Human Molecular Genetics, 2005, Vol. 14, No. 5
doi:10.1093/hmg/ddi064
Advance Access published on January 20, 2005
679–691
Neuronal dysfunction in a polyglutamine
disease model occurs in the absence of
ubiquitin –proteasome system impairment
and inversely correlates with the degree of
nuclear inclusion formation
Aaron B. Bowman1,3, Seung-Yun Yoo2,{, Nico P. Dantuma4 and Huda Y. Zoghbi1,2,3,*
1
Department of Molecular and Human Genetics, 2Department of Neuroscience and 3Howard Hughes Medical
Institute, Baylor College of Medicine, Houston, TX 77030, USA and 4Department of Cell and Molecular Biology,
Karolinska Institute, Stockholm, Sweden
Received January 3, 2005; Revised January 10, 2005; Accepted January 13, 2005
The accumulation of protein deposits in neurons, in vitro proteasome assays and over-expression studies
suggest that impairment of the ubiquitin – proteasome system (UPS) may be a common mechanism of pathogenesis in polyglutamine diseases such as Huntington disease and spinocerebellar ataxias (SCAs). Using a
knock-in mouse model that recapitulates the clinical features of human SCA7, including selective neuronal
dysfunction, we assessed the UPS at cellular resolution using transgenic mice that express a green fluorescent protein (GFP)-based reporter substrate (UbG76V-GFP) of the UPS. The levels of the reporter remained
low during the initial phase of disease, suggesting that neuronal dysfunction occurs in the presence of a
functional UPS. Late in disease, we observed a significant increase in reporter levels specific to the most vulnerable neurons. Surprisingly, the basis for the increase in UbG76V-GFP protein can be explained by a corresponding increase in Ub G76V-GFP mRNA in the vulnerable neurons. An in vitro assay also showed normal
proteasome proteolytic activity in the vulnerable neurons. Thus, no evidence for general UPS impairment
or reduction of proteasome activity was seen. The differential increase of UbG76V-GFP among individual
neurons directly correlated with the down-regulation of a marker of selective pathology and neuronal dysfunction in SCA7. Furthermore, we observed a striking inverse correlation between the neuropathology
revealed by this reporter and ataxin-7 nuclear inclusions in the vulnerable neurons. Altogether, these data
show a protective role against neuronal dysfunction for polyglutamine nuclear inclusions and exclude
significant impairment of the UPS as a necessary step for polyglutamine neuropathology.
INTRODUCTION
Expansion of the glutamine-encoding CAG triplet repeats in nine
different genes is the cause of several degenerative neurological
diseases (Huntington disease, SCA-1, -2, -3, -6, -7 and -17,
DRPLA and SBMA) (1). A common feature of polyglutamine
diseases is the neuronal accumulation of the mutant protein in
nuclear or cytoplasmic inclusions (2,3). Similarly, expression of
proteins containing long polyglutamine tracts in mice leads to
protein aggregation, inclusions and neuronal dysfunction, thus
providing an important model system for the study of disease
pathology. These models have shown that the rate of neurological
dysfunction is proportional to the length of the polyglutamine
tract and expression level of the expanded protein. In fact,
many investigators take advantage of this fact and use polyglutamine tracts much longer than normally observed in human
disease, or over-express the mutant proteins, to observe neuropathology in the short lifespan of the mouse.
*To whom correspondence should be addressed. Tel: þ1 7137986558; Fax: þ1 7137988728; Email: [email protected]
{
Present address: Department of Neurology, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA 02115, USA.
Human Molecular Genetics, Vol. 14, No. 5 # Oxford University Press 2005; all rights reserved
680
Human Molecular Genetics, 2005, Vol. 14, No. 5
Involvement of the ubiquitin –proteasome system (UPS) is
strongly implicated in polyglutamine diseases (4). Polyglutamine inclusions and aggregates contain ubiquitin, molecular
chaperones and components of the proteasome, suggesting
that the mutant proteins are abnormally folded and may be
targeted for degradation by the UPS (5 – 7). The continuous
accumulation of these proteins suggests that they may be
resistant to degradation by the UPS. A relationship
between the UPS and polyglutamine diseases was bolstered
by research showing that genes in the UPS and protein
folding pathways are modifiers of neurodegeneration in
disease models and by recent work that provides indirect
evidence of UPS impairment in human patients (8 –13).
These observations have culminated in the hypothesis that
impairment of the UPS may underlie the polyglutamine
neurotoxicity and pathogenesis.
Support for the role of UPS impairment in polyglutamine
neuropathology has been indirectly substantiated by two
major findings. First, investigators using a GFP-based reporter
of UPS activity observed near-complete impairment of the
UPS by polyglutamine inclusions in an over-expression cell
culture model (14). Impairment of UPS activity was confirmed
in other over-expression cell culture models by examination of
endogenous proteasome targets and in vitro proteasome
activity assays (15,16). Second, recent work demonstrates
that eukaryotic proteasomes fail to degrade polyglutamine
sequences in vitro (17). These authors suggest that the
failure of the proteasome to release undigested polyglutamine
repeats may interfere with proteasome function. This idea is
supported by recent work showing inefficient degradation of
over-expressed polyglutamine proteins and sequestration of
the proteasome by polyglutamine protein aggregates in cultured cells (18). However, contradicting data have been presented, showing that soluble polyglutamine-containing
proteins with a strong degradation signal are efficiently
degraded in cultured cells regardless of glutamine tract
length and without proteasome impairment (19 – 21).
Additionally, conditional polyglutamine disease models
show rapid clearance of polyglutamine inclusions from
neurons (22,23). Furthermore, measurement of extractable
proteasome activity from transgenic mice expressing a
mutant huntingtin fragment and from a knock-in Huntington
disease model has revealed no change in proteasome activity
by in vitro proteasome assays (24,25). Issues of soluble
versus insoluble polyglutamine protein, expression level,
protein and cellular context, and the nature of the degradation
signal may justify these discrepancies. Most importantly, none
of these studies examined UPS activity in the context of the
vulnerable neuronal populations in vivo. Although measurement of extractable proteasome activity in the earlier mentioned mouse models comes close, the vulnerable neurons
represented a minority of cells from the extracted tissues.
Therefore, if proteasome impairment is restricted to the vulnerable neurons, then the reduction to extractable proteolytic
activity would be small and likely undetectable. Moreover,
besides reduction of proteasome activity, many other events
can potentially impair the complex UPS under pathological
conditions (26). To resolve these issues, it is necessary to
examine UPS activity specifically in the vulnerable neuronal
population during disease progression.
Spinocerebellar ataxia 7 (SCA7) is an autosomal dominant
neurodegenerative disease characterized by ataxia, retinal
degeneration and brainstem dysfunction. SCA7 is caused by
the expansion of the glutamine-encoding CAG repeat in
SCA7. In normal individuals, the repeat length ranges from
4 to 35 repeats, whereas affected individuals have between
37 and 460 repeats (27 –29). As with other polyglutamine diseases, the age of onset is inversely proportional to repeat
length. Repeat lengths less than 60 typically cause adult
onset disease, repeat lengths greater than 60 cause juvenile
onset and repeat lengths greater than 200 cause infantile
onset. Clinically, infantile SCA7 is among the most severe
forms of polyglutamine diseases and is associated with very
large repeat expansions (27,29).
