Human Molecular Genetics, 2003, Vol. 12, No. 7
DOI: 10.1093/hmg/ddg074
749–757
Aggresomes protect cells by enhancing the
degradation of toxic polyglutamine-containing
protein
J. Paul Taylor1,*, Fumiaki Tanaka1, Jon Robitschek1, C. Miguel Sandoval1, Addis Taye1,
Silva Markovic-Plese2 and Kenneth H. Fischbeck1
1
Neurogenetics Branch, National Institute of Neurological Disorders and Stroke, National Institutes of Health,
10 Center Drive, Building 10, Room 3B-14, Bethesda, MD 20892-1250, USA and 2Neuroimmunology Branch,
National Institute of Neurological Disorders and Stroke, National Institutes of Health, Bethesda, MD, USA
Received November 21, 2002; Revised and Accepted January 21, 2003
Expression of misfolded protein in cultured cells frequently leads to the formation of juxtanuclear inclusions
that have been termed ‘aggresomes’. Aggresome formation is an active cellular response that involves
trafficking of the offending protein along microtubules, reorganization of intermediate filaments and
recruitment of components of the ubiquitin proteasome system. Whether aggresomes are benevolent or
noxious is unknown, but they are of particular interest because of the appearance of similar inclusions in
protein deposition diseases. Here we present evidence that aggresomes serve a cytoprotective function and
are associated with accelerated turnover of mutant proteins. We show that mutant androgen receptor (AR),
the protein responsible for X-linked spinobulbar muscular atrophy, forms insoluble aggregates and is toxic
to cultured cells. Mutant AR was also found to form aggresomes in a process distinct from aggregation.
Molecular and pharmacological interventions were used to disrupt aggresome formation, revealing their
cytoprotective function. Aggresome-forming proteins were found to have an accelerated rate of turnover, and
this turnover was slowed by inhibition of aggresome formation. Finally, we show that aggresome-forming
proteins become membrane-bound and associate with lysosomal structures. Together, these findings
suggest that aggresomes are cytoprotective, serving as cytoplasmic recruitment centers to facilitate
degradation of toxic proteins.
INTRODUCTION
Correct folding requires proteins to assume one particular
structure from a constellation of possible but incorrect
conformations. The failure of polypeptides to adopt their
proper structure is a major threat to cell function and viability.
Consequently, elaborate systems have evolved to protect cells
from the deleterious effects of misfolded proteins. Molecular
chaperones associate with nascent polypeptides as they emerge
from the ribosome, promoting correct folding and preventing
untoward interactions. A large fraction of newly translated
proteins nonetheless fails to fold correctly, generating a
substantial burden of defective polypeptide (1). These proteins
are degraded primarily by the ubiquitin–proteasome system
(UPS), a multicomponent system that identifies and degrades
unwanted proteins (2). Such protective mechanisms face a
particularly difficult challenge when disease-causing mutations
lead to excess production of misfolded protein. This challenge
may be met by efficient removal of the offending product by
degradation, as is the case for mutant forms of the cystic
fibrosis transmembrane conductance regulator (CFTR) (3).
Alternatively, misfolded protein may aggregate and accumulate
in and around cells. The deposition of misfolded protein in
intraneuronal inclusions has emerged as a common feature
underlying many neurodegenerative diseases, suggesting that
these diseases may share common mechanisms relating to the
failure of neurons to cope with excess misfolded protein (4).
Expression of mutant proteins in cultured cells frequently
leads to the formation of juxtanuclear inclusion bodies that have
been termed ‘aggresomes’ (5,6). These structures develop at a
specific location within the cell, the microtubule organizing
center (MTOC). In addition to the misfolded protein itself,
which is typically ubiquitinated, components of the ubiquitin–
proteasome pathway are enriched in aggresomes (7).
*To whom correspondence should be addressed. Tel: þ1 3014359288; Fax: þ1 3014803365; Email: [email protected]
750
Human Molecular Genetics, 2003, Vol. 12, No. 7
Aggresome formation is dependent on dynein-mediated, minus
end-directed movement along microtubules, and inhibition of
this microtubule function blocks aggresome formation (6,8).
The development of aggresomes is intimately related to the cell’s
capacity to handle misfolded protein. Inhibition of proteasome
function provokes aggresome formation, suggesting that these
structures develop when a threshold of misfolded protein is
exceeded (6). Aggresomes are of particular interest because they
resemble the intracellular inclusions found in neurodegenerative
diseases such as Parkinson’s disease, amyotrophic lateral
sclerosis and the polyglutamine diseases. It is not known
whether aggresomes are benevolent or noxious, and this
parallels the debate as to whether neuronal inclusions in
neurodegenerative diseases are helpful or harmful. An elegant
reporter system designed to monitor proteasome function
provided evidence that aggresome formation is accompanied
by a defect in proteasome-mediated degradation (9). However, it
is unclear whether aggresome formation in this setting is
causative or reactive, or whether the proteasomal defect would
be lesser or greater in the absence of aggresome formation.
