Date post: | 21-Feb-2023 |
Category: |
Documents |
Upload: | independent |
View: | 0 times |
Download: | 0 times |
Caspase-mediated proteolysis of the polyglutamine disease protein
ataxin-3
Sarah J. Shoesmith Berke,*,� Francisca A. Flores Schmied,� Ewout R. Brunt,� Lisa M. Ellerby§and Henry L. Paulson*,�
*Neuroscience Graduate Program and the �Department of Neurology, University of Iowa, Iowa City, Iowa, USA
�Department of Neurology, University Hospital Groningen, Groningen, The Netherlands
§Buck Institute for Research in Aging, Novato, California, USA
Abstract
Spinocerebellar ataxia type-3, also known as Mach-
ado-Joseph Disease, is one of many inherited neurodegen-
erative disorders caused by polyglutamine-encoding CAG
repeat expansions in otherwise unrelated disease genes.
Polyglutamine disorders are characterized by disease protein
misfolding and aggregation; often within the nuclei of affected
neurons. Although the precise mechanism of polyglutamine-
mediated cell death remains elusive, evidence suggests that
proteolysis of polyglutamine disease proteins by caspases
contributes to pathogenesis. Using cellular models we now
show that the endogenous spinocerebellar ataxia type-3 dis-
ease protein, ataxin-3, is proteolyzed in apoptotic paradigms,
resulting in the loss of full-length ataxin-3 and the corres-
ponding appearance of an approximately 28-kDa fragment
containing the glutamine repeat. Broad-spectrum caspase
inhibitors block ataxin-3 proteolysis and studies suggest that
caspase-1 is a primary mediator of cleavage. Site-directed
mutagenesis experiments eliminating three, six or nine
potential caspase cleavage sites in the protein suggest
redundancy in the site(s) at which cleavage can occur, as
previously described for other disease proteins; but also map
a major cleavage event to a cluster of aspartate residues
within the ubiquitin-binding domain of ataxin-3 near the poly-
glutamine tract. Finally, caspase-mediated cleavage of
expanded ataxin-3 resulted in increased ataxin-3 aggregation,
suggesting a potential role for caspase-mediated proteolysis
in spinocerebellar ataxia type-3 pathogenesis.
Keywords: ataxin-3, caspase, proteolysis, Machado-Joseph
disease, polyglutamine disease, spinocerebellar ataxia type-3.
J. Neurochem. (2004) 89, 908–918.
Evidence from many experimental systems links the clea-
vage of polyglutamine (polyQ) disease proteins by cellular
proteases to disease pathogenesis. First, polyQ protein
fragments often aggregate and cause cell death more readily
than their full-length protein counterparts, and are sufficient
to produce disease in various cellular, invertebrate and
vertebrate models (Satyal et al. 2000; Wellington et al. 2000;
Morley et al. 2002; Sanchez et al. 2003). Second, activated
caspases have been detected in numerous models of disease,
and inhibition of caspase activity suppresses disease protein
proteolysis and subsequent pathogenesis in vitro and in vivo
(Ellerby et al. 1999a; Ellerby et al. 1999b; Kim et al. 1999;
Sanchez et al. 1999; Li et al. 2000; Wellington et al. 2000).
Finally, in some polyQ diseases proteolytic fragments have
been observed in susceptible regions of human disease brain
tissue, with the appearance of such fragments preceding the
onset of disease phenotypes (Gafni and Ellerby 2002; Garden
et al. 2002; Wellington et al. 2002).
Caspases are cysteine proteases that cleave proteins at
specific aspartate residues (Earnshaw et al. 1999). Caspase
Received November 25, 2003; revised manuscript received December
18, 2003; accepted December 22, 2003.
Address correspondence and reprint requests to Sarah J Shoesmith
Berke, University of Iowa, Department of Neurology, 240 A EMRB,
Paulson Laboratory, Iowa City, Iowa 52242.
E-mail: [email protected]
Abbreviations used: ALLN, N-acetyl-leucinyl-leucyl-norleucinal; CI,
caspase inhibitor; DMSO, dimethylsulfoxide; GAPDH, glyceraldehyde-
3-phosphate dehydrogenase; HD, Huntington Disease; PARP, poly-
adenosine ribose phosphatase; polyQ, polyglutamine; SCA3/MJD,
Spinocerebellar ataxia type-3/Machado-Joseph Disease; ST, staurospo-
rine.
Journal of Neurochemistry, 2004, 89, 908–918 doi:10.1111/j.1471-4159.2004.02369.x
908 � 2004 International Society for Neurochemistry, J. Neurochem. (2004) 89, 908–918
involvement in polyQ pathogenesis appears to be complex,
varying in the details among the polyQ diseases. An initiator
role is supported by studies in a mouse model of Huntington
Disease (HD), in which caspase-derived proteolytic frag-
ments were observed prior to behavioral or pathological
signs of disease (Wellington et al. 2002). Continued pro-
duction of toxic polyQ fragments may further propagate
cellular toxicity and aggregation (Ellerby et al. 1999a;
Paulson 2000). Low-grade chronic caspase activation during
disease progression could also have secondary effects on
other cellular targets. Finally, caspases may accelerate
apoptosis leading to the neuronal demise that occurs in this
class of diseases (Ellerby et al. 1999b; Sanchez et al. 1999).
Several polyQ disease proteins are known targets for
caspase cleavage, including huntingtin (caspases -2, -3, -6
and -7), the androgen receptor (caspases -1, -3, -7 and -8)
and atrophin-1 (caspases -1, -3, -7 and -8) (Wellington
et al. 1998; Ellerby et al. 1999a; Ellerby et al. 1999b;
Kim et al. 1999; LaFevre-Bernt and Ellerby 2003).
Caspase involvement varies among the polyQ diseases,
in part because the disease proteins are highly dissimilar,
with the number and position of potential caspase sites
differing for each protein, and in part because different
caspases may be activated in different polyQ diseases.
Moreover, cell-type dependent cleavage by specific casp-
ases or other proteases (Troy and Salvesen 2002) may
explain some of the selective degeneration seen in these
diseases (Zoghbi and Orr 2000), as evidenced by a recent
report in HD showing that caspase-mediated fragments
were found in susceptible human and mouse brain regions
(Wellington et al. 2002).
In this report, we address the possible role of proteolytic
cleavage in spinocerebellar ataxia type-3 (also known as
Machado-Joseph disease or SCA3/MJD). SCA3/MJD is the
most common dominantly inherited ataxia and one of at least
nine neurodegenerative polyQ diseases (Paulson 2000). As in
other polyQ diseases, there is regional selectivity to pathol-
ogy. Although mutant ataxin-3, the SCA3/MJD disease
protein, is expressed widely throughout the brain, only
specific brain regions degenerate, including neuronal popu-
lations within the globus pallidus, midbrain, pons and spinal
cord. While many polyQ proteins are known targets for
caspase-mediated cleavage, characterization of the caspase
cleavage of ataxin-3 has not been thoroughly investigated.
Wellington et al. (Wellington et al. 1998) showed a modest
production of ataxin-3 fragments by caspases in vitro. In
contrast, Evert et al. (1999) reported that mutant ataxin-3 is
not prone to fragmentation in stably transfected cell lines.