We have generated a knock-in model of SCA7 by inserting
266 glutamine-encoding CAG repeats (instead of the usual
five CAG repeats) into the endogenous SCA7 mouse gene
(30). This polyglutamine disease model displays high fidelity
to the pathological features of the human disease including
expression of a single mutant allele by endogenous control
elements, a polyglutamine repeat size within the range
observed in humans, considerable overlap in the specific vulnerable neurons and shared symptoms with patients who have
similar repeat sizes (30). The mice (Sca7266Q/5Q) develop
molecular changes in the retina at 4 weeks and functional
impairments of the retina and cerebellum at 5 –6 weeks. The
visual and neurological functions of these mice progressively
deteriorate; by 14 weeks (C57Bl/6J genetic background), the
animals die due to brainstem dysfunction. There is a gradual
accumulation of insoluble ataxin-7 aggregates concomitant
with disease progression. Like other polyglutamine diseases,
ataxin-7 nuclear inclusions form in both affected and unaffected cells of the brain and retina. The cone and rod photoreceptor neurons, whose nuclei are found in the outer
nuclear layer (ONL) of the retina, are the most sensitive to
SCA7 pathology. In fact, although ataxin-7 nuclear inclusions
eventually form in all three nuclear layers of the retina, significant dysfunction, apoptosis and cell loss are confined to the
photoreceptors. A striking decrease of abundant photoreceptor
specific transcripts, first in the cones and then in the rods, precedes cell loss and detectable functional impairment of the
photoreceptors and is the earliest known pathological
feature. Furthermore, disease progression is marked by the
severe and rapid loss of the photoreceptor outer segments, a
decrease in photoreceptor inner segment (IS) mass and ONL
photoreceptor cell count, with little to no detectable loss of
cells, both neurons and glia, located in the other layers of
the retina. The retinal neuropathology of SCA7 presents a
unique opportunity to study polyglutamine neurotoxicity in a
tissue where the vulnerable neurons are a majority (.70%)
of the cells (31).
A prominent mechanism by which UPS impairment has
been proposed to occur in polyglutamine disease is through
direct blockage of the 20S core proteasome by polyglutamine
sequences (17,32). The mechanism suggests that the longer the
polyglutamine tract, the more severe the proteasome impairment might be. Thus under this mechanism, infantile SCA7
with its very large polyglutamine tract would be the most
sensitive disease for detecting such an impairment. The
UbG76V-GFP transgenic mouse has been developed as an
Human Molecular Genetics, 2005, Vol. 14, No. 5
in vivo reporter of the UPS (33). The UbG76V-GFP reporter
protein is an N-terminal ubiquitin mutant (UbG76V) in frame
with the enhanced green fluorescent protein (GFP), widely
expressed at high levels by a chimeric CMV-IE enhancer,
chicken b-actin promoter. The reporter is rapidly processed
and degraded by the cellular UPS, so steady state levels are
very low under wild-type conditions. Impairment to the UPS
in vivo causes an accumulation of the UbG76V-GFP reporter
protein, sensitive to both the extent and duration of inhibition
(33). Alterations in core proteasome activity and general UPS
activity are detectable by this reporter, whereas impairment to
some target-specific UPS components (e.g. a specific ubiquitin
ligase not active on the reporter) may not be detected (33,34).
Here, we use this reporter to assess impairment of the UPS at
cellular resolution in a mouse model of the polyglutamine
disease SCA7.
RESULTS
Using an in vivo reporter to assess the UPS in SCA7
Male Sca7266Q/5Q (henceforth simplified to Sca7266Q) mice
were crossed with female UbG76V-GFP/1þ transgenic mice
to generate experimental animals for assessment of the UPS.
As an initial assessment of the UPS, UbG76V-GFP/1þ/2
Sca75Q/5Q (henceforth simplified to UbG76V-GFPþ), UbG76VGFP/1þ/2 Sca7266Q/5Q (henceforth simplified to UbG76VGFPþ Sca7266Q) and UbG76V-GFP/12/2 (henceforth called
non-transgenic) littermates were analyzed by western blot
for the UbG76V-GFP protein. We found a commercially
available, affinity-purified, GFP antibody (Novus Biologicals),
which could detect steady state levels of UbG76V-GFP protein
in homogenates of normal transgenic mouse retina by western
blot analysis (Fig. 1A and B, first lanes). Analysis of the UPS
reporter levels in retinas of 6-week-old animals (early stage of
pathogenesis) revealed no significant difference between
UbG76V-GFPþ and UbG76V-GFPþ Sca7266Q animals
(Fig. 1A). However, by 13 weeks (terminal stage of pathogenesis), there was a significant increase in reporter levels in
UbG76V-GFPþ Sca7266Q retinas (Fig. 1B). Thus, although
this experiment argued against a general blockade of the
UPS in the initial phase of the SCA7 pathology, it raised the
possibility of an impairment to the UPS at the final stages of
disease.
To assess changes in UbG76V-GFP levels in specific layers
of the retina, we used laser confocal microscopy on 16 mm
frozen sections of retina to observe the native GFP fluorescence. We can detect UbG76V-GFP protein by its native fluorescence in UbG76V-GFPþ mouse retina as visualized by
comparison with the non-transgenic control (Fig. 2A). We
observed a significant increase in GFP fluorescence, specifically in the ONL of 13-week-old UbG76V-GFPþ Sca7266Q
mice (Fig. 2A). To confirm the increase of UbG76V-GFP reporter protein, we used laser confocal microscopy on replicate
16 mm frozen retina sections stained for GFP by immunofluorescence visualized with Cy3 to avoid overlap with
the emission spectrum of native GFP fluorescence. Again,
13-week-old UbG76V-GFPþ Sca7266Q mice showed a specific
increase in reporter levels in the ONL of the retina when compared with UbG76V-GFPþ littermates (Fig. 2B). Comparison of
681
Figure 1. UbG76V-GFP reporter levels in total retina. Representative western
blots for UbG76V-GFP and Gapdh in total retinal protein extracts. Genotype
is indicated above each lane; animals are littermates. (A) Levels of UbG76VGFP are unaltered in retinas of animals at 6 weeks of age, early in SCA7
pathogenesis. (B) Levels of UbG76V-GFP are elevated in retinas of animals
at 13 weeks of age, late in SCA7 pathogenesis.
the UbG76V-GFP images with Toto-3 nuclear stained images
revealed that UbG76V-GFP is both cytoplasmic and nuclear,
but excluded from the more densely nuclear stained heterochromatic core of the ONL photoreceptors.
To determine how much UbG76V-GFP protein levels change
during SCA7 pathogenesis and to quantify the difference, we
have measured native GFP fluorescence levels in retina of
UbG76V-GFPþ, UbG76V-GFPþ Sca7266Q and non-transgenic
littermate sets at 5, 7, 9, 11, 13 and 14 weeks of age
(Sca7266Q animals die at 13– 14 weeks). To optimize quantitative measurements, samples in each set were prepared and
analyzed side-by-side in identical fashion. Although GFP
immunofluorescence gave a higher signal-to-noise ratio,
native fluorescence was determined to be optimal for quantitative analysis due to significantly lower signal variability.
The fluorescent signal from the UbG76V-GFP reporter in
UbG76V-GFPþ and UbG76V-GFPþ Sca7266Q animals was
detected significantly higher than the background signal of
the non-transgenic control for all data sets (t-test,
P , 0.005) (Fig. 3A and B). Analysis across time points
revealed an increase in UbG76V-GFP protein levels in the
ONL over the course of pathogenesis (Fig. 3A). In earlystage pathogenesis (5 and 7 weeks), reporter protein levels
were not significantly higher in the UbG76V-GFPþ Sca7266Q
mice when compared with UbG76V-GFPþ mice. By midstage pathogenesis (9 and 11 weeks), UbG76V-GFPþ
Sca7266Q ONL reporter levels were consistently and significantly elevated, on average 1.6-fold above UbG76V-GFPþ
ONL reporter levels (t-test, P , 0.001). By late-stage pathogenesis (13 and 14 weeks), UbG76V-GFPþ Sca7266Q ONL
reporter levels were further elevated from UbG76V-GFPþ
Sca7266Q mid-stage levels and were on average 3.2-fold
above UbG76V-GFPþ ONL levels. In contrast, inner nuclear
layer (INL) levels were not substantially different between
UbG76V-GFPþ and UbG76V-GFPþ Sca7266Q animals over the
time course of pathogenesis (Fig. 3B). Elevation of UbG76VGFP levels over the course of disease was confirmed in an
independent Sca7266Q line of mixed C57Bl/6J and 129SV/
EV background (data not shown). We also observed an
increase in UbG76V-GFP protein in the IS of UbG76V-GFPþ
Sca7266Q animals when compared with UbG76V-GFPþ
animals (Fig. 2A and B). However, the loss of IS mass in
the Sca7266Q animals made the follow-up measurements of
this difference problematic.