Spinobulbar muscular atrophy (SBMA) is an X-linked
neurodegenerative disorder characterized by progressive loss
of motor neurons in the brainstem and spinal cord. The disease
is caused by an expanded polyglutamine tract in the amino
terminal domain of the androgen receptor (10). Polyglutaminecontaining N-terminal fragments of the androgen receptor
accumulate and form inclusions in degenerating motor
neurons. SBMA is one of nine known ‘polyglutamine
diseases’, and as such is a typical protein deposition disease.
We used a model of mutant androgen receptor toxicity to study
the fundamental nature of polyglutamine inclusions and to
evaluate their role in polyglutamine-mediated toxicity. We
demonstrate that (i) inclusions of mutant androgen receptor
exhibit key features of aggresomes, (ii) aggresome/inclusion
formation is a distinct process from aggregation, (iii) inhibiting
aggresome/inclusion formation (but not aggregation) enhances
the cytotoxicity of mutant androgen receptor, (iv) aggresomeforming proteins exhibit an enhanced rate of turnover, and (v)
inclusions composed of mutant androgen receptor associate
with lysosomal structures, implicating autophagy as a possible
route of degradation.
full-length AR. Importantly, the expression of this truncated
form reproduces many aspects of the aggregation and toxicity of
SBMA (12–15). The inclusions observed here are similar to
those described for other polyglutamine containing proteins
(Fig. 1A). Within 48 h 60% of transfected cells contained
inclusions and by 96 h post-transfection many cells harboring
inclusions assumed apoptotic morphology with fragmented
nuclei. Constructs with shorter repeats had a greater lag time
before inclusions appeared and a lower percentage of transfected
cells that contained inclusions (not shown). The majority of
inclusions were cytoplasmic and in a juxtanuclear location,
while a small fraction were intranuclear (Fig. 1B). The ratio of
cytoplasmic to nuclear inclusions varied to some extent with cell
type. We saw no difference in the rate, frequency and location
of inclusions formed by GFP-tagged protein when compared
with untagged proteins detected by immunofluorescence.
Polyglutamine length-dependent inclusion formation also
occurred upon expression of full-length androgen receptor,
albeit with less frequency, with no apparent toxicity (Fig. 2C).
Western blot of extracts prepared from transfected cells showed
that a substantial fraction of AR-112Q is retained in the well, the
stacking gel, and the interface of the stacking and resolving
gel, correlating with the formation of insoluble aggregates as
has been described for expanded polyglutamine-containing
huntingtin, ataxin-1 and ataxin-3 (16–18).
To evaluate cell viability, we used fluorescence-activated cell
sorting (FACS) analysis to assess the ability of cells to exclude
propidium iodide. The use of GFP-tagged constructs allowed
us to examine specifically the transfected population of cells.
Truncated AR with expanded polyglutamine was found to be
cytotoxic, with about 30% of AR-112Q-expressing cells dead
by 96 h post-transfection, comparable to the toxicity of the proapoptotic protein Bax (Fig. 1D). We observed comparable
toxicity in a variety of neuronal and non-neuronal cell lines
including the motor neuronal cell line MN-1 (not shown). This
general toxicity of expanded polyglutamine has been widely
observed; the cell-type specificity of polyglutamine disease is
probably determined by protein context. In the case of the
androgen receptor, for example, this might relate to the higher
level of mutant protein expression in motor neurons or to an
increased propensity of motor neurons to generate toxic
fragments.
RESULTS AND DISCUSSION
Cytoplasmic polyglutamine inclusions are aggresomes
Polyglutamine expansion in the androgen receptor leads
to inclusion formation, aggregation and cytotoxicity
The inclusions generated by AR with expanded polyglutamine
exhibited characteristics of aggresomes. Nearly all of
the AR-112Q inclusions were found in close juxtaposition to
the MTOC, as identified by staining for g-tubulin, a major
MTOC component (Fig. 2A). Furthermore, AR-112Q aggresome formation was associated with a striking redistribution of
intermediate filaments from the normal lattice network into a
cage-like structure around the inclusion. In neuronal cells, AR112Q aggresomes were encased by a thick cortex of neurofilaments (Fig. 2B), while in HEK-293 cells the aggresomes were
encased by vimentin (not shown). As has been observed for
prototypical aggresomes, inclusions of AR-112Q were found in
association with components of the ubiquitin-proteasome
system, including ubiquitin (Fig. 2D–F), Hsp70 (Fig. 2G–I),
and the 20 S proteasome subunit (not shown).