Although there have been no descriptions of ataxin-3
fragments in SCA3/MJD human disease tissue, a recently
developed SCA3/MJD mouse model suggests the accumu-
lation of proteolytic fragments (Goti and Colomer 2003). A
possible explanation for these divergent results is that
caspase cleavage depends on protein–protein interactions
(Earnshaw et al. 1999), which may be affected by cellular
environment or by the overexpression of ataxin-3.
Here we show that ataxin-3 is a target for caspase-
mediated cleavage in cell culture. Full-length ataxin-3 is
cleaved during staurosporine (ST)-induced apoptosis in all
cell types tested and leads to the appearance of an ataxin-3
fragment that contains the glutamine repeat in at least one
cell type. In our cellular model, this proteolysis is mediated
predominantly by caspase-1, and site-directed mutagenesis
experiments narrows the major cleavage site to three
aspartate residues clustered in the ubiquitin binding region
of ataxin-3 near the polyQ domain. Interestingly, aggregation
of mutant ataxin-3 was increased in ST treated cells, and
blocked by caspase inhibition, consistent with caspase
involvement in SCA3/MJD pathogenesis.
Experimental procedures
Cell culture and transfection
Methods for cell culture of Cos-7, HeLa and PC6-3 cells have been
described previously (Chai et al. 1999a, b). NIH3T3 cells were
grown in Delbecco’s modified Eagle’s medium (DMEM; Gibco
BRL, Grand Island, NY, USA) supplemented with 10% FBS
(HyClone, Logan, UT, USA) and 100 U/mL/500 lg/mL penicillin/
streptomycin (P/S; Gibco BRL). NT2 cells were grown in Opti-
minimum essential medium (MEM; Gibco BRL) supplemented with
5% FBS and 1% P/S. All cells were maintained at 37�C with 5%
CO2. PC63 cells were differentiated using 100 ng/mL nerve growth
factor (NGF; Promega, Madison, WI, USA) maintained in 4% HS,
2%FBS with 1% P/S. For transient transfections, 50% confluent
dishes were transfected with 1 lg of expression plasmid DNA using
LipofectAMINE PLUSTM Reagent (Invitrogen, Carlsbad, CA,
USA).
Apoptotic paradigms and treatments
Twenty-four hours after plating, apoptosis was induced by incuba-
ting cells with 1 lM ST in serum free media (ST; Sigma; St. Louis,
MO, USA). Control cells were incubated with equivalent volumes
of the vehicle, dimethylsulfoxide (DMSO). To block caspase
activation, Z-VAD-FMK or specific caspase inhibitors to caspases
-1, -2, -3, -5, -6, -8 and -9 were used at 100 lM and 10 lM,respectively (Calbiochem, San Diego, CA, USA). Lactacytsin was
used to inhibit proteasome activity, and N-acetyl-leucinyl-leucyl-
norleucinal (ALLN) to block calpain activity at 10 lM and 100 lM,respectively (Calbiochem). All inhibitors were added at the time of
ST treatment except ALLN, which was added 1 h prior to ST
treatment.
Western blot analysis
Cell lysates were collected 24 h after ST treatment. Both live and
dead cells from the same condition were harvested into Laemmli
buffer (50 mM Tris pH 6.8, 100 mM dithiothreitol, 2% SDS, 0.1%
bromophenol blue, 10% glycerol) and then pooled. Lysates were
analyzed by western blot as described previously (Chai et al.
1999b). Ataxin-3 was detected with rabbit polyclonal antibod-
ies, anti-ataxin-3 (1 : 20 000) and 146 pa (1 : 200; gift from
Caspase cleavage of ataxin-3 909
� 2004 International Society for Neurochemistry, J. Neurochem. (2004) 89, 908–918
V. Colomer, Johns Hopkins University) mouse monoclonal anti-
bodies, anti-1H9, and anti-2B6 (1 : 2000; gift from Y. Trottier,
Louis Pasteur University), and rabbit polyclonal antibodies anti-HA,
or anti-myc (1 : 1000; Santa Cruz, Santa Cruz, CA, USA). Polya-
denosine ribose phosphatase (PARP) expression was detected using a
rabbit polyclonal antibody, anti-PARP (1 : 500; Santa Cruz), and the
loading control glyceraldehyde-3-phosphate dehydrogenase (GAP-
DH) was assessed with anti-GAPDH mouse monoclonal antibody
(1 : 1000; Research Diagnostics Inc., Flanders, NJ, USA). Secondary
antibodies were peroxidase-conjugated goat anti-rabbit or mouse
antisera (1 : 15 000; Jackson Laboratories, West Grove, PA, USA).
Site-directed mutagenesis
Site-directed mutagenesis was performed using the QuikChange Kit
(Stratagene, La Jolla, CA, USA) to generate the pCMX-HA-ataxin-3
(Q35) D225N, pCMX-HA-ataxin-3(Q35)D225/228 N and pCMX-
HA-ataxin-3(Q35)-6mut constructs (Ellerby et al. 1999a, b; LaFe-
vre-Bernt and Ellerby 2003). The pFLAG-CMV6a-ataxin-3(Q28)-
3mut and pFLAG-CMV6a-ataxin-3(Q35)-9mut constructs were
generated by PCR amplification using pCDNA3-ataxin-3(Q28)
and pCMX-HA-ataxin-3(Q35)-6mut as templates, respectively. The
generated PCR fragments were then subcloned into the pFLAG-
CMV6a vector (Sigma) and sequence verified.
Plasmids
The plasmids, pcDNA3-HA ataxin-3(Q28) and pcDNA3-HA ataxin-
3(Q84) were used to generate doubly tagged ataxin-3 constructs.
Normal and expanded ataxin-3 epitope tagged with myc at the
C-terminus were generated by PCR amplification of an Eco0109
I/Apa I fragment with pcDNA3-HA ataxin-3(Q28) as the template,
then subcloning this fragment into pcDNA3-HA ataxin-3(Q28) and
pcDNA-HA-ataxin-3(Q84). All constructs were sequence verified.
Human disease tissue
We looked for ataxin-3 fragments in two different SCA3/MJD
disease brains. The first brain, designated MJD-LL, was from a
31-year-old female in late stages of SCA3/MJD, described in
(Paulson et al. 1997a). The CAG repeat lengths were approximately
29 and 82 repeats. The second SCA3/MJD brain, S017-01, was
from a 34 years old male who died in advanced-stage disease with
cytosine-adenine-guanine (CAG) repeat lengths of 20 and 75. Tissue
samples were collected at 22 h for MJD-LL and 4 h for S017-01.
Normal disease brain tissue was from a 58 years old female who
died of non-neurologic causes, with a postmortem interval of 16 h.
For western blot analysis, lysates were collected in 50 mM Tris
(pH ¼ 7.5) supplemented with 1% triton-X-100, 1 mM EDTA and
protease inhibitors, pepstatin, leupeptin, and PMSF. The tissue was
then homogenized centrifuged at 75 600 g for 30 min, after which
soluble and pelleted samples were separated for further analysis. For
the soluble protein, the protein concentration was determined using
Protein Assay Reagent (Bio-Rad, Hercules, CA, USA), and 100 lgof protein was lysed in Laemmli buffer. To the pelleted protein, a
similar volume of 50 mM Tris was added, the pellet resuspended,
and half of the sample was boiled at 100�C for 4 min. Both pelleted
samples were then heated at 37�C for 10 min, sonicated, and protein
concentration determined. The desired amount of protein was then
lysed in Laemmli buffer.