682
Human Molecular Genetics, 2005, Vol. 14, No. 5
Figure 2. UbG76V-GFP reporter levels are increased specifically in photoreceptors. The sum images of z-section confocal image scans of retina cryosections of
littermate animal sets. The IS, ONL and inner nuclear layer (INL) are indicated on the left; genotype and age are indicated on the right. Photoreceptor nuclei
reside in the ONL. (A) Left panels in blue show Toto-3 nuclear stain image; right panels in green show native GFP fluorescence. There is a specific increase
in UbG76V-GFP levels in the ONL. (B) Left panels in blue show Toto-3 nuclear stain image; right panels in green show anti-GFP immunofluorescence using a
Cy3-conjugated secondary antibody for visualization. There is a specific increase in UbG76V-GFP levels in the ONL.
Human Molecular Genetics, 2005, Vol. 14, No. 5
683
abundant photoreceptor specific transcripts are significantly
down-regulated in the Sca7266Q mice, we wanted to determine
whether changes in the levels of the reporter mRNA were
influencing our measurement of UPS activity. Our predominant concern was that Ub G76V-GFP mRNA levels would be
decreased in the Sca7266Q mice leading to an underestimate
of UPS impairment. The Ub G76V-GFP mRNA levels were
similar between UbG76V-GFPþ and UbG76V-GFPþ Sca7266Q
animals at an early stage of pathogenesis (5 weeks).
However, contrary to expectations, we observed a significant
increase in UbG76V-GFP mRNA levels in late-stage pathogenesis (12.5 weeks) in UbG76V-GFPþ Sca7266Q animals by northern blot (Fig. 4A). Furthermore, the difference in Ub G76V-GFP
mRNA levels in total retina at late stages of pathogenesis
matched, or were perhaps slightly greater than, the difference
in UbG76V-GFP protein levels for total retina at late stages
determined by both confocal microscopy and western blot
(Fig. 4B). Thus, the levels of reporter mRNA reflected the
levels of reporter protein at both early and late stages of pathogenesis. To determine whether the increase in Ub G76V-GFP
mRNA levels was specific to the ONL, like the increase in
UbG76V-GFP protein levels was, we performed in situ hybridization on 16 mm frozen retinal sections. We observed a striking increase in Ub G76V-GFP mRNA levels in the ONL and IS
(where the protein production machinery of the photoreceptor
cells resides) in UbG76V-GFPþ Sca7266Q animals (Fig. 5). The
changes in Ub G76V-GFP mRNA levels were remarkably
similar to the changes in UbG76V-GFP protein levels.
An independent assay confirms normal levels of
proteasome activity
Figure 3. UbG76V-GFP levels over SCA7 pathogenesis in the ONL and INL.
Quantitative confocal image analysis of native fluorescent signal in independent littermate animal sets over the course of SCA7 pathogenesis. Mean fluorescent signal relative to the UbG76V-GFPþ animal are plotted, with error bars
indicating standard error (n . 15 for UbG76V-GFPþ and UbG76V-GFPþ
Sca7266Q samples; n . 10 for non-transgenic control samples). The data are
grouped for each independent set of three animals, with animal age indicated
below each group. The levels of fluorescence (indicating UbG76V-GFP reporter
levels) in UbG76V-GFPþ (5Q) and UbG76V-GFPþ Sca7266Q (266Q) animals
were detected significantly (two-tail t-test, P , 0.005) above the non-transgenic control (N-T), for all data sets shown. (A) The levels of fluorescence
in the ONL were consistently and significantly (two-tail t-test, P , 0.001)
increased in 266Q over 5Q for data sets 9 weeks of age and older. (B) The
levels of fluorescence in the INL were not substantially elevated in 266Q
mice when compared with 5Q mice over pathogenesis.
An increase in UbG76V-GFP mRNA levels accounts for the
increase in protein levels
Steady state levels of the UbG76V-GFP protein are determined
by a balance between protein production and protein degradation. For changes in UbG76V-GFP protein levels to be an
accurate measure of UPS activity, changes in protein production must be minimal relative to changes in protein degradation in the samples being compared. Because many
The change in UbG76V-GFP protein levels observed in
UbG76V-GFPþ Sca7266Q animals was matched by a change
in Ub G76V-GFP mRNA levels, indicating no significant
impairment of UPS activity. We considered the possibility
that proteasome activity may have been reduced in the
Sca7266Q retina, but left unnoticed in the reporter mice
because it did not reach the critical threshold for functional
impairment of the UPS. To test this possibility, we measured
the extractable chymotrypsin-like proteasome activity in
wild-type and Sca7266Q littermates at the final stages of
pathogenesis. Two independent measurements of extractable
proteasome activity failed to detect a decrease in the proteasome activity in the retina (Fig. 6). In fact, a slight elevation
in proteasome activity was observed in both sample pairs.
This assay also failed to detect a decrease in proteasome
activity in another vulnerable tissue of Sca7266Q animals, the
cerebellum (data not shown).
UbG76V-GFP levels as a marker of polyglutamine
neuropathology
A marker of SCA7 pathogenesis and photoreceptor dysfunction is the progressive down-regulation of photoreceptor
specific transcripts, such as Rhodopsin (30). As the increase
in UbG76V-GFP mRNA and protein levels observed is specific
to the vulnerable photoreceptor neurons, while less vulnerable
neurons of the retina show no substantial change, UbG76V-GFP
reporter levels seem to be a marker of selective neuropathology.
684
Human Molecular Genetics, 2005, Vol. 14, No. 5
segment, staining in the ONL is detectable in both wild-type
and mutant retina (data not shown). The photoreceptors with
the highest levels of UbG76V-GFP reporter consistently displayed relatively low levels of rhodopsin (Fig. 7A), whereas
photoreceptors with the highest levels of rhodopsin consistently displayed relatively low levels of UbG76V-GFP (Fig. 7B).
To quantify this relationship, the mean fluorescent signal of rhodopsin and UbG76V-GFP for individual photoreceptors in the
ONL was measured for ONL cells with the highest intensity
of rhodopsin signal and ONL cells with highest intensity of
UbG76V-GFP signal (see Materials and Methods). The 95%
confidence interval for both measures of the two classes of
cells was plotted on an XY graph to illustrate the inverse
relationship (Fig. 7C). The mean UbG76V-GFP and mean rhodopsin signals were inversely correlated and significantly
different (t-test, P , 0.0001) between the class of photoreceptors with the brightest rhodopsin staining and the class
of photoreceptors with the brightest UbG76V-GFP signal.
Therefore, the increase in UbG76V-GFP reporter levels in
individual photoreceptors directly correlates with the downregulation of rhodopsin, a marker of selective neuropathology
and neuronal dysfunction.
Strong inverse relationship between UbG76V-GFP detected
neuropathology and ataxin-7 nuclear inclusions
Figure 4. Ub G76V-GFP mRNA levels in total retina. (A) Representative northern blots for Ub G76V-GFP and Gapdh mRNA in total retinal RNA extracts at 5
and 12.5 weeks of age. Levels of Ub G76V-GFP mRNA are unaltered early in
SCA7 pathogenesis, but elevated late in SCA7 pathogenesis. Genotype is indicated above each lane; animals of the same age are littermates. (B) Quantification of Ub G76V-GFP mRNA levels by northern blot phosphoimager analysis
relative to Gapdh versus levels of UbG76V-GFP protein by western blot densitometry relative to Gapdh and quantitative confocal image analysis. The mean
levels in total retina from UbG76V-GFPþ Sca7266Q (266Q) animals were set
relative to UbG76V-GFPþ (5Q) animals, which were defined as the unitary
value; error bars indicate standard error of the relative values. The increase
in UbG76V-GFP protein levels late in pathogenesis in UbG76V-GFPþ
Sca7266Q animals correlated with an increase in mRNA levels.