Truncated androgen receptor containing a 19-glutamine repeat
(AR-19Q-GFP) showed a diffuse pan-cellular pattern of
expression that remained unchanged up to 1 week following
transient transfection (Fig. 1A). In contrast, when the polyglutamine tract was expanded to 112 residues, the protein
formed compact, spherical inclusions almost immediately after
expression was evident, 15 h post-transfection. The use of
truncated AR constructs is supported by the finding of only Nterminal epitopes within inclusions of SBMA motor neurons,
suggesting that proteolysis of full-length AR occurs during the
course of the disease (11). Thus, the truncated form of AR
represents a potential proteolytic fragment of the mutant
Human Molecular Genetics, 2003, Vol. 12, No. 7
751
Figure 1. AR with expanded polyglutamine forms cytoplasmic, juxtanuclear inclusions, aggregates by western blot and is cytotoxic. (A) AR-19Q-GFP is found in
a diffuse, pan-cellular distribution 48 h post-transfection, and the cells retain normal morphology 96 h post-transfection. AR-112Q-GFP is found in compact, spherical inclusions that distort the nuclear architecture. After 96 h many AR-112Q-GFP transfected cells show nuclear fragmentation characteristic of apoptosis. Inset:
higher magnification of an inclusion-bearing cell undergoing nuclear fragmentation. Red scale bar ¼ 20 mm. (B) Polyglutamine inclusions form both in the nucleus
and in the cytoplasm of MN-1 cells and HEK-293 cells, but cytoplasmic inclusions predominate. White scale bar ¼ 5 mm. (C) AR-19Q migrates as a monomer in
SDS–PAGE, while a substantial fraction of AR-112Q is retained in the well and the stacking gel. (D) Polyglutamine length-dependent cell death is evident upon
expression of AR-112Q-GFP in HEK-293 cells. Cell viability was determined by propidium iodide exclusion within the transfected cell population and quantitated
by FACS analysis.
Prototypical aggresomes are formed by two misfolded
proteins: GFP-250 (a GFP-tagged fragment of the vesicleassociated protein p115) (8) and the DF508 mutant of the
CFTR (6,8). When co-expressed with GFP-250 or DF508
CFTR, AR-112Q formed closely apposed, yet distinct inclusions (Fig. 2J–L and M–O, respectively). Thus, the proteins
segregated into distinct, homogeneous inclusions, both adjacent to the MTOC, rather than co-mingling in a single
aggresome. The formation of distinct, homotypic aggresomes
have been previously observed when unrelated misfolded
proteins are co-expressed (19).
Key to aggresome formation is dynein-mediated retrograde
transport along microtubules (1). Nanomolar concentrations of
nocodazole resulted in a pronounced reduction in the formation
of inclusions by AR-112Q. In this dose range, nocodazole
blocks microtubule function by altering dynamic instability,
demonstrating that AR-112Q aggresome formation is dependent on microtubule function (20). In nocodazole-treated cells,
the expanded polyglutamine proteins were distributed diffusely
throughout the cytoplasm (Fig. 2P) or coalesced into multiple,
peripheral spiculated bodies (Fig. 2Q). The inhibitory effect of
nocodazole on inclusion formation was dose-dependent, with
2 mg/ml resulting in 80% reduction in inclusion formation
(Fig. 2R).
Minus end-directed movement along microtubules is mediated
by dynein motors, which bind their cargo via a large macromolecular intermediate called dynactin (21). Overexpression of
p50/dynamitin, a subunit of dynactin, disrupts dynactin function
and results in a failure of dynein-mediated, minus end-directed
movement along microtubules (22). This feature has enabled use
of p50/dynamitin in evaluating dynein-mediated movement of
cargo within cultured cells. With this approach, we observed a
dose-dependent decrease in cytoplasmic inclusion formation
by AR-112Q, demonstrating that polyglutamine inclusion
formation was not only microtubule-dependent, but also
mediated by dynein motor function (Fig. 2S).
Aggresome formation is cytoprotective
The function of aggresomes and their role in toxic protein
diseases is unknown. The ability to manipulate aggresome
formation with both molecular and pharmacological interventions offered the opportunity to address this central question.
We found that increasing doses of nocodazole resulted in a
dose-dependent exacerbation of AR-112Q toxicity (Fig. 3A).
Similarly, we found that overexpression of p50/dynamitin
exacerbated the toxicity of AR-112Q (Fig. 3B). The doses of
nocodazole and p50/dynamitin we used strongly suppressed
aggresome formation (Fig. 2M and N), but were not toxic to
cells expressing AR-19Q (Fig. 3A and B). The inverse
relationship between inclusion-formation and AR112 toxicity
is consistent with a cytoprotective function for these structures.
