE was induced in Thy1-YFP+
mice and
in vivo MRI was collected followed by optical clearing of both brains
and spinal cords using the “passive clearing” CLARITY technique.
Materials and methods
Overview
EAE was induced in Thy1.1-YFP+
C57BL/6 mice and typical clinical
disease was observed, with signs beginning at day 12 and a mean
peak score of 4.1 at day 21 (Fig. 1). In vivo MRIs were collected from
five mice with EAE 25 days after disease induction and from six ageand
sex-matched healthy controls. The mice were then sacrificed,
CLARITY performed on their CNS, and 3D confocal images collected.
Animals
Male B6.Cg-Tg(Thy1-YFP)HJrs/J mice (Jackson Laboratories, Bar Harbor,
ME) between 8 and 10 weeks of age were used for all experiments.
Mice were maintained in a 12 h dark/light cycle with access to food and
water ad libitum. All procedures were performed in accordance to the
guidelines of the National Institutes of Health and the Chancellor's
Animal Research Committee of the University of California, Los Angeles
Office for the Protection of Research Subjects.
Experimental autoimmune encephalomyelitis
Mice were immunized s.c. with MOG peptide 35–55 (300 μg/
mouse) and Mycobacterium tuberculosis (500 μg/mouse) emulsified in
complete Freund's adjuvant, in a volume of 0.1 ml/mouse over the
right-draining inguinal and axillary lymph nodes. One week later a
booster immunization was delivered s.c. over the contralateral lymph
nodes. Pertussis toxin (500 ng/mouse; List Biological Laboratories,
Campbell, CA) was injected intraperitoneally on days 0 and 2 (Liu
et al., 2003; Suen et al., 1997). EAE was graded on a scale of 0–5,
based on difficulty with ambulation (Pettinelli and McFarlin, 1981).
MRI acquisition
Mice were anesthetized with isoflurane and their heads secured
with bite and ear bars. Respiration rate was monitored and the mice
were maintained at 37 °C using a circulating water pump. In vivo magnetic
resonance imaging was performed using a 200 mm horizontal
bore 7.0 T Bruker imaging spectrometer with a micro-imaging gradient
insert with a maximum gradient strength of 100 G/cm and 30 mm birdcage
RF coil (Bruker Instruments, Billerica, MA). An actively decoupled
quadrature surface coil array was used for signal reception and a 72mm
birdcage coil was used for transmission. Images were acquired
and reconstructed using ParaVision 5.1 software. Imaging parameters
were as follows: rapid-acquisition with relaxation enhancement
(RARE) sequence, matrix dimensions = 256 × 128 × 64; field of
view = 3.84 cm × 1.92 cm × 0.96 cm; repetition time (TR) =
3500 ms; apparent time to echo (apparent TE) = 32 ms; echo train
length = 16; and total scan time = 37 min. Spatial resolution was
150 μm3
per voxel.
MRI analysis
Images were skull-stripped using the Brain Surface Extractor (BSE)
and residual non-brain signal was removed by a single operator (RDS)
manually editing the masks using BrainSuite 11a (Shattuck and Leahy,
2001) and bias-field inhomogeneities were removed using the N3 correction.
After inhomogeneity correction, a minimum deformation atlas
(MDA) was produced as described (MacKenzie-Graham et al., 2009).
Images were aligned to the MDA using Alignlinear (AIR) (Woods et al.,
1998a,b) to permit the direct comparison of images in a standard
space, correcting both gross size differences and gross intensity differences,
yet preserving anatomically significant local changes that a
non-linear warp may erase. Following creation of this atlas, cerebral cortices
were manually labeled on the atlas. The labels were then warped
onto the individual spatially normalized images and manually corrected
by a single operator (AMG) to produce standardized estimates of gray
matter volumes in individual subjects. An intensity rescaling cost function
was used in order to intensity-normalize the images. All automated
image processing was performed using the LONI Pipeline Processing Environment
(Rex et al., 2003) on an 8-processor core Mac Pro computer
(Apple, Cupertino, CA).
Fig. 1. Experimental autoimmune encephalomyelitis. Graph of the mean clinical disease
score in EAE (green) and healthy control (black) mice plotted against disease duration.
Mean peak disease reached a score of 4.1 at day 21.
626 R.D. Spence et al. / NeuroImage 101 (2014) 625–632
Cerebral cortex labels were based on the Mouse Atlas Project 2003
mouse brain atlas (MacKenzie-Graham et al., 2004). For clarity and consistency,
our cerebral cortex label was bounded ventrally by the plane
inferior to the most anterior point of the corpus callosum at midline. Importantly,
this label contained the somatosensory regions (primary and
secondary) and the motor cortex (primary and secondary) (MacKenzieGraham
et al., 2012). Additional anatomical information was obtained
from the Franklin and Paxinos mouse brain atlas (Franklin and
Paxinos, 2008).
To determine the reproducibility of our in vivo MRI data, 10 wildtype
female C57BL/6J mice were scanned in vivo and brain volumes
were measured. The average brain volume was 436 mm3
with a standard
error of 2.3 mm3
(95% confidence interval, 432–440; data not
shown). To establish scan–rescan reliability a single wild-type female
C57BL/6J mouse was imaged in vivo four times on separate days. The
average brain volume was 438 mm3
with a standard error of 1.8 mm3
(95% confidence interval, 434–442; data not shown).