Examination of the UbG76V-GFP protein pattern revealed a
non-uniform accumulation in the ONL. To determine
whether the variation in UbG76V-GFP levels reflects differences in the degree of neuronal dysfunction and pathology
between individual neurons, we compared UbG76V-GFP
levels with the levels of rhodopsin in photoreceptors by
immunofluoresence. By late (14 weeks)-stage pathogenesis,
Rhodopsin mRNA levels have been reduced for several
weeks, allowing sufficient time for protein levels to decline.
Although the majority of rhodopsin is located in the outer
Nuclear inclusions are a hallmark feature of polyglutamine
disease. To determine how the abundance of the UbG76VGFP reporter in individual photoreceptor neurons of the
ONL correlated with the accumulation of inclusions, we
stained the retinal sections for ataxin-7 nuclear inclusions by
immunofluorescence. We observed a strong inverse correlation between UbG76V-GFP protein levels and the size and
intensity of ataxin-7 nuclear inclusions from mid (9 weeks)
to late (14 weeks) stages of pathogenesis. For example, at
mid-stage pathogenesis, the photoreceptor nuclei with the
highest intensity nuclear inclusions had the weakest UbG76VGFP signal (Fig. 8A), whereas the cells with the strongest
UbG76V-GFP signal had weaker staining or lacked ataxin-7
nuclear inclusions (Fig. 8A). The confounding effect of multiple cells on the same z-axis makes visualization of this observation in two-dimensional reconstructions of the confocal
images difficult, as a photoreceptor cell with a bright nuclear
inclusion may lie behind a cell with a bright UbG76V-GFP
signal. Volume rendering three-dimensional reconstructions
of the ataxin-7 nuclear inclusions and UbG76V-GFP signal
revealed the robustness of this observation (data not shown).
Examination of three-dimensional reconstructions revealed
that photoreceptors with the brightest ataxin-7 nuclear
inclusions invariably had relatively weak UbG76V-GFP reporter
levels, whereas photoreceptors with the strongest UbG76VGFP reporter levels invariably had the weakest stained or
lacked detectable ataxin-7 nuclear inclusions altogether.
This inverse correlation is in stark contrast with the ataxin-7
staining pattern in the less vulnerable cells of the INL,
where the UbG76V-GFP reporter levels were unchanged
from wild-type and independent of the nuclear inclusions
(data not shown). To quantify the inverse correlation in the
photoreceptors, the mean fluorescent signal of ataxin-7 and
UbG76V-GFP for individual photoreceptors in the ONL was
Human Molecular Genetics, 2005, Vol. 14, No. 5
685
Figure 5. Ub G76V-GFP mRNA levels are specifically increased in photoreceptors. Representative in situ hybridization for Ub G76V-GFP mRNA (purple) with an
eosin counter stain (pink). IS, ONL and INL are indicated on the left of each panel; genotype and age are indicated on the right. Photoreceptor nuclei reside in the
ONL, and photoreceptor protein translation occurs in the IS. Both layers show a specific increase in Ub G76V-GFP mRNA levels.
measured for ONL cells with high intensity ataxin-7
nuclear inclusions and ONL cells with high intensity
UbG76V-GFP signal at mid (9 weeks) and late (14 weeks)
stages of pathogenesis (see Materials and Methods). The
95% confidence interval for both measures of the two
classes was plotted on an XY graph to illustrate the inverse
relation (Fig. 8B). The mean UbG76V-GFP and mean ataxin-7
signals were inversely correlated and significantly different
686
Human Molecular Genetics, 2005, Vol. 14, No. 5
Figure 6. Proteasome activity is not decreased in an in vitro proteasome assay.
Chymotrypsin-like proteasome activity in total retinal protein extracts of wildtype (5Q) and Sca7266Q (266Q) animals in two littermate pairs. Arbitrary in
vitro activity relative to total protein concentration of each extract is
graphed, with error bars indicating the sum of the linear regression errors.
No impairment in proteasome activity was detected.
(t-test, P , 0.0001 at both 9 and 14 weeks) between the class
of photoreceptors with the brightest ataxin-7 nuclear
inclusions and the class of photoreceptors with the brightest
UbG76V-GFP signal.
DISCUSSION
We have examined the UPS in a mouse model of the polyglutamine disease SCA7 (specifically the infantile form); in this
model, the real-time progression of disease and selective neuronal vulnerability is similar to what is seen in infant patients
with similar polyglutamine expansions (30). We have examined the neurons most vulnerable to this disease with an in
vivo reporter that assays UPS activity at the resolution of individual cells. Because the increase in UbG76V-GFP protein
levels were more than matched by an increase in Ub G76VGFP mRNA levels, and because in vitro assays of extractable
chymotrypsin-like proteasome activity were at normal levels,
in fact possibly higher in the mutants, our experiments rule
out significant impairment of the UPS in the Sca7266Q
model. Furthermore, no changes in UbG76V-GFP reporter
levels are observed at 5 and 7 weeks of age, despite documented photoreceptor dysfunction as measured by electrophysiological and molecular studies (30). It is difficult to determine
the minimal level of UPS impairment necessary to detect an
increase in UbG76V-GFP reporter levels, as relating independent measures of proteasome activity (e.g. extractable
chymotrypsin-like proteasome activity) to changes in UPS
activity on in vivo substrates is problematic (26). However,
previous work has shown that the UbG76V-GFP transgenic
reporter line is sensitive to both the time and degree of
proteasome impairment (33). Twenty hours after treatment
with proteasome inhibitors at the concentration necessary to
Figure 7. UbG76V-GFP levels in the ONL are a marker of neuronal dysfunction. The down-regulation of Rhodopsin in photoreceptors is a marker of pathology and neuronal dysfunction in SCA7. The photoreceptor cells of the ONL
were analyzed by confocal imaging of 14-week-old UbG76V-GFPþ Sca7266Q
retina cryosections stained by immunofluorescence for rhodopsin and
UbG76V-GFP. (A) Representative single optical section of a photoreceptor
cell classified to the group of cells having the brightest UbG76V-GFP staining.
Single channel and merged images are shown. (B) Representative single
optical section of a photoreceptor cell classified to the group of cells having
the brightest rhodopsin staining. Single channel and merged images are
shown. (C) A two-point XY plot showing the quantitative confocal measurement of 14-week-old UbG76V-GFPþ Sca7266Q retinas, plotting the arbitrary
mean rhodopsin signal (Y-axis) and the arbitrary mean anti-GFP immunofluorescent signal (X-axis) from two groups of photoreceptors, those with the
brightest rhodopsin staining (red square) versus those with the brightest
UbG76V-GFP signal (green circle). Error bars indicate the 95% confidence
interval for the mean value of the rhodopsin signal (error bars along the Yaxis) and the UbG76V-GFP signal (error bars along the X-axis) for individual
photoreceptors of each group (n ¼ 22). Photoreceptor cells with the highest
levels of UbG76V-GFP exhibit the weakest levels of rhodopsin, whereas photoreceptor cells with the least degree of rhodopsin down-regulation show the
lowest levels of UbG76V-GFP. A strong inverse correlation in both rhodopsin
and UbG76V-GFP signals is observed between the two groups.