The aggresome-forming proteins GFP-250 and DF508 CFTR
were not toxic to cells in culture (Fig. 3C and D, only GFP-250
shown), demonstrating that aggresomes themselves are not
inherently toxic. However, inhibiting aggresome formation
with either nocodazole or p50/dynamitin resulted in marked
toxicity in cells expressing GFP-250 (Fig. 3C and D) or DF508
752
Human Molecular Genetics, 2003, Vol. 12, No. 7
Figure 2. Cytoplasmic inclusions of polyglutamine are aggresomes. (A) Virtually all cytoplasmic inclusions formed by AR-112Q (green) are found at the MTOC as
shown by close apposition to staining for g-tubulin (red). (B) Inclusion formation (green) is accompanied by a collapse of the neurofilament network into a dense cortex
around the inclusion (red). (C) Expression of full-length androgen receptor with expanded polyglutamine also leads to the formation of cytoplasmic, juxtanuclear inclusions, albeit with less frequency. (D–F) Polyglutamine inclusions (green) stain positively for ubiquitin (red). (G–I) Polyglutamine inclusions (green) stain richly for
HSP70 (red). Cotransfection of AR-112Q (red) with DF518 CFTR (green, J–L) or GFP-250 (green, M–O) results in the formation of adjacent, independent, homotypic
inclusions. Polyglutamine inclusion formation is blocked by nocodazole and overexpression of p50/Dynamitin (P, Q) and the effect is dose-dependent (R, S). Scale
bar ¼ 5 mm.
CFTR (not shown). These observations clearly dissociate
aggresome/inclusion formation from cytotoxicity and support
the notion that aggresomes protect cells from the deleterious
effects of misfolded proteins.
Aggregation is distinct from inclusion formation
The term ‘inclusion’ describes abnormal intracellular structures
observed histologically, while the term ‘aggregate’ implies
a biochemical phenomenon. Aggregation-prone proteins
associated with neurodegenerative diseases are frequently
found in inclusions, prompting the frequent (and inappropriate)
interchangeable use of the terms ‘aggregate’ and ‘inclusion’.
This confusion has complicated the debate as to whether
inclusions themselves are cytotoxic. The distinction between
aggregation and inclusion formation is underscored by the
observation that these processes can be experimentally
dissociated. Microtubule inhibition resulted in a marked
reduction in inclusion formation by AR-112Q (Fig. 2R and
S). Despite this reduction in inclusions, western blot revealed
that aggregation of AR-112Q persists (Fig. 4A). Interestingly,
reduction of inclusion formation was associated with an
increase in the steady-state level of AR-112Q monomer.
Thus, it is likely that the toxicity associated with expanded
polyglutamine-containing proteins is mediated by monomeric
or microaggregated species rather than large histologically
detectable inclusions (Fig. 4B).
Inclusion formation is associated with an accelerated
rate of mutant protein turnover
The increased steady-state levels of monomeric AR-112Q with
microtubule inhibition, with no apparent change in the amount
of aggregated AR-112Q, suggested that the rate of AR-112Q
turnover was reduced in the absence of inclusions. To address
this directly, we performed pulse-chase experiments to
Human Molecular Genetics, 2003, Vol. 12, No. 7
753
Figure 3. Blocking aggresome formation exacerbates the toxicity of expanded polyglutamine and reveals the inherent toxicity of other aggregation-prone proteins.
(A) Polyglutamine toxicity is exacerbated in a dose-dependent manner by nocodazole. The dose range of 5–100 ng/ml alters microtubule dynamics and blocks
aggresome formation, but is non-toxic to cells expressing AR-19Q. (B) Overexpression of p50/dynamitin also results in a dose-dependent exacerbation of polyglutamine toxicity. (C, D) Expression of neither GFP-250 nor 250-GFP is cytotoxic in the absence of microtubule inhibition. However, blocking aggresome formation with either nocodazole (C) or p50/dynamitin (D) reveals the inherent toxicity of GFP-250.
determine the half-life of aggresome-forming proteins in the
presence and absence of microtubule inhibition. The same
approach was used to look at the impact of microtubule
inhibition on proteins not associated with inclusions. AR-112Q
was found to have a half-life of 1.8 h, substantially shorter
than the half-life of AR-19Q, which was 2.8 h (Fig. 5A).
Similarly, full-length AR with an expanded glutamine tract has
been shown to have an accelerated rate of turnover relative to
wild-type AR (23). Under conditions of microtubule inhibition
the half-life of AR-112Q was prolonged to approximately that
of AR-19Q, which was unaffected by nocodazole. Similarly,
GFP-250 had a short half-life of 1.6 compared with 3 h for
250-GFP, and microtubule inhibition prolonged the half-life of
GFP-250 to approximately that of 250-GFP (Fig. 5B). These
observations indicate that the turnover rates of inclusion-prone
proteins may be particularly sensitive to microtubule inhibition,
perhaps because the inclusions serve as cytoplasmic recruitment centers to facilitate degradation.