Results
Endogenous ataxin-3 is proteolyzed during apoptosis
Caspases cleave many polyQ disease proteins, in some cases
producing cleavage products that may contribute to disease
pathogenesis (Ellerby et al. 1999a; Ellerby et al. 1999b;
Wellington et al. 2002). Evidence for cleavage of ataxin-3
has only been demonstrated in vitro with recombinant
disease protein and purified caspases (Wellington et al.
1998). We analyzed the protein sequence of ataxin-3 and
found nine potential caspase cleavage sites (Fig. 1), seven
amino-terminal to the polyQ domain and two in the
Fig. 1 Potential caspase cleavage site
aspartate residues in ataxin-3 are highly
conserved. Shown are the amino acid
sequences of human, mouse and rat ataxin-
3. The originally published C-terminus of
human ataxin-3 is also shown (Orig C;
Kawaguchi et al. 1994; Schmidt et al.
1998). Red boxes highlight nine potential
caspase cleavage sites. *Highlights aspar-
tates that were mutated to asparagines or
alanine in subsequent experiments. The
blue and green lines mark two identified
ubiquitin interaction motifs in ataxin-3.
910 S. J. Shoesmith Berke et al.
� 2004 International Society for Neurochemistry, J. Neurochem. (2004) 89, 908–918
C-terminus of the protein (Kawaguchi et al. 1994), one of
which is only present in an alternatively spliced form of
ataxin-3 (Schmidt et al. 1998). At least eight of these
potential cleavage sites are conserved across mammalian
species, suggesting that caspase-mediated proteolysis of
ataxin-3 may be evolutionarily conserved. Some of these
sites may also be conserved due to their importance in the
ubiquitin binding domain of ataxin-3 (Chai et al. 2003;
Donaldson et al. 2003).
This large number of potential caspase cleavage sites
prompted us to study the proteolysis of endogenous ataxin-3
in cells undergoing apoptotic cell death.We induced apoptosis
in cells with staurosporine (ST), which initiates apoptosis by
inhibiting a broad spectrum of cellular kinases. Since caspase
cleavage is influenced by protein–protein interactions and cell
type (Troy and Salvesen 2002), we examined five different
immortalized cell lines and primary cerebellar neurons for ST-
induced cleavage of endogenous ataxin-3. After 24 h treat-
ment with 1 lM ST, cells were lysed and the lysates analyzed
for ataxin-3 cleavage by western blot analysis. The degree of
ataxin-3 proteolysis varied between cell types (Fig. 2).
NIH3T3 mouse fibroblast and NT2 human neural cells
showed complete proteolysis of full-length ataxin-3 while
Cos-7 primate epithelial cells showed a marked reduction.
HeLa human cervical cells, differentiated PC6-3 neural cells
(a derivative of rat PC12 cells) and primary cerebellar granule
neurons showed a modest reduction of ataxin-3. In Cos-7
cells, loss of full-length ataxin-3 was accompanied by the
appearance of an approximately 28-kDa fragment that often
appeared as a doublet. This fragment was recognized by both
monoclonal (1H9) and polyclonal (anti-ataxin-3) ataxin-3
antibodies (Fig. 4a). Thus, ataxin-3 is cleaved in a cell-type
dependent manner during apoptosis in cell culture.
The appearance of an ataxin-3 fragment in Cos-7 cells
undergoing apoptosis is consistent with the toxic fragment
hypothesis, a model of disease pathogenesis in which the
production of polyQ fragments accelerates toxicity in part by
promoting misfolding or aggregation (Ellerby et al. 1999a;
Paulson 2000). To further investigate ataxin-3 proteolysis
and its potential relationship to disease, we performed
additional studies in Cos-7 cells because this cell type
exhibited both the loss of full-length ataxin-3 and the
consistent appearance of a readily observed fragment. A ST
dose–response curve showed that 1–10 lM ST was effective
at inducing ataxin-3 cleavage and the appearance of this
fragment (Fig. 3a). At much higher concentrations (100 lM),cell death occurred rapidly via necrosis rather than apoptosis
(data not shown), which may explain the relative resistance
of full-length ataxin-3 to cleavage at this concentration
(Fig. 3a). Using an optimal concentration of ST (1 lM), weperformed a time course of ataxin-3 proteolysis, assessing
cleavage and the appearance of the fragment at 0, 2, 6, 12
and 24 h (Fig. 3b). Appearance of the 28 kDa ataxin-3
fragment was optimal at 24 h. Therefore, treatment with
1 lM ST for 24 h was used in all subsequent experiments.
Ataxin-3 proteolysis is caspase-dependent, mediated
predominantly by caspase-1
Caspases are activated during ST treatment (Brophy et al.
2002; Caballero-Benitez and Moran 2003), suggesting that
they might mediate ataxin-3 cleavage. We investigated the
involvement of caspases by treating Cos-7 cells with the
broad-spectrum caspase inhibitor, Z-VAD-FMK. Z-VAD-
FMK (100 lM) inhibited both ataxin-3 cleavage and the
appearance of the 28-kDa fragment (Fig. 4a) suggesting that
one or more caspases were responsible for ST-induced
Fig. 2 Endogenous ataxin-3 is cleaved in various cell types during ST
induced apoptosis. Cells were treated with ST or vehicle control
(DMSO) for 24 h, after which live and dead cells were pooled and
lysed for western blot analysis. Shown here are western blot results
obtained with lysates from various non-neuronal (HeLa, Cos-7,
NIH3T3) and neuronal (hNT2, and differentiated PC6-3 cells) cell lines
and primary rat cerebellar granule neurons. Full-length ataxin-3 is
marked by an arrowhead (c), and the 28 kDa ataxin-3 fragment is
marked by an asterisk (*). The differing apparent molecular weights of
ataxin-3 reflect the divergent polyglutamine tract lengths in various
mammalian species.
Caspase cleavage of ataxin-3 911
� 2004 International Society for Neurochemistry, J. Neurochem. (2004) 89, 908–918
ataxin-3 cleavage. To determine which caspase(s) mediated
ataxin-3 cleavage, cells were treated with lower concentra-
tions of relatively specific inhibitors (10 lM) to caspases -1,-2, -3, -5, -6, -8, and -9. The caspase-1 inhibitor was most
effective at inhibiting ataxin-3 proteolysis and fragment
production (Fig. 4b), while inhibitors of caspases -5 and -6
led to mild inhibition, and inhibitors of caspases -2, -3, -8 and
-9 had little or no effect. As a positive control for caspase
activation in Cos-7 cells, we compared cleavage of the
known caspase target, PARP (Caballero-Benitez and Moran
2003) in the same cells (Fig. 4a and b, middle panel). For
these experiments, PARP cleavage was not completely
blocked by the specific or broad-spectrum caspase inhibitors
because of the lower concentrations (10 lM) used, whichwere necessary to reduce cross reactivity between the specific
inhibitors.