Human Molecular Genetics, 2005, Vol. 14, No. 5
687
Figure 8. Ataxin-7 nuclear inclusions are inversely correlated with UbG76V-GFP levels. (A) Representative section of a sum image of five slices from z-section
confocal image scan of the ONL from a 9-week-old UbG76V-GFPþ Sca7266Q retina cryosection stained by immunofluorescence for ataxin-7 nuclear inclusions
and native GFP fluorescence of UbG76V-GFP. Photoreceptors with the brightest ataxin-7 nuclear inclusions have low UbG76V-GFP levels, whereas those with the
brightest UbG76V-GFP levels have either weakly stained or undetectable ataxin-7 nuclear inclusions. Single channel and merged images are shown. (B) A
two-point XY plot showing the quantitative confocal measurement of 9- and 14-week-old UbG76V-GFPþ Sca7266Q retinas, plotting the arbitrary mean ataxin7 signal (Y-axis) and the arbitrary mean native GFP fluorescent signal (X-axis) from two groups of photoreceptors, those with the brightest nuclear inclusions
(red square) versus those with the brightest UbG76V-GFP signal (green circle). Error bars indicate the 95% confidence interval for the mean value of the ataxin-7
signal (error bars along the Y-axis) and the UbG76V-GFP signal (error bars along the X-axis) for individual photoreceptors of each group (9 weeks, n ¼ 25; 14
weeks, n ¼ 14). A strong inverse correlation in both ataxin-7 and UbG76V-GFP signals is observed between the two groups.
detect statistically significant changes in extractable proteasome activity, many mouse tissues showed substantial increases
in UbG76V-GFP reporter levels; moreover, proteasome impairment in primary cultures showed .10-fold increase after
16 h. Because photoreceptor dysfunction and pathology
persist for at least 9 weeks (from 5 weeks of age till death at
14 weeks of age) in the Sca7266Q mutant line, this allows substantial time for potential reporter accumulation. Furthermore,
our analysis indicated that we could detect very small
changes, such as an increase of just 1.5-fold in ONL UbG76VGFP protein levels, at a high statistical confidence. The high
sensitivity of our assay, coupled with the ability to examine
UPS activity at the cellular level, should have greatly facilitated
our ability to detect impairments to the UPS (34). It is thereby
remarkable that we fail to detect significant UPS impairment
in the most vulnerable neuronal population even at the terminal
stages of pathogenesis.
Although there could be UPS impairment in the Sca7266Q
model below detection of the UbG76V-GFP reporter, it is noteworthy that another proteasome targeted GFP reporter easily
detected UPS impairment upon over-expression of a polyglutamine containing protein in cultured cells (14). Additionally,
significant impairment of extractable chymotrypsin-like
in vitro proteasome activity was detected after 3 days of
expression of polyglutamine proteins in cultured cells
(15,16,25). We employed similar methods of UPS activity
determination and failed to detect impairment. In vitro
assays from other cell culture and mouse models have also
failed to detect impairment of proteasome activity
(24,25,35). Specifically, investigation of a knock-in model of
Huntington disease did not reveal impairment to proteasome
activity (25). However, cell-type specific proteasome impairment could not be ruled out as a mitigating factor, as the vulnerable neurons represent a minority of cells in the brain
688
Human Molecular Genetics, 2005, Vol. 14, No. 5
regions examined (24,25). Therefore, by examining UPS
activity specifically in the vulnerable neuronal population,
using established methods comparable to those previously
used to show impairment in over-expression models, we
conclude that significant UPS impairment is not required
or detected in an authentic model of SCA7 polyglutamine
neuropathology. Although we obviously cannot exclude the
possibility of UPS impairment in other polyglutamine diseases
or other polyglutamine genetic models (e.g. transgenic overexpression models), current theory as to the mechanism of
proteasome impairment does not provide a reason to justify
UPS impairment being disease specific. The mechanisms
used to explain proteasome impairment focus on the polyglutamine tract itself (3,4,17,18,32). Thus, showing that any
polyglutamine disease can occur in the absence of detectable
UPS impairment is broadly applicable.
The approximately equivalent changes in UbG76V-GFP
protein and mRNA indicate that the UbG76V-GFP protein
levels are an indirect measure of Ub G76V-GFP transcription
(or possibly mRNA stability). Interestingly, another GFP
reporter targeted for proteasomal degradation has been developed as an in vivo reporter of transcriptional activity (36).
Li et al. showed that the short half-life of their GFP reporter
was instrumental in detecting changes in promoter activity, as
the high steady state levels found with wild-type GFP masked
increases due to transcriptional induction. It is tempting to
speculate that the short half-life of the UbG76V-GFP reporter
in our in vivo model aided in the detection of small transcriptional aberrations. It is intriguing that the earliest sign of pathology in the Sca7266Q mice is a down-regulation of abundant
photoreceptor specific transcripts (30,37). The SCA7 gene
has recently been demonstrated to encode a subunit of the
GCN5 histone acetyltransferase complex, STAGA, implicating its normal function to be in transcriptional regulation
(38). The fact that transcriptional changes are the earliest
event in SCA7 pathogenesis suggests that the mechanism for
Ub G76V-GFP reporter mRNA changes in SCA7 is likely to
be transcriptional. In fact, a transcriptional pathology has
been implicated for numerous polyglutamine diseases (39).
We propose two possible scenarios to explain the increase in
Ub G76V-GFP mRNA levels. One possibility is that neuronal
stress, secondary to SCA7 neuropathology, is causing an
increase in expression of the transgenic reporter. Another
possibility is that the down-regulation of cell-specific transcription in the vulnerable neurons might free up the general
transcriptional machinery, increasing its availability for
expression of the transgenic reporter. This idea is consistent
with our observation that photoreceptors with the highest
level of reporter exhibit significantly lower levels of rhodopsin
than neighboring cells with lower levels of the reporter.
There has been considerable debate regarding the role of
polyglutamine inclusion bodies in neuropathology (2,3). Previous work has shown that nuclear inclusions are not required
for cell death or neuropathology, and genetic conditions that
lead to decreased inclusion body formation in cultured
neurons or transgenic mice, correlate with increased cell
death and neuropathology (8,40 – 42). However, these studies
could not control for susceptibility differences between
neuronal subtypes or the altered genetic conditions of the
neuronal populations examined. A recent study has elegantly
demonstrated that polyglutamine nuclear inclusion formation
promotes cell survival in neurons transfected with a polyglutamine GFP fusion protein (43). Given the protection against
cell death in cultured neurons, this poses the question of
whether nuclear inclusions also provide protection in vivo or
from the selective neuronal dysfunction that gradually develops in disease and precedes neuronal death. In the Sca7266Q
model, significant photoreceptor dysfunction occurs prior to
cell loss (30). To determine whether nuclear inclusions play
a protective role against neuronal dysfunction in vivo, and at
the level of individual neurons, we examined the relationship
between the ataxin-7 nuclear inclusions and UbG76V-GFP
reporter levels. UbG76V-GFP reporter accumulation is a sensitive in vivo marker of disease pathology, given its specific and
progressive accumulation in the vulnerable neurons, and the
correlation between this accumulation and the loss of rhodopsin,
a marker of photoreceptor dysfunction and neuropathology
(Figs 3 and 7). We found a striking inverse correlation
between nuclear inclusion load and the degree of neuropathology between individual photoreceptors (Fig. 8 and data not
shown). Importantly, because we observe this effect among
neighboring neurons of the same subtype, this protective
effect cannot be simply explained by differences in the susceptibility or genetic state of the neuronal population. Lastly,
ataxin-7 nuclear inclusion formation antagonizes neuronal
dysfunction throughout pathogenesis, such that at any stage
of disease, there is an inverse correlation between inclusions
and neuropathology. Thus, within the vulnerable neuronal
population, the cells with more extensive inclusion formation exhibit reduced pathology relative to those cells with a
lesser degree of inclusion formation. Unfortunately, nuclear
inclusion formation is insufficient to halt disease, as the neuropathology gradually overwhelms the photoreceptors despite
continued inclusion formation.
Polyglutamine aggregation reactions are kinetically driven
by a nucleation process, and the aggregation process can
deplete soluble pools of the polyglutamine protein (2,3,43).