A role for aggresomes in accelerating the turnover of
misfolded proteins is suggested by their composition. In
addition to sequestering misfolded protein, aggresomes
extensively recruit from cytosolic pools of proteolytic machinery. As they grow, aggresomes become enriched with ubiquitin
and major components of the 26 S proteasome including the
20 S proteosomal subunit and the PA700 and PA28 proteosomal activator complexes (7). Aggresomes are also rich in
cellular chaperones that participate in the presentation of
misfolded proteins to the proteasome complex including
members of the Hsp40 family (Hdj1 and Hdj2), a member of
the chaperonin family (TCP1), and Hsp70 (7,8). Based on these
observations, aggresomes may represent a site of highly
concentrated proteasome activity, protecting the cell in the
setting of excess misfolded protein. Alternatively, aggresomes
may provide access to an alternative degradative pathway in
situations where the capacity of the ubiquitin-proteasome
system is exceeded (24). Autophagy is the primary cellular
means of disposing of large structures such as organelles. Large
structures are first engulfed by autophagosomes, which then
fuse with lysosomes allowing degradation by lysosomal
hydrolases. Evidence supporting a role for autophagy in the
degradation of aggregation-prone polyglutamine-containing
and polyalanine-containing proteins has recently been reported
(25). Consistent with these findings, electron microscopy
revealed relatively large accumulations of electron-dense
material in double-membraned structures adjacent to nuclei in
transfected HEK-293 cells (Fig. 6A). The large doublemembraned structures were frequently found fused with
lysosomes (Fig. 6B). Only infrequently did we observe nonmembrane bound spherical clusters of electron-dense material,
as previously described for aggresomes formed by CFTR and
GFP-250 (6,8). The size, frequency and location of these
double-membraned structures correlated with the size, frequency and location of AR-112Q inclusions observed by
immunofluorescence, and the presence of AR-112Q protein
754
Human Molecular Genetics, 2003, Vol. 12, No. 7
Figure 4. Blocking aggresome formation does not prevent polyglutamine aggregation. (A) High molecular weight aggregate of AR-112Q persists in the presence
of nocodazole despite a block in aggresome formation, underscoring the distinction between the histological inclusions (aggresomes) and biochemical aggregates.
In contrast, the steady-state levels of AR-112Q monomer are increased in a dose-dependent manner by nocodazole. (B) It is likely that the toxicity associated with
expanded polyglutamine-containing proteins is mediated by monomeric or microaggregated species rather than large histologically detectable inclusions.
within these structures was verified by immunolabeling
(Fig. 6C). In addition, labeled material was found in association
with lysosomal structures in the periphery of the cytoplasm.
Frequently, positively labeled small densities consistent with
microaggregates were found adjacent to lysosomes (Fig. 6D)
and in some cases within lysosomes (Fig. 6E). These
observations suggest that lysosomal-mediated degradation
may contribute to the turnover of expanded polyglutamine.
Concentration of protein aggregates within inclusions may
facilitate elimination of harmful proteins by multiple degradative pathways, such as the ubiquitin–proteasome system and
the lysosomal system, and this mechanism may be relevant to a
wide array of neurodegenerative diseases. Recognition of the
protective role of inclusions is of particular relevance in
ongoing efforts at identifying therapeutic interventions for
protein deposition diseases—and caution that elimination of
inclusions should not be used as a surrogate marker of
therapeutic efficacy. Indeed, compounds that increase the
process of inclusion formation may even be beneficial.
green fluorescent protein (EGFP). A variety of polyglutamine
repeat lengths was generated by spontaneous expansions and
contractions of the CAG repeat within the bacterial strain
DH5a. pBAX and pBAX-GFP were a gift from Richard Youle
(NINDS/NIH, Bethesda, MD, USA). p50/dynamitin was a gift
from Richard Vallee (University of Massachussetts, Worcester,
MA, USA). p250-GFP and pGFP-250 were gifts from Elizabeth
Sztul (University of Alabama at Birmingham, AL, USA).
Cell culture and transfections
HEK-293 cells and MN-1 cells (mouse motor neuronneuroblastoma cells) (26) were maintained in DME medium
(Gibco BRL) with 10% fetal bovine serum (Gibco BRL). Cells
were plated at a density of 60% in six-well culture dishes the
day prior to transfection. Transfection with 1.0 mg of pCMVARNGFP or pBax-GFP DNA was carried out using
Lipofectamine Plus following the manufacturers protocol
(Gibco BRL). For the experiments indicated, growth media
was supplemented with 5–100 ng/ml nocodazole (Sigma).
MATERIALS AND METHODS
Immunofluorescence microscopy
DNA constructs
HEK-293 and MN-1 cells were grown on coverslips or two-well
chamber-slides (Nunc). At the times indicated the cells were
fixed in 4% paraformaldehyde for 15 min at room temperature
then washed three times in phosphate-buffered saline (PBS).
Cells were permeabilized with 0.1% Triton X-100 in PBS for
15 min, rinsed with PBS, and incubated for 30 min in blocking
pCMV-ARNGFP plasmids (N ¼ 19 or 112 CAG repeats) were
generated by subcloning an Xba1/EcoR1 fragment of
pAR65DHA (13) into pEGFP-N1 (Clontech), generating a
fusion cDNA of truncated androgen receptor and enhanced
Human Molecular Genetics, 2003, Vol. 12, No. 7
755
Figure 5. Microtubule inhibition prolongs the half-life of aggresome-forming proteins. (A) AR-112Q (solid circles, solid line) has a short half-life of 1.8 h compared with 2.8 h for AR-19Q (open triangles, solid line). Microtubule inhibition prolongs the half-life of AR-112Q (Solid circles, dashed line) to approximately
that of AR-19Q (open triangles, dashed line), which is unaffected by nocodazole. Similarly, (B) GFP-250 (solid diamonds, solid line) exhibits a short half-life of
1.6 compared with 3 h for 250-GFP (open squares, solid line). Nocodazole treatment prolongs the half-life of GFP-250 to approximately that of 250-GFP.
