Behavioral abnormalities precede neuropathological markers in rats transgenic for Huntington's disease

Abstract

Huntington's disease (HD) is caused by an expanded CAG repeat leading to the synthesis of an aberrant protein and to the formation of polyglutamine (polyQ)-containing inclusions and aggregates. Limited information is available concerning the association of neuropathological markers with the development of behavioral markers in HD. Using a previously generated transgenic rat model of HD (tgHD rat), we performed association studies on the time-course of behavioral symptoms (motor function, learning, anxiety) and the appearance of striatal atrophy, 1C2 immunopositive aggregates and polyQ recruitment sites, a precursor to these aggregates. At the age of 1 month, tgHD rats exhibited reduced anxiety and improved motor performance, while at 6 months motor impairments and at 9 months cognitive decline occurred. In contrast, polyQ recruitment sites appeared at around 6-9 months of age, indicating that HD-like behavioral markers preceded the appearance of currently detectable neuropathological markers. Interestingly, numerous punctate sites containing polyQ aggregates were also seen in areas receiving afferents from the densely recruiting regions suggesting either transport of recruitment-competent aggregates to terminal projections where initially 1C2 positive aggregates were formed or different internal properties of neurons in different regions. Furthermore, striatal atrophy was observed at the age of 12 months. Taken together, our findings support the hypothesis of a dynamic process leading to region- and age-specific polyQ recruitment and aggregation. The dissociation of onset between behavioral and neuropathological markers is suggestive of as yet undetected processes, which contribute to the early phenotype of these HD transgenic rats.

Full text

Behavioral abnormalities precede
neuropathological markers in rats transgenic
for Huntington’s disease
Huu Phuc Nguyen
1,4
, Philipp Kobbe
1
, Henning Rahne
1
, Till Wo
¨
rpel
1
, Burkard Ja
¨
ger
2
,
Michael Stephan
1
, Reinhard Pabst
1
, Carsten Holzmann
3
, Olaf Riess
4
, Hubert Korr
5
,
Orsolya Ka
´
ntor
5
, Elisabeth Petrasch-Parwez
6
, Ronald Wetzel
7
, Alexander Osmand
7
and Stephan von Ho
¨
rsten
1,8,
*
1
Department of Functional and Applied Anatomy and
2
Department of Psychosomatics and Psychotherapy, Medical
School of Hannover, Germany,
3
Department of Medical Genetics, University of Rostock, Germany,
4
Department of
Medical Genetics, University of Tu
¨
bingen, Germany,
5
Department of Anatomy and Cell Biology, RWTH Aachen
University, Germany,
6
Department of Neuroanatomy and Molecular Brain Research, University of Bochum, Germany,
7
Department of Medicine, University of Tennessee, Knoxville, TN, USA and
8
Experimental Therapy,
Franz-Penzoldt-Center, Friedrich-Alexander-University of Erlangen-Nu
¨
rnberg, Germany
Received July 5, 2006; Revised August 30, 2006; Accepted September 12, 2006
Huntington’s disease (HD) is caused by an expanded CAG repeat leading to the synthesis of an aberrant
protein and to the formation of polyglutamine (polyQ)-containing inclusions and aggregates. Limited infor-
mation is available concerning the association of neuropathological markers with the development of beha-
vioral markers in HD. Using a previously generated transgenic rat model of HD (tgHD rat), we performed
association studies on the time-course of behavioral symptoms (motor function, learning, anxiety) and the
appearance of striatal atrophy, 1C2 immunopositive aggregates and polyQ recruitment sites, a precursor
to these aggregates. At the age of 1 month, tgHD rats exhibited reduced anxiety and improved motor perform-
ance, while at 6 months motor impairments and at 9 months cognitive decline occurred. In contrast, polyQ
recruitment sites appeared at around 6–9 months of age, indicating that HD-like behavioral markers preceded
the appearance of currently detectable neuropathological markers. Interestingly, numerous punctate sites
containing polyQ aggregates were also seen in areas receiving afferents from the densely recruiting regions
suggesting either transport of recruitment-competent aggregates to terminal projections where initially 1C2
positive aggregates were formed or different internal properties of neurons in different regions. Furthermore,
striatal atrophy was observed at the age of 12 months. Taken together, our findings support the hypothesis of
a dynamic process leading to region- and age-specific polyQ recruitment and aggregation. The dissociation
of onset between behavioral and neuropathological markers is suggestive of as yet undetected processes,
which contribute to the early phenotype of these HD transgenic rats.
INTRODUCTION
Huntington’s disease (HD) is an autosomal dominantly
inherited neurodegenerative disorder. It is caused by an
expanded CAG trinucleotide repeat (.38) leading to the
synthesis of an aberrant protein (mutant huntingtin) with
an expanded N-terminal polyglutamine (polyQ) tract
(1,2). Clinically, the disease presents with progressive
emotional, motor and cognitive disturbances until death
within 15–20 years. No effective treatment is presently
available.
The development of therapies for HD requires preclinical
testing of drugs in animal models that reproduce the dysfunc-
tion and specific regional pathology observed in HD. In order
# The Author 2006. Published by Oxford University Press. All rights reserved.
For Permissions, please email: journals.permissions@oxfordjournals.org
*To whom correspondence should be addressed. Tel: þ49 91318523504; Fax: þ49 91318523502; Email: stephan.v.hoersten@ze.uni-erlangen.de
Human Molecular Genetics, 2006, Vol. 15, No. 21 3177–3194
doi:10.1093/hmg/ddl394
Advance Access published on September 19, 2006
at Universitaetsbibliothek Rostock on September 10, 2012http://hmg.oxfordjournals.org/Downloaded from

to increase the chance of identifying successful treatments, it
has been widely accepted that a compound should show posi-
tive outcomes in more than one animal model, and should be
reproducible in more than one laboratory prior to consider-
ation for use in clinical trials (3). Furthermore, a positive
outcome in more than one species would be very helpful to
bring promising compounds into clinical trials. Therefore,
we have recently generated a transgenic rat model of HD
with 51 CAG repeats (4). This HD rat model closely
resembles the human HD phenotype exhibiting emotional
disturbance, motor deficits and cognitive decline. In order
to take full advantage of this rat model of HD, a precise
characterization of the onset and progression of various beha-
vioral parameters as well as power calculations—useful for
the later determination of treatment benefits and the design
of therapeutic trials—are necessary. Using several behavioral
tests, here we present a detailed evaluation of onset and
time-course of a quantifiable phenotype that is present in
these rats concerning all three key systems affected in HD:
motor function, emotional condition and cognitive capabili-
ties. So far, only a few studies on learning and memory def-
icits in transgenic mouse models of HD, mainly in R6/2 mice
(5–7), and more recently in YAC128 mice (8) have been
published. This may be due to the fact that behavioral test
paradigms of cognition are sophisticated, more labor-
intensive and more difficult in mice. Moreover, most of
these tests were originally designed for rats (9,10). Therefore,
a rat model of HD represents a unique tool to examine learn-
ing and memory in the time-course of HD. We chose the
radial maze test of spatial learning in order to avoid con-
founding effects of motor function and high stress levels
which were observed in R6/2 mice in the Morris water
maze (5). Other early symptoms of HD are behavioral
changes and psychiatric symptoms that have hardly been
studied in animal models. There are only two studies, in
which anxiety changes and explorative behavior in R6/2
mice (11) and in N171-82Q mice (12) have been investi-
gated. But these studies have used the elevated plus maze
and the open field test that may cause problems when used
repeatedly in the same animals. Therefore, it is important
to choose a test paradigm that has been well validated and
allows repeated testing. We solved this problem using the
social interaction test of anxiety, which can be conducted
repeatedly within the same subjects (13,14).
In addition, this transgenic rat model reproduces the neu-
ropathological hallmarks of HD, namely the formation of
intranuclear and neuropil polyQ aggregates, which can be
detected by antibodies such as 1C2 (15) and EM48 (16).
However, the role that aggregation plays in the pathogenesis
of HD has been highly controversial, ranging from being
central to disease pathogenesis, to a benign epiphenomenon,
or even to being neuroprotective (17– 23). So far, attempts to
correlate the presence of aggregates with the onset of a phe-
notype have been complicated by the difficulties in detecting
and quantifying small aggregated forms of mutant huntingtin.
Using a new method based on the recruitment of tagged
polyQ peptides into existing aggregates, we have been able
to identify previously undetected aggregated huntingtin in
human HD brains (24,25). These structures have been
termed aggregation foci (AF) and are distinct from those
previously identified by immunohistochemical techniques.
With this new method, we evaluated the possible correlation
of onset and progression of behavioral phenotype with the
appearance of AF and 1C2-reactive aggregates. Furthermore,
we have determined striatal volume in tgHD rats at different
ages in order to detect the onset of striatal atrophy and to
obtain a possible quantitative outcome measure for thera-
peutic studies.
Our findings are consistent with previous studies suggesting
that behavioral abnormalities may precede the appearance of
detectable aggregates and striatal atrophy. But for the first
time this has been shown in another species than the mouse.
Furthermore, we demonstrate that detectable AF appears
shortly after the onset of behavioral symptoms suggesting
that, yet, undefined processes may contribute to the early phe-
notype of these HD transgenic rats. With the rat being superior
to the mouse for several approaches such as cognitive testing,
electrophysiology or neuroimaging, our findings provide a
valuable baseline for future studies.
RESULTS
Accelerod test
Motor coordination and balance of rats were measured using
an accelerod and the performance is displayed as rotations
per minute (rpm) and time (s) (Fig. 1). All rats used in this
study acquired the accelerod test quickly and reached a
stable level of performance before testing. Throughout the
study, wild-type rats showed a constant performance on the
accelerod. Transgenic rats exhibited a better performance
than wild-type littermates over the first 4 months of age
with a significant increase in motor coordination and balance
capabilities at 1 month of age. However, there was a slowly
progressive decline in performance of heterozygous transgenic
rats after 2 months of age, with poorer performance than wild-
type littermates at 6 months of age and a significantly poorer
performance than wild-type littermates at 8 and 9 months of
age. Homozygous transgenic animals showed a more severe
deterioration of motor coordination and balance capabilities
after 1 month of age, with poorer performance than heterozy-
gous and wild-type littermates at 5 months of age and a signifi-
cantly impaired motor coordination and balance capabilities
compared to heterozygous and wild-type littermates at 6, 7,
8 and 9 months of age.
The significance of the difference between groups was
confirmed by two-factorial ANOVA for repeated measure-
ments, which revealed a significant interaction between gen-
otype and rpm (F
16,272
¼ 5.11, P , 0.001) and between
genotype and time (F
16,272
¼ 5.19, P , 0.001). One-factorial
ANOVA showed a significant effect for genotype at 1
month (F
2,34
¼ 3.79, P , 0.05 for rpm and F
2,34
¼ 3.76,
P , 0.05 for time), 6 months (F
2,34
¼ 3.97, P , 0.05 for
rpm and F
2,34
¼ 3.64, P , 0.05 for time), 7 months
(F
2,34
¼ 4.95, P , 0.05 for rpm and F
2,34
¼ 5.19, P , 0.05
for time), 8 months (F
2,34
¼ 9.47, P , 0.001 for rpm and
F
2,34
¼ 9.40, P , 0.001 for time) and 9 months of age
(F
2,34
¼ 12.25, P , 0.001 for rpm and F
2,34
¼ 9,23,
P , 0,001 for time).
3178 Human Molecular Gen etics, 2006, Vol. 15, No. 21
at Universitaetsbibliothek Rostock on September 10, 2012http://hmg.oxfordjournals.org/Downloaded from

Beam walking test
The beam walking test was used to compare the fine motor
coordination and balance capabilities of tgHD rats and
control animals. Different levels of task difficulty were
achieved by varying the shape and cross-section of the
beams. We found no difference in the set of young rats
(age of 1– 6 months) on all beams tested (data not shown).
In the group of adult animals (age of 7 months, tested until
12 months of age), homozygous transgenic rats showed
significant difficulty in traversing the smallest round beam
beginning at the age of 8 months, as measured by their
increased latency to traverse the 1.5 cm wide beam, compared
with control rats (Fig. 2A). Heterozygous tgHD rats were sig-
nificantly slower in traversing the smallest beam only at the
age of 12 months. Starting at 9 months of age, homozygous
tgHD rats also made significantly more footslips than wild-
type rats on the smallest round beam, the frequency of
which increased with age (Fig. 2B). With increasing age,
this was also seen in heterozygous tgHD rats, however, this
only reached significance at the age of 11 months.
The significance of these observations was confirmed by
two-factorial ANOVA for repeated measurements, which
showed a significant effect of genotype on both the latency
and the number of footslips made while traversing the smallest
round beam (latency: F
2,31
¼ 2.09, P , 0.05; footslips:
F
2,31
¼ 2.73, P , 0.01).
Social interaction test
The social interaction test was used to assess anxiety-related
changes of socio-positive behaviors by analyzing the time
spent in active social interaction of two experimental subjects
within a novel environment. Figure 3 illustrates the time that
wild-type littermates, heterozygous and homozygous trans-
genic rats spent in social interaction. At all ages tested, trans-
genic rats spent significantly more time in social interaction
than their wild-type littermates indicating a reduced anxiety.
One-factorial ANOVA showed a significant effect for geno-
type at 1 month (F
2,39
¼ 67.90, P , 0.001), 2 months
(F
2,39
¼ 15.91, P , 0.001) and 7 months (F
2,39
¼ 5.27,
P , 0.05) of age.
Elevated plus maze test
At all ages tested, tgHD rats spent a significantly higher per-
centage of time on the open arms than did their wild-type lit-
termates (see Fig. 4A; ANOVA, genotype: 3 months:
F
2,32
¼ 7.84, P , 0.01; 6 months: F
2,29
¼ 5.55, P , 0.01;
9 months: F
2,36
¼ 7.67, P , 0.01; 12 months: F
2,30
¼ 6.09,
P , 0.01) indicating a reduced anxiety. Furthermore, as
shown in Figure 4B, heterozygous and homozygous transgenic
rats did not differ in the total number of arm entries at all
testing points, compared with control animals (ANOVA,
genotype: 3 months: F
2,32
¼ 2.30; 6 months: F
2,29
¼ 2.12;
9 months: F
2,36
¼ 2.67; 12 months: F
2,30
¼ 2.03, all not sig-
nificant, ns). Therefore, in this behavioral test, tgHD rats did
not exhibit an altered motor activity when compared with
control rats.
Radial maze
Spatial learning was assessed in an eight-arm radial maze. The
results are shown in Figure 5A–D. When given 10 min for
Figure 1. Balance and motor coordination on the accelerod. The means +
SEM of the maximum speed (in rpm) reached at each testing age is shown
on the upper panel. The chart below illustrates the latency to fall (in
seconds) at each testing age (from 1 to 9 months). Transgenic animals
showed a significantly better performance than controls at 1 month of age.
However, subsequently, there was a progressive decline in performance of
transgenic rats, with homozygous rats exhibiting a more severe deterioration.
Heterozygous rats were significantly worse than wild-type littermates on this
motor function test at 8 months, whereas homozygous rats displayed a signifi-
cantly worse performance already at 6 months. Asterisks indicate significant
differences between wild-type control and HD transgenic rats (

P , 0.05,

P , 0.01,

P , 0.001,

P , 0.0001).
Human Molecular Genetics, 2006, Vol. 15, No. 21 3179
at Universitaetsbibliothek Rostock on September 10, 2012http://hmg.oxfordjournals.org/Downloaded from

free investigation of the maze, transgenic HD exhibited
explorative behavior comparable to their wild-type littermates.
There was no major difference in preference for certain angles
when choosing arms at all time-points tested (Fig. 5A).
Activity, measured by total number of arm entries was not sig-
nificantly changed (data not shown). Arm bias scores and
angle bias scores, calculated as described by Holter et al.
(26), also revealed no significant differences between HD
transgenic rats and controls, indicating that there were no per-
severative tendencies or uneven distribution of locomotor
activity across the maze (data not shown). Therefore, the
transgenic animals have sufficient general motor, cognitive
and sensory abilities to master this learning task.
In experiment 2, the reinforced alternation task, rats were
tested on their working memory (WM), i.e. the ability to
retain the information on which arms they had visited
before in the ongoing trial. As illustrated in Figure 5B,
there was no significant difference between transgenic
animals and controls at 6 and 9 months of age (ANOVA,
genotype: F
2,26
¼ 1.38 at 6 months and F
2,26
¼ 0.14 at
9 months of age, both ns), whereas at the age of 12
months, transgenic HD rats made significantly more WM
errors than the controls in this task (ANOVA, genotype:
F
2,31
¼ 5.44, P , 0.01).
In experiment 3, the allocentric reversal task, rats were
additionally tested on their reference memory (RM), i.e. the
ability to learn to visit only the baited arms, in order to
assess whether spatial learning deficits result only from a
deficient WM or whether an impaired RM also contributes
to the cognitive phenotype. The results are shown in
Figure 5C and D. At the age of 6 months, control and HD
transgenic rats committed similar amounts of RM errors on
all testing days (Fig. 5D). ANOVA for repeated measurements
revealed no significant effect for the factor genotype
(F
2,26
¼ 0.79, ns) and no significant interaction between gen-
otype and testing day (F
2,234
¼ 1.66, ns) was evident.
However, there was a tendency in homozygous rats to
commit more WM errors than heterozygous and wild-type
rats on several testing days (Fig. 5C), although ANOVA for
Figure 2. Balance and motor coordination in the beam walking test. The
latency to walk across a raised beam (A) and the number of footslips made
(B) were recorded on each trial. Performance on the smallest round beam is
shown. TgHD rats exhibit a progressive decline in beam walking ability
with age. Asterisks indicate significant differences between wild-type
control and HD transgenic rats (

P , 0.05,

P , 0.01,

P , 0.001,

P , 0.0001).
Figure 3. Anxiety-related behavior in the social interaction test. At all ages
tested, transgenic rats spent significantly more time in social interaction
than control rats. Asterisks indicate significant differences between wild-
type control and HD transgenic rats (

P , 0.05,

P , 0.001,

P , 0.0001).
3180 Human Molecular Gen etics, 2006, Vol. 15, No. 21
at Universitaetsbibliothek Rostock on September 10, 2012http://hmg.oxfordjournals.org/Downloaded from

repeated measurements showed no significant effect for the
factor genotype (F
2,26
¼ 1.41, ns) and no significant inter-
action between genotype and testing day (F
2,234
¼ 1.04, ns).
By 9 months of age, all three groups still made comparable
amounts of RM errors (Fig. 5D, ANOVA for repeated
measurement, genotype: F
2,26
¼ 0.36, ns; genotype  testing
day: F
2,234
¼ 1.67, ns), whereas transgenic HD rats committed
significantly more WM errors than the controls (Fig. 5C,
ANOVA for repeated measurement, genotype: F
2,26
¼ 4.96,
P , 0.05). The interaction of the factors genotype and
testing day was also significant (F
2,234
¼ 1.681, P , 0.05).
A significant effect was found on day 1 (F
2,26
¼ 7.2,
P , 0.01), day 2 (F
2,26
¼ 4.8, P , 0.05), day 5 (F
2,26
¼ 3.4,
P , 0.05) and day 6 (F
2,26
¼ 3.9, P , 0.05). As shown in
Figure 5C, this confirms that transgenic rats performed
worse than the control rats, particularly at the beginning.
At 12 months of age, transgenic HD rats showed a learning
deficit as before, as they committed significantly more WM
errors than their wild-type littermates (Fig. 5C, ANOVA for
repeated measurements, genotype: F
2,31
¼ 8.12, P , 0.01),
whereas there was no significant interaction between genotype
and testing day (F
2,279
¼ 1.21, ns). But this time, the perform-
ance of HD transgenic rats also declined in terms of RM
errors. HD transgenic rats, especially the homozygous rats
committed significantly more RM errors compared to the con-
trols (ANOVA for repeated measurements, genotype:
F
2,31
¼ 12.51, P , 0.001). No interaction between genotype
and testing day was found (F
2,279
¼ 1.18, ns).
Two-way active avoidance
Associative learning was tested in the two-way active avoid-
ance test. Performance on this task is shown in Figure 6. In
this learning test, tgHD rats seemed to acquire the task even
better than control rats, however, there was no significant
difference between transgenic animals and controls at
8 months of age (ANOVA for repeated measurements,
genotype  avoidance reactions: F
2,28
¼ 1.01, ns). It can
also be clearly seen from Figure 6 that all groups showed
associative learning since active avoidance reactions increased
from testing day 1 to day 8 (ANOVA for repeated measure-
ments, test day: F
7
¼ 20.31, P , 0.001).
Aggregates and AF
In order to characterize the onset and distribution of AF and
polyQ aggregates in the HD rat model, brain sections of trans-
genic and wild-type rats of different ages were stained with
1C2 antibody and AF were detected in adjacent sections
using a novel synthetic peptide (bPEGQ30). Aggregation
sites and aggregates were prominent in cells and projections
of the basal ganglia, particularly in the olfactory tubercle
and the nucleus accumbens, in several thalamic nuclei as
well as in the hypothalamus, in the substantia nigra pars com-
pacta and the ventral tegmental area, and in the subependymal
and the caudal caudate-putamen (Fig. 7). Moderate reactivity
was identified in the olfactory bulb, in various cortical
regions, notably in the piriform cortex (Fig. 7A–D) and
deeper layers of cortex, in the diagonal band, medial genicu-
late nucleus and superior colliculus. The presence of
polyQ-containing species, documented by weak
1C2-immunoreactivity at similar sites in adjacent sections in
all areas where recruitment activity was observed, confirmed
that these AF contained polyQ (Fig. 7b, d and A through P
inserts). However, many neuropil aggregates were unable to
further recruit synthetic peptide, as there was a striking
Figure 4. Anxiety-related behavior in the EPM test. The percentage of time
spent on the open arms of the elevated plus-maze was significantly higher
in tgHD rats, tested at 3, 6, 9 and 12 months of age compared with the
control group indicating a reduced anxiety (A). Asterisks indicate significant
differences between wild-type control and HD transgenic rats (

P , 0.05,

P , 0.01,

P , 0.001). Motor activity in the EPM test was not altered
in tgHD rats as shown by the total number of arm entries which did not
differ significantly from control animals (B).
Human Molecular Genetics, 2006, Vol. 15, No. 21 3181
at Universitaetsbibliothek Rostock on September 10, 2012http://hmg.oxfordjournals.org/Downloaded from

Figure 5. (A) Exploratory behavior in the eight-arm radial maze. When given 10 min for free investigation of the maze, transgenic HD exhibited explorative
behavior comparable to their wild-type littermates. There was no major difference in preference for certain angles when choosing arms at all tested time-points.
(B) WM errors in the reinforced alternation task of the radial maze. The ability to retain and manipulate the mnemonic information to guide ongoing behavior
was evaluated. There was no difference between transgenic and wild-type rats in this test at the age of 6 and 9 months. Whereas at 12 months of age, the number
of WM errors was significantly increased in transgenic rats. Asterisks indicate significant differences between wild-type control and HD transgenic rats
(

P , 0.05,

P , 0.01). (C) WM errors in the allocentric reversal experiment in the radial maze. Re-entries into arms already visited within a
trial are counted as WM errors. There was a change of baited arms on day 6. At 6 months of age, no difference in WM between transgenic and wild-type
rats was detectable. Whereas at 9 and 12 months of age, WM was impaired in transgenic rats indicated by significantly higher WM errors. Asterisks indicate
significant differences between wild-type control and HD transgenic rats (

P , 0.05,

P , 0.01). (D) RM errors in the allocentric reversal experiment in the
radial maze. The first entries into never-rewarded arms were counted as RM errors. There was a change of baited arms on day 6. RM errors on the first testing day
were not submitted to analysis since animals could not have developed an RM before being subjected to the test. There was no significant difference in RM
between transgenic animals and controls at 6 and 9 months of age. Whereas 12-month-old transgenic HD rats committed significantly more RM errors
(ANOVA for repeated measurements, genotype: F
2,31
¼ 12.5, P ¼ 0.0001).
3182 Human Molecular Gen etics, 2006, Vol. 15, No. 21
at Universitaetsbibliothek Rostock on September 10, 2012http://hmg.oxfordjournals.org/Downloaded from

qualitative difference between recruitment activity and 1C2
immunoreactivity in most areas. Thus, some regions may
present with dense AF (Fig. 7a, d), while corresponding
areas in adjacent sections only show weak 1C2-reaction
product (Fig. 7b, d). Many small 1C2-reactive polyQ aggre-
gates in the cortical neuropil were arrayed in a linear
manner (Fig. 7d), reminiscent of human HD neuropil aggre-
gates, and these were largely undetectable in recruitment pro-
tocols (Fig. 7c). No staining was observed in wild-type rats at
all time-points studied (data not shown) or in earlier ages of
transgenic rats than shown for selected areas.
PolyQ recruitment into structures analogous to the cyto-
plasmic AF earlier identified in human HD brains were readily
detected in tgHD rat brains at 9 months of age, whereas at the
age of 6 months only weak AF reactivity was visible. Also at
6 months only few 1C2-reactive aggregates were present,
appearing by 9 and 10 months in thalamus (Fig. 7C, G) and
in the substantia nigra pars compacta (Fig. 7D, H), with very
little reactivity in caudate-putamen (Fig. 7E) and only modest
expression in cortical areas (Fig. 7F). At 12 months there was
strong reactivity in the thalamus (Fig. 7K), in the substantia
nigra pars compacta (Fig. 7L), a wider distribution in the
cortex (Fig. 7J), with reactivity appearing in the caudate-
putamen (Fig. 7I) and in other regions of the basal ganglia. At
24 months, the tgHD rats showed strong reactivity in all of
the above-mentioned areas, again most dramatically in the
caudate-putamen (Fig. 7M), the thalamus (Fig. 7O) and the sub-
stantia nigra pars compacta (Fig. 7P) and extensive expression in
the cortex (Fig. 7N).
Numerous punctate sites containing 1C2-reactive polyQ
aggregates were found particularly in the lateral olfactory
tract (Fig. 8B), in the neuropil of the lateral globus pallidus
(Fig. 8C), the ventral pallidum (Fig. 8D) and the substantia
nigra pars reticulata (Fig. 8E, F), primarily areas receiving
afferents from the most densely recruiting regions. PolyQ con-
taining ‘neuropil aggregates’ were also seen distinctively dis-
tributed in layer I of certain cortical regions (Fig. 8A). These
observations suggest that in this tgHD model, polyQ aggre-
gates formed in the cytoplasm of projecting neurons may be
in the process of being transported to the projections areas
of these cells; these aggregates appear to have a markedly
reduced ability to recruit further polyQ compared with cyto-
plasmic AF.
A summary of distribution and appearance of 1C2-reactive
aggregates with a semi-quantitative assessment is given in
Table 1, however, by 15 months of age small numbers of
1C2-reactive neuropil aggregates were found widely scattered
in many subcortical regions and only those regions with
characteristic and localized high densities at 18 months are
shown.
In order to compare the distribution of AF and aggregates
identified by 1C2 with previously detected aggregates in our
transgenic rat model (4), we have stained adjacent brain sec-
tions of 21-months-old tgHD rats with polyclonal EM48
(EM48). We had to use EM48, which is not commercially
available, as monoclonal EM48 has failed to detect aggregates
in our rat model in previous studies (Nguyen, Osmand,
Ka´ntor, von Ho¨rsten, unpublished observations). Figure 9
provides an overview of polyQ recruitments sites (A,D),
EM48 (B) and 1C2 (C) immunoreactivity at the level of the
olfactory tubercle of 21-month-old tg rat brains,  +2 m
M
rostral of Bregma. Semi-quantitative scores of the three differ-
ent tools for characterizing aggregates and aggregation are
provided in Table 1, illustrating basically an overlap of distri-
bution patterns. However, 1C2 immunoreactivity was the most
reliable approach, while bPEGQ30 aggregation recruitment
sites appeared more sensitive in several regions. In general,
reactivity with EM48 was weaker in all regions compared
with 1C2 staining or polyQ recruitment. Most of EM48
immunoreactive products appeared as punctate labeling in
the striatum, nucleus accumbens, olfactory tubercle, thalamus,
medial and lateral geniculate nucleus. Fewer aggregates were
detected in olfactory bulb, stria terminalis, lateral septal
nucleus, ventral pallidum and globus pallidus, substantia
nigra, amygdala, inferior and superior colliculus whereas
only very weak or no aggregates were seen in cortex and
hippocampus.
Striatal volume
In order to determine whether the most affected brain region in
HD patients, namely the striatum, is also compromised in
tgHD rats, volume estimates of the striatum were calculated
using stereological approach. There was no detectable differ-
ence in striatal volume at 3 and 6 months of age in tgHD
rats compared with wild-type littermates (Fig. 10). By 9
months of age, there was a decrease in striatal volume in
tgHD rats, which, however, did not reach significance
(Fig. 10). But at the age of 12 and 15 months, striatal
volume was significantly decreased in transgenic animals
when compared with the controls (Fig. 10, at 12 months:,
P , 0.05; at 15 months:, P , 0.05).
Figure 6. Associative learning in the two-way active avoidance test. All
groups showed associative learning indicated by increasing active avoidance
reactions from testing day 1 to day 8. However, there was no significant
difference between transgenic animals and controls at 8 months of age in
this learning task.
Human Molecular Genetics, 2006, Vol. 15, No. 21 3183
at Universitaetsbibliothek Rostock on September 10, 2012http://hmg.oxfordjournals.org/Downloaded from

Figure 7. (a– d) AF and 1C2-reactive neuropil aggregates in a 24-month-old heterozygous HD rat. (a) AF revealed with bPEGQ30 and (b) neuropil aggregates
(1C2), are from adjacent sections of layer III of visual cortex; (c) (bPEGQ30) and (d) (1C2) are from layer III of entorhinal cortex. Arrows in (d) mark aggregates
arranged in a linear manner. Polyglutamine recruitment reveals primarily cytoplasmic sites (a and c), while 1C2 reacts strongly with neuropil aggregates (arrows
in d) and only weakly with cytoplasmic sites (arrowheads in b and d). (A–P) Time-course of polyQ recruitment sites and 1C2-reactive aggregates. These images
are of bPEGQ30 recruitment with the adjacent 1C2 stained section as an insert in each panel that covers one-quarter of the area. (A,B) are from a 6-month-old
homozygous HD rat (+/+); (C,D) are from a heterozygous 9-month-old rat; (E – H) are from a 10-month-old homozygote; (I –L) are from a 12-month-old homo-
zygote and M–P are from a 24-month-old heterozygote. (E, I and M) are from caudate-putamen showing appearance of recruitment after 10 months and increas-
ing reactivity with age with dense nuclear staining with 1C2 seen in I and M; (F, J and N) are from motor cortex, showing a similar time of appearance to that
seen in cortex; (A, C, G and O) are from the centromedial nucleus of thalamus showing weak diffuse recruitment at 6 months, increasing by 9 months and with
increasing numbers of neuropil aggregates at 10, 12 and 24 months of age; (B, D, H and L) are from substantia nigra pars compacta showing weak diffuse
recruitment at 6 months with widespread appearance by 9 months and a progressive increase in recruitment activity and 1C2 reactivity. All images are the
same magnification with the bar in (A) ¼ 50 mm.
3184 Human Molecular Gen etics, 2006, Vol. 15, No. 21
at Universitaetsbibliothek Rostock on September 10, 2012http://hmg.oxfordjournals.org/Downloaded from

Power calculation
In order to characterize the observed effect size (ES, standar-
dized mean difference between two groups) of all parameters
quantified, we calculated the minimal total number of animals
that would have been sufficient to detect the effects observed
(Table 2). The calculation was based on the effect sizes that
we gained from each of our criterion variables. Low inter-
animal variability was observed for striatal volume and the
social interaction test, when only three animals were needed
to detect the differences (indicated as % improvement) in
our studies. In case of the radial maze test, up to 18 rats
would have been necessary to confirm the effects observed,
whereas in the accelerod test only seven animals would have
been sufficient.
In a second approach, we performed a power analysis on the
phenotyping data in order to estimate the number of animals
per group that we would need to detect a 5, 10, 20 or 40%
rescue at a significance of P , 0.05 with 80% power
(Table 3). A very small number of animals is required in a
therapeutic trial where the goal is to determine a 40% rescue
with the striatal volume at 15 months as endpoint, namely
only one animal. For a less effective rescue of 5%, only 13
animals are required to determine a difference using striatal
volume as an endpoint. As variability increases, the number
of animals necessary to determine a significant effect on that
particular phenotype also rises. For example, at least 11
tgHD rats have to be used to detect an improvement of 40%
and even 643 rats for a 5% rescue in the radial maze test para-
digm. However, in the social interaction test, a 40% ameliora-
tion can be already determined using only three animals. And
for a less effective rescue of 20%, only 11 rats are required.
This is especially attractive for designing therapeutic trials,
as we can observe changes in anxiety in the social interaction
test already at 1 month of age.
DISCUSSION
A first aim of this study was to carefully monitor the onset
and progression of major HD-like symptoms in our trans-
genic rat model of HD providing a basis for future pharma-
cological testing. Using a battery of behavioral tests, we have
achieved this for all key systems affected in HD. However,
there were some surprising and novel findings. First, trans-
genic rats showed a significant better performance on the
accelerod than their wild-type littermates at 1 month of
age, followed by a slow decline of motor function and result-
ing in a significantly impaired performance at 6 months of
age for homozygous HD rats and at 8 months of age for het-
erozygous rats (Fig. 1). This kind of biphasic motor pheno-
type does not completely resemble findings in R6/2 mice
(5,27), HD knock-in mice (28,29) and in YAC mouse
models expressing full-length mutant Huntingtin (30,31), as
the initial improved accelerod performance has not been
observed in mice models of HD. Secondly, striking
changes in anxiety were detected in transgenic HD rats
already at the age of 1 month. Although this reduced
anxiety-like phenotype is concordant with observations in
R6/2 (11) and R6/1 mice (32) and thus, anxiety-related beha-
viors may be used to assess therapeutic effects of new com-
pounds, these emotional changes in HD transgenic rats occur
much earlier than expected. Our observations represent a
novel finding with unknown significance. One may speculate
that in young transgenic animals, as yet undefined processes
such as repair mechanisms interfere with accelerod perform-
ance and similarly the early anxiolytic-like phenotype may
reflect ongoing mechanism of repair by e.g. increased
expression of neurotrophic factors, many of which act
anxiolytic-like. The early phenotype in young transgenic
HD rats may also point to the hypothesis of ‘hypercompensa-
tion’ which have emerged very recently in other neurodegen-
erative disorders such as Amyotrophic lateral sclerosis
(ALS). Hampton et al. (33) have reported an athletic gait
in presymptomatic SOD G93A mice (model for ALS) with
greater stride length and longer stride duration compared
with controls, whereas in symptomatic mice shortened
stride length and increased paw placement angles demon-
strated gait disturbances. Furthermore, unpublished data (per-
sonal communications with Thomas G. Hampton, Beth Israel
Deaconess Medical Center, Harvard Medical School) show a
‘supernormal’ gait in presymptomatic R6/2 mice and an
impaired gait in symptomatic mice, both similar to the obser-
vations in the ALS mice. Therefore, consistent with obser-
vations by the Hampton and coworkers in ALS mice, it
might be that presymptomatically, we need not look for a
deficit, but for evidence of compensatory hyperexcitability
of neurons. Nevertheless, especially the early onset of these
behavioral changes in transgenic HD rat offers an attractive
opportunity to use them as a parameter in therapeutic
approaches. In case of ALS, this has been shown for propra-
nolol which mitigates the presymptomatic ‘athletic’ gait in
Figure 8. 1C2-reactive microaggregates and neuropil aggregates in fiber tract
and projection areas. All images are the same magnification with the bar in
5(A) ¼ 50 mm. (A) is from the visual cortex in 10 months +/+. (B) is from
the lateral olfactory tract in a 24 months +/2 old rat. (C) and (D) are from
lateral globus pallidum and ventral pallidum, in a 12 month +/+; (E) and
(F) are sections from substantia nigra pars reticulata at 12 months +/+ and
24 months +/2.
Human Molecular Genetics, 2006, Vol. 15, No. 21 3185
at Universitaetsbibliothek Rostock on September 10, 2012http://hmg.oxfordjournals.org/Downloaded from

SOD1 G93A mice and also extends lifespan in these mice
[personal communications with Thomas G. Hampton (33)].
Furthermore, our data indicate that onset of cognitive
decline in HD transgenic rats occurs between the age of 6
and 9 months and worsens with age. The early manifestation
of spatial WM deficits is in accord with observations in
human HD (34). At the mild to moderate stages of HD,
patients show a progressive deterioration in attention, execu-
tive function and immediate memory, whereas other cognitive
functions such as general cognition and delayed recall memory
do not significantly deteriorate in early stages (35). This pattern
may be analogous to our findings in HD transgenic rats with
Table 1. Summary of distribution and appearance of 1C2-reactive aggregates with semi-quantitative assessment (only those regions with characteristic and
localized high densities at 18 months are shown) and comparison with polyQ recruitments sites and EM48 at the age of 21 months
Semi-quantitative score of 1C2 and EM48 immunoreactivity as well as polyQ recruitment site in tgHD rat brains across different brain structures.
Score: 2 , negative; +/2 , weak, but definite; +, positive; ++, strong; +++, very strong.
3186 Human Molecular Gen etics, 2006, Vol. 15, No. 21
at Universitaetsbibliothek Rostock on September 10, 2012http://hmg.oxfordjournals.org/Downloaded from

an early appearance of short-term memory deficits and a
delayed onset of long-term memory dysfunction. Unlike in
R6 mice where evidence of visual dysfunction and retinal neu-
rodegeneration has been shown (36,37), and therefore beha-
vioral tests based on visual acuity of R6 mice should be
carefully interpreted, we did not observe retinal degeneration
in heterozygous and homozygous HD transgenic rats until
18 months of age (Petrasch-Parwez, unpublished observation).
Thus, the radial maze test of spatial learning reliably deter-
mines differences between HD transgenic and wild-type rats
and can be used for evaluating beneficial effects of
therapeutic agents.
However, it was surprising that cognitive deficits were
detected shortly after the onset of motor symptoms, whereas
in HD patients and HD mice it seems that cognitive impair-
ment occurred before motor dysfunction (7,8,38). Our obser-
vations of a delayed onset of cognitive impairment in the
radial maze test were confirmed in the two-way active avoid-
ance test, whereas at 8 months of age no difference in associ-
ative learning was evident (Fig. 6). This is also concordant
with findings in other cognitive tests such as the choice reac-
tion time task, whereas at 15 months only moderate cognitive
impairment was seen, which became significant at 20 months
of age in tgHD rats (Cao et al., unpublished observations).
Figure 9. Comparison of polyQ recruitment sites (A, D), EM48 (B) and 1C2 (C) immunoreactivity at the level of the olfactory tubercle of 21-month-old tg rat
brains. Bars represent 200 mm. Abbreviations of structures: CPu, caudate-putamen (striatum); ec, external capsule; Acb, accumbens nucleus; ac, anterior
commissure; Pir, piriform cortex; lo, lateral olfactory tract; Tu, olfactory tubercle.
Human Molecular Genetics, 2006, Vol. 15, No. 21 3187
at Universitaetsbibliothek Rostock on September 10, 2012http://hmg.oxfordjournals.org/Downloaded from

Nevertheless, we cannot rule out that with more sensitive
behavioral tests, cognitive decline could be detected in tgHD
rats earlier than so far.
A second goal of this study was to examine the regional and
temporal distribution of neuropathological markers in relation
to behavioral abnormalities occurring over time. Figure 11
summarizes the temporal association of our behavioral and
neuropathological findings. Initial studies in our transgenic
rat model had revealed the presence of nuclear and neuropil
huntingtin aggregates in the striatum and to a lesser extent
in the cortex of transgenic HD rats at the age of 12 months
and older (4). We have now extended the sensitivity of detec-
tion of polyQ aggregates dramatically by using 1C2 antibody
after aggressive antigen retrieval and biotin tyramide amplifi-
cation. In addition, we have attempted to uncover so far non-
detected forms of polyQ aggregates by applying a new method
based on the recruitment of tagged polyQ peptides into exist-
ing reactive polyQ aggregates (25). These polyQ recruitment
sites may represent cytoplasmic precursors of the mature neu-
ropil aggregates seen in the cortex in human HD (24). AF first
appeared in thalamus, substantia nigra pars compacta and deep
layers of cortex and only later in the caudate-putamen (Fig. 8).
However, polyQ aggregates and recruitment sites did not
appear in significant numbers before the age of 9 months,
which argue against a primary role of aggregates and AF in
the earliest manifestations of the mutation. Therefore, it is
likely that yet undetected processes contribute to the early
Figure 10. Striatal volume estimates. Striatal volume was calculated using a
stereological approach. Homozygous transgenic rats showed no signsificant
difference in striatal volume at 3, 6 and 9 months of age. A significant decrease
in striatal volume was seen at 12 and 15 months of age. Asterisks indicate
significant differences between wild-type control and HD transgenic rats
(

P , 0.05).
Table 2. Power for 1

3 ANOVAs (independent groups)
Criterion Age
(months)
ES %
Improvement
Power
%
% Minimal
N
Accelerod
(time)
9 0.72 58.7 80 0.05 7
Accelerod
(rpm)
9 0.93 62.9 80 0.05 5
WM errors
in radial
maze
12 0.41 33.6 80 0.05 18
RM errors
in radial
maze
12 0.46 60.1 80 0.05 15
Social
interaction
2 1.35 37.9 80 0.05 3
Beam walk 7–12 0.48 33.3 80 0.05 15
Striatal
volume
15 1.85 16.1 80 0.05 3
Overview of the minimal total number of animals to detect significant
differences between the investigated groups with regard to effect size
(ES) of criterion variable. %impr., % of improvement in the control
group compared with the transgenic group.
Figure 11. Summary of findings. Onsets of behavioral abnormalities are
shown above the time bar. At 1 month of age, we found reduced anxiety in
transgenic rats indicated by an arrow pointing downwards as well as improved
motor performance on the accelerod illustrated by an arrow pointing up. At
6 months of age, homozygous rats were significantly worse than controls in
the accelerod. Cognitive impairment was seen in 9-month-old transgenic
rats. The appearance of neuropathological markers is illustrated beneath the
time bar. AF and 1C2 positive aggregates were both detectable at 9 months
of age.
Table 3. Power analyses for quantitative phenotypes in tgHD rats
Phenotype 5%
rescue
10%
rescue
20%
rescue
40%
rescue
Accelerod (time) 709 178 45 12
Accelerod (rpm) 391 98 25 7
WM errors in radial maze 643 161 41 11
RM errors in radial maze 1071 221 67 17
Social interaction 164 41 11 3
Beam walk 767 192 48 12
Striatal volume at 15 months 13 4 1 1
Power analyses determines the minimal number of tgHD rats necessary to
detect a significant (P , 0.05) difference in treated versus untreated
animals if one predict an 80% chance of discerning a 5, 10, 20 or 40%
rescue of the various quantitative phenotypes. It was supposed to inves-
tigate identical groups within a repeated-measurement-design (two
measurements) with SD and baseline values. Animal numbers are
hypothetical results with regard to statistical power and have to be sub-
jected to further aspects of experimental designs. Power analyses to
detect hypothetical improvements in therapeutic studies.
3188 Human Molecular Gen etics, 2006, Vol. 15, No. 21
at Universitaetsbibliothek Rostock on September 10, 2012http://hmg.oxfordjournals.org/Downloaded from

phenotype of HD transgenic rats and that a simple model of
aggregate causality may not be adequate. However, since we
must assume that AF are sites of ongoing aggregation and
are comprised of mutant huntingtin or its fragments,
1C2-reactive aggregates seen at 9 months may result from pro-
cesses that have been ongoing or turning over, albeit at an
undetectable level, perhaps for some period of time. Thus,
there remains no a priori reasons to exclude the possibility
that polyQ-mediated aggregation processes underlie these
early phenotypic events.
Recently, in vitro data have been published suggesting that
readily visible mutant polyQ inclusion bodies are protective
by reducing the amount of a hypothetically toxic and diffusely
distributed form of mutant huntingtin (39). Although these
findings are persuasive, there are some limitations in transpos-
ing these findings to processes in vivo where aggregation
formation takes place over many months or even years. In
addition, these experiments were performed with expressed
exon 1 fragments, rather than longer fragments or full-length
huntingtin. Furthermore, the toxicity of mutant huntingtin in
these primary striatal cells may, in part, reflect the tendency
of these cells to accumulate intranuclear inclusions rather
than form neuropil aggregates. Therefore, they do not rule
out the possibility that toxicity arises from early precursors
to inclusion bodies such as AF (24) or microaggregates that
have been described in HD KI mice (28,29).
More recently, Slow et al. (23) reported that a mouse model
(the shortstop mouse), expressing an N-terminal human htt
fragment corresponding to the regions encoded by exons 1
and 2 and with an expanded polyglutamine repeat of 128 glu-
tamines, under the control of the endogenous human promoter,
does not display any behavioral abnormalities or evidence for
neurodegeneration. These results were taken to indicate that
htt inclusions are not pathogenic in vivo. However, several
questions remained unanswered. For example, the authors
could not adequately explain why the ‘shortstop’ mice do
not show the same phenotype as the R6/2 mice, a model
expressing only the exon1-encoded sequence. There are indi-
cations that there may be additional differences between the
‘shortstop’ and other full-length HD mouse models, since
the distribution of aggregates and AF is radically different in
the ‘shortstop’ (Osmand, unpublished observations). The
difference between ‘shortstop’ mice and other HD animal
models, expressing a longer htt fragment or full-length htt,
could be attributable to the absence of those regions of htt
that allow interactions with cytoskeletal elements that lead
to the redistribution of aggregates into axons and dendrites
in vulnerable neuronal populations.
Nevertheless, there is increasing evidence, that aggregates
are not necessarily cytotoxic or cytolytic. But it is unclear
whether the presence of aggregates or aggregating species is
merely compromising cells or whether cells are only affected
in some deleterious manner when the aggregates are redistribu-
ted into some specific cell compartment. It seems reasonable to
acknowledge that the majority of aggregates may be inconse-
quential, since aggregates are seen in HD animal models in
largely normal neurons. Indeed, when aggregation events are
cytotoxic there will probably be very little pathological evi-
dence other than diffuse cell loss, particularly given the ease
of clearance of aggregates and the reversal of pathology in
conditional mouse models (40). Perhaps, only in much older
animals does the accumulation becomes noticeable, presumably
at a stage when clearance mechanisms have become over-
whelmed. It is important to make the distinction between aggre-
gates, a pathological marker, and aggregation, an intrinsic
mechanism in their formation. Because one does not see aggre-
gates at some particular point in time does not mean that aggre-
gation has not occurred or is not occurring and hence polyQ
aggregation may remain the central pathogenic process in HD.
Understanding the dynamics of aggregation in vivo, may yet
be the key to elucidating its link to toxicity. With new
methods, we have been able to locate sites of aggregation,
which have not been previously described. One surprising
finding is that aggregates and foci appeared in the olfactory
tubercle, the nucleus accumbens, thalamus and substantia
nigra pars compacta substantially before they were detected
in cortical areas or the caudate-putamen, providing an anatom-
ical correlate to the early onset of emotional changes. More-
over, it appeared that in this model small polyQ aggregates
formed in the cytoplasm of projecting neurons were readily
transported to the projection areas of these cells (Fig. 5),
rather than accumulated locally as large neuropil aggregates.
This is interesting since ultrastructural examination of neuro-
pil aggregates in knock-in mice revealed that they were associ-
ated with axonal degeneration (41) potentially leading to
defective neuronal interaction, abnormal synaptic transmission
and impaired supply of growth factors. It will be important to
compare the many existing models of HD not only from the
viewpoint of which rodent model best reflects the human
disease, but also, given the paucity of data on HD, where
one might look for novel pathologies in HD.
To further examine whether the HD-related neuropathologi-
cal phenotype exhibited in tgHD rats recapitulates the changes
observed in the human disease, we have measured striatal
volume at different ages (Fig. 10). Our findings of striatal
atrophy in tgHD rats starting at 12 months of age and the
observation of neuronal cell loss at the same age (42)
confirm that this rat model exhibit progressive, quantitative
phenotypes and demonstrate the potential usefulness of this
model in therapeutic trials. This is further illustrated by the
power analysis estimations (Tables 2 and 3). The phenotype
with the lowest variability, striatal volume, requires only
four animals to determine a 10% rescue at 15 months of
age. The behavioral phenotypes show greater variability, but
because of their early onset, tests such as the social interaction
offer an attractive opportunity to use them as a parameter in
therapeutic approaches. For a less robust therapeutic effect
(20%), only 11 animals are needed in this test. Using the
data reported in this manuscript, it is now possible to design
preclinical therapeutic trials ensuring an adequate number of
animals to properly assess promising compounds.
MATERIALS AND METHODS
Animals
The transgenic rats were generated as described previously (4).
The transgenic HD rat expresses 727 amino acids of the HD gene
with 51 CAG repeats (cDNA position 324 –2321 corresponding
to 22% of full length), which are under the control of 886 bp of
Human Molecular Genetics, 2006, Vol. 15, No. 21 3189
at Universitaetsbibliothek Rostock on September 10, 2012http://hmg.oxfordjournals.org/Downloaded from

the rat huntingtin promoter (position 900 to 15). The genetic
background of the transgenic HD rat is Sprague– Dawley
strain (outbred) derived, which is now in a process of permanent
inbreeding. A colony of transgenic Sprague– Dawley rats was
established at the central animal facilities, Hannover Medical
School and the line was maintained by backcrossing. Tail tips
were removed from all rats at the age of 3 weeks and the geno-
type was confirmed by Southern blot analysis.
The present study used male homozygous, heterozygous
and wild-type littermate control rats from litters all born
within 2 days of each other. Two rats of randomized genotype
were housed together. All rats were tested within the dark
phase of a 12 h light/dark cycle for the social interaction, elev-
ated plus maze, beam walking, two-way active avoidance and
accelerod tests. Radial maze tests were conducted in the light
cycle. All research and animal care procedures were approved
by the district government, Hannover, Germany, and per-
formed according to international guidelines for the use of
laboratory animals.
Accelerod test
To determine fore and hind limb motor coordination and
balance, an Accelerod for rats (TSE-Systems, Bad Homburg,
Germany) was used (4). This apparatus consisted of a base
platform and a rotating rod of 7 cm diameter with a
non-skid surface. When operated in the acceleration modus,
the rotor would accelerate from 4 to 40 rpm in a period of
5 min, monitored by a bar-graph type of speed indicator
placed on the front panel. The accelerod has been shown to
be more sensitive than the rotarod in detecting motor function
deficits (43) and to produce more consistent results (44). Male
homozygous (+/+; n ¼ 10), heterozygous (+/2; n ¼ 15) and
wild-type (2/2 ; n ¼ 12) rats were tested monthly between
the age of 1 and 9 months. Before testing, rats were trained
twice a day on three consecutive days. By this time, a
steady baseline level of performance was attained. Rats
started with 4 rpm subsequently accelerating every 30 s gradu-
ally up to 40 rpm in 5 min. The times spent on the rod before
falling off and the maximum speed level reached were
recorded.
Beam walking test
The beam walk paradigm was used to assess motor coordi-
nation and balance by measuring the ability of the animals
to traverse a graded series of 1 m long wooden beams,
which were elevated 60 cm above the ground. The ground
was covered with 20 cm thick cellular foam. At the end of
the starting platform, a white, brightly lit shelter was installed,
whereas the shelter of the target platform at the end of the
wooden beam was painted black and only lit dimly. Two
sets of male tgHD rats, either young (1–6-month-old;
n ¼ 14 2 /2, 18+/2, 12+/+) or adult (6-month-old;
n ¼ 10 2 /214+/2 , 10+/+) rats were evaluated in this motor
task. These two sets of animals were followed up for
6 months, respectively. From the various beams, four rec-
tangular and three round were chosen. Young animals were
trained and tested on beams being a rectangular 1.6 cm wide
beam, a rectangular 0.9 cm wide beam and two round beams
with a diameter of 0.8 cm and 1.0 cm. The adult animals
had to traverse a rectangular 2.3 cm wide beam, a rectangular
1.2 cm wide beam and a round beam with a diameter of
1.5 cm. Thus, the difficulty to balance across increased from
beam to beam and beams were adapted, as the animals grow
bigger. Animals were trained for 5 days before baseline evalu-
ation at the age of either 1 or 6 months of age and then tested
without further training once per month. Animals had to cross
each beam twice and latency to traverse as well as number of
footslips were recorded. In case of dropping off the beam,
animals were replaced immediately. For both runs, the
average scores were calculated.
Social interaction test of anxiety (SI test)
The SI test was carried out according to the method described
by S. File with minor modifications as described previously
(13,14,45). The animals tested were male homozygous (+/+;
n ¼ 10), heterozygous (+/2; n ¼ 18) and wild-type (2/2;
n ¼ 14) rats. Animals were tested at the age of 1, 2 and
7 months. Two genotype-matched rats were removed from
their home cages and exposed to a novel test environment.
The rats were always from different cages. To avoid cohort
removal effects, rats of the same cage were not tested on the
same day. The test arena (squared open field of 50  50 cm)
was placed inside a sound isolation box (46). The illumination
of the open field of 1.3 lux was provided by a red photo bulb
(Philips PF712E). The behavior of the two rats was monitored
online by a video camera placed inside the sound isolation
box. The duration of a test was 10 min and the following par-
ameters were scored: duration of time spent sniffing, follow-
ing, crawling under and over the other rat. Passive body
contact as resting and sleeping was not recorded. The sum
of social interaction time of two rats was calculated and
used for statistics.
Elevated plus maze test
The elevated plus maze (EPM) test is one of the most widely
used tests for assessing anxiety in small rodents. It is based on
the aversion of rodents to open spaces and height. The EPM
apparatus (TSE Systems, Bad Homburg, Germany) consists
of four arms (50 cm  10 cm) arranged in the shape of a
cross and was used as previously reported (46). Two opposing
arms are covered with 40 cm high walls (‘closed arms’); the
other two opposing arms have no walls (‘open arms’). The
maze is elevated from the floor (70 cm) and lit by red photo
light (Philips PF712E; 1.3 lux). The red light bulb was
placed 30 cm above the maze in a way that the closed arms
were in shade. Furthermore, the maze is equipped with light
beam sensors that enable computerized measurement of
EPM performance. The experiment was started by placing
the rat on the central platform, with its head facing one of
the closed arms. Monitoring system was then activated and
beam interruptions were monitored for 5 min. The maze was
carefully cleaned after each rat exposure. The following par-
ameters were calculated: total number of arm entries (TA);
entries to closed arms (CA); entries to open arms (OA);
percentage frequency of entries to open arms (%OA:
OA  100/TA); total trial duration (TT) (300 s); duration of
3190 Human Molecular Gen etics, 2006, Vol. 15, No. 21
at Universitaetsbibliothek Rostock on September 10, 2012http://hmg.oxfordjournals.org/Downloaded from

stay in closed arms (closed time; CLT); percentage share of
CLT in total arms-stay duration (CLT  100/AT); duration
of stay in open arms (open time; OT); and percentage share
of OT in total arm-stay duration (% OT: OT  100/AT). An
increase of the time spent on the open arms is interpreted as
an anxiolytic response and a decrease of this parameter an
anxiogenic response, whereas the number of entries into
closed arms provides an indication of general activity (47).
To avoid confounding effects of habituation (loss of con-
struct sensitivity via reduction of novelty) four different sets
of male rats were tested at the age of 3, 6, 9 and 12 months.
Set 1 (at the age of 3 months): homozygous (+/+): n ¼ 10,
heterozygous (+/2): n ¼ 15, wild-type (2/2 ): n ¼ 10.
Set 2 (at the age of 6 months): homozygous (+/+): n ¼ 9,
heterozygous (+/2): n ¼ 13, wild-type (2/2 ): n ¼ 10.
Set 3 (at the age of 9 months): homozygous (+/+): n ¼ 10,
heterozygous (+/2): n ¼ 17, wild-type (2/2 ): n ¼ 12.
Set 4 (at the age of 12 months): homozygous (+/+): n ¼ 9,
heterozygous (+/2): n ¼ 13, wild-type (2/2 ): n ¼ 11.
Radial maze test
Spatial learning and memory were assessed in an automated
sensor-equipped radial maze for rats (TSE, Bad Homburg,
Germany). The radial maze consisted of an octagonal central
area from which eight arms (550  150  225 mm;
L  W  H) radiated outwards, like spokes around a hub. It
was elevated above the ground and made of grey plastic
walls and bottom. Each arm was individually marked with a
number (not visible for the animal). At the distal end of
each arm a food cup was placed. If the arm was to be
baited, a food pellet (precision pellets for rodents: Campden
Instruments Ltd., Loughborough, UK) was placed in the cup.
The platform and all arms were covered with clear acrylic
lids allowing spatial orientation at extra-maze cues. During
all experiments, the maze was kept in a constant position.
At the entrance of each arm (10 cm from the central square)
and inside each food cup an infrared sensor was located
serving to monitor the animals’ transfers and the removal of
the pellet from the cup. The protocol used in the present
study has been adapted from the work of Ho¨lscher and
Schmidt (48) and has been repeatedly used with modifications
in previous studies (4,46). Three days prior to the beginning of
testing all animals were fed for only an hour per day. To avoid
confounding effects of habituation to the radial maze and of
previous, repeated learning trials, three different sets of rats
were tested at the age of 6, 9 and 12 months.
Set 1 (at the age of 6 months): homozygous (+/+): n ¼ 11,
heterozygous (+/2): n ¼ 10, wild-type (2/2 ): n ¼ 8.
Set 2 (at the age of 9 months): homozygous (+/+): n ¼ 8,
heterozygous (+/2): n ¼ 12, wild-type (2/2 ): n ¼ 9.
Set 3 (at the age of 12 months): homozygous (+/+): n ¼ 10,
heterozygous (+/2): n ¼ 14, wild-type (2/2 ): n ¼ 10.
Experiment 1 (day 1): exploratory behavior. The rat was put
into the center of the eight-arm maze and was allowed to
freely investigate the maze for 10 min. Orientation was poss-
ible due to extra-maze cues such as doors, shelves and marks
on the walls. The sequence of arm entries was recorded. The
frequency distribution of angles between consecutively
entered arms was evaluated. One trial per animal was given.
The arms were not baited.
Experiment 2 (day 2): reinforced alternation. The animal
started in a randomly chosen arm. The remaining seven arms
were baited with a food pellet. After all seven pellets had
been collected the trial was terminated. The number of arm
visits required per trial to collect all seven pellets was evaluated.
Returning into a previously visited arm was recorded as a WM
error. One trial per animal was given. Experiment 3 (day 3–12):
allocentric reversal experiment without intra-atrial delay. Four
randomly chosen arms were baited. They were not changed
for the first 5 days. On the sixth day different arms were
baited. Starting arms were changed randomly among unbaited
arms after each trial. Orientation was allocentric because of
external cues in the room visible to the animal (marks on the
walls, shelves, door, etc.). Egocentric information could not
be used due to change of starting arm. When analyzing the
errors made in this experiment, it was differentiated between
WM errors (multiple entries of baited arms within one run)
and RM errors (entry of an unbaited arm). Four trials per
animal per day were given.
Two-way active avoidance (shuttle box) test
Two-way active avoidance is a type of conditioning that
results in associative learning. Essentially, the animals learn
to avoid a signaled noxious stimulus (electrical foot shock—
signaled by light or sound) by initiating a specific locomotor
response (moving to another compartment). The number of
correct avoidance reactions, i.e. transfer into the ‘safe’ com-
partment after the signal and before application of the aversive
stimulus (electrical shock), is evaluated.
We used the TSE Shuttle box system (Technical and Scien-
tific Equipment GmbH, Bad Homburg, Germany), which
allows active and passive avoidance experiments to be
carried out with small laboratory animals. It consisted of
two shuttle boxes, a control unit, an IBM-compatible compu-
ter with special interface and the Shuttle box software. Each
shuttle box was divided into two compartments by a wall
with a central inverted U-shaped opening. Foot shock could
be applied via the grid rods of the box for each compartment
separately. By means of the control unit and dedicated soft-
ware, the experiments were run completely automatically.
Rats were transported from the housing colony to the testing
room at least 2 h before testing was to begin. To avoid aggres-
sive encounters with animals that had yet to be trained, already
tested rats were placed in a holding cage. After each animal
having completed testing, all surfaces of the avoidance appar-
atus were cleaned with a 70% ethanol solution to eliminate
any odors, faecal deposits and urine.
One adaptation session of 5 min for free ambulation in the
shuttle box was given to the animals just before the first
testing session in order to familiarize them with the learning
environment. All rats were submitted to 20 avoidance trials
each day, for eight consecutive days. The interval between
two avoidance trials was variable, ranging from 10 to 50 s.
The unconditioned stimulus, namely the electrical shock,
was set to 0.8 mA. As conditioned stimulus we chose light
in the ‘start area’. In this case, the light is switched on in
that compartment of the box in which the rat was located.
Human Molecular Genetics, 2006, Vol. 15, No. 21 3191
at Universitaetsbibliothek Rostock on September 10, 2012http://hmg.oxfordjournals.org/Downloaded from

After 10 s, an electrical foot shock was given for 5 s with the
light remaining on during this period. If the animal crossed the
barrier to the opposite compartment during the first 10 s when
the light in the ‘starting area’ was presented, the light stimulus
was terminated and no shock was delivered (avoidance
response). A crossing response during shock terminated the
light stimulus and the shock (escape response). If the animal
failed to cross during the entire light-shock trial, the light
and the shock were switched off after 15 s. The number of
avoidance reactions was recorded and analyzed.
One set of male rats was tested in the two-way active
avoidance (shuttle box) test at the age of 8 months: homozy-
gous (+/+): n ¼ 9, heterozygous (+/2): n ¼ 13, wild-type
(2/2): n ¼ 9.
Statistical analysis
Data were subjected to one- or two-way ANOVA with one
between-subject factor (genotype) and with repeated measure-
ments on one or more factors depending on the test used. The
Fisher PLSD test was used for post hoc comparison. A critical
value for significance of P , 0.05 was used throughout the
study. All data represent means + SEM.
Power calculation
Two types of power calculations were performed: (a) in order
to calculate the minimal number of animals, which would
have been hypothetically necessary on each outcome
measure to detect the resulting differences between genotypes
in our experimental approach; (b) in order to determine the
number of animals that would have been required to detect a
potential benefit of 5, 10, 20 and 40% of a hypothesized treat-
ment with the given values and from our approach. In both
types, we assumed a P-value of 0.05 and a power of 80%. Cal-
culations were performed using PASS module of NCSS 2000
statistical package (NCSS 2000, Kaysville, Utah).
Immunohistochemistry
Under isoflurane inhalation anesthesia (Isofluran-Lilly; Lilly
GmbH, Giessen, Germany) rats were transcardially perfused
with 60 ml saline, followed by 400 ml 4% paraformaldehyde
with 0.4% picric acid in 0.16
M phosphate buffer solution
(pH 7.2). Brains were removed, post-fixed in the same fixative
for 12 h and then placed stepwise into a 10, 20 and 30%
buffered sucrose solution at 48C overnight.
Multiple perfusion-fixed rat brains, from control, homozy-
gotic and heterozygotic animals ranging in age from 3 to
24 months, were embedded in a single gelatin block,
post-fixed, and freeze-cut at 40 mm. The use of MultiBrain
TM
embedding (Neuroscience Associates, Knoxville TN) enables
the simultaneous processing of both control and transgenic
animal brain sections of various ages under identical con-
ditions and expedited an examination of the full extent of
the distribution of polyQ recruitment and polyQ in the brain
of these animals.
AF were detected in freeze-cut serial sections using a novel
QKKQ30KK peptide (bPEGQ30) including an N-terminal Gln
analog in which the Y(gamma)-glutamyl side chain contained
a biotinylated polyethylene glycol moiety. The peptide was
disaggregated and purified as previously described (49). The
recruitment and elongation was performed for 16 h at 25 n
M
on free-floating 40 mm sections and the bound biotin detected
by a sensitive histochemical method involving a single round
of biotin tyramide amplification (50) and Ni
+þ
/DAB/H
2
O
2
detection of the avidin-biotinylated peroxidase complex
(Vector Laboratories, Burlingame, CA, USA).
PolyQ aggregates were detected on adjacent sections with
1C2 (MAB1547, Chemicon, Temecula, CA, USA), a mono-
clonal antibody, which reacts solely with the polyQ tract. Fol-
lowing aggressive antigen retrieval using 98% formic acid, the
tissue was incubated with the 1C2 (1:40 000), which was sub-
sequently detected using rat-absorbed biotinylated anti-mouse
IgG (Vector Laboratories), a single round of biotin tyramide
amplification as above, and Ni
+þ
/DAB/H
2
O
2
.
Sections were mounted on gelatin-coated glass slides and
lightly counterstained with thionin prior to coverslipping and
digital images of mounted sections were obtained using a
SpotRT camera mounted on a Leica RB microscope. Through-
focus sequences of images were processed as local contrast
composites using ImagePro Plus software (Media Cybernetics,
Silver Spring, MD, USA).
Immunohistochemistry for EM48 positive aggregates was
performed on free-floating sections according to standard
protocols using perfused rat brains ranging from 6 to
21 months of age (4,42). In brief, after blocking the
endogenous peroxidases with H
2
O
2
(0.04%) sections were
treated with normal sera prior to the addition of the
primary antibodies to block non-specific binding sites.
This was followed by an incubation with rabbit polyclonal
EM48, specific to the N-terminal region of huntingtin
(1:200; Dr Xiao-Jiang Li, Emory University, Atlanta,
USA). Incubation with primary antibody lasted for 48 h at
48C. Subsequently, sections were incubated with
biotinylated anti rabbit-IgG (1:600; Vector Laboratories,
Burlingame, CA, USA) for 2 h at room temperature. Sec-
tions were then reacted with preformed avidin– biotin per-
oxidase complexes (1:200), for 2 h at room temperature
(Vector Laboratories). Immunoreactivity was visualized by
3,3
0
-diaminobenzidine (DAB, 0.2%) in the presence of
0.007% hydrogen peroxide until a dark brown reaction
product was evident. Finally, the sections were mounted
onto APES (3-aminopropyl-triethoxysilan, Sigma – Aldrich,
Steinheim, Germany) coated slides, dehydrated and cover-
slipped with DePeX (Serva, Heidelberg, Germany).
Quantitative morphological analysis
All quantitative analyses were performed blind with respect to
genotype. A second set of rat brains, from control animals and
homozygous tgHD rats, ranging in age from 3 to 15 months
and with n¼ 3 for each group, was perfusion-fixed as described
earlier. Brains were then embedded in gelatin blocks, post-
fixed and freeze-cut to coronal sections of 1  60 mm fol-
lowed by 5  40 mm sections using MultiBrain
TM
embedding
(Neuroscience Associates, Knoxville, TN, USA). The 60 mm
sections were selected, mounted on gelatinized glass slides,
dried, defatted with 70% ethanol, and stained with cresyl
3192 Human Molecular Gen etics, 2006, Vol. 15, No. 21
at Universitaetsbibliothek Rostock on September 10, 2012http://hmg.oxfordjournals.org/Downloaded from

violet (Merck, Darmstadt, Germany). Slides were coverslipped
using DePeX (Serva, Heidelberg, Germany).
Quantitative analysis was performed with a stereology
workstation, consisting of a modified light microscope
(Eclipse 80i; Nikon, Tokyo, Japan), Nikon objectives (Plan
Apo, 2,NA¼ 1.0; Plan 10,NA¼ 0.35; Plan Apo VC
60, oil, NA ¼ 1.4), motorized specimen stage for automatic
sampling (Ma¨rzha¨user, Wetzlar, Germany), electronic micro-
cator (Heidenhain, Traunreut, Germany), CCD color camera
(Microfire, Optronics), PC with frame grabber board (Flash-
point Intrigue light, Integral Technologies, Indianapolis, IN,
USA) and stereology software (Stereo Investigator, Micro-
BrightField, Williston, VT, USA).
Striatal volume was investigated in a certain part of the
neostriatum (42) lying between 10.6 mm (where the corpus
callosum crosses the midline for the first time) and 8.2 mm
rostral to the inter-aural line (where the fornix enters the
diencephalon) (51). Dorsal and lateral boundaries were
defined by the corpus callosum. Medially, the lateral ventri-
cle bound the region of interest. A line drawn between the
ventral tip of the lateral ventricle and the rhinal fissure
defined the ventral border of the striatum. After exactly
tracing the boundary of the chosen brain area on video
images displayed on the monitor, the volumes of the
chosen brain areas were calculated with Cavalieri’s principle
(52). Note: tissue from wild-type and tgHD rats at the same
time-point (e.g. 3 months) was treated identically; however,
there was some experimental variability between time-points
(e.g. variability between tissue from 3 and 12 month time-
points), making volume comparisons between time-points
invalid.
Photomicrographs were produced by digital photography
using a Nikon DXM 1200F digital camera (Nikon, Tokyo,
Japan) attached to an Olympus AX 70 microscope and
ACT-1 software (Nikon). Images were processed with
Imaris imaging software (Bitplane, Zurich, Switzerland) and
converted into grey scales using Adobe Photoshop CS2 soft-
ware. Only minor adjustments of contrast and brightness
were made, which in no case altered the appearance of the
original materials.
ACKNOWLEDGEMENTS
This study was supported by grants from the Cure
Huntington’s Disease Initiative, the Hereditary Diseases and
High Q Foundation and by the START-Program of the
Faculty of Medicine, RWTH Aachen.
Conflict of Interest statement. I have had no involvements that
might raise the question of bias in the work reported or in the
conclusions, implications or opinions stated.
REFERENCES
1. The Huntington’s Disease Collaborative Research Group (1993) A novel
gene containing a trinucleotide repeat that is expanded and unstable on
Huntington’s disease chromosomes. Cell, 72, 971– 983.
2. Zuhlke, C., Riess, O., Schroder, K., Siedlaczck, I., Epplen, J.T., Engel, W.
and Thies, U. (1993) Expansion of the (CAG)n repeat causing
Huntington’s disease in 352 patients of German origin. Hum. Mol. Genet.,
2, 1467 –1469.
3. Bates, G.P. and Hockly, E. (2003) Experimental therapeutics in
Huntington’s disease: are models useful for therapeutic trials? Curr. Opin.
Neurol., 16, 465– 470.
4. von Horsten, S., Schmitt, I., Nguyen, H.P., Holzmann, C., Schmidt, T.,
Walther, T., Bader, M., Pabst, R., Kobbe, P., Krotova, J. et al. (2003)
Transgenic rat model of Huntington’s disease. Hum. Mol. Genet., 12,
617–624.
5. Luesse, H.G., Schiefer, J., Spruenken, A., Puls, C., Block, F. and
Kosinski, C.M. (2001) Evaluation of R6/2 HD transgenic mice for
therapeutic studies in Huntington’s disease: behavioral testing and impact
of diabetes mellitus. Behav. Brain Res., 126, 185– 195.
6. Murphy, K.P., Carter, R.J., Lione, L.A., Mangiarini, L., Mahal, A., Bates,
G.P., Dunnett, S.B. and Morton, A.J. (2000) Abnormal synaptic plasticity
and impaired spatial cognition in mice transgenic for exon 1 of the human
Huntington’s disease mutation. J. Neurosci., 20, 5115–5123.
7. Lione, L.A., Carter, R.J., Hunt, M.J., Bates, G.P., Morton, A.J. and
Dunnett, S.B. (1999) Selective discrimination learning impairments in
mice expressing the human Huntington’s disease mutation. J. Neurosci.,
19, 10428– 10437.
8. Van Raamsdonk, J.M., Pearson, J., Slow, E.J., Hossain, S.M.,
Leavitt, B.R. and Hayden, M.R. (2005) Cognitive dysfunction precedes
neuropathology and motor abnormalities in the YAC128 mouse model of
Huntington’s disease. J. Neurosci., 25, 4169– 4180.
9. D’Hooge, R. and De Deyn, P.P. (2001) Applications of the Morris water
maze in the study of learning and memory. Brain Res. Brain Res. Rev., 36,
60–90.
10. Whishaw, I.Q., Metz, G.A., Kolb, B. and Pellis, S.M. (2001) Accelerated
nervous system development contributes to behavioral efficiency in the
laboratory mouse: a behavioral review and theoretical proposal. Dev.
Psychobiol., 39, 151– 170.
11. File, S.E., Mahal, A., Mangiarini, L. and Bates, G.P. (1998) Striking
changes in anxiety in Huntington’s disease transgenic mice. Brain Res.,
805, 234 – 240.
12. Klivenyi, P., Bende, Z., Hartai, Z., Penke, Z., Nemeth, H., Toldi, J. and
Vecsei, L. (2006) Behaviour changes in a transgenic model of
Huntington’s disease. Behav. Brain Res., 169, 137– 141.
13. File, S.E. and Seth, P. (2003) A review of 25 years of the social interaction
test. Eur. J. Pharmacol., 463, 35 –53.
14. Kask, A., Nguyen, H.P., Pabst, R. and von Horsten, S. (2001) Factors
influencing behavior of group-housed male rats in the social interaction
test: focus on cohort removal. Physiol. Behav., 74, 277 –282.
15. Trottier, Y., Devys, D., Imbert, G., Saudou, F., An, I., Lutz, Y., Weber, C.,
Agid, Y., Hirsch, E.C. and Mandel, J.L. (1995) Cellular localization of the
Huntington’s disease protein and discrimination of the normal and
mutated form. Nat. Genet., 10, 104 –110.
16. Li, S.H. and Li, X.J. (1998) Aggregation of N-terminal huntingtin is
dependent on the length of its glutamine repeats. Hum. Mol. Genet., 7,
777–782.
17. Saudou, F., Finkbeiner, S., Devys, D. and Greenberg, M.E. (1998)
Huntingtin acts in the nucleus to induce apoptosis but death does not
correlate with the formation of intranuclear inclusions. Cell, 95, 55– 66.
18. Klement, I.A., Skinner, P.J., Kaytor, M.D., Yi, H., Hersch, S.M.,
Clark, H.B., Zoghbi, H.Y. and Orr, H.T. (1998) Ataxin-1 nuclear
localization and aggregation: role in polyglutamine-induced disease in
SCA1 transgenic mice. Cell, 95, 41 –53.
19. Kuemmerle, S., Gutekunst, C.A., Klein, A.M., Li, X.J., Li, S.H.,
Beal, M.F., Hersch, S.M. and Ferrante, R.J. (1999) Huntington aggregates
may not predict neuronal death in Huntington’s disease. Ann. Neurol., 46,
842–849.
20. Kim, M., Lee, H.S., La Foret, G., McIntyre, C., Martin, E.J., Chang, P.,
Kim, T.W., Williams, M., Reddy, P.H., Tagle, D. et al. (1999) Mutant
huntingtin expression in clonal striatal cells: dissociation of inclusion
formation and neuronal survival by caspase inhibition. J. Neurosci., 19,
964–973.
21. Di Figlia, M., Sapp, E., Chase, K.O., Davies, S.W., Bates, G.P.,
Vonsattel, J.P. and Aronin, N. (1997) Aggregation of huntingtin in
neuronal intranuclear inclusions and dystrophic neurites in brain. Science,
277, 1990– 1993.
22. Davies, S.W., Turmaine, M., Cozens, B.A., Di Figlia, M., Sharp, A.H.,
Ross, C.A., Scherzinger, E., Wanker, E.E., Mangiarini, L. and Bates, G.P.
(1997) Formation of neuronal intranuclear inclusions underlies the
neurological dysfunction in mice transgenic for the HD mutation. Cell, 90,
537–548.
Human Molecular Genetics, 2006, Vol. 15, No. 21 3193
at Universitaetsbibliothek Rostock on September 10, 2012http://hmg.oxfordjournals.org/Downloaded from

23. Slow, E.J., Graham, R.K., Osmand, A.P., Devon, R.S., Lu, G., Deng, Y.,
Pearson, J., Vaid, K., Bissada, N., Wetzel, R. et al. (2005) Absence of
behavioral abnormalities and neurodegeneration in vivo despite
widespread neuronal huntingtin inclusions. Proc. Natl Acad. Sci. USA,
102, 11402 –11407.
24. Osmand, A., Berthelier, V. and Wetzel, R. (2002) Identification of
aggregation foci, intracellular neuronal structures in the neocortex in
Huntington’s disease capable of recruiting polyglutamine. Soc. Neurosci.
Abstr., 293,6.
25. Osmand, A., Berthelier, V. and Wetzel, R. (2006) Imaging polyglutamine
deposits in brain tissue. Meth. Enzymol., 412, 106–122.
26. Holter, S.M., Tzschentke, T.M. and Schmidt, W.J. (1996) Effects of
amphetamine, morphine and dizocilpine (MK-801) on spontaneous
alternation in the 8-arm radial maze. Behav. Brain Res., 81, 53 –59.
27. Carter, R.J., Lione, L.A., Humby, T., Mangiarini, L., Mahal, A.,
Bates, G.P., Dunnett, S.B. and Morton, A.J. (1999) Characterization of
progressive motor deficits in mice transgenic for the human Huntington’s
disease mutation. J. Neurosci., 19, 3248 –3257.
28. Menalled, L.B., Sison, J.D., Dragatsis, I., Zeitlin, S. and Chesselet, M.F.
(2003) Time course of early motor and neuropathological anomalies in a
knock-in mouse model of Huntington’s disease with 140 CAG repeats.
J. Comp. Neurol., 465, 11–26.
29. Menalled, L.B., Sison, J.D., Wu, Y., Olivieri, M., Li, X.J., Li, H.,
Zeitlin, S. and Chesselet, M.F. (2002) Early motor dysfunction and
striosomal distribution of huntingtin microaggregates in Huntington’s
disease knock-in mice. J. Neurosci., 22, 8266 – 8276.
30. Hodgson, J.G., Agopyan, N., Gutekunst, C.A., Leavitt, B.R., LePiane, F.,
Singaraja, R., Smith, D.J., Bissada, N., McCutcheon, K., Nasir, J. et al.
(1999) A YAC mouse model for Huntington’s disease with full-length
mutant huntingtin, cytoplasmic toxicity, and selective striatal
neurodegeneration. Neuron, 23, 181–192.
31. Slow, E.J., van Raamsdonk, J., Rogers, D., Coleman, S.H., Graham, R.K.,
Deng, Y., Oh, R., Bissada, N., Hossain, S.M., Yang, Y.Z. et al. (2003)
Selective striatal neuronal loss in a YAC128 mouse model of Huntington
disease. Hum. Mol. Genet., 12, 1555 –1567.
32. Naver, B., Stub, C., Moller, M., Fenger, K., Hansen, A.K., Hasholt, L. and
Sorensen, S.A. (2003) Molecular and behavioral analysis of the R6/1
Huntington’s disease transgenic mouse. Neuroscience, 122, 1049–1057.
33. Hampton, T.G., Kale, A.S., McCue, S., Glazier, S., Hampton, M. and
Amende, I. (2005) Gait deterioration in the SOD G93A mouse model of
ALS. Soc. Neurosci. Abstr., 131, 13.
34. Lawrence, A.D., Sahakian, B.J., Hodges, J.R., Rosser, A.E., Lange, K.W.
and Robbins, T.W. (1996) Executive and mnemonic functions in early
Huntington’s disease. Brain, 119, 1633 –1645.
35. Ho, A.K., Sahakian, B.J., Brown, R.G., Barker, R.A., Hodges, J.R.,
Ane, M.N., Snowden, J., Thompson, J., Esmonde, T., Gentry, R. et al.
(2003) Profile of cognitive progression in early Huntington’s disease.
Neurology, 61, 1702– 1706.
36. Petrasch-Parwez, E., Habbes, H.W., Weickert, S.,
Lobbecke-Schumacher, M., Striedinger, K., Wieczorek, S., Dermietzel,
R. and Epplen, J.T. (2004) Fine-structural analysis and connexin
expression in the retina of a transgenic model of Huntington’s disease.
J. Comp. Neurol., 479, 181–197.
37. Helmlinger, D., Yvert, G., Picaud, S., Merienne, K., Sahel, J.,
Mandel, J.L. and Devys, D. (2002) Progressive retinal degeneration and
dysfunction in R6 Huntington’s disease mice. Hum. Mol. Genet., 11,
3351–3359.
38. Lawrence, A.D., Hodges, J.R., Rosser, A.E., Kershaw, A.,
ffrench-Constant, C., Rubinsztein, D.C., Robbins, T.W. and Sahakian, B.J.
(1998) Evidence for specific cognitive deficits in preclinical Huntington’s
disease. Brain, 121, 1329–1341.
39. Arrasate, M., Mitra, S., Schweitzer, E.S., Segal, M.R. and Finkbeiner, S.
(2004) Inclusion body formation reduces levels of mutant huntingtin and
the risk of neuronal death. Nature, 431, 805– 810.
40. Yamamoto, A., Lucas, J.J. and Hen, R. (2000) Reversal of neuropathology
and motor dysfunction in a conditional model of Huntington’s disease.
Cell, 101, 57– 66.
41. Li, H., Li, S.H., Yu, Z.X., Shelbourne, P. and Li, X.J. (2001) Huntingtin
aggregate-associated axonal degeneration is an early pathological event in
Huntington’s disease mice. J. Neurosci., 21, 8473– 8481.
42. Ka´ntor, O., Yasin Temel, Y., Holzmann, C., Raber, K., Nguyen, H.P.,
Cao, C., Tu
¨
rkoglu, H.O
¨
., Rutten, B.P.F., Visser-Vandewalle, V.,
Steinbusch, H.W.M. et al. (2006) Selective striatal neuron loss and
alterations in behavior in line with impaired striatal activity in
Huntington’s disease transgenic rats. Neurobiol. Dis., 22, 538 –547.
43. Bogo, V., Hill, T.A. and Young, R.W. (1981) Comparison of accelerod
and rotarod sensitivity in detecting ethanol- and acrylamide-induced
performance decrement in rats: review of experimental considerations of
rotating rod systems. Neurotoxicology, 2, 765– 787.
44. Jones, B.J. and Roberts, D.J. (1968) The quantitative measurement of
motor in co-ordination in naive mice using an accelerating rotarod.
J. Pharm. Pharmacol., 20, 302 –304.
45. Kask, A., Nguyen, H.P., Pabst, R. and Von Horsten, S. (2001)
Neuropeptide YY1 receptor-mediated anxiolysis in the dorsocaudal lateral
septum: functional antagonism of corticotropin-releasing
hormone-induced anxiety. Neuroscience, 104, 799– 806.
46. Karl, T., Hoffmann, T., Pabst, R. and von Horsten, S. (2003) Extreme
reduction of dipeptidyl peptidase IV activity in F344 rat substrains is
associated with various behavioral differences. Physiol. Behav., 80,
123–134.
47. Pellow, S., Chopin, P., File, S.E. and Briley, M. (1985) Validation of
open:closed arm entries in an elevated plus-maze as a measure of anxiety
in the rat. J. Neurosci. Meth., 14, 149– 167.
48. Holscher, C. and Schmidt, W.J. (1994) Quinolinic acid lesion of the rat
entorhinal cortex pars medialis produces selective amnesia in allocentric
working memory (WM), but not in egocentric WM. Behav. Brain Res.,
63, 187 –194.
49. Chen, S. and Wetzel, R. (2001) Solubilization and disaggregation of
polyglutamine peptides. Protein Sci., 10, 887– 891.
50. Adams, J.C. (1992) Biotin amplification of biotin and horseradish
peroxidase signals in histochemical stains. J. Histochem. Cytochem., 40,
1457–1463.
51. Paxinos, G. and Watson, C. (1982) The Rat Brain in Sterotaxic
Coordinates, Academic Press, San Diego, USA.
52. Cavalieri, B. (1635) Geometria Indivisibilibus Continuorum (reprinted
1966 as Geometria Degli Indivisibilii), Bononiae: Typis Clementis
Ferronij. Unione Tipografico-Editrice Torinese, Torino.
3194 Human Molecular Gen etics, 2006, Vol. 15, No. 21
at Universitaetsbibliothek Rostock on September 10, 2012http://hmg.oxfordjournals.org/Downloaded from