Cell Motility and the Cytoskeleton 53:26 38 (2002)

Cytoplasmic Dynein/Dynactin Mediates the

Assembly of Aggresomes
Jennifer A. Johnston,* Michelle E. Illing, and Ron R. Kopito
Department of Biological Sciences, Stanford University, Stanford, California
Aggresomes are pericentrosomal cytoplasmic structures into which aggregated,
ubiquitinated, misfolded proteins are sequestered. Misfolded proteins accumulate
in aggresomes when the capacity of the intracellular protein degradation machinery is exceeded. Previously, we demonstrated that an intact microtubule cytoskeleton is required for the aggresome formation [Johnston et al., 1998: J. Cell Biol.
143:18831898]. In this study, we have investigated the involvement of microtubules (MT) and MT motors in this process. Induction of aggresomes containing
misfolded F508 CFTR is accompanied by a redistribution of the retrograde
motor cytoplasmic dynein that colocalizes with aggresomal markers. Coexpression of the p50 (dynamitin) subunit of the dynein/dynactin complex prevents the
formation of aggresomes, even in the presence of proteasome inhibitors. Using in
vitro microtubule binding assays in conjunction with immunogold electron microscopy, our data demonstrate that misfolded F508 CFTR associate with
microtubules. We conclude that cytoplasmic dynein/dynactin is responsible for the
directed transport of misfolded protein into aggresomes. The implications of these
findings with respect to the pathogenesis of neurodegenerative disease are discussed. Cell Motil. Cytoskeleton 53:26 38, 2002. 2002 Wiley-Liss, Inc.
Key words: aggresome; dynein; proteolysis; centrosome; microtubule

INTRODUCTION

Aggresomes are pericentriolar cytoplasmic structures in which aggregated, multiubiquitinated misfolded

proteins are sequestered [Johnston et al., 1998]. Aggresomes, which appear to be a general response to conditions where a cells capacity to degrade malfolded proteins is exceeded, closely resemble ubiquitin-rich
cytoplasmic inclusion bodies that are pathognomonic
for degenerating neurons in Parkinsons disease, amyotrophic lateral sclerosis and other neurodegenerative diseases [Mayer et al., 1989]. To study the mechanism by
which aggresomes form in cells may help illuminate the
cellular events which underlie these neurological disorders. In particular, the restriction of aggresomes to the
vicinity of the microtubule organizing center (MTOC)
and the dependence of aggresome formation on microtubules strongly suggested a central role for the microtubule (MT) cytoskeleton in the cellular response to
protein aggregation [Johnston et al., 1998].
Microtubules are linear polymers of tubulin that
form polarized tracks extending from the centrosome and
2002 Wiley-Liss, Inc.

radiate outward toward the periphery of eukaryotic cells.

A wealth of literature establishes that the MT cytoskeleton is responsible for most of the directed movement of
material inside of cells including: the dynamic assembly
of membrane-bounded organelles like the Golgi apparatus [Lippincott-Schwartz, 1998], the endoplasmic reticulum [Terasaki et al., 1986], the intracellular movement
of mitochondria [Gotoh et al., 1985], lysosomes and
Abbreviations used: ALLN, N-Acetyl-L-leucyl-L-leucyl-L-norleucinal; CFTR, cystic fibrosis transmembrane conductance regulator;
MTOC, microtubule organizing center; MT, microtubule.
Contract grant sponsor: National Institutes of Health; Contract grant
number: DK43994.
*Correspondence to: Dr. Jennifer A. Johnston, 800 Gateway Blvd.,
South San Francisco, CA 94980.
E-mail: jjohnston@elanpharma.com
Received 12 November 2001; Accepted 8 April 2002
Published online 22 July 2002 in Wiley InterScience (www.
interscience.wiley.com). DOI: 10.1002/cm.10057