Corresponding pulse-chase blots are shown below.
buffer (2% goat serum, 0.1% Triton X-100, 0.2% dry milk, and
1% bovine serum albumin). The cells were then incubated with
primary antibody in blocking buffer for 1 h, washed three times
with PBS, followed by incubation with secondary antibody in
blocking buffer for 30 min. All cells were stained with 1 mg/ml
Hoechst 33342 (Sigma) in the final PBS wash for 10 min.
Rabbit polyclonal antibody which recognizes human androgen
receptor (N20, Santa Cruz Biotechnologies) was used at a
1 :100 dilution. g-Tubulin monoclonal antibody (clone GTU-88
from Sigma) was used at 1: 10 000 dilution. Neurofilament 200
monoclonal antibody (clone NE14 from Sigma) was used at
1 :40 dilution. Ubiquitin monoclonal antibody NCL-Ubiqm
(Novocastra) was used at 1 :50 dilution. Texas Red-conjugated
goat anti-rabbit and rabbit anti-mouse secondary antibodies
(Fischer Scientific) were used at 1: 100 dilution. Deconvolved
images were produced on an Olympus microscope using a 60 water-immersion objective and Deltavision software (Applied
Precision) on a Silicon Graphics workstation. Multiple 0.2 mm
optical sections were analyzed.
were fixed with 4% paraformaldehyde in 0.1 M phosphate
buffer for 45 min, washed three times with 0.1 M phosphate
buffer, then blocked and permeablized with PBS þ 5% goat
serum and 0.1% saponin for 1 h. The slides were then
incubated with rabbit polyclonal antibody N20 (1:1000,
Santa Cruz Biotechnologies) which recognizes the Nterminus of human androgen receptor in PBS þ 5% normal
goat serum þ 0.1% saponin for 1 h at room temperature.
The cells were then washed three times with PBS þ 1%
goat serum, followed by three washes with PBS þ 5% dry
milk. The cells were then incubated with 4 ml anti-rabbit
Nanogold (Nanoprobes) in 1 ml of PBS þ 5% dry milk for
1 h at room temperature, followed by one wash in
PBS þ 5% dry milk, then three washes in PBS. The
samples were then fixed a second time in 2% glutaraldehyde in PBS for 30 min, washed three times in PBS, rinsed
with deionized water, and subjected to staining and silver
enhancement as described (27). After dehydration, embedding and sectioning, the samples were examined on a JEOL
1200EX electron microscope.
Electron microscopy
HEK-293 cells were plated at 60% density on permanox
chamber slides (Nunc) and transfected with pAR-112Q as
described above. Ninety-six hours post-transfection, the cells
Western blotting and pulse-chase
Forty-eight hours post-transfection, HEK-293 cells were
harvested by scraping in lysis buffer [2% SDS, 10 mM Tris
756
Human Molecular Genetics, 2003, Vol. 12, No. 7
Figure 6. Ultrastructural analysis demonstrates that AR-112Q associates with double membraned structures and with lysosomes. (A) A large double membrane
bound structure (In) adjacent to the nucleus (N) is filled with electron-dense material. Inset, the double membranes of the inclusion (arrow) and nucleus (arrowhead)
are seen. (B) The large membrane-bound structures were frequently observed to be fused with lysosomes. (C) Immunogold labeling using an antibody directed to
the N-terminus of the androgen receptor reveals labeled material within membrane-bound structures. (D, E) Immunolabeled material was frequently observed in
small cytoplasmic densities in the periphery of the cell in association with lysosomes.
pH 6.8, 10% glycerol, Complete protease inhibitor cocktail
(Roche)] and sonicated with two 5 s pulses. Fifty micrograms
of extract was boiled for 5 min and resolved by 12% SDS–
PAGE. Proteins were transferred to an Immobilon-P membrane
(Millipore) for 90 min at 150 mA. The membranes were
blocked in 5% milk, and probed with anti-AR N-20 polyclonal
antibody (1: 1000, Santa Cruz Biotechnologies). After three
10 min washes in Tris-buffered saline (TBS) with 0.5% Tween
20, HRP-labeled anti-rabbit antibody secondary (1:3000, Santa
Cruz Biotechnologies) was applied for 45 min. After three
10 min washes in TBS þ 0.5% Tween 20 and three 5 min
washes in TBS, ECL (Amersham/Pharmacia) was used for
detection. For pulse-chase experiments, 48 h post-transfection,
HEK-293 cells were washed twice with methionine-free
DMEM (Gibco) and incubated in methionine-free DMEM for
10 min. The cells were then incubated with pulse media
(methionine-free DMEM supplemented with 2 mCi of [35S]
labeled methionine) for 10 min, followed by incubation in
chase media (methionine-free DMEM supplemented with
15 mg/l unlabeled methionine). At time zero and multiple time
points post-chase the cells were harvested by scraping, pelleted
and lysed by sonication in 500 ml of 1 PBS containing 0.1%
Triton X-100 and Complete protease inhibitor cocktail
(Roche). One milligram of lysates was used for immunoprecipitation with 30 ml of agarose-conjugated rabbit polyclonal
anti-GFP antibody (catalog no. sc-8334 AC, Santa Cruz
Biotechnologies) in NET-gel buffer in a total volume of 1 ml
as described (28). The resuspended immunoprecipitate was
subjected to SDS–PAGE and transferred to an Immobilon-P
membrane (Millipore) for 90 min at 150 mA. A Storm 560
phosphorimager (Molecular Dynamics) was used to detect
radioactivity, and the signal was quantified using ImageQuant
1.2 software. The average and standard deviation of three
experiments is plotted. The half-life of AR protein in each
experiment was estimated using the equation derived from a
logarithmic plot of percentage of protein remaining versus time
using Microsoft Excel software.
FACS survival assay
To measure directly the toxicity of expanded polyglutamine in
the context of truncated androgen receptor we used a
fluorescence-activated cell sorting (FACS)-based survival
assay. The use of GFP-fused constructs allowed specific
analysis of the transfected population. At 48, 72 or 96 h after
transient transfection cells were harvested with trypsin, gently
pelleted by centrifugation and resuspended in PBS with 1%
serum on ice at a concentration 106/ml. The cells were stained
with propidium iodide 1 mg/ml (PI, Sigma), gently vortexed
and incubated for 15 min at room temperature in the dark. For
each sample 50 000 non-gated events were acquired (Beckman
Coulter XL instrument and software package used for
analysis). The results were expressed as a percentage of PInegative (viable) cells (FL-2 channel) relative to total GFP
positive (transfected) cells (FL-1 channel).
ACKNOWLEDGEMENTS
Electron microscopy was performed with the help and advice
of the NINDS EM Facility. We thank Richard Vallee, Elizabeth
Sztul and Diane Merry for contributing the p50/dynamitin,
Human Molecular Genetics, 2003, Vol. 12, No. 7
p250-GFP and pAR65DHA expression constructs, respectively.
J.P.T. is supported by NINDS Career Transition Award K22NS44125-01. J.R. and C.M.S. were participants in the Howard
Hughes Medical Institute medical student scholars program.
REFERENCES
1. Schubert, U., Anton, L.C., Gibbs, J., Norbury, C.C., Yewdell, J.W. and
Bennink, J.R. (2000) Rapid degradation of a large fraction of newly
synthesized proteins by proteasomes. Nature, 404, 770–774.
2. Hershko, A. and Ciechanover, A. (1998) The ubiquitin system. A. Rev.
Biochem., 67, 425–479.
3. Ward, C.L., Omura, S. and Kopito, R.R. (1995) Degradation of CFTR by
the ubiquitin-proteasome pathway. Cell, 83, 121–127.
4. Taylor, J.P., Hardy, J. and Fischbeck, K.H. (2002) Toxic proteins in
neurodegenerative disease. Science, 296, 1991–1995.
5. Kopito, R.R. (2000) Aggresomes, inclusion bodies and protein aggregation.
Trends Cell Biol., 10, 524–530.
6. Johnston, J.A., Ward, C.L. and Kopito, R.R. (1998) Aggresomes: a cellular
response to misfolded proteins. J. Cell Biol., 143, 1883–1898.
7. Wigley, W.C., Fabunmi, R.P., Lee, M.G., Marino, C.R., Muallem, S.,
DeMartino, G.N. and Thomas, P.J. (1999) Dynamic association
of proteosomal machinery with the centrosome. J. Cell Biol., 145, 481–490.
8. Garcia-Mata, R., Bebock, Z., Sorscher, E.J. and Sztul, E.S. (1999)
Characterization and dynamics of aggresome formation by a cytosolic
GFP-chimera. J. Cell Biol., 146, 1239–1254.
9. Bence, N.F., Sampat, R.M. and Kopito, R.R. (2001) Impairment of the
ubiquitin–proteasome system by protein aggregation. Science, 292,
1552–1555.
10. La Spada, A.R., Wilson, E.M., Lubahn, D.B., Harding, A.E. and
Fischbeck, K.H. (1991) Androgen receptor gene mutations in X-linked
spinal and bulbar muscular atrophy. Nature, 352, 77–79.