CLARITY
All tissue were optically cleared using the CLARITY protocol (Chung
et al., 2013) (http://clarityresourcecenter.org/) modified for “passive
clearing” (Tomer et al., 2014). Briefly, EAE and healthy control mice
were anesthetized and perfused with 25 ml of ice cold 0.1 M PBS followed
by 25 ml of ice cold hydrogel. Each brain and spinal cord were quickly
dissected and placed in 20 ml of cold hydrogel in a 50 ml conical tube for
48 h at 4 °C. After 48 h the 50 ml tube lids were twisted open, yet not
removed, and placed in a desiccation chamber (McMaster-Carr, Santa
Fe Springs, CA). The desiccation chamber was then attached to a vacuum
for 10 min to allow for a negative pressure to form inside. Immediately
after the 10 min the chamber was purged with nitrogen gas. The
desiccation chamber lid was removed and the 50 ml tube lid was quickly
screwed shut to prevent exposure to the outside air. The 50 ml tubes
were then placed in a 37 °C water bath for 4–5 h to allow for adequate
polymerization of the hydrogel and tissue. After the hydrogel was polymerized,
excess gel was removed by placing the tissue in a kimwipe
(Kimberly-Clark, Neenah, WI) and gently rubbing. The whole brain
was then cut sagittally down the midline to generate two hemispheres.
Both hemispheres were placed into one 50 ml tube with clearing solution
(200 mM boric acid, 4% SDS, pH 8.5) and placed in an incubator
with agitation at 37 °C. Each spinal cord was placed into its own 50 ml
tube. Every 24 h the clearing solution was replaced with fresh clearing
solution until the tissue had completely cleared. This process relies entirely
on passive diffusion to remove all lipids from the tissue, the incubation
time required was substantial: 5 weeks for the brain and
3 weeks for the spinal cord. This process, known as “passive clearing”
(Tomer et al., 2014) was used instead of electrophoretic clearing
(Chung et al., 2013). Passive clearing does not require dedicated electrophoretic
equipment and can easily be scaled to allow complete clearing
of all brains and spinal cords simultaneously with high reproducibility
and low variation between subjects. Once clear, tissue was submerged
in FocusClear (CelExplorer Labs, Hsinchu, Taiwan) for 24 h at 37 °C to
allow for refractive index matching. Tissue was then placed in a
35 mm glass bottom dish (Ted Pella, Redding CA) with a thin barrier
of BluTack putty (Bostick, Fort Wayne, IN) around the tissue to prevent
tissue and FocusClear from moving. Tissue was then immediately used
for microscopy.
Microscopy
Confocal laser scanning microscopy was performed at the CNSI
Advanced Light Microscopy/Spectroscopy Shared Resource Facility at
UCLA. A Leica TCS-SP5 AOBS confocal microscope was used in combination
with Leica Application Suite Advanced Fluorescence (LAS AF) software
(Leica Microsystems, Inc., Buffalo Grove, IL). The resonant scanner
was used at a speed of 8000 Hz. 5× images of the spinal cord and
cerebral cortex were taken with a resolution of 512 × 512 and a line average
of 8. The 5× objective used was a Lecia HC PL FLUOTAR with an NA
of 0.15. The dimensions of each voxel were 1.5 μm × 1.5 μm × 44.31 μm.
10× images of the spinal cord and cerebral cortex were taken with a
Fig. 2. Thy1-YFP expression in the spinal cord. A 3D image stack of Thy1-YFP expression in
intact cervical and thoracic spinal cord in a healthy control (CON) (A) and an EAE mouse
(B) imaged at 5× with a pixel resolution of 1.5 μm × 1.5 μm × 44.31 μm. (C) Total spinal
cord (SP) Thy1-YFP expression was measured in a 10 mm long spinal cord. EAE mice demonstrated
a statistically significant reduction of 20% in spinal cord Thy1-YFP expression
compared to healthy controls. n = 5. *p = 0.0099. T-test analysis. Error bars indicate SEM.
627R.D. Spence et al. / NeuroImage 101 (2014) 625–632
resolution of 512 × 512 and a line average of 8. The 10× objective used
was a Leica HCX PL FLUOTAR with an NA of 0.3. The dimensions of each
voxel were 760 nm × 760 nm × 11.02 μm. All images were captured
using the same acquisition parameters, specifically exposure time. LAS
AF stitched the images together to form large high-resolution images.
Vaa3D64 was used to make to make bitmap still frames of image stacks
(http://penglab.janelia.org). MPEG Streamclip was used to make mp4
movies from the bitmap still frames (http://www.squared5.com).
CLARITY analysis
To measure the total Thy1-YFP expression in the spinal cord and cerebral
cortex, individual 5× images were manually outlined in ImageJ
(http://rsbweb.nih.gov/ij/) by a single operator (RDS). Images were
then converted to greyscale and set to a universal threshold for all subjects.
Percent expression was then measured in each outlined image.
The number of ovoid positive axons was measured by randomly choosing
25 axons that had a contiguous length of at least 5 mm from our 5×
images. An axon was considered ovoid positive if it had at least one
ovoid along its length. The number of ovoids per axon was measured
by choosing 25 ovoid positive axons that had a contiguous length of at
least 5 mm. The number of ovoids present on each axon was then manually
counted. The number of axons that terminated in end bulbs was
measured by randomly choosing 25 axons from the dorsal corticospinal
tract. An axon was considered end bulb positive if it ended in a swollen
bulb and had no other visible connection. The number of cortical layer V
pyramidal neurons was counted in 5× images of an entire hemisphere
of the brain. For clarity and consistency, the cerebral cortex boundaries
were the same as those used in MRI, bounded ventrally by the plane
inferior to the most anterior point of the corpus callosum at midline. Importantly,
this label contained somatosensory (primary and secondary)
and motor cortex (primary and secondary) (MacKenzie-Graham et al.,
2012). This area was outlined in each image of the 3D stacks and then
converted to greyscale. Images were then set to a universal threshold
for all subjects. The Image-based Tool for Counting Nuclei (ITCN;
http://rsb.info.nih.gov/ij/plugins/itcn.html) plugin was then used
to count only the somas of layer V pyramidal neurons. Dendrite length
was measured in layer V pyramidal neurons in the somatosensory
and motor cortices by randomly choosing 25 neurons and manually
tracing the dendrites in three dimensions from the soma to their most
distal extent near the dorsal surface of the cortex in ImageJ from our
10× images.