Microtubule Motors and Aggresome Formation

endosomes [Matteoni and Kreis, 1987], and the movement of chromosomes at mitosis [Barton and Goldstein,
1996]. MTs function as a dynamic cellular scaffold upon
which force-generating motor proteins move directionally to transport their cargo. Motor proteins fall into two
categories: anterograde motors (mainly kinesins), which
move cargo from the centrosome to the cell periphery,
and retrograde motors such as cytoplasmic dynein, which
move cargo in the opposite direction. Motor proteins of
both types traditionally have been considered to carry
membrane bounded cargo (such as organelles and vesicles) [Gotoh et al., 1985]. However, the demonstration
that cytoplasmic dynein mediates retrograde transport of
membrane-free viral capsids [Sodeik et al., 1997] and the
delivery of MT into axons [Ahmad et al., 1998] establishes that the MT transport apparatus is not limited to
membrane-bounded cargo.
Cytoplasmic dynein usually conducts cargo in association with dynactin, a 20S complex consisting of at
least 9 polypeptides that appear to contribute an essential
role in linking cargo to the dynein motor [Karki and
Holzbaur, 1999; Eckley et al., 1999; King and Schroer,
2000]. The p50 subunit of dynactin (dynamitin) has been
shown to be directly involved in the attachment of cargo
to dynein in a variety of systems. Overexpression of
dynamitin dissociates the dynactin-dynein complex and
disrupts cytoplasmic dynein-dependent events including
chromosome alignment and spindle organization in mitosis [Echeverri et al., 1996], localization of membranous
organelles in interphase [Burkhardt et al., 1997], and
axonal transport of MT into axons [Ahmad et al., 1998].
The p150glued protein is another subunit of the dynactin
complex and has been shown to interact with dynein in a
manner that is critical for dynein function [Vaughan and
Vallee, 1995; Boylan et al., 2000]. An active area of
investigation concerns the determination of how the dynein/dynactin complex interacts with specific cargo.
We have previously reported that the pericentriolar
accumulation of ubiquitinated and aggregated protein
into aggresomes requires an intact MT cytoskeleton
[Johnston et al., 1998]. In the absence of MTs, multiple
small foci of aggregated protein are distributed randomly
throughout the cytoplasm, suggesting that aggresome
formation requires directed transport on MT. In the
present report, we have investigated the involvement of
MT motors in aggresome formation. Our data indicate
that aggresome formation is accompanied by a partial
redistribution of dynein, p50 dynamitin, and p150glued
protein to aggresomes. The overexpression of dynamitin
blocks aggresome formation. Together with data showing that misfolded protein aggregates co-sediment with
MT in vitro, our data establish a central role for the
dynein/dynactin complex in the cellular response to protein aggregation.

27

MATERIALS AND METHODS

Antibodies and Reagents

C. Echeverri and R. Vallee (Worcester Institute,

MA) generously donated p50 dynamitin cDNA and antibodies. Polyclonal antibodies to the C-terminus of
CFTR [Ward and Kopito, 1994] were affinity purified by
high pH elution from a CFTR peptide column. AlphamannosidaseII antibodies were the gift of M. Farquhar
[Velasco et al., 1993]; pericentrin polyclonal antibodies
were the gift of T. Stearns (Stanford University, CA).
Other antibodies used: monoclonal dynein IC 70.1 and
monoclonal kinesin IBII (Sigma, St. Louis, MO); antip150, p50 (dynamitin) Chemicon (Temecula, CA); fluorophore conjugated secondary antibodies Alexa 488 and
Alexa 546 were purchased from Molecular Probes (Eugene, OR). HEK 293 cells were transfected by the
method of calcium phosphate precipitation, or were stably expressing GFP-F508 CFTR as described [Johnston
et al., 1998]. Purified tubulin was purchased from Cytoskeleton, Inc (Denver, CO). The proteasome inhibitors
epoxomicin, clasto-lactacystin--lactone, MG-132, and
ALLN were purchased lyophilized from Boston Biochem., Inc. (Cambridge, MA) and were reconstituted in
DMSO.
Immunoblotting

Cell pellets from washed and transfected HEK cells

were lysed in 250 l of ice-cold IPB buffer (10 mM
Tris-HCl [pH 7.5], 5 mM EDTA, 1% NP-40, 0.5%
deoxycholate, 150 mM NaCl) plus protease inhibitors
(100 M TLCK, 100 M TPCK, 1 mM PMSF) for 30
min on ice. Insoluble material was recovered by centrifugation at 13,000g for 15 min and solubilized in 50 l 10
mM Tris-HCl, 1% SDS for 10 min at room temperature.
After addition of 200 l IPB, samples were sonicated for
20 s with a tip sonicator. Cell fractions, normalized for
total protein, were separated on 7.5% SDS-PAGE and
electroblotted. Chemiluminescent detection was carried
out with the Renaissance detection kit (New England
Nuclear).
Microtubule Assembly Assays