These results suggested that ST-induced ataxin-3 proteo-
lysis is mediated by caspases, with caspase-1 being a primary
mediator in Cos-7 cells. However, ST treatment can also
activate other cellular proteases including calpains and the
proteasome, both of which have been implicated in the
proteolysis of other polyQ disease proteins (Kim et al. 2001;
Gafni and Ellerby 2002; Goffredo et al. 2002). To test
whether these other proteases played a role in ataxin-3
cleavage, we treated Cos-7 cells with ST and a specific
proteasome inhibitor (lactacystin) or an inhibitor that blocks
both calpains and the proteasome (ALLN). Neither lactacys-
tin nor ALLN suppressed ataxin-3 cleavage, indicating that
the proteasome complex and calpains do not cleave ataxin-3
in this apoptotic paradigm [Fig. 4c,d; staining with
anti-ataxin-3 showed no change in fragment production
(data not shown)].
Domain mapping of the ataxin-3 fragment
To characterize the ataxin-3 fragment, we performed
antibody profiling with monoclonal and polyclonal atax-
in-3 antibodies (Fig. 5a). The fragment was recognized by
a polyclonal antibody directed against the entire protein,
anti-ataxin-3 (Paulson et al. 1997a), and more weakly by
the monoclonal antibody 1H9 (Trottier et al. 1998). A
polyclonal antibody also recognized the fragment, p146a
(V Colomer, unpublished results) directed against the
originally published C-terminus, but not by monoclonal
antibody 2B6 (data not shown) (Trottier et al. 1998).
These results suggested that a carboxy-terminal portion of
ataxin-3 comprises the fragment. Currently available
antibodies, however, did not allow us to determine
whether either the N- or C-termini of ataxin-3 were
present in the fragment. Moreover, because we are
analyzing endogenous ataxin-3 in Cos-7 cells, which
contains a very short polyQ domain, we could not use
the polyQ-specific antibody 1C2 to determine whether the
fragment contains polyQ.
To further define the cleavage fragment, we transfected
Cos-7 cells with normal (28 glutamines, Q28) or mutant
(Q68 and Q84) ataxin-3 constructs doubly tagged at the
N-terminus with a HA epitope and at the C-terminus
with a myc-epitope (designated HA-ataxin-3(Q28)-myc,
HA-ataxin-3(Q68)-myc, and HA-ataxin-3(Q84)-myc;
Fig. 5(b). As shown in Fig. 5(c), the proteolytic ataxin-3
fragment contained the polyQ region as evidenced both by
an incrementally larger ataxin-3 fragment corresponding to
increasing polyQ lengths, and by its recognition with the
monoclonal antibody 1C2 (Fig. 5c third panel). The
ataxin-3 fragment was also recognized by anti-myc
(Fig. 5c second panel) but not anti-HA antibody (data
not shown), indicating that the fragment contains the
C-terminus of ataxin-3. These results were further con-
firmed in separate experiments probing first with either
1C2 or anti-myc.
Interestingly, in cells overexpressing normal or mutant
ataxin-3, we also noted that significant fragment production
(a)
(b)
Fig. 3 ST induced ataxin-3 cleavage occurs in a dose- and time-
dependent manner. (a) The proteolysis of ataxin-3 observed with
increasing concentrations of ST treatment for 24 h in Cos-7 cells. The
relative resistance of ataxin-3 to cleavage at 100 lM ST is likely due to
the rapid necrotic death that is known to occur at that concentration.
(b) A time course of ataxin-3 cleavage with 1 lM ST treatment. Optimal
concentrations chosen for subsequent experiments were 1 lM ST
treatment for 24 h. As a positive control for caspase activity the blots
were probed with anti-PARP (middle panels), and for loading control
with anti-GAPDH (lower panels).
912 S. J. Shoesmith Berke et al.
� 2004 International Society for Neurochemistry, J. Neurochem. (2004) 89, 908–918
occurred even without ST treatment. This too could be
blocked with caspase inhibitors (Fig. 5c). While RT-PCR
analysis showed that the originally published ataxin-3 C-
terminus (Kawaguchi et al. 1994) was the only ataxin-3
species expressed in Cos-7 cells, overexpression of ataxin-3
containing the alternative C-terminus (Schmidt et al. 1998)
also produced significant fragment production in the absence
of ST treatment (data not shown). This suggests that
overexpressed ataxin-3 either activates caspases or is more
accessible to cleavage by basal cellular proteases.
Intriguingly, when cells were transfected with expanded
ataxin-3 constructs and subsequently treated with ST, an
increase in ataxin-3 aggregation was observed in the stacking
region of the gel (Fig. 6). This increased aggregation was
suppressed by treatment with Z-VAD-FMK (100 lM), sug-gesting it is a caspase-mediated event. Indeed, treatment with
Z-VAD-FMK reduced ataxin-3 aggregation even in the
absence of the apoptotic stressor ST, suggesting that caspase
cleavage of ataxin-3 contributes to the formation of ataxin-3
aggregates in this cell model (Fig. 6; compare lanes 5 with 8,
9 with 12).
Site-directed mutagenesis of ataxin-3 caspase cleavage
sites
Caspase cleavage sites are four amino acid motifs marked by
an essential aspartate residue, termed the P1 aspartate, at
which cleavage occurs (Earnshaw et al. 1999). Mutation of
the P1 aspartate to an asparagine prevents caspase recogni-
tion and subsequent cleavage (Ellerby et al. 1999a; Ellerby
et al. 1999b; Wellington et al. 2000). The size of the ataxin-3
fragment, together with analysis of likely caspase sites in
ataxin-3, suggested that two predicted cleavage sites corres-
ponding to P1 aspartates at amino acid positions 225 and 228
of ataxin-3 were strong candidate targets (sites LDED and
DEED in Fig. 5a). To test their role in proteolysis, we created
ataxin-3 constructs in which one or both of these aspartates
were mutated to asparagine residues. We transfected Cos-7
cells with wild-type ataxin-3(Q35) or mutated ataxin-3
constructs (ataxin-3(Q35) D225N or D225/228N) and one
day later treated cells with 1 lM ST for 24 h. As shown in
Fig. 7(a), ataxin-3(Q35) and ataxin-3(Q35) D225/228N were
both cleaved during ST-induced apoptosis, suggesting that
the sites at 225 and 228 either do not represent major
cleavage sites or that their elimination promotes cleavage at
nearby secondary sites.
For some polyQ disease proteins, elimination of multiple
P1 aspartate residues is required to block caspase cleavage
(Wellington et al. 1998; 2000). Reasoning that similar caspase
site redundancy may exist in ataxin-3, we created an ataxin-3
construct in which all nine potential amino-terminal aspartate
residues (145, 171, 208, 217, 225, 228, 241, 244 and 248)
were mutated to asparagines or alanine (aspartate 217), termed
(a) (b)
(c) (d)
Fig. 4 Ataxin-3 cleavage is caspase-
dependent and primarily mediated by ca-
spase-1, not the proteasome or calpains.
(a) ST-induced proteolysis of full-length
ataxin-3 (c) and the appearance of the
28 kDa fragment (c) is blocked by treat-
ment with 100 lM Z-VAD-FMK (CI), a broad
spectrum caspase inhibitor. (b) Cells trea-
ted with ST and a panel of inhibitors to
specific caspases (10 lM) or Z-VAD-FMK
(10 lM). In panels c and d, ST-induced
ataxin-3 proteolysis was not blocked by
treatment with the proteasome and calpain
inhibitor, ALLN (100 lM), or the specific
proteasome inhibitor, lactacystin (10 lM).
The fragment is not present in panels c and
d because the 1H9 antibody only weakly
recognizes the fragment. Blots were also
probed with anti-PARP (panels second from
the bottom) and anti-GAPDH (lower pan-
els).
Caspase cleavage of ataxin-3 913
� 2004 International Society for Neurochemistry, J. Neurochem. (2004) 89, 908–918
ataxin-3(Q35)-9mut. We transfected ataxin-3(Q28) and atax-
in-3(Q35)-9mut into Cos-7 cells and 24 h later treated with ST.
Ataxin-3(Q35)-9mut was resistant to caspase-mediated clea-
vage during ST treatment (Fig. 7b). To further define which
region of ataxin-3 was primarily involved in cleavage, we
created ataxin-3 constructs with either the first 6 aspartates
(145, 171, 208, 217, 225, 228) or the last 3 aspartates (241,
244, 248) mutated to asparagines or alanine (aspartate 217).
These constructs were expressed in Cos-7 cells, which were
(a)
(b)
(c)
Fig. 5 Ataxin-3 antibody recognition sequences and generated con-
structs. (a) A schematic diagram of ataxin-3 highlighting potential
caspase cleavage sites (A–G) as well as the epitope locations for
ataxin-3 monoclonal antibodies 2B6 and 1H9 and polyclonal antibody
146 pa. The ST-induced ataxin-3 fragment is recognized by 146 pa
and weakly by 1H9, but not by 2B6. (Not drawn to scale; slash marks
indicate that N-terminus is longer than depicted in diagram). (b)
pcDNA-HA-ataxin-3-myc constructs used for the transient transfec-
tions in panel C. Panel C, Cos-7 cells were transfected with the
pcDNA3-HA-ataxin-3-myc constructs and 24 h later treated with ST
(1 lM) and/or Z-VAD-FMK (100 lM). The ataxin-3 fragment (c)
increases in size with corresponding repeat lengths. The fragments
are immunostained by anti-myc, which recognizes the extreme C-
terminus of the proteins (second panel), and 1C2 antibody, which
recognizes expanded polyglutamine tracts (third panel). Anti-PARP
and anti-GAPDH were performed as controls (bottom two panels).
Fig. 6 Aggregation of mutant ataxin-3 is reduced by broad spectrum
caspase inhibitors. Cos-7 cells were transiently transfected with the
normal (Q28) or mutant (Q68 and Q84) ataxin-3 constructs described
in Fig. 5(b). High molecular weight SDS-insoluble ataxin-3 aggregate
running in the stacking gel (b) are increased during 1 lM ST treat-
ment, and suppressed by 100 lM Z-VAD-FMK. The small degree of
mutant ataxin-3 aggregation occurring even in the absence of ST is
also reduced by caspase inhibitor treatment.
(a)
(b)
Fig. 7 Site-directed mutagenesis partially blocked proteolysis of full-
length ataxin-3. Panel A shows results from Cos-7 cells transiently
transfected with either normal ataxin-3(Q35), or ataxin-3 constructs
with D225N or D225/228N. Neither mutation blocked ataxin-3 pro-
teolysis during 1 lM ST treatment. Full-length ataxin-3 (c); ataxin-3
fragment (c). Z-VAD-FMK treatment, 100 lM. In Panel B cells were
transfected with either normal ataxin-3 or ataxin-3 with 3, 6 or 9
aspartate to asparagine or alanine mutations. Ataxin-3(Q28)-3mut and
ataxin-3(Q35)-9mut were relatively resistant to proteolysis, whereas
ataxin-3(Q35)-6mut was readily proteolyzed.
914 S. J. Shoesmith Berke et al.
� 2004 International Society for Neurochemistry, J. Neurochem. (2004) 89, 908–918
then treated with ST 24 h later. As shown in Fig. 7(b), ataxin-
3(Q28)-3mut was relatively resistant to cleavage, similar to
that of ataxin-3(Q35)-9mut, whereas ataxin-3(Q35)-6mut was
efficiently proteolyzed. These results suggest that a preferred
cleavage site for ST-induced ataxin-3 proteolysis resides
within the cluster of potential caspase sites corresponding to
aspartates 241, 244, and 248.
Proteolytic fragments are not observed in SCA3/MJD
human disease tissue
The above results establish that ataxin-3 is a target for
caspase-mediated cleavage during apoptosis in cultured cells.
To characterize the role of ataxin-3 proteolysis in SCA3/MJD
disease, we looked for evidence of ataxin-3 cleavage in brain
tissue from two SCA3/MJD patients. The first individual
died of end stage SCA3/MJD (Paulson et al. 1997b) while
the second, S01-017, was in advanced-stage SCA3/MJD yet
died of other causes. We were unable to detect ataxin-3
fragments, in either soluble or pelleted fractions, in any of the
brain regions tested. These included the globus pallidus,
midbrain, and pons, three regions known to be susceptible to
degeneration in SCA3/MJD (Fig. 8; data not shown).
Because the observed ataxin-3 fragment in cells contains
the polyQ domain, it is possible that a similar fragment in
human brain would be incorporated into insoluble aggregates
and therefore would not be detectable by western blot
analysis.
Discussion
Here we have presented evidence that the polyQ disease
protein, ataxin-3, is a substrate for caspase-mediated
proteolysis in tissue culture. Using ST to induce apoptosis
in cultured cells, we established that: (1) both endogenous
ataxin-3 and exogenous, overexpressed ataxin-3 are proteo-
lytic targets during apoptosis, resulting in production of a
polyQ-containing fragment of ataxin-3; (2) cleavage is
caspase-mediated, primarily by caspase-1 in Cos-7 cells,
with no evidence for calpain or proteasomal involvement;
(3) apoptotic cleavage of ataxin-3 is abolished by mutating
all nine potential caspase cleavage sites amino terminal to the
polyQ domain, and significantly inhibited by mutating a
cluster of three sites within a ubiquitin interaction motif near
the polyQ domain; (4) caspase-mediated proteolysis of
expanded ataxin-3 leads to increased formation of SDS-
insoluble aggregates; and (5) this aggregation can be
suppressed by treatment with caspase inhibitors. Taken
together, our findings are consistent with a disease model
in which caspase-mediated proteolysis of ataxin-3 contri-
butes to SCA3/MJD pathogenesis.