11. Li, M., Miwa, S., Kobayashi, Y., Merry, D.E., Yamamoto, M., Tanaka, F.,
Doyu, M., Hashizume, Y., Fischbeck, K.H. and Sobue, G. (1998) Nuclear
inclusions of the androgen receptor protein in spinal and bulbar muscular
atrophy. Ann. Neurol., 44, 249–254.
12. Abel, A., Walcott, J., Woods, J., Duda, J. and Merry, D.E. (2001)
Expression of expanded repeat androgen receptor produces neurologic
disease in transgenic mice. Hum. Mol. Genet., 10, 107–116.
13. Merry, D.E., Kobayashi, Y., Bailey, C.K., Taye, A.A. and Fischbeck, K.H.
(1998) Cleavage, aggregation and toxicity of the expanded androgen
receptor in spinal and bulbar muscular atrophy. Hum. Mol. Genet., 7,
693–701.
14. Diamond, M.I., Robinson, M.R. and Yamamoto, K.R. (2000) Regulation of
expanded polyglutamine protein aggregation and nuclear localization by the
glucocorticoid receptor. Proc. Natl Acad. Sci. USA, 97, 657–661.
757
15. McCampbell, A., Taylor, J.P., Taye, A.A., Robitschek, J., Li, M., Walcott, J.,
Merry, D.E., Chai, Y., Paulson, H.L., Sobue, G. et al. (2000) CREB-binding
protein sequestration by expanded polyglutamine. Hum. Mol. Genet., 9,
2197–2202.
16. Davies S.W., Turmaine, M., Cozens, B.A., DiFiglia, M., Sharp, A.H.,
Ross, C.A., Scherzinger, E., Wanker, E.E., Mangiarini, L. and Bates, G.P.
(1997) Formation of neuronal intranuclear inclusions underlies the
neurological dysfunction in mice transgenic for the HD mutation. Cell, 90,
537–548.
17. Paulson H.L., Perez, M.K., Trottier, Y., Trojanowski, J.Q., Subramony, S.H.,
Das, S.S., Vig, P., Mandel, J.L., Fischbeck, K.H. and Pittman, R.N. (1997)
Intranuclear inclusions of expanded polyglutamine protein in spinocerebellar
ataxia type 3. Neuron, 19, 333–344.
18. Klement I.A., Skinner, P.J., Kaytor, M.D., Yi, H., Hersch, S.M., Clark, H.B.,
Zoghbi, H.Y. and Orr, H.T. (1998) Ataxin-1 nuclear localization and
aggregation: role in polyglutamine-induced disease in SCA1 transgenic mice.
Cell, 95, 41–53.
19. Rajan, R.S., Illing, M.E., Bence, N.F. and Kopito, R.R. (2001) Specificity
in intracellular protein aggregation and inclusion body formation. Proc. Natl
Acad. Sci. USA, 98, 13060–13065.
20. Vasquez, R.J., Howell, B., Yvon, A.M., Wadsworth, P. and
Cassimeris, L. (1997) Nanomolar concentrations of nocodazole alter
microtubule dynamic instability in vivo and in vitro. Mol. Biol. Cell, 8,
973–985.
21. Hirokawa, N. (1998) Kinesin and dynein superfamily proteins and the
mechanism of organelle transport. Science, 279, 519–526.
22. Burkhardt, J.K., Echeverri, C.J., Nilsson, T. and Vallee, R.B. (1997)
Overexpression of the dynamitin (p50) subunit of the dynactin complex
disrupts dynein-dependent maintenance of membrane organelle
distribution. J. Cell Biol., 139, 469–484.
23. Lieberman, A.P., Harmison, G., Strand, A.D., Olson, J.M. and
Fischbeck, K.H. (2002) Altered transcriptional regulation in cells
expressing the expanded polyglutamine androgen receptor. Hum. Mol.
Genet., 11, 1967–1976.
24. Garcia-Mata, R., Gao, Y.-S. and Sztul, E.S. (2002) Hassles with taking out
the garbage: aggravating aggresomes. Traffic, 3, 388–396.
25. Ravikumar, B., Duden, R. and Rubinsztein, D.C. (2002) Aggregate-prone
proteins with polyglutamine and polyalanine expansions are degraded by
autophagy. Hum. Mol. Genet., 11, 1107–1117.
26. Salazar-Grueso E.F., Kim, S. and Kim, H. (1991) Embryonic mouse spinal
cord motor neuron hybrid cells. Neuroreport, 2, 505–508.
27. Tanner, V., Ploug, T. and Tao-Cheng, J.-H. (1996) Subcellular
localization of SV2 and other secretory vesicle components in
PC12 cells by an efficient method of pre-embedding EM
immunocytochemistry for cell cultures. J. Histochem. Cytochem.,
144, 1481–1488.
28. Ausubel, F.M., Brent, R., Kingston, R.E., Moore, D.D., Seidman, J.G.,
Smith, J.A. and Struhl, K. (1990) Detection and analysis of proteins
expressed from cloned genes. In Current Protocols in Molecular Biology.
John Wiley, New York.