Microtubules were assembled from 0.25 0.5

mg/ml purified tubulin in G-PEM Buffer (80 mM Pipes,
1 mM MgCl2, 1 mM EGTA) at room temperature by the
addition of 10 M paclitaxel (Taxol) in DMSO (Sigma)
and 25 M GTP. After 20 min at room temperature,
GFP-F508 CFTR cell extract was added (1/3 volume of
MT assembly reaction) and incubated at 16C for 112 h
(routinely 1-h incubations were used, with occasional
12-h overnight incubations; no observable difference in
associated CFTR after 1 h). GFP-F508 CFTR cell extracts were prepared as follows: 90% confluent T75

28

Johnston et al.

flasks of cell lines stable expressing GFP-F508 CFTR

were treated with 25 M MG-132 for 57 h. Cells were
harvested by release into 1 ml PBS/4mM EDTA, 3,000g
spin and then resuspended in 500 l of G-PEM 1%
NP-40. Extraction was allowed to proceed on ice for 20
min, then centrifuged at 4C for 5 min at 3,000g. The
supernatant from the 3,000g spin was then centrifuged
for 20 min at 15,000g. The resulting material is extract
that was added to MT assembly assays. Protein not used
immediately was quick frozen in liquid nitrogen and
stored at 80C. After 16C incubation for MT affinity
reactions, mixtures were underlain with a 10% sucrose
cushion made in G-PEM containing taxol and GTP (except for the control reactions, which do not contain any
taxol or GTP), and centrifuged at 10,000g for 15 min at
room temperature. Collected MTs were resuspended in
equal volume of G-PEM, 10 l from both Supernatant
and Pellet fractions were removed to which equal volumes of 2 sample buffer was added for analysis by
SDS-PAGE.
Electron Microscopy

Microtubule assembly assays were performed as

described above with the exception that assay mixtures
were fixed with 2% glutaraldehyde (EMS, Fort Washington, PA) for 20 min at room temperature prior to
sedimentation through 10% sucrose. Microtubule pellets
from in vitro MT assembly assays were resuspended in
100 l of G-PEM buffer. Formvar-Ni 200mesh grids
(EMS, Inc.) were floated on 20 l of fixed MTs for 5 min
at room temperature. Nonspecific binding was blocked
by 5% BSA/PBS for 10 min, followed by 25 min in
primary antibody in PBS/1%BSA. Following five PBS/
1%BSA washes, grids were floated for 20 min on precleared 15-nm gold conjugated secondary antibodies
(EMS, Inc.) diluted 1:10 in PBS/BSA. Subsequently, five
PBS washes and five water washes, a 5-min post fix in
2% glutaraldehyde followed by five more water washes
were carried out. Washed grids were stained with 1%
uranyl acetate for 10 min at room temperature, air dried,
and observed with a JOEL transmission electron microscope at 60 kV.
Immunofluorescence Microscopy

In all experiments, HEK 293 cells grown on no. 1

coverslips were fixed in 20 methanol for 6 min. After
fixation, cells were washed 5 in PBS, followed by 10
min in 5% BSA. Primary antibodies were added to each
coverslip in 1% BSA/PBS and incubated for 30 min to
2 h at room temperature. Cells were washed in PBS and
stained for 3 min with 10 g/ml bisbenzimide. Following
one final wash in PBS, secondary antibodies conjugated
to fluorophore were added for 20 45 min at room temperature. Cells were washed again 5 in PBS, and then

mounted in 50%glycerol/50% PBS on microscope slides

and viewed with a Zeiss Axiovert, using Metamorph
software (Universal Imaging) and a Princeton Instruments CCD to collect and analyze images. Adobe Photoshop v6.0 (San Jose, CA) was used to prepare final
images for publication.