Importantly, much of our analysis was performed on
endogenous ataxin-3. Thus, our studies establish that, at
normal expression levels in an intact cellular milieu, ataxin-3
is readily proteolyzed by caspases. The only prior evidence
that ataxin-3 is a proteolytic target came from an in vitro
analysis of recombinant proteins, in which several other
polyQ proteins were more robust targets than ataxin-3
(Wellington et al. 1998). We found that ataxin-3 is suscep-
tible to cleavage both in neurons and in various non-neuronal
cell lines. While our results agree with an earlier in vitro
study showing that purified ataxin-3 protein is cleaved by
recombinant caspase-1 (Wellington et al. 1998), we were
unable to detect caspase-3 mediated cleavage in cell culture.
Additionally, our observed fragment in Cos-7 cells was larger
(28 kDa) than the proteolytic fragment noted in vitro
[14 kDa; (Wellington et al. 1998)], which may reflect altered
proteolysis due to the cellular environment. In the cell,
endogenous ataxin-3 presumably engages in a variety of
protein–protein interactions (Wang et al. 2000; Chai et al.
2001; Doss-Pepe et al. 2003) that may modulate its suscep-
tibility to caspases. Indeed, caspase-mediated events are
known to be influenced both by cell type and by protein–
protein interactions (Troy and Salvesen 2002). Further
supporting the importance of cellular context is the fact that
cleavage of ataxin-3 and the appearance of a detectable
cleavage product were not identical across the various cell
lines we tested.
Our studies with ataxin-3 constructs mutated at potential
caspase cleavage sites suggest redundancy in the sites at
which cleavage may occur, as has been described for
huntingtin (Wellington et al. 2000). Of the mutated con-
structs we tested, ataxin-3 mutated at three closely spaced
sites proved to be most resistant to cleavage in Cos-7 cells,
indicating that a favored cleavage event occurs at aspartates
241, 244 and/or 248. This is intriguing because these three
aspartate residues reside within, or next to, a recently
Fig. 8 Proteolytic fragments are not observed in SCA3/MJD human
disease brain tissue. Tissue lysates from different brain regions of a
patient with SCA3/MJD (patient S01-017). Normal (.) and mutant (c)
ataxin-3 protein is present, however, no fragment is observed. Brain
regions are as follows, frontal cortex (FC), caudate (Ca), occipital
cortex (OC), parietal cortex (PC), cingulated gyrus (CG), putamen (P),
substantia nigra (SN), cerebellar cortex (CC), and vermis (V). The
band appearing at approximately 33 kDa is non-specific staining from
the polyclonal ataxin-3 antibody.
Caspase cleavage of ataxin-3 915
� 2004 International Society for Neurochemistry, J. Neurochem. (2004) 89, 908–918
discovered ubiquitin interaction motif (Donaldson et al.
2003). Ataxin-3 is now known to be an ubiquitin binding
protein, and likely participates in ubiquitin dependent
pathways (Burnett et al. 2003; Chai et al. 2003; Donaldson
et al. 2003; Doss-Pepe et al. 2003). Cleavage within the
ubiquitin binding domain would eradicate this integral
biochemical property of ataxin-3, thereby likely altering
ubiquitin-dependent events in the cell. During apoptosis,
proteins serving vital cellular functions are often targets for
caspase-mediated cleavage, which in turn accelerates the
process of cellular death. As more is learned about the
cellular functions of ataxin-3 and the precise role played by
ubiquitin binding, the effect of ataxin-3 cleavage on the sick
or dying cell may become clear. Moreover, our observation
that overexpressed ataxin-3 underwent proteolysis even in
the absence of ST treatment suggests ataxin-3 levels may be
closely regulated within the cell. Future studies may define a
relationship between ubiquitin binding of ataxin-3 and its
regulation by proteolysis.
When cells were treated with an apoptotic stressor,
expanded ataxin-3 showed increased aggregation. Import-
antly, in our studies aggregation of expanded ataxin-3 could
be suppressed by treatment with a broad-spectrum caspase
inhibitor, suggesting that caspase-mediated cleavage of
mutant ataxin-3 promotes its aggregation in cells. To
determine whether this proteolysis directly contributes to
SCA3/MJD pathogenesis, further studies will be required in
animal models employing ataxin-3 constructs engineered to
lack the relevant caspase cleavage sites.
Although we were unable to detect ataxin-3 fragments
in disease brain, this does not rule out the involvement of
proteolytic cleavage in SCA3/MJD pathogenesis. As
suggested by the current study and previous reports (Ikeda
et al. 1996; Mangiarini et al. 1996; Paulson et al. 1997b;
Wellington et al. 2000; Kim et al. 2001; Sanchez et al.
2003), polyQ-containing ataxin-3 fragments are prone to
aggregation. In the chronic disease state, such fragments
may be highly insoluble and hence undetectable by
standard SDS-gel immunoblot analysis. In diseased brain,
moreover, soluble ataxin-3 fragments could be degraded
more rapidly and thus be present at low steady state levels.
Our current antibodies and detection methods may not be
sensitive enough to detect low amounts of an ataxin-3
fragment in disease tissue, especially if it is insoluble.
Moreover, in HD, protein extraction utilizing formic acid
was required to detect some polyQ fragments (Lunkes
et al. 2002), suggesting more stringent conditions may be
necessary to detect ataxin-3 fragments in SCA3/MJD
human disease tissue. The presence of ataxin-3 fragments
in an SCA3/MJD mouse model notwithstanding (Goti and
Colomer 2003), it is also possible that ataxin-3 is not
cleaved during SCA3/MJD disease pathogenesis, in which
case no ataxin-3 fragments would be detected in human
disease tissue.
Although a detectable ataxin-3 fragment was not observed
in most cell types, full-length ataxin-3 was readily proteo-
lyzed during apoptosis in all cell types tested. This is
intriguing in light of the fact that ataxin-3 is now known to be
a ubiquitin-binding protein with ubiquitin –protease activity
(Burnett et al. 2003; Chai et al. 2003; Donaldson et al.
2003; Doss-Pepe et al. 2003). Perturbations of the ubiquitin-
proteasome pathway have been shown to accelerate polyQ
disease pathogenesis in cellular and animal models (Berke and
Paulson 2003). Thus, regardless of whether detectable ataxin-
3 fragments are produced, ataxin-3 proteolysis under
conditions of cellular or apoptotic stress may be deleterious
to the cell.
In our cellular model, ataxin-3 cleavage was prevented by
broad-spectrum caspase inhibitors and reduced by specific
caspase inhibitors. Caspase inhibitors have been proposed as
a potential therapy for polyQ diseases and some studies in
mouse models support their potential utility in slowing
disease progression (Ona et al. 1999). However, since
apoptosis is necessary during development and may contrib-
ute to normal brain functioning (Troy and Salvesen 2002),
treatment with caspase inhibitors would need to be tempor-
ally and spatially regulated to be feasible. Our current results
show that site-directed mutagenesis of specific aspartate
residues could only partially block ataxin-3 cleavage. It will
be important to extend this analysis to animal models of
SCA3/MJD to assess whether specific caspase inhibitors can
block ataxin-3 processing and toxicity in disease neurons.
The development of therapeutic strategies for SCA3/MJD
and other polyQ diseases will require further characterization
of the role of caspases in pathogenesis. Caspase activity
profiles should be conducted in SCA3/MJD human disease
tissue and animal models to determine whether caspase
activation plays a primary or secondary role in disease
progression. Although the studies described here in Cos-7
cells suggest that particular attention should be paid to
caspase-1 activity, molecular events in disease neurons may
well differ. Finally, dominant negative caspase constructs,
beginning with caspase-1, could be expressed in various cell
and animal models of SCA3/MJD to determine whether they
can suppress aspects of the disease phenotype, as has been
shown in models of HD (Ona et al. 1999).
Acknowledgements
We thank V. Colomer (Johns Hopkins University) for communica-
ting results prior to publication and for the polyclonal ataxin-3
antibody 146 pa. SJSB is supported by the National Institutes of
Health grant NS043076 and National Institutes of Health/National
Institute on Aging Grant T32 AG 00214, Interdisciplinary Research
Training Program on Aging, University of Iowa. HLP and FFS are
supported by National Institutes of Health grant NS38712. LME is
supported by National Institutes of Health grant NS40251A and the
Muscular Dystrophy Association. We are also grateful to the
916 S. J. Shoesmith Berke et al.
� 2004 International Society for Neurochemistry, J. Neurochem. (2004) 89, 908–918
Hereditary Disease Foundation and to the Huntington’s Disease
Society of America for their support of these studies. Thanks for
technical support from Jessica Young and Stephanie Propp.
References
Berke S. J. and Paulson H. L. (2003) Protein aggregation and the
ubiquitin proteasome pathway: gaining the upper hand on neuro-
degeneration. Curr. Opin. Genet Dev 13, 253–261.
Brophy V. A., Tavare J. M. and Rivett A. J. (2002) Treatment of COS-7
cells with proteasome inhibitors or gamma-interferon reduces the
increase in caspase 3 activity associated with staurosporine-in-
duced apoptosis. Arch. Biochem. Biophys. 397, 199–205.
Burnett B., Li F. and Pittman R. N. (2003) The polyglutamine neuro-
degenerative protein ataxin-3 binds polyubiquitylated proteins and
has ubiquitin protease activity. Hum. Mol Genet 12, 3195–3205.
Caballero-Benitez A. and Moran J. (2003) Caspase activation pathways
induced by staurosporine and low potassium: role of caspase-2.
J. Neurosci. Res. 71, 383–396.
Chai Y., Berke S. S., Cohen R. E. and Paulson H. L. (2003) Poly-ubiquitin
binding by the polyglutamine disease protein ataxin-3 links its
normal function to protein surveillance pathways. J Biol Chem. 279,
3605–3611.
Chai Y., Koppenhafer S. L., Bonini N. M. and Paulson H. L. (1999a)
Analysis of the role of heat shock protein (Hsp) molecular chap-
erones in polyglutamine disease. J. Neurosci. 19, 10338–10347.
Chai Y., Koppenhafer S. L., Shoesmith S. J., PereZ. M. K. and Paulson
H. L. (1999b) Evidence for proteasome involvement in polyglu-
tamine disease: localization to nuclear inclusions in SCA3/MJD
and suppression of polyglutamine aggregation in vitro. Hum Mol
Genet 8, 673–682.
Chai Y., Wu L., Griffin J. D. and Paulson H. L. (2001) The role of
protein composition in specifying nuclear inclusion formation in
polyglutamine disease. J. Biol. Chem. 276, 44889–44897.
Donaldson K. M., Li W., Ching K. A., Batalov S., Tsai C. C. and
Joazeiro C. A. (2003) Ubiquitin-mediated sequestration of normal
cellular proteins into polyglutamine aggregates. Proc. Natl Acad.
Sci. USA 100, 8892–8897.
Doss-Pepe E. W., Stenroos E. S., Johnson W. G. and Madura K. (2003)
Ataxin)3 interactions with rad23 and valosin-containing protein
and its associations with ubiquitin chains and the proteasome are
consistent with a role in ubiquitin-mediated proteolysis. Mol Cell
Biol. 23, 6469–6483.
Earnshaw W. C., Martins L. M. and Kaufmann S. H. (1999) Mammalian
caspases: structure, activation, substrates, and functions during
apoptosis. Annu. Rev. Biochem. 68, 383–424.
Ellerby L. M., Andrusiak R. L., Wellington C. L., Hackam A. S., Propp
S. S., Wood J. D., Sharp A. H., Margolis R. L., Ross C. A.,
Salvesen G. S., Hayden M. R. and Bredesen D. E. (1999a) Clea-
vage of atrophin-1 at caspase site aspartic acid 109 modulates
cytotoxicity. J. Biol. Chem. 274, 8730–8736.
Ellerby L. M., Hackam A. S., Propp S. S., Ellerby H. M., Rabizadeh S.,
Cashman N. R., Trifiro M. A., Pinsky L., Wellington C. L.,
Salvesen G. S., Hayden M. R. and Bredesen D. E. (1999b) Ken-
nedy’s disease: caspase cleavage of the androgen receptor is a
crucial event in cytotoxicity. J. Neurochem. 72, 185–195.
Evert B. O., Wullner U., SchulZ. J. B., Weller M., Groscurth P., Trottier
Y., Brice A. and Klockgether T. (1999) High level expression of
expanded full-length ataxin-3 in vitro causes cell death and for-
mation of intranuclear inclusions in neuronal cells. Hum Mol Genet
8, 1169–1176.
Gafni J. and Ellerby L. M. (2002) Calpain activation in Huntington’s
disease. J. Neurosci. 22, 4842–4849.
Garden G. A., Libby R. T., Fu Y. H., Kinoshita Y., Huang J., Possin D. E.,
Smith A. C., MartineZ. R. A., Fine G. C., Grote S. K., Ware C. B.,
EinumD.D.,MorrisonR. S., Ptacek L. J., Sopher B. L. and La Spada
A. R. (2002) Polyglutamine-expanded ataxin-7 promotes non-cell-
autonomous purkinje cell degeneration and displays proteolytic
cleavage in ataxic transgenic mice. J. Neurosci. 22, 4897–4905.
Goffredo D., Rigamonti D., Tartari M., De Micheli A., Verderio C.,
Matteoli M., Zuccato C. and Cattaneo E. (2002) Calcium-
dependent cleavage of endogenous wild-type huntingtin in primary
cortical neurons. J. Biol. Chem. 277, 39594–39598.
Goti D. and Colomer V. (2003) A mutant ataxin-3 cleavage fragment
was detected in brain of MJD patients, characterized in MJD
transgenic mouse brain, and toxic to transfected cells. Society for
Neuroscience, Abstract 308.1.
Ikeda H., Yamaguchi M., Sugai S., Aze Y., Narumiya S. and Kakizuka A.
(1996) Expanded polyglutamine in the Machado-Joseph disease
protein induces cell death in vitro and in vivo. Nat Genet 13, 196–
202.
Kawaguchi Y., Okamoto T., Taniwaki M., Aizawa M., Inoue M., Kat-
ayama S., Kawakami H., Nakamura S., Nishimura M., Akiguchi I.
and et al. (1994) CAG expansions in a novel gene for Machado-
Joseph disease at chromosome 14q32.1. Nat Genet 8, 221–228.