The inverse relationship between the nuclear inclusions and
UbG76V-GFP reporter levels is consistent with the neurotoxic
species being either monomeric polyglutamine proteins or oligomers. This finding supports recent in vitro and cell culture
observations pointing to a mechanism of transcription factor
impairment, mediated through soluble polyglutamine proteins
(44). Inhibition of polyglutamine aggregation is being pursued
as a therapeutic approach to polyglutamine disease. Our findings do not necessarily obviate a therapeutic benefit for this
approach as long as decreasing the aggregate load is
accompanied by a global decrease in total cellular polyglutamine protein. It is obviously critical to ensure that the
process of eliminating aggregates is not enhancing the
soluble toxic pool of the mutant protein. In all likelihood,
the relationship between the inclusions and concentration of
the toxic polyglutamine species is complex and may explain
the beneficial effect of some aggregation inhibitors in polyglutamine models (45,46). However, the robust in vivo inverse
correlation we observed strongly excludes a causative role
for polyglutamine nuclear inclusions in neuronal dysfunction
and selective neuronal vulnerability. In summary, the data presented clarify the relationship between nuclear inclusions and
polyglutamine-induced pathology in vulnerable neurons and
Human Molecular Genetics, 2005, Vol. 14, No. 5
reveal, in an authentic disease model, that despite an established disease state, no impairment of the UPS is detected.
MATERIALS AND METHODS
Mouse genetics
Both the Sca7266Q/5Q/C1 line (30) and UbG76V-GFP/1 line (33)
used in this study have been backcrossed with C57Bl/6J. Heterozygous male Sca7266Q mice (7 –10-week-old) were crossed
with two or three heterozygous female UbG76V-GFPþ mice
(2 – 6-month-old) to generate experimental animals. Genotyping of the progeny was performed by PCR using the primer
sets as previously described (30,33). Genotyping for all
animals was confirmed after dissections. Data sets always
included three animals, littermate paired UbG76V-GFPþ and
UbG76V-GFPþ Sca7266Q animals, plus a non-transgenic age
matched animal that was almost always a littermate.
Western blot
Total retinal protein was extracted from dissected retina by
Dounce homogenization in TSTE buffer [150 mM NaCl,
50 mM Tris pH7.5, 0.5% Triton X-100, 2 mM EDTA and 1
protease inhibitor cocktail (Sigma P-8340)], incubated on ice
for 4 –6 h, centrifuged at 14 000 g and supernatant pulled to
a fresh tube, stored at 2808C till use. Equal protein was
loaded and analyzed by standard western blot protocols. An
affinity-purified polyclonal antibody against GFP (Novus Biologicals Cat No. 600 –308) was used at a 1: 1000 dilution. A
mouse monoclonal antibody against Gapdh (Advanced
Immuno Chemical Cat No. RGM2) was used at a 1: 10 000
dilution. Western blots were developed using Pierce
SuperSignal West Dura Extended Duration Substrate, according to manufacture’s directions. Relative levels of full-length
UbG76V-GFP protein (47) to Gapdh protein were determined
by densitometry.
Preparation of retinal cryosections for confocal analysis
All steps of tissue preparation, sectioning, slide preparation
and staining were done side-by-side in identical fashion,
with common reagents for each data set. The complete dissection of retina in cold PBS was performed in ,5 min and retina
placed in 4% paraformaldehyde on ice in the dark (diluted
from 16% solution, Electron Microscopy Sciences), until all
three pairs of retinas were completed. The retinas were fixed
at 48C in the dark for 4 –6 h with the fixative changed one
time. The fixative was replaced with 30% sucrose overnight
at 48C in the dark to equilibrate and cryoprotect. The samples
were mounted in OCT and stored at 2808C until sectioning.
The 16 mm cryosections were cut with minimal lighting and
slides covered to protect from light, while completing all
samples. Slides were dried at room temperature in the dark
for 1 –4 h, then placed at 2808C. For native GFP analysis,
slides were removed from 2808C and allowed to dry for
20 min in the night before analysis. Slides were washed
together in PBT (PBS with 0.1% Triton X-100), stained with
Toto-3 (Molecular Probes) at a 1 : 1000 dilution for 5 min,
washed with PBT three to five times, briefly washed with
689
PBS, then coverslipped with Vectashield-Dapi (Vector) and
stored at 48C in the dark until the next morning for confocal
analysis. Immunofluorescence staining for UbG76V-GFP or
ataxin-7 nuclear inclusions was performed as previously
described, except slides were kept in dark whenever possible
(30). The rabbit anti-GFP antibody (Novus) was used at
1 : 1000, the mouse anti-rhodopsin antibody (Chemicon) was
used at 1 : 200, the 1261 anti-ataxin-7 antibody (48,49) was
used at 1: 200 and the anti-rabbit and anti-mouse (Cy-3 or
Alexa488) conjugated secondary antibodies (Jackson
Immunochemical) were used at 1 : 400.
Quantitative confocal image collection and data analysis
Image collection was done in a single sitting for each data set
on a Zeiss LSM 510 confocal microscope. Identical confocal
settings were used for all samples within a data set, the settings were adjusted for the data set to the UbG76V-GFPþ
Sca7266Q animal sample to avoid pixel saturation and
achieve optimal detection. Laser power was set to give
minimal bleaching during z-section collection. Image collection of an area was done only once, and selection of the
area to collect was done with the GFP excitation laser off to
avoid collection bias. Z-section image collection was done
at constant step interval (0.6 mm), through the full-depth of
the 16 mm cryosection for the each channel, with slice count
noted and kept within 15% variability for all image collections
in the set. Quantitative analysis of the fluorescence signal was
done with ImageJ software, using the sum image of the entire
z-stack. Selection of the ‘region-of-interest’ (ROI) to measure
was decided by marking the ROI from the Toto-3 nuclear stain
channel image, then measuring the GFP channel in the
matched image. Regions of the image were excluded where
the nuclear stain revealed damage to the tissue from
slide preparation, otherwise the entire retinal layer of interest
(ONL, INL or total retina) was selected as the ROI. For analysis of total retinal levels, all layers were included in the ROI
except the outer segment because of auto-fluorescence. The
mean value of the ROI in the GFP channel was determined
for each image collected. Then the average of means with
standard error for each genotype of the data set was determined; the sample values for each genotype were then set relative to the UbG76V-GFPþ sample of each data set, defined as
the unitary value. Analysis of independently analyzed replicate slides gave ,20% variability in the determined relative
values (data not shown). To quantify the inverse correlation
between UbG76V-GFP (Alexa488) and rhodopsin (Cy3) immunofluorescent signals, photoreceptors from single optical sections of multiple z-section images from a single animal were
used; for the inverse correlation of UbG76V-GFP native fluorescence and Ataxin-7 nuclear inclusions (Cy3), photoreceptors
from constant optical depth sum images of multiple z-section
images from a single animal were used. The inverse correlation between UbG76V-GFP and rhodopsin or ataxin-7
nuclear inclusions was independently repeated with at least
two animals. In choosing ONL cells with the highest intensity
Cy3 staining or ONL cells with the highest intensity UbG76VGFP signal, we excluded cells when another photoreceptor in
the same Z-axis had a high intensity signal in either channel
that overlapped on the optical section image. To ensure that
690
Human Molecular Genetics, 2005, Vol. 14, No. 5
all detectable ataxin-7 aggregates in a selected cell would be
measured, only cells entirely contained within the optical
section image were measured. The mean fluorescent signals
for the individual photoreceptors of each class were measured
for both the UbG76V-GFP and Cy3 channel images, using a circular ROI selection in ImageJ software. All statistical analysis
of data to determine the mean value, standard error, 95% confidence interval of the mean and t-test (two sample assuming
unequal variances) were performed using Microsoft Excel.
Northern blot
Total retina RNA was prepared using both retinas from individual animals, using a Trizol (Invitrogen) extraction with
Dounce homogenization, according to the manufacture’s
instructions. Northern blot was performed as previously
described (30). The Gapdh probe is a 486 bp PCR fragment
of the mouse cDNA; the GFP probe was from the plasmid
pEGFP-N1 (Invitrogen). The relative levels of Ub G76V-GFP
mRNA and Gapdh mRNA were determined by phosphoimager
analysis.
In situ hybridization
The protocol for in situ hybridization on retina with a DIGlabeled anti-sense RNA probe was adapted for 16 mm cryosections, but otherwise performed as described (50). The EGFP
probe, subcloned from the pEGFP-N1 vector into pBluescript
II SK, was transcribed with T3 RNA polymerase from linearized vector to avoid vector sequence contamination. Tissue
was prepared as described earlier for confocal analysis.
In vitro proteasome activity
The chymotrypsin-like activity of total retina extracted proteasomes from Sca7266Q and wild-type littermates were measured
using the fluorogenic substrate succinyl-Leu-Leu-Val-TyrAMC (Affiniti) in conditions very similar to those previously
described (14). Both retinas of each animal were extracted in
400 ml of proteasome activity buffer [10 mM Tris pH 7.5,
1 mM EDTA, 20% glycerol, 2 mM ATP, 4 mM DTT, 0.5%
Triton X-100 and 1 protease inhibitor cocktail (Sigma
P-8340)] on ice by dounce homogenization, spun at 14 000g,
supernatant collected on ice and assayed within 4 h. The
substrate was used at 130 mM , and digestion was monitored
over a period of 2 h with five time points on a fluorometer
(Hoefer Pharmacia Biotech TKO 100). Proteasome independent digestion was very low as monitored by digestion of
the substrate after 2 h in the presence of 50 mM lactacystin.
Total protein concentration was determined by the Biorad
protein assay using linear regression of a five point standard
curve with a standardized BSA solution. Proteasome activity
is reported as the ratio of the arbitrary units of fluorescence
generated per minute, determined by linear regression of the
five time points, over the total protein concentration for each
extract, with error bars reporting the sum of the regression
errors.
ACKNOWLEDGEMENTS
We thank Richard Atkinson for expert advice with confocal
imagery and 3D reconstructions, Xiuqian Mu and William
Klein for valuable advice on retinal in situ and Sukeshi Vaishnav
for in situ reagents. We appreciate Jean-Louis Mandel, Gael
Yvert and Dominique Helmlinger for sharing the 1261
ataxin-7 antibody and Juan Botas, Hung-Kai Chen, Adriano
Flora, Noah Shroyer and Hiroshi Tsuda for critical comments
on the manuscript and helpful suggestions. A.B.B. was an
HHMI postdoctoral fellow of the Life Sciences Research
Foundation and is currently a postdoctoral fellow of the
Hereditary Disease Foundation. H.Y.Z. is an investigator of
the Howard Hughes Medical Institute. This work was also
supported by grants from the NIH: NINDS NS27699 (to
H.Y.Z.) and NICHD HD24064 (to the Gene expression core
of BCM-MRRC).
REFERENCES
1. Ross, C.A. (2002) Polyglutamine pathogenesis: emergence of unifying
mechanisms for Huntington’s disease and related disorders. Neuron, 35,
819–822.
2. Ross, C.A. and Poirier, M.A. (2004) Protein aggregation and
neurodegenerative disease. Nat. Med., 10(suppl.), S10– S17.
3. Michalik, A. and van Broeckhoven, C. (2003) Pathogenesis of
polyglutamine disorders: aggregation revisited. Hum. Mol. Genet., 12,
R173– R186.
4. Ciechanover, A. and Brundin, P. (2003) The ubiquitin proteasome system
in neurodegenerative diseases: sometimes the chicken, sometimes the egg.
Neuron, 40, 427 –446.
5. Schmidt, T., Lindenberg, K.S., Krebs, A., Schols, L., Laccone, F., Herms,
J., Rechsteiner, M., Riess, O. and Landwehrmeyer, G.B. (2002) Protein
surveillance machinery in brains with spinocerebellar ataxia type 3:
redistribution and differential recruitment of 26S proteasome subunits and
chaperones to neuronal intranuclear inclusions. Ann. Neurol., 51,
302–310.
6. Stenoien, D.L., Cummings, C.J., Adams, H.P., Mancini, M.G., Patel, K.,
DeMartino, G.N., Marcelli, M., Weigel, N.L. and Mancini, M.A. (1999)
Polyglutamine-expanded androgen receptors form aggregates that
sequester heat shock proteins, proteasome components and SRC-1, and
are suppressed by the HDJ-2 chaperone. Hum. Mol. Genet., 8, 731–741.
7. Chai, Y., Koppenhafer, S.L., Shoesmith, S.J., Perez, M.K. and Paulson,
H.L. (1999) Evidence for proteasome involvement in polyglutamine
disease: localization to nuclear inclusions in SCA3/MJD and suppression
of polyglutamine aggregation in vitro. Hum. Mol. Genet., 8, 673 –682.
8. Cummings, C.J., Reinstein, E., Sun, Y., Antalffy, B., Jiang, Y.,
Ciechanover, A., Orr, H.T., Beaudet, A.L. and Zoghbi, H.Y. (1999)
Mutation of the E6-AP ubiquitin ligase reduces nuclear inclusion
frequency while accelerating polyglutamine-induced pathology in SCA1
mice. Neuron, 24, 879–892.
9. Fernandez-Funez, P., Nino-Rosales, M.L., de Gouyon, B., She, W.C.,
Luchak, J.M., Martinez, P., Turiegano, E., Benito, J., Capovilla, M.,
Skinner, P.J. et al. (2000) Identification of genes that modify
ataxin-1-induced neurodegeneration. Nature, 408, 101–106.
10. Chan, H.Y., Warrick, J.M., Andriola, I., Merry, D. and Bonini, N.M.
(2002) Genetic modulation of polyglutamine toxicity by protein
conjugation pathways in Drosophila. Hum. Mol. Genet., 11, 2895–2904.
11. Cummings, C.J., Mancini, M.A., Antalffy, B., DeFranco, D.B., Orr, H.T.
and Zoghbi, H.Y. (1998) Chaperone suppression of aggregation and
altered subcellular proteasome localization imply protein misfolding in
SCA1. Nat. Genet., 19, 148–154.
12. Cummings, C.J., Sun, Y., Opal, P., Antalffy, B., Mestril, R., Orr, H.T.,
Dillmann, W.H. and Zoghbi, H.Y. (2001) Over-expression of inducible
HSP70 chaperone suppresses neuropathology and improves motor
function in SCA1 mice. Hum. Mol. Genet., 10, 1511–1518.
13. De Pril, R., Fischer, D.F., Maat-Schieman, M.L., Hobo, B., de Vos, R.A.,
Brunt, E.R., Hol, E.M., Roos, R.A. and van Leeuwen, F.W. (2004)
Human Molecular Genetics, 2005, Vol. 14, No. 5
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
Accumulation of aberrant ubiquitin induces aggregate formation and cell
death in polyglutamine diseases. Hum. Mol. Genet., 13, 1803–1813.
Bence, N.F., Sampat, R.M. and Kopito, R.R. (2001) Impairment of the
ubiquitin– proteasome system by protein aggregation. Science, 292,
1552–1555.
Nishitoh, H., Matsuzawa, A., Tobiume, K., Saegusa, K., Takeda, K.,
Inoue, K., Hori, S., Kakizuka, A. and Ichijo, H. (2002) ASK1 is essential
for endoplasmic reticulum stress-induced neuronal cell death triggered by
expanded polyglutamine repeats. Genes Dev., 16, 1345–1355.
Jana, N.R., Zemskov, E.A., Wang, G. and Nukina, N. (2001) Altered
proteasomal function due to the expression of polyglutamine-expanded
truncated N-terminal huntingtin induces apoptosis by caspase activation
through mitochondrial cytochrome c release. Hum. Mol. Genet., 10,
1049–1059.
Venkatraman, P., Wetzel, R., Tanaka, M., Nukina, N. and Goldberg, A.L.
(2004) Eukaryotic proteasomes cannot digest polyglutamine sequences
and release them during degradation of polyglutamine-containing
proteins. Mol. Cell, 14, 95– 104.
Holmberg, C.I., Staniszewski, K.E., Mensah, K.N., Matouschek, A. and
Morimoto, R.I. (2004) Inefficient degradation of truncated polyglutamine
proteins by the proteasome. EMBO J., 23, 4307–4318.
Verhoef, L.G., Lindsten, K., Masucci, M.G. and Dantuma, N.P. (2002)
Aggregate formation inhibits proteasomal degradation of polyglutamine
proteins. Hum. Mol. Genet., 11, 2689–2700.
Michalik, A. and van Broeckhoven, C. (2004) Proteasome degrades
soluble expanded polyglutamine completely and efficiently. Neurobiol.
Dis., 16, 202 –211.
Kaytor, M.D., Wilkinson, K.D. and Warren, S.T. (2004) Modulating
huntingtin half-life alters polyglutamine-dependent aggregate formation
and cell toxicity. J. Neurochem., 89, 962–973.
Zu, T., Duvick, L.A., Kaytor, M.D., Berlinger, M.S., Zoghbi, H.Y., Clark,
H.B. and Orr, H.T. (2004) Recovery from polyglutamine-induced
neurodegeneration in conditional SCA1 transgenic mice. J. Neurosci., 24,
8853–8861.
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.
Diaz-Hernandez, M., Hernandez, F., Martin-Aparicio, E., Gomez-Ramos,
P., Moran, M.A., Castano, J.G., Ferrer, I., Avila, J. and Lucas, J.J. (2003)
Neuronal induction of the immunoproteasome in Huntington’s disease.
J. Neurosci., 23, 11653–11661.
Zhou, H., Cao, F., Wang, Z., Yu, Z.X., Nguyen, H.P., Evans, J., Li, S.H.
and Li, X.J. (2003) Huntingtin forms toxic NH2-terminal fragment
complexes that are promoted by the age-dependent decrease in
proteasome activity. J. Cell. Biol., 163, 109–118.
Lindsten, K. and Dantuma, N.P. (2003) Monitoring the ubiquitin/
proteasome system in conformational diseases. Ageing Res. Rev., 2,
433 –449.
van de Warrenburg, B.P., Frenken, C.W., Ausems, M.G., Kleefstra, T.,
Sinke, R.J., Knoers, N.V. and Kremer, H.P. (2001) Striking anticipation in
spinocerebellar ataxia type 7: the infantile phenotype. J. Neurol., 248,
911 –914.
Johansson, J., Forsgren, L., Sandgren, O., Brice, A., Holmgren, G. and
Holmberg, M. (1998) Expanded CAG repeats in Swedish spinocerebellar
ataxia type 7 (SCA7) patients: effect of CAG repeat length on the clinical
manifestation. Hum. Mol. Genet., 7, 171 –176.
Benton, C.S., de Silva, R., Rutledge, S.L., Bohlega, S., Ashizawa, T. and
Zoghbi, H.Y. (1998) Molecular and clinical studies in SCA-7 define a
broad clinical spectrum and the infantile phenotype. Neurology, 51,
1081–1086.
Yoo, S.-Y., Pennesi, M.E., Weeber, E.J., Xu, B., Atkinson, R., Chen, S.,
Armstrong, D.L., Wu, S.M., Sweatt, J.D. and Zoghbi, H.Y. (2003) SCA7
knockin mice model human SCA7 and reveal gradual accumulation of
mutant ataxin-7 in neurons and abnormalities in short-term plasticity.
Neuron, 37, 383–401.
Young, R.W. (1985) Cell proliferation during postnatal development of
the retina in the mouse. Brain Res., 353, 229 –239.
Goellner, G.M. and Rechsteiner, M. (2003) Are Huntington’s and
polyglutamine-based ataxias proteasome storage diseases?
Int. J. Biochem. Cell. Biol., 35, 562 –571.
691
33. Lindsten, K., Menendez-Benito, V., Masucci, M.G. and Dantuma, N.P.
(2003) A transgenic mouse model of the ubiquitin/proteasome system.
Nat. Biotechnol., 21, 897–902.
34. Hernandez, F., Diaz-Hernandez, M., Avila, J. and Lucas, J.J. (2004)
Testing the ubiquitin–proteasome hypothesis of neurodegeneration
in vivo. Trends Neurosci., 27, 66–69.
35. Ding, Q., Lewis, J.J., Strum, K.M., Dimayuga, E., Bruce-Keller, A.J.,
Dunn, J.C. and Keller, J.N. (2002) Polyglutamine expansion, protein
aggregation, proteasome activity, and neural survival. J. Biol. Chem., 277,
13935–13942.
36. Li, X., Zhao, X., Fang, Y., Jiang, X., Duong, T., Fan, C., Huang, C.C. and
Kain, S.R. (1998) Generation of destabilized green fluorescent protein as a
transcription reporter. J. Biol. Chem., 273, 34970–34975.
37. La Spada, A.R., Fu, Y.H., Sopher, B.L., Libby, R.T., Wang, X., Li, L.Y.,
Einum, D.D., Huang, J., Possin, D.E., Smith, A.C. et al. (2001)
Polyglutamine-expanded ataxin-7 antagonizes CRX function and induces
cone–rod dystrophy in a mouse model of SCA7. Neuron, 31, 913–927.
38. Helmlinger, D., Hardy, S., Sasorith, S., Klein, F., Robert, F., Weber, C.,
Miguet, L., Potier, N., van-Dorsselaer, A., Wurtz, J.M. et al. (2004)
Ataxin-7 is a subunit of GCN5 histone acetyltransferase-containing
complexes. Hum. Mol. Genet., 13, 1257–1265.
39. Sugars, K.L. and Rubinsztein, D.C. (2003) Transcriptional abnormalities
in Huntington disease. Trends Genet., 19, 233 –238.
40. 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.
41. 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.
42. Watase, K., Weeber, E.J., Xu, B., Antalffy, B., Yuva-Paylor, L.,
Hashimoto, K., Kano, M., Atkinson, R., Sun, Y., Armstrong, D.L. et al.
(2002) A long CAG repeat in the mouse Sca1 locus replicates SCA1
features and reveals the impact of protein solubility on selective
neurodegeneration. Neuron, 34, 905–919.
43. 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.
44. Schaffar, G., Breuer, P., Boteva, R., Behrends, C., Tzvetkov, N., Strippel,
N., Sakahira, H., Siegers, K., Hayer-Hartl, M. and Hartl, F.U. (2004)
Cellular toxicity of polyglutamine expansion proteins: mechanism of
transcription factor deactivation. Mol. Cell, 15, 95 –105.
45. Pollitt, S.K., Pallos, J., Shao, J., Desai, U.A., Ma, A.A., Thompson, L.M.,
Marsh, J.L. and Diamond, M.I. (2003) A rapid cellular FRET assay of
polyglutamine aggregation identifies a novel inhibitor. Neuron, 40,
685–694.
46. Apostol, B.L., Kazantsev, A., Raffioni, S., Illes, K., Pallos, J., Bodai, L.,
Slepko, N., Bear, J.E., Gertler, F.B., Hersch, S. et al. (2003) A cell-based
assay for aggregation inhibitors as therapeutics of polyglutamine-repeat
disease and validation in Drosophila. Proc. Natl Acad. Sci. USA, 100,
5950–5955.
47. Dantuma, N.P., Lindsten, K., Glas, R., Jellne, M. and Masucci, M.G.
(2000) Short-lived green fluorescent proteins for quantifying ubiquitin/
proteasome-dependent proteolysis in living cells. Nat. Biotechnol., 18,
538–543.
48. Lindenberg, K.S., Yvert, G., Muller, K. and Landwehrmeyer, G.B. (2000)
Expression analysis of ataxin-7 mRNA and protein in human brain:
evidence for a widespread distribution and focal protein accumulation.
Brain Pathol., 10, 385 –394.
49. Yvert, G., Lindenberg, K.S., Picaud, S., Landwehrmeyer, G.B., Sahel, J.A.
and Mandel, J.L. (2000) Expanded polyglutamines induce
neurodegeneration and trans-neuronal alterations in cerebellum and retina
of SCA7 transgenic mice. Hum. Mol. Genet., 9, 2491–2506.
50. Mu, X., Beremand, P.D., Zhao, S., Pershad, R., Sun, H., Scarpa, A., Liang,
S., Thomas, T.L. and Klein, W.H. (2004) Discrete gene sets depend on
POU domain transcription factor Brn3b/Brn-3.2/POU4f2 for their
expression in the mouse embryonic retina. Development, 131,
1197–1210.