RESULTS

To examine the role of molecular motors in aggresome formation, we used indirect immunofluorescence
microscopy to assess the distribution of the anterograde
motor kinesin (Fig. 1) and the retrograde motor cytoplasmic dynein (Fig. 2) in HEK cells stably expressing GFPF508 CFTR and exposed to proteasome inhibitors. This
F508 mutant membrane protein has previously been
shown to be quantitatively incapable of folding and is
normally rapidly degraded in a proteasome- and ubiquitin-dependent fashion [Jensen et al., 1995; Ward et al.,
1995]. We previously reported that overexpression of
F508 CFTR leads spontaneously to the formation of
aggresomes containing ubiquitinated misfolded F508
CFTR molecules that have been extracted or dislocated
from the ER membrane [Johnston et al., 1998]. Exposure
of these cells to proteasome inhibitor (epoxomicin [Sin,
1999], lactacystin [Fenteany and Schreiber, 1998], MG132 [Bogyo et al., 1997], or ALLN [Vinitsky et al.,
1992]) leads to formation of massive aggresomes within
8 16 h [Johnston et al., 1998]. Epoxomicin is currently
the most specific inhibitor available [Kim, 1999; Sin,
1999]. Large CFTR immunopositive aggresomes, evident in epoxomicin treated GFP-F508 CFTR expressing cells (Figs. 1C, 2D), were also strongly stained with
antibody to cytoplasmic dynein (Fig. 2C) but not with
antibody to kinesin (Fig. 1D). By contrast, cytoplasmic
dynein exhibited a diffuse, somewhat punctate distribution throughout the cytoplasm of untreated untransfected
HEK cells (Fig. 2G) as well as HEK cells stably expressing GFP-F508 CFTR (Fig. 2A). Pericentrin staining in
Figure 2H demonstrates the cellular location of the centrosome, and a comparison to Figure 2G suggests that
dynein is not normally concentrated at the centrosome
area (merged image in Fig. 2I). When untransfected HEK
cells were exposed overnight to epoxomicin, a significant
portion of the cells cytoplasmic dynein was found to be
strongly colocalized with pericentrin staining (Fig. 2J
L). This redistribution of dynein was accompanied with
the distortion of the nuclear envelope that is highly
characteristic of aggresome formation [Johnston et al.,
1998]. These data suggest that proteasome inhibition
redistributed dynein, but not kinesin, to the area occupied
by aggresomes, where dynein colocalizes both with centrosomal markers and with misfolded aggregated protein.

Microtubule Motors and Aggresome Formation

29

Fig. 1. Kinesin does not redistribute to aggresomes. HEK cells stably expressing GFP CFTR (AD) and
imaged for GFP fluorescence (A,C) or indirect immunofluorescence using kinesin antibodies (B,D).
Vehicle treated cells are shown in A,B; C,D demonstrate cells treated with 3 M epoxomicin. Bar 15
m.

Moreover, the data in Figure 2JL indicates this redistribution is not dependent on CFTR expressing cells.
Although GFP-F508 molecules that accumulate
in aggresomes are misfolded, and largely insoluble in
non-denaturing detergent [Johnston et al., 1998] (Fig.
3B, compare lanes 3,4 with 5,6), neither the mobility nor
the detergent solubility of dynein was affected by pro-

teasome inhibition (Fig. 3A compare lanes 1, 2 with 7, 8).

These data suggest that dynein is not a substrate for the
aggresome pathway.
Retrograde transport of cargo by cytoplasmic dynein usually occurs in the context of a dynein/dynactin
complex, which, in addition to the force-generating dynein heavy chain, contains components like dynamitin

30

Johnston et al.

Figure 2.

Microtubule Motors and Aggresome Formation

31

Fig. 2. Cytoplasmic dynein redistributes into aggresomes. AF:

HEK cells stably expressing GFP-F508 CFTR were exposed for 12 h
to vehicle (A,B) or 3 M epoxomicin (CF) to induce aggresome
formation. Cells were imaged for GFP fluorescence (B,D,F) together
with monoclonal antibodies to dynein (A,C,E). Indirect immunofluorescence reveals strong co-localization of GFP-F508 CFTR with
dynein in the area of GFP-F508 CFTR aggresome formation (compare C and D). AD: Imaged using 100 objective to show high
resolution. E,F: Imaged using 40 objective to demonstrate that the

redistribution of dynein after proteasome inhibition occurs in all of the

cells in a field of view. Arrowheads demonstrate a couple of aggresomes in the field of cells. GL: Untransfected HEK cells were treated
with 5 M lactacystin (JL) or vehicle (GI) for 12 h and labeled with
antibodies to dynein (G,J) or pericentrin (H,K). Nuclei were imaged
with bisbenzimide (GL). The merged images demonstrate punctate
staining of dynein with small, if any, overlap with pericentrin (I) and
a distinct colocalization of dynein with pericentrin after proteasome
inhibitor treatment (L). Bar 15 m.