Kim M., Lee H. S., LaForet G., McIntyre C., Martin E. J., Chang P., Kim
T. W., Williams M., Reddy P. H., Tagle D., Boyce F. M., Won L.,
Heller A., Aronin N. and DiFiglia M. (1999) Mutant huntingtin
expression in clonal striatal cells: dissociation of inclusion for-
mation and neuronal survival by caspase inhibition. J. Neurosci.
19, 964–973.
Kim Y. J., Yi Y., Sapp E., Wang Y., Cuiffo B., Kegel K. B., Qin Z. H.,
Aronin N. and DiFiglia M. (2001) Caspase 3-cleaved N-terminal
fragments of wild-type and mutant huntingtin are present in normal
and Huntington’s disease brains, associate with membranes, and
undergo calpain-dependent proteolysis. Proc. Natl Acad. Sci. USA
98, 12784–12789.
LaFevre-Bernt M. A. and Ellerby L. M. (2003) Kennedy’s Disease:
phosphorylation of the polyglutamine-expanded form of androgen
receptor regulates its clevage by caspase-3 and enhances cell death.
J. Biol. Chem. 278, 34918–34924.
Li S. H., Lam S., Cheng A. L. and Li X. J. (2000) Intranuclear huntingtin
increases the expression of caspase-1 and induces apoptosis. Hum
Mol Genet 9, 2859–2867.
Lunkes A., Lindenberg K. S., Ben-Haiem L., Weber C., Devys D.,
Landwehrmeyer G. B., Mandel J. L. and Trottier Y. (2002) Pro-
teases acting on mutant huntingtin generate cleaved products that
differentially build up cytoplasmic and nuclear inclusions.Mol Cell
10, 259–269.
Mangiarini L., Sathasivam K., Seller M., Cozens B., Harper A.,
Hetherington C., Lawton M., Trottier Y., Lehrach H., Davies S. W.
and Bates G. P. (1996) Exon 1 of the HD gene with an expanded
CAG repeat is sufficient to cause a progressive neurological phe-
notype in transgenic mice. Cell 87, 493–506.
Morley J. F., Brignull H. R., Weyers J. J. and Morimoto R. I. (2002)
The threshold for polyglutamine-expansion protein aggregation
and cellular toxicity is dynamic and influenced by aging in
Caenorhabditis elegans. Proc. Natl Acad. Sci. USA 99, 10417–
10422.
Ona V. O., Li M., Vonsattel J. P., Andrews L. J., Khan S. Q., Chung W.
M., Frey A. S., Menon A. S., Li X. J., Stieg P. E., Yuan J., Penney
J. B., Young A. B., Cha J. H. and Friedlander R. M. (1999)
Inhibition of caspase-1 slows disease progression in a mouse model
of Huntington’s disease. Nature 399, 263–267.
Paulson H. L. (2000) Toward an understanding of polyglutamine neu-
rodegeneration. Brain Pathol 10, 293–299.
Caspase cleavage of ataxin-3 917
� 2004 International Society for Neurochemistry, J. Neurochem. (2004) 89, 908–918
Paulson H. L., Das S. S., Crino P. B., PereZ. M. K., Patel S. C.,
Gotsdiner D., Fischbeck K. H. and Pittman R. N. (1997a) Mach-
ado-Joseph disease gene product is a cytoplasmic protein widely
expressed in brain. Ann. Neurol. 41, 453–462.
Paulson H. L., PereZ. M. K., Trottier Y., Trojanowski J. Q., Subra-
mony S. H., Das S. S., Vig P., Mandel J. L., Fischbeck K. H. and
Pittman R. N. (1997b) Intranuclear inclusions of expanded
polyglutamine protein in spinocerebellar ataxia type 3. Neuron
19, 333–344.
Sanchez I., Mahlke C. and Yuan J. (2003) Pivotal role of oligomerization
in expanded polyglutamine neurodegenerative disorders. Nature
421, 373–379.
Sanchez I., Xu C. J., Juo P., Kakizaka A., Blenis J. and Yuan J. (1999)
Caspase-8 is required for cell death induced by expanded poly-
glutamine repeats. Neuron 22, 623–633.
Satyal S. H., Schmidt E., Kitagawa K., Sondheimer N., Lindquist S.,
Kramer J. M. and Morimoto R. I. (2000) Polyglutamine aggregates
alter protein folding homeostasis in Caenorhabditis elegans. Proc.
Natl Acad. Sci. USA 97, 5750–5755.
Schmidt T., Landwehrmeyer G. B., Schmitt I., Trottier Y., Auburger G.,
Laccone F., Klockgether T., Volpel M., Epplen J. T., Schols L. and
Riess O. (1998) An isoform of ataxin-3 accumulates in the nucleus
of neuronal cells in affected brain regions of SCA3 patients. Brain
Pathol 8, 669–679.
Trottier Y., Cancel G., An-Gourfinkel I., LutZ. Y., Weber C., Brice A.,
Hirsch E. and Mandel J. L. (1998) Heterogeneous intracellular
localization and expression of ataxin-3. Neurobiol. Dis 5, 335–
347.
Troy C. M. and Salvesen G. S. (2002) Caspases on the brain. J. Neu-
rosci. Res. 69, 145–150.
Wang G., Sawai N., Kotliarova S., Kanazawa I. and Nukina N. (2000)
Ataxin-3, the MJD1 gene product, interacts with the two human
homologs of yeast DNA repair protein RAD23, HHR23A and
HHR23B. Hum Mol Genet 9, 1795–1803.
Wellington C. L., Ellerby L. M., Gutekunst C. A., Rogers D., Warby S.,
Graham R. K., Loubser O., van Raamsdonk J., Singaraja R., Yang
Y. Z., Gafni J., Bredesen D., Hersch S. M., Leavitt B. R., Roy S.,
Nicholson D. W. and Hayden M. R. (2002) Caspase cleavage of
mutant huntingtin precedes neurodegeneration in Huntington’s
disease. J. Neurosci. 22, 7862–7872.
Wellington C. L., Ellerby L. M., Hackam A. S., Margolis R. L., Trifiro
M. A., Singaraja R., McCutcheon K., Salvesen G. S., Propp S. S.,
Bromm M., Rowland K. J., Zhang T., Rasper D., Roy S., Thorn-
berry N., Pinsky L., Kakizuka A., Ross C. A., Nicholson D. W.,
Bredesen D. E. and Hayden M. R. (1998) Caspase cleavage of
gene products associated with triplet expansion disorders generates
truncated fragments containing the polyglutamine tract. J. Biol.
Chem. 273, 9158–9167.
Wellington C. L., Singaraja R., Ellerby L., Savill J., Roy S., Leavitt
B., Cattaneo E., Hackam A., Sharp A., Thornberry N., Nichol-
son D. W., Bredesen D. E. and Hayden M. R. (2000) Inhibiting
caspase cleavage of huntingtin reduces toxicity and aggregate
formation in neuronal and nonneuronal cells. J. Biol. Chem. 275,
19831–19838.
Zoghbi H. Y. and Orr H. T. (2000) Glutamine repeats and neurodegen-
eration. Annu. Rev. Neurosci. 23, 217–247.
918 S. J. Shoesmith Berke et al.
� 2004 International Society for Neurochemistry, J. Neurochem. (2004) 89, 908–918