Statistical analysis
MRI and CLARITY data were analyzed using a Welch's t-test. Regression
analysis was also performed and significance was determined at
the 95% confidence interval. All statistical analyses were performed
using Prism 6 software (Graphpad, La Jolla, CA).
Fig. 3. Spinal cord pathology. Lateral view of a 3D image stack of the dorsal corticospinal tract from cervical to thoracic segments from a healthy control (CON) (A) and an EAE mouse (B/C).
Note the absence of axonal ovoids in the healthy control compared to the EAE mouse (arrows). (C/D) Dorsal corticospinal tract Thy1-YFP+
axons ended in axonal bulbs in the EAE mouse
(arrows). (E) 61% of axons from the EAE mice had at least one ovoid on each axon. (F) Ovoid positive axons contained an average of 22 ovoids. (G) 8% of axons from the EAE mice had at
least one axonal end bulb. Virtually no ovoids nor end bulbs were observed in healthy control mice. n = 5. *p = 0.0003, **p = 0.002, ***p = 0.0033. T-test analysis. Error bars indicate SEM.
628 R.D. Spence et al. / NeuroImage 101 (2014) 625–632
Results
Axonal ovoids and end bulbs in the spinal cord
Intact 10 mm long spinal cord sections containing cervical and thoracic
segments from mice with EAE and healthy controls were optically
cleared using CLARITY and imaged (Figs. 2A/B, Movie 1). The total
amount of Thy1-YFP expression was measured and a loss of 20% of
Thy1-YFP fluorescence was observed in EAE mice compared to healthy
controls (Fig. 2C) (p = 0.0099).
A distinguishing feature of axonal damage in both MS and EAE is the
presence of axonal ovoids. These ovoids are thought to represent disruption
of fast axonal transport (Koo et al., 1990; Sherriff et al., 1994)
or alterations of axonal microtubules that lead to Wallerian degeneration
(Erturk et al., 2007; Imitola et al., 2011). Large numbers of axonal
ovoids were observed in the dorsal corticospinal tract of the optically
cleared spinal cords in EAE mice (Figs. 3A–D). Since CLARITY permits
the examination of pathology in intact tissue, the number of axonal
ovoids in dorsal corticospinal tract axons over a 5 mm length in three dimensions
was quantified. 61% of axons in the dorsal corticospinal tract
exhibited ovoids in EAE mice, with virtually none present in normal
controls (Fig. 3E) (p = 0.0003). The number of ovoids present on each
ovoid positive axon was counted next. Surprisingly, ovoid positive
axons contained an average of 22.2 ± 3.1 ovoids over a length of
5 mm, with some axons containing more than 60 ovoids (Fig. 3F). The
number of transected axons that ended in a swollen end bulb was
then measured. 8% of axons in the dorsal corticospinal tract terminated
in end bulbs in EAE mice, with virtually none in healthy control mice
(Fig. 3G) (p = 0.0033). While most ovoids appeared to be clustered
near areas of axonal transection, ovoids were also present outside
areas of clustering. Similarly, axonal end bulbs appeared most commonly
near clusterings of ovoids. However, axonal transection was also observed
outside of ovoid clusters, albeit to a lesser degree.
Cortical pathology
The observation of axonal pathology in the dorsal corticospinal tract
led to the examination of cortical pathology in the same mice. Intact cerebral
hemispheres from mice with EAE and healthy controls were optically
cleared using CLARITY and imaged (Figs. 4A/B, Movie 2). The
total amount of Thy1-YFP expression in the cerebral cortex was measured
and a loss of 17% was observed in EAE mice compared to healthy
controls (Fig. 4G) (p = 0.0038). Previous work by Porrero et al. had
shown that injection of a fluorescent tracer in layer V cells of the primary
motor and somatosensory cortex led to high expression of the fluorescent
label in the dorsal corticospinal tract of the thoracic spinal
cord, demonstrating that the axons of layer V pyramidal neurons project
through the dorsal corticospinal tract (Porrero et al., 2010). Given the
observation of axonal ovoids and end bulbs in the dorsal corticospinal
tract in EAE, the total number of layer V Thy1-YFP+
pyramidal neurons
in the cerebral cortices was counted. Healthy control mice had a mean
of 33,553 ± 274.9 layer V pyramidal neurons, whereas mice with
EAE had a mean of 31,157 ± 605.5 (Figs. 4C/D), a difference of 7%
(Fig. 4H) (p = 0.013).
Fig. 4. Cortical pathology. A 3D image stack of Thy1-YFP+
neurons in the cerebral cortex and hippocampus from healthy control (CON) (A) and EAE mice (B). Layer V pyramidal neurons in
healthy control (C) and EAE mice (D). Apical dendrites (arrows) from healthy control (E) and EAE mice (F). (G) EAE mice demonstrated a 17% decrease in Thy1-YFP expression in the
cerebral cortex compared to healthy control mice. (H) EAE mice demonstrated a loss of 7% of layer V pyramidal neurons compared to healthy control mice. (I) EAE mice demonstrated
a decrease in dendrite length of 25% compared to healthy control mice. n = 5. *p = 0.0038, **p = 0.013, ***p = 0.0001. T-test analysis. Error bars indicate SEM.
629R.D. Spence et al. / NeuroImage 101 (2014) 625–632
In addition to an overall decrease in the number of cortical layer V
pyramidal neurons, apical dendrite lengths also appeared to be decreased
in the surviving neurons in EAE mice compared to healthy controls
(Figs. 4E/F). The lengths of the apical dendrites of layer V pyramidal
neurons in the somatosensory and motor cortices were measured in
EAE mice and healthy controls. Healthy control mice had a mean apical
dendrite length of 595.2 ± 11.48 um, whereas EAE mice had a mean of
444.7 ± 30.81 um, a decrease of 25% (Fig. 4I) (p = 0.0001).
Cortical volume loss
Prior to sacrificing the mice, in vivo MRI scans at day 25 of EAE were
performed (Fig. 5A). Volumetry was used to determine if there were any
cortical volume differences between EAE and healthy controls. The volume
of the cerebral cortex in healthy control mice had a mean of 73.5 ±
0.77 mm3
and mice with EAE had a mean of 68.3 ± 0.99 mm3
, a difference
of 7% (Fig. 5B)(p = 0.0032), consistent with our previous findings
of cerebral cortical atrophy in C57BL6 mice with EAE (MacKenzieGraham
et al., 2012).
Cortical volume loss is correlated with axonal end bulbs in the spinal cord
The relationship between cortical atrophy by MRI and underlying
pathology by CLARITY was determined by regression analyses performed
on our in vivo MRI cortical volume data with our layer V pyramidal
cell counts, axonal ovoid, and end bulb CLARITY data. Interestingly, a
very strong direct correlation between cortical volume and cortical layer
V pyramidal neuron number was observed (R2
= 0.83, p = 0.032). Further,
a very strong negative correlation between in vivo MRI cortical volume
and axonal end bulbs in cord was also seen (R2
= 0.81, p = 0.040).
However, there was no correlation between in vivo MRI cortical volume
and axonal ovoids in cord (R2
= 0.01, p = 0.86). Together this showed
that smaller cerebral cortex volume by MRI is directly related to
permanent damage as measured by loss of layer V pyramidal cells in
the cerebral cortex and axonal end bulbs in the dorsal corticospinal
tract of the spinal cord.
Discussion
In MS, inflammatory activity and white matter lesions are only
weakly correlated with clinical disability (Brex et al., 2002). Conversely,
gray matter atrophy correlates strongly with clinical disease progression
(Rudick and Trapp, 2009; Vigeveno et al., 2012) and is thought to
be a biomarker for neurodegeneration (Gold and Voskuhl, 2009;
Vigeveno et al., 2012). To address the cellular mechanisms that drive
gray matter atrophy and neurodegeneration we must go beyond MRI
and use histology to observe the pathologies that underlie the changes
detected by MRI.
In the last decade, MS research has focused on the role of axonal pathology
and neuronal loss as the cause of permanent clinical disability,
with axonal loss now accepted as a major cause of irreversible disability
in MS. Axonal loss has been shown to occur in the settings of both acute
inflammatory demyelination (Trapp et al., 1998) and chronic demyelination
(Bjartmar et al., 2000; Lovas et al., 2000). In fact, recent animal
studies have shown not only axonal loss, but also neuronal loss in the
setting of inflammatory demyelination (Bannerman and Hahn, 2007;
Bannerman et al., 2005; MacKenzie-Graham et al., 2009, 2012;
Mangiardi et al., 2011; Ziehn et al., 2010).
Until now histological studies have focused on a sparse sampling of
the tissue in question, as it was not previously possible to do a complete
evaluation of these tissues 3-dimensionally throughout the entire CNS.
With CLARITY, imaging cell populations in 3D in intact tissue is possible
(Chung et al., 2013; Tomer et al., 2014). For example, a common pathological
characteristic of axons in both MS and EAE is the presence of
ovoids and end bulbs (Ferguson et al., 1997; Imitola et al., 2011; Trapp
et al., 1998). Yet, despite their hypothesized importance (Ferguson
et al., 1997; Imitola et al., 2011; Trapp et al., 1998), it has been difficult
to fully appreciate their frequency or relationship to each other along
an axon since most pathology on the spinal cord was performed
on transverse sections that bisect individual axons (Bannerman and
Hahn, 2007; Bannerman et al., 2005; Shriver and Dittel, 2006). Likewise,
physically sectioning the spinal cord into longitudinal slices can cause
data to be lost since axons will come into and out of the plane of section
(Erturk and Bradke, 2013).
Erturk and colleagues demonstrated the utility of optical clearing
methods for the analysis of axonal injury and regeneration. 3DISCO
(3-dimensional imaging of solvent cleared organs) (Erturk et al.,
2012a) permitted them to evaluate the capacity of axons to regenerate
after an injury in the spinal cord by visualizing the trajectories of the
regenerating axons in 3-dimensions through the spinal cord (Erturk
et al., 2012b). Similarly, using CLARITY we were able to measure the
number of axonal ovoids and end bulbs in axons for distances over 5
mm. Our results demonstrated that axonal ovoids and end bulbs are
common in the dorsal corticospinal tract in mice with EAE. By following
axons through the spinal cord we could determine the number of ovoids
per axon and the number of axonal end bulbs in the dorsal corticospinal
tract. Furthermore, we were able to visualize and count all Thy1-YFP+
layer V pyramidal neurons in the cerebral cortex, not just an estimate
based on counting neurons in widely-spaced serial sections. This is the
underlying strength of optical clearing methods, the ability to observe
phenomena in 3D that were previously only glimpsed at using physical
sectioning methods in 2D.
A very recently developed approach, CUBIC (clear, unobstructed
brain imaging cocktails and computational analysis), combines rapid,
extensive tissue clearing with the ability to perform immunohistochemistry
(Susaki et al., 2014). Future studies may benefit from the use of this
promising new method, allowing the combination of inherent fluorescence
with antibody probing, all in the context of an intact brain. This
will permit investigators to ask more insightful questions in order to
Fig. 5. Cortical volume loss in EAE. (A) A minimum deformation atlas comprising the
in vivo MRI scans of both healthy control (CON) and EAE mice demonstrating the extent
of the cerebral cortex delineation (red). (B) Cortical volume loss of 7% was seen in EAE
mice versus healthy controls. n = 5. *p = 0.0032. T-test analysis. Error bars indicate SEM.
630 R.D. Spence et al. / NeuroImage 101 (2014) 625–632
more fully understand the mechanisms that underlie gray matter
atrophy.
In vivo MRI in mice has become a promising and widely available
technique to examine brain morphology (Bock et al., 2006; Delatour
et al., 2006). Its quantitative three-dimensional format makes it an accurate
tool to measure volume (Badea et al., 2007; Kovacevic et al., 2005;
MacKenzie-Graham et al., 2009). MRI has permitted the repeated measure
of neuroanatomical structures to evaluate the progression of volume
loss in these structures over time in vivo (Lau et al., 2008;
MacKenzie-Graham et al., 2012; Zhang et al., 2010). In MS, gray matter
atrophy in the brain is measured using MRI to evaluate disease progression.
Using in vivo MRI, we demonstrated that EAE mice had decreased
cortical volumes, consistent with our previous report (MacKenzieGraham
et al., 2012). Here we used both CLARITY and in vivo MRI in
the same experimental animals, permitting us to determine that the
number of axonal end bulbs in the spinal cord is strongly correlated
with the volume of cerebral cortical gray matter. Importantly, axonal
ovoids in the spinal cord did not show a correlation with cerebral cortex
volume, suggesting that axonal damage short of transection is insufficient
to cause gray matter loss in the cortex. This is consistent with recent
findings demonstrating reversible axon damage in EAE (Nikic
et al., 2011). That said, it has also been postulated that the presence of
axonal ovoids will eventually lead to axonal loss and Wallerian degeneration
(Erturk et al., 2007; Imitola et al., 2011). Together these observations
suggest that there may be a critical window whereby axons may
be damaged but not yet transected. If a neuroprotective treatment
were initiated within this critical window, cortical atrophy may be
averted.
Conclusions
The use of both MRI and CLARITY provides a unique opportunity to
explore the relationship between gray matter atrophy and the pathologies
that underlie it. Here, the combination of these techniques allowed
us to appreciate that there is a relationship between cortical atrophy
and axonal end bulbs, but not axonal ovoids, an insight that suggests
that there may be a window of reversibility in axonal damage. The combination
of CLARITY with immunohistochemistry (Chung et al., 2013;
Tomer et al., 2014) may permit further interrogation of the CNS in EAE
to determine the extent of inflammation or demyelination or in fact
any number of other processes. Most importantly, the use of both MRI
and CLARITY as applied herein to EAE now sets a precedent for a novel
approach to better discern neuropathological processes in intact tissues
in 3D in a variety of neurodegenerative diseases.
Supplementary data to this article can be found online at http://dx.
doi.org/10.1016/j.neuroimage.2014.07.017.
Acknowledgments
This work was generously supported by the National Multiple
Sclerosis Society (NMSS) grant FG 1759A1/1 and the Conrad N. Hilton
Foundation. We thank Dr. Laurent Bentolila and Dr. Matthew Schibler
for their invaluable assistance with confocal microscopy. We thank
Dr. Stephen Sawiak for developing the pulse sequence used for magnetic
resonance imaging. We thank Dr. Rhonda Voskuhl for the insightful
discussion and review of the manuscript. We thank those that have contributed
to the CLARITY forum (http://forum.claritytechniques.org).
References
Amato, M.P., Bartolozzi, M.L., Zipoli, V., Portaccio, E., Mortilla, M., Guidi, L., Siracusa, G.,
Sorbi, S., Federico, A., De Stefano, N., 2004. Neocortical volume decrease in relapsingremitting
MS patients with mild cognitive impairment. Neurology 63, 89–93.
Badea, A., Ali-Sharief, A.A., Johnson, G.A., 2007. Morphometric analysis of the C57BL/6J
mouse brain. Neuroimage 37, 683–693.
Bannerman, P.G., Hahn, A., 2007. Enhanced visualization of axonopathy in EAE using
thy1-YFP transgenic mice. J. Neurol. Sci. 260, 23–32.
Bannerman, P.G., Hahn, A., Ramirez, S., Morley, M., Bonnemann, C., Yu, S., Zhang, G.X.,
Rostami, A., Pleasure, D., 2005. Motor neuron pathology in experimental autoimmune
encephalomyelitis: studies in THY1-YFP transgenic mice. Brain 128, 1877–1886.
Bermel, R.A., Bakshi, R., 2006. The measurement and clinical relevance of brain atrophy in
multiple sclerosis. Lancet Neurol. 5, 158–170.
Bjartmar, C., Kidd, G., Mork, S., Rudick, R., Trapp, B.D., 2000. Neurological disability correlates
with spinal cord axonal loss and reduced N-acetyl aspartate in chronic multiple
sclerosis patients. Ann. Neurol. 48, 893–901.
Bo, L., Vedeler, C.A., Nyland, H., Trapp, B.D., Mork, S.J., 2003. Intracortical multiple sclerosis
lesions are not associated with increased lymphocyte infiltration. Mult. Scler. 9,
323–331.
Bock, N.A., Kovacevic, N., Lipina, T.V., Roder, J.C., Ackerman, S.L., Henkelman, R.M., 2006. In
vivo magnetic resonance imaging and semiautomated image analysis extend the
brain phenotype for cdf/cdf mice. J. Neurosci. 26, 4455–4459.
Brex, P.A., Ciccarelli, O., O'Riordan, J.I., Sailer, M., Thompson, A.J., Miller, D.H., 2002.
A longitudinal study of abnormalities on MRI and disability from multiple sclerosis.
N. Engl. J. Med. 346, 158–164.
Calabrese, M., Atzori, M., Bernardi, V., Morra, A., Romualdi, C., Rinaldi, L., McAuliffe, M.J.,
Barachino, L., Perini, P., Fischl, B., Battistin, L., Gallo, P., 2007. Cortical atrophy is relevant
in multiple sclerosis at clinical onset. J. Neurol. 254, 1212–1220.
Chard, D., Miller, D., 2009. Grey matter pathology in clinically early multiple sclerosis:
evidence from magnetic resonance imaging. J. Neurol. Sci. 282, 5–11.
Chard, D.T., Griffin, C.M., Parker, G.J., Kapoor, R., Thompson, A.J., Miller, D.H., 2002. Brain
atrophy in clinically early relapsing-remitting multiple sclerosis. Brain 125, 327–337.
Charil, A., Dagher, A., Lerch, J.P., Zijdenbos, A.P., Worsley, K.J., Evans, A.C., 2007. Focal cortical
atrophy in multiple sclerosis: relation to lesion load and disability. Neuroimage
34, 509–517.
Chung, K., Wallace, J., Kim, S.Y., Kalyanasundaram, S., Andalman, A.S., Davidson, T.J.,
Mirzabekov, J.J., Zalocusky, K.A., Mattis, J., Denisin, A.K., Pak, S., Bernstein, H.,
Ramakrishnan, C., Grosenick, L., Gradinaru, V., Deisseroth, K., 2013. Structural and
molecular interrogation of intact biological systems. Nature 497, 332–337.
Cross, A.H., O'Mara, T., Raine, C.S., 1993. Chronologic localization of myelin-reactive cells
in the lesions of relapsing EAE: implications for the study of multiple sclerosis. Neurology
43, 1028–1033.
Dalton, C.M., Chard, D.T., Davies, G.R., Miszkiel, K.A., Altmann, D.R., Fernando, K., Plant, G.T.,
Thompson, A.J., Miller, D.H., 2004. Early development of multiple sclerosis is associated
with progressive grey matter atrophy in patients presenting with clinically isolated
syndromes. Brain 127, 1101–1107.
De Stefano, N., Matthews, P.M., Filippi, M., Agosta, F., De Luca, M., Bartolozzi, M.L., Guidi, L.,
Ghezzi, A., Montanari, E., Cifelli, A., Federico, A., Smith, S.M., 2003. Evidence of early
cortical atrophy in MS: relevance to white matter changes and disability. Neurology
60, 1157–1162.
Delatour, B., Guegan, M., Volk, A., Dhenain, M., 2006. In vivo MRI and histological evaluation
of brain atrophy in APP/PS1 transgenic mice. Neurobiol. Aging 27, 835–847.
Dodt, H.U., Leischner, U., Schierloh, A., Jahrling, N., Mauch, C.P., Deininger, K., Deussing,
J.M., Eder, M., Zieglgansberger, W., Becker, K., 2007. Ultramicroscopy: threedimensional
visualization of neuronal networks in the whole mouse brain. Nat.
Methods 4, 331–336.
Edwards, S.G., Gong, Q.Y., Liu, C., Zvartau, M.E., Jaspan, T., Roberts, N., Blumhardt, L.D.,
1999. Infratentorial atrophy on magnetic resonance imaging and disability in multiple
sclerosis. Brain 122 (Pt 2), 291–301.
Erturk, A., Bradke, F., 2013. High-resolution imaging of entire organs by 3-dimensional
imaging of solvent cleared organs (3DISCO). Exp. Neurol. 242, 57–64.
Erturk, A., Hellal, F., Enes, J., Bradke, F., 2007. Disorganized microtubules underlie the formation
of retraction bulbs and the failure of axonal regeneration. J. Neurosci. 27,
9169–9180.
Erturk, A., Becker, K., Jahrling, N., Mauch, C.P., Hojer, C.D., Egen, J.G., Hellal, F., Bradke, F.,
Sheng, M., Dodt, H.U., 2012a. Three-dimensional imaging of solvent-cleared organs
using 3DISCO. Nat. Protoc. 7, 1983–1995.
Erturk, A., Mauch, C.P., Hellal, F., Forstner, F., Keck, T., Becker, K., Jahrling, N., Steffens, H.,
Richter, M., Hubener, M., Kramer, E., Kirchhoff, F., Dodt, H.U., Bradke, F., 2012b.
Three-dimensional imaging of the unsectioned adult spinal cord to assess axon regeneration
and glial responses after injury. Nat. Med. 18, 166–171.
Ferguson, B., Matyszak, M.K., Esiri, M.M., Perry, V.H., 1997. Axonal damage in acute multiple
sclerosis lesions. Brain 120 (Pt 3), 393–399.
Fisher, E., Rudick, R.A., Simon, J.H., Cutter, G., Baier, M., Lee, J.C., Miller, D., WeinstockGuttman,
B., Mass, M.K., Dougherty, D.S., Simonian, N.A., 2002. Eight-year follow-up
study of brain atrophy in patients with MS. Neurology 59, 1412–1420.
Franklin, K.B.J., Paxinos, G., 2008. The Mouse Brain in Stereotaxic Coordinates, 3rd ed.
Academic Press, New York.
Gold, S.M., Voskuhl, R.R., 2009. Estrogen treatment in multiple sclerosis. J. Neurol. Sci. 286,
99–103.
Hama, H., Kurokawa, H., Kawano, H., Ando, R., Shimogori, T., Noda, H., Fukami, K., SakaueSawano,
A., Miyawaki, A., 2011. Scale: a chemical approach for fluorescence imaging
and reconstruction of transparent mouse brain. Nat. Neurosci. 14, 1481–1488.
Iannucci, G., Minicucci, L., Rodegher, M., Sormani, M.P., Comi, G., Filippi, M., 1999. Correlations
between clinical and MRI involvement in multiple sclerosis: assessment
using T(1), T(2) and MT histograms. J. Neurol. Sci. 171, 121–129.
Imitola, J., Cote, D., Rasmussen, S., Xie, X.S., Liu, Y., Chitnis, T., Sidman, R.L., Lin, C.P., Khoury,
S.J., 2011. Multimodal coherent anti-Stokes Raman scattering microscopy reveals
microglia-associated myelin and axonal dysfunction in multiple sclerosis-like lesions
in mice. J. Biomed. Opt. 16, 021109.
Ke, M.T., Fujimoto, S., Imai, T., 2013. SeeDB: a simple and morphology-preserving optical
clearing agent for neuronal circuit reconstruction. Nat. Neurosci. 16, 1154–1161.
Koo, E.H., Sisodia, S.S., Archer, D.R., Martin, L.J., Weidemann, A., Beyreuther, K., Fischer, P.,
Masters, C.L., Price, D.L., 1990. Precursor of amyloid protein in Alzheimer disease
631R.D. Spence et al. / NeuroImage 101 (2014) 625–632
undergoes fast anterograde axonal transport. Proc. Natl. Acad. Sci. U. S. A. 87,
1561–1565.
Kovacevic, N., Henderson, J.T., Chan, E., Lifshitz, N., Bishop, J., Evans, A.C., Henkelman, R.M.,
Chen, X.J., 2005. A three-dimensional MRI atlas of the mouse brain with estimates of
the average and variability. Cereb. Cortex 15, 639–645.
Lau, J.C., Lerch, J.P., Sled, J.G., Henkelman, R.M., Evans, A.C., Bedell, B.J., 2008. Longitudinal
neuroanatomical changes determined by deformation-based morphometry in a
mouse model of Alzheimer's disease. Neuroimage 42, 19–27.
Liu, H.B., Loo, K.K., Palaszynski, K., Ashouri, J., Lubahn, D.B., Voskuhl, R.R., 2003. Estrogen
receptor alpha mediates estrogen's immune protection in autoimmune disease. J.
Immunol. 171, 6936–6940.
Lovas, G., Szilagyi, N., Majtenyi, K., Palkovits, M., Komoly, S., 2000. Axonal changes in
chronic demyelinated cervical spinal cord plaques. Brain 123 (Pt 2), 308–317.
MacKenzie-Graham, A., Lee, E.F., Dinov, I.D., Bota, M., Shattuck, D.W., Ruffins, S., Yuan, H.,
Konstantinidis, F., Pitiot, A., Ding, Y., Hu, G., Jacobs, R.E., Toga, A.W., 2004. A multimodal,
multidimensional atlas of the C57BL/6 J mouse brain. J. Anat. 204, 93–102.
MacKenzie-Graham, A., Tiwari-Woodruff, S.K., Sharma, G., Aguilar, C., Vo, K.T., Strickland,
L.V., Morales, L., Fubara, B., Martin, M., Jacobs, R.E., Johnson, G.A., Toga, A.W., Voskuhl,
R.R., 2009. Purkinje cell loss in experimental autoimmune encephalomyelitis.
Neuroimage 48, 637–651.
MacKenzie-Graham, A., Rinek, G.A., Avedisian, A., Gold, S.M., Frew, A.J., Aguilar, C., Lin, D.R.,
Umeda, E., Voskuhl, R.R., Alger, J.R., 2012. Cortical atrophy in experimental autoimmune
encephalomyelitis: in vivo imaging. Neuroimage 60, 95–104.
Mangiardi, M., Crawford, D.K., Xia, X., Du, S., Simon-Freeman, R., Voskuhl, R.R., TiwariWoodruff,
S.K., 2011. An animal model of cortical and callosal pathology in multiple
sclerosis. Brain Pathol. 21, 263–278.
Nikic, I., Merkler, D., Sorbara, C., Brinkoetter, M., Kreutzfeldt, M., Bareyre, F.M., Bruck, W.,
Bishop, D., Misgeld, T., Kerschensteiner, M., 2011. A reversible form of axon damage
in experimental autoimmune encephalomyelitis and multiple sclerosis. Nat. Med.
17, 495–499.
Pender, M.P., 1987. Demyelination and neurological signs in experimental allergic encephalomyelitis.
J. Neuroimmunol. 15, 11–24.
Peterson, J.W., Bo, L., Mork, S., Chang, A., Trapp, B.D., 2001. Transected neurites, apoptotic
neurons, and reduced inflammation in cortical multiple sclerosis lesions. Ann. Neurol.
50, 389–400.
Pettinelli, C.B., McFarlin, D.E., 1981. Adoptive transfer of experimental allergic encephalomyelitis
in SJL/J mice after in vitro activation of lymph node cells by myelin basic protein:
requirement for Lyt 1+ 2- T lymphocytes. J. Immunol. 127, 1420–1423.
Pirko, I., Lucchinetti, C.F., Sriram, S., Bakshi, R., 2007. Gray matter involvement in multiple
sclerosis. Neurology 68, 634–642.
Porrero, C., Rubio-Garrido, P., Avendano, C., Clasca, F., 2010. Mapping of fluorescent
protein-expressing neurons and axon pathways in adult and developing Thy1eYFP-H
transgenic mice. Brain Res. 1345, 59–72.
Rex, D.E., Ma, J.Q., Toga, A.W., 2003. The LONI pipeline processing environment.
Neuroimage 19, 1033–1048.
Rudick, R.A., Trapp, B.D., 2009. Gray-matter injury in multiple sclerosis. N. Engl. J. Med.
361, 1505–1506.
Sailer, M., Fischl, B., Salat, D., Tempelmann, C., Schonfeld, M.A., Busa, E., Bodammer, N.,
Heinze, H.J., Dale, A., 2003. Focal thinning of the cerebral cortex in multiple sclerosis.
Brain 126, 1734–1744.
Sanfilipo, M.P., Benedict, R.H., Weinstock-Guttman, B., Bakshi, R., 2006. Gray and white
matter brain atrophy and neuropsychological impairment in multiple sclerosis. Neurology
66, 685–692.
Shattuck, D.W., Leahy, R.M., 2001. Automated graph-based analysis and correction of cortical
volume topology. IEEE Trans. Med. Imaging 20, 1167–1177.
Sherriff, F.E., Bridges, L.R., Gentleman, S.M., Sivaloganathan, S., Wilson, S., 1994. Markers of
axonal injury in post mortem human brain. Acta Neuropathol. 88, 433–439.
Shriver, L.P., Dittel, B.N., 2006. T-cell-mediated disruption of the neuronal microtubule
network: correlation with early reversible axonal dysfunction in acute experimental
autoimmune encephalomyelitis. Am. J. Pathol. 169, 999–1011.
Suen, W.E., Bergman, C.M., Hjelmstrom, P., Ruddle, N.H., 1997. A critical role for
lymphotoxin in experimental allergic encephalomyelitis. J. Exp. Med. 186, 1233–1240.
Susaki, E.A., Tainaka, K., Perrin, D., Kishino, F., Tawara, T., Watanabe, T.M., Yokoyama, C.,
Onoe, H., Eguchi, M., Yamaguchi, S., Abe, T., Kiyonari, H., Shimizu, Y., Miyawaki, A.,
Yokota, H., Ueda, H.R., 2014. Whole-brain imaging with single-cell resolution using
chemical cocktails and computational analysis. Cell 157, 726–739.
Tomer, R., Ye, L., Hsueh, B., Deisseroth, K., 2014. Advanced CLARITY for rapid and highresolution
imaging of intact tissues. Nat. Protoc. 9, 1682–1697.
Trapp, B.D., Peterson, J., Ransohoff, R.M., Rudick, R., Mork, S., Bo, L., 1998. Axonal transection
in the lesions of multiple sclerosis. N. Engl. J. Med. 338, 278–285.
van den Elskamp, I.J., Boden, B., Dattola, V., Knol, D.L., Filippi, M., Kappos, L., Fazekas, F.,
Wagner, K., Pohl, C., Sandbrink, R., Polman, C.H., Uitdehaag, B.M., Barkhof, F., 2010. Cerebral
atrophy as outcome measure in short-term phase 2 clinical trials in multiple
sclerosis. Neuroradiology 52, 875–881.
Vigeveno, R.M., Wiebenga, O.T., Wattjes, M.P., Geurts, J.J., Barkhof, F., 2012. Shifting imaging
targets in multiple sclerosis: from inflammation to neurodegeneration. J. Magn.
Reson. Imaging 36, 1–19.
Woods, R.P., Grafton, S.T., Holmes, C.J., Cherry, S.R., Mazziotta, J.C., 1998a. Automated
image registration: I. General methods and intrasubject, intramodality validation. J.
Comput. Assist. Tomogr. V22, 139–152.
Woods, R.P., Grafton, S.T., Watson, J.D., Sicotte, N.L., Mazziotta, J.C., 1998b. Automated
image registration: II. Intersubject validation of linear and nonlinear models. J.
Comput. Assist. Tomogr. 22, 153–165.
Wujek, J.R., Bjartmar, C., Richer, E., Ransohoff, R.M., Yu, M., Tuohy, V.K., Trapp, B.D., 2002.
Axon loss in the spinal cord determines permanent neurological disability in an animal
model of multiple sclerosis. J. Neuropathol. Exp. Neurol. 61, 23–32.
Zhang, J., Peng, Q., Li, Q., Jahanshad, N., Hou, Z., Jiang, M., Masuda, N., Langbehn, D.R.,
Miller, M.I., Mori, S., Ross, C.A., Duan, W., 2010. Longitudinal characterization of
brain atrophy of a Huntington's disease mouse model by automated morphological
analyses of magnetic resonance images. Neuroimage 49, 2340–2351.
Zhu, B., Luo, L., Moore, G.R., Paty, D.W., Cynader, M.S., 2003. Dendritic and synaptic pathology
in experimental autoimmune encephalomyelitis. Am. J. Pathol. 162, 1639–1650.
Ziehn, M.O., Avedisian, A.A., Tiwari-Woodruff, S., Voskuhl, R.R., 2010. Hippocampal CA1
atrophy and synaptic loss during experimental autoimmune encephalomyelitis,
EAE. Lab. Invest. 90, 774–786.
632 R.D. Spence et al. / NeuroImage 101 (2014) 625–632