that are important for regulating the specific coupling of

cargo [Karki and Holzbaur, 1999]. To determine the role
of the dynactin complex in the transport of aggregates of
misfolded protein during aggresome formation, we examined the effect of dynamitin expression on GFPF508 distribution. Transfection of GFP-F508 CFTR
HEK cell lines with dynamitin cDNA (Fig. 4) abrogated
the accumulation of GFP-F508 in aggresomes following proteasome inhibition. Instead, the GFP-F508 protein was found in a punctate, diffuse cytoplasmic pattern
resembling that observed for F508 following proteasome inhibition in the presence of nocodazole [Johnston
et al., 1998]. These results were consistent across ten
independent experiments (n 2,000 cells analyzed), at
concentrations of p50 plasmid from 112 g. At all
doses of cDNA transfection, the efficacy of dynamitin
expression was assessed by evaluating dispersion of the
Golgi apparatus (judged by the distribution of mannosidase II; data not shown), which requires dynein/dynactin
dependent transport to preserve its dynamic structural
integrity [Presley et al., 1997]. At the level of resolution

of this study, it is impossible to determine if the dispersed

GFP-F508 CFTR microaggregates were associated
with dynamitin. We did, however, notice that the cellular
p150glued (Fig. 5) and dynamitin (Fig. 4A,D) subunits of
the dynactin complex redistributed to the aggresome in a
manner similar to dynein (Fig. 1C) after proteasome
inhibition. For example, a comparison of the distribution
of endogenous dynamitin in GFP-F508 HEK cells before (Fig. 4A) and after epoxomicin treatment (Fig. 4D)
demonstrates that the dynamitin redistributes to the same
area as GFP-F508 CFTR aggresome formation (Fig.
4F). These data are consistent with our previous
[Johnston et al., 1998] finding that transport on MT is
required for the formation of aggresomes and these data
establish an essential role for dynein/dynactin in aggresome formation.
To further explore the role of MT and molecular
motors in aggresome formation, we assessed the binding
of F508 to MT in vitro (Fig. 6). Detergent extracts were
prepared from GFP-F508 expressing cells that had been
exposed to proteasome inhibitor for 57 h, conditions in

32

Johnston et al.

Fig. 3. Solubility of cytoplasmic

dynein is unaffected by aggresome
formation. A: HEK cells (HEK,
lanes 1,2) or GFP-F508 CFTR
HEK cells (F508, lanes 3 8)
were incubated overnight in vehicle (lanes 1 4) or in the presence
of 10 M nocodazole (lanes 5,6) or
10 M ALLN (lanes 7,8). Detergent-soluble (s, odd numbered
lanes) or insoluble (I, even numbered lanes) fractions were separated on 7.5% SDS-PAGE and
probed with antibody to dynein intermediate chain 70.1. B: Similar
experiment probed with antibody
to CFTR.

which pericentriolar aggresomes are evident, but prior to

the point at which they condense into massive, detergentinsoluble aggresomes. To assess GFP-F508 binding to
MT, these extracts were incubated with taxol-stabilized
MT. GFP-F508 binding to MT was assessed by immunoblot analysis of supernatant and pellet fractions following sedimentation of the mixture through a 10%
sucrose cushion. Under these conditions, a proportion of
GFP-F508 was found in the pellet fraction, suggesting
that the GFP-F508 had cosedimented with MT (Fig.
6A, lanes 4,5). When the assay was conducted under
conditions which do not support MT polymerization
(omission of taxol/GTP) GFP-F508 was found in the
supernatant fraction (Fig. 6A, lanes 2,3). Although we
did not observe 100% of the GFP-F508 binding to MT
in this assay, we consistently observed at least 50%
binding. It will be interesting to use this assay to investigate the specific requirements for GFP-F508 CFTR
aggregated protein binding to MT. This result demonstrates that the sedimentation of GFP-F508 was due to
binding to MT, not to GFP-F508 condensation into a
dense aggregate that sedimented non-specifically after
incubation. MT association of dynein was also observed
in the presence of taxol (Fig. 6A, bottom lanes 4,5), but
not in the absence (Fig. 6A, bottom lanes 2,3). Together,
these data suggest that misfolded protein from proteasome-treated cells sediments with MT under conditions
in which dynein also sediments with MTs.
The interaction of GFP-F508 with MT was also
evaluated by negative stain immunogold electron microscopy (Fig. 6B). Pellet fractions from binding assays as in
Figure 6A were first incubated with polyclonal antibody

to CFTR [Ward and Kopito, 1994] followed by secondary antibody conjugated to 15 nm colloidal gold. Transmission electron microscopy of negative-stain preparations revealed that MT were visible as the major
component of these pellets, by morphological criteria.
Moreover, immunogold-labeling experiments with antitubulin antibodies show that the linear polymers are
largely composed of tubulin (data not shown). In immunogold labeling experiments for CFTR in these preparations, a large proportion of the CFTR-immunoreactivity
was associated with the MT, confirming the specificity of
GFP-F508 binding to MT indicated by our immunoblot
experiments (Fig. 6B, panel a). In data from three independent experiments, 10 12 random fields were photographed and gold particles were counted: 70% of gold
particles were MT associated, while 30% were found
to be not associated with MTs. This immunoreactive
material was frequently found in clusters containing 15
gold particles and was usually associated with what appear to be amorphous proteinaceous attachments that
project from the sides of assembled MTs (see Fig. 6C).
These data demonstrate that the CFTR that sediments
through 10% sucrose in the presence of taxol is associated with polymerized MTs.
DISCUSSION

The accumulation of aggregates of ubiquitinated

misfolded protein into a single large cytoplasmic inclusion body is a hallmark of cellular pathology associated
with neurodegenerative disease [Mayer et al., 1989] and
a general response of animal cells to the presence of

Microtubule Motors and Aggresome Formation

Fig. 4. Co-expression of dynamitin blocks GFP-F508 aggresome

formation. GFP-F508 CFTR HEK cells were transfected with dynamitin for 12 h, and then were treated overnight with either DMSO
(AC) or 3 M epoxomicin (DF) to induce aggresome formation.
Cells were prepared for indirect immunofluorescence with antibody to
dynamitin (A,D), imaged for GFP fluorescence (B, E), and the images

33

merged (A,B: C and D,E: F). Arrow in DF indicates a cell that is

overexpressing dynamitin, and does not contain an aggresome of
GFP-F508 CFTR. Note the other cells in the field that do not
overexpress dynamitin form aggresomes. A comparison of A and D
reveals that the endogenous dynamitin redistributes to aggresomes
after proteasome inhibition. Bar 15 m.

34

Johnston et al.

Fig. 5. Aggresome formation is accompanied by a redistribution of dynactin subunit p150glued. HEK

cells stably expressing GFP-F508 CFTR (AH) and imaged for GFP fluorescence (B,D,F,H) or indirect
immunofluorescence using p150glued antibodies (A,C,E,G). AD: Vehicle treated cells. EH: Cells treated
with 3 M epoxomicin for 12 h. A, B, E, F were imaged using a 100 objective, C,D,G,H were imaged
using 40 objective. Arrows in G, H indicate a group of cells that contain GFP-F508 CFTR aggresomes
immunopositive for p150glued. Bar 15 m.

undegradable protein. Such inclusion bodies can be induced experimentally by overexpression of aggregationprone protein or by directly inhibiting the proteasome
[Johnston et al., 1998]. In these cases, aggregated proteins accumulate in a pericentriolar structure, encircled in

a collapsed intermediate filament network, which we

have termed the aggresome [Johnston et al., 1998]. The
concentration of aggregated proteins near the MTOC and
the dependence of this process on an intact MT cytoskeleton led us to propose a model in which the retrograde

Microtubule Motors and Aggresome Formation

Figure 5.

transport of misfolded protein on MT tracks is required

for aggresome formation. In this study, we confirm that
hypothesis and demonstrate the direct involvement of the
minus end directed molecular motor cytoplasmic dynein
in aggresome formation.
The dominant negative effect of dynamitin coexpression on aggresome formation argues strongly that
dynein mediates aggresome formation in conjunction
with the dynactin complex (Fig. 4). Dynein/dynactin has
been shown to be responsible for the minus end directed
movement of organelles [Presley et. al., 1997], vesicles
[Burkhardt et. al., 1997], viral capsids [Sodeik et. al.,
1997], and MT [Ahmad et. al., 1998] in eukaryotic cells.
We previously established that the F508 molecules that
accumulate in aggresomes are free of membrane, either

35

(Continued.)

because they were never inserted into the ER or, more

likely, because they were dislocated from the ER [Bebok
et al., 1998; Jensen et al., 1995; Johnston et al., 1998;
Ward et al., 1995]. Our data, therefore, suggest that the
dynein/dynactin motor complex is also able to deliver
membrane-free protein aggregates to the MTOC. This
conclusion is supported by a report in which dynamitin
co-expression blocks the sequestration of a cytosolic
GFP chimera into aggresome-like structures [GarciaMata et al., 1999]. However, in that study, direct binding
of substrate to microtubules was not examined, nor was
the distribution of microtubule motor proteins after proteasome inhibition. The data in this study indicate that
dynein colocalizes with aggresomes after proteasome
inhibition (Fig. 2). It is possible that a dynamic pool of

36

Johnston et al.

rapidly degraded dynein molecules is spared from degradation by the presence of proteasome inhibitor and thus
enters the aggresome pathway. However, dynein remains
soluble and non-ubiquitinated even under prolonged aggresome-promoting conditions (Fig. 3A). These data
suggest that dynein present in aggresomes reflect a
change in steady-state distribution of dynein as a result of
an increased flux of misfolded protein to the aggresome,
rather than dynein itself becoming a substrate of the
aggresome pathway. The observed redistribution of dynein into aggresomes may reflect an increased net retrograde transport of dynein (together with misfolded substrate) combined with a reduced rate of dissociation of
dynein/dynactin/substrate complexes at the MTOC. The
accumulation of dynein in the vicinity of the aggresomes
may reflect a traffic jam created by the large retrograde
flux of protein aggregate and the deposition of intermediate filaments around the MTOC [Johnston et al., 1998].
Alternatively, it is possible that the redistribution of
dynein into aggresomes is an indirect consequence of the
disruption of proteasome activity. It has recently been
shown that the formation of aggresome in cells can result
in a cell cycle block at G2-M transition [Bence et. al.,
2001]. Several components of the dynein/dynactin complex are phosphoproteins [Holleran et. al., 1998] and it is
possible that disregulation of one or more kinases or
phosphatases in response to proteasome inhibition could
lead to a change in dynein distribution. Future experiments will be required to distinguish among these possibilities.
In regard to the dynein/dynactin complex, is clear
that dynein (Fig. 2), dynamitin (Fig. 4), and p150glued
(Fig. 5) are be found to be similarly redistributed to the
area of F508 aggresome formation after proteasome
inhibition. Components of MT motors can be found on
the surfaces of vesicles [Waterman-Storer et al., 1997],
and it will be interesting in future studies to determine if
the non-membrane associated, dispersed aggregates are
bound to MT motor proteins. The MT sedimentation
assay described in Figure 6A will be useful to determine
if the dispersed CFTR aggregates that occur after dynamitin overexpression are associated with any components of the dynein/dynactin complex, as well as to
determine if dynamitin overexpression prevents in vitro
binding of CFTR to MTs. The molecular details of how
dynein/dynactin recognizes and binds misfolded protein,
and how these signals differ from those involved in
organelle and vesicle movement, remain to be investigated. The cell-free MT binding assay described in this
study provides a biochemically tractable system with
which to investigate the molecular requirements of aggresome formation.
Whatever the mechanism by which dynein becomes redistributed to aggresomes, our observations

Fig. 6. Misfolded CFTR co-sediments with microtubules in vitro.

A: Detergent extracts of proteasome inhibited GFP-F508 CFTR
expressing HEK cells were incubated with bovine brain tubulin followed by sedimentation through a sucrose cushion. Total detergent
extract (lane 1) Supernatant (sup, lanes 2,4) and pellet fractions (pel,
lanes 3,5) were prepared as described in Materials and Methods and
subjected to SDS-PAGE and immunoblotting with antibodies to CFTR
(top) or dynein intermediate chain 70.1 (bottom). Extracts were
incubated with tubulin under conditions that either prevent (taxol,
lanes 2,3) or promote (taxol, lanes 4,5) MT polymerization. B: Pellet
fractions from incubations described in A were labeled with antibody
to CFTR followed by secondary antibody coupled to 15-nm colloidal
gold. Specimens were prepared for negative stain transmission electron microscopy as described in Materials and Methods. a: Representative micrograph demonstrating specific staining of CFTR antibodies
along MTs. b: dense clusters of MTs do not entrap aggregates of
GFP-508 nonspecifically; c: High-magnification view of GFP-508
aggregates attached to MTs. Bar 100 nm.

have important implications for understanding the mechanism by which aggresome-like inclusion bodies may be
linked to cell death associated with neurodegenerative
disease. Sequestration of dynein in an aggresome-like
inclusion could lead to depletion of the pool of dynein
available for axonal transport; defective axonal transport
has been observed in models of amyotrophic lateral sclerosis [Zhang et al., 1997] and Alzheimers disease
[Burke et al., 1990]. Because dynein and dynactin are
required for transport of MT into axons [Ahmad et al.,
1998], it is possible that the reduction of available pools
of retrograde motors could lead to progressive depletion
of MT from the axon, resulting in further impairment of
axonal transport.
Although it remains to be established whether or
not mislocalization of cytoplasmic dynein to cytoplasmic
inclusion bodies occurs in degenerating neurons, or indeed whether such sequestration is linked to pathogene-

Microtubule Motors and Aggresome Formation

Figure 6.

(Continued.)

37

38

Johnston et al.

sis, the data presented in this study establish a central role

for the dynein-dynactin complex in the formation of
cellular aggresomes. This finding adds the cellular response to protein misfolding and aggregation, the formation of pericentrosomal aggresomes, to the growing list
of functions ascribed to this retrograde motor protein.
ACKNOWLEDGMENTS

We are grateful to R.B. Vallee and C.J. Echeverri for

providing the dynamitin plasmid and antibodies, Tim
Stearns for pericentrin antibody, M.G. Farquhar for antibody to Golgi mannosidase II, and Roger Sloboda for
helpful discussions. This work was supported by a grant
from the National Institutes of Health (DK43994) to R.R.K.
REFERENCES
Ahmad FJ, Echeverri CJ, Vallee RB, Baas PW. 1998. Cytoplasmic
dynein and dynactin are required for the transport of microtubules into the axon. J Cell Biol 140:391 401.
Barton NR, Goldstein LS. 1996. Going mobile: microtubule motors
and chromosome segregation. Proc Natl Acad Sci USA93:
17351742.
Bebok Z, Mazzochi C, King SA, Hong JS, Sorscher EJ. 1998. The
mechanism underlying cystic fibrosis transmembrane conductance regulator transport from the endoplasmic reticulum to the
proteasome includes Sec61beta and a cytosolic, deglycosylated
intermediary. J Biol Chem 273:2987329878.
Bence N, Sampat RM, Kopito RR. 2001. Impairment of the ubiquitinproteasome system by protein aggregation. Science 292:15521555.
Bogyo M, McMaster JS, Gaczynska M, Tortorella D, Goldberg AL,
Ploegh H. 1997. Covalent modification of the active site threonine of proteasomal beta subunits and the Escherichia coli
homolog HslV by a new class of inhibitors. . Proc Natl Acad
Sci USA 94:6629 6634.
Boylan K, Serr M, Hays T. 2000. A molecular genetic analysis of the
interaction between the cytoplasmic dynein intermediate chain and
the glued (dynactin) complex. Mol Biol Cell 11:37913803.
Burke WJ, Park DH, Chung HD, Marshall GL, Haring JH, Joh TH.
1990. Evidence for decreased transport of tryptophan hydroxylase in Alzheimers disease. Brain Res 537:83 87.
Burkhardt JK, Echeverri CJ, Nilsson T, Vallee RB. 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.
Echeverri CJ, Paschal BM, Vaughan KT, Vallee RB. 1996. Molecular
characterization of the 50-kD subunit of dynactin reveals function for the complex in chromosome alignment and spindle
organization during mitosis. J Cell Biol 132:617 633.
Eckley DM, Gill SR, Melkonian KA, Bingham JB, Goodson HV,
Heuser JE, Schroer, TA. 1999. Analysis of dynactin subcomplexes reveals a novel actin-related protein associated with the
arp1 minifilament pointed end. J Cell Biol 147:307320.
Fenteany G, Schreiber SL. 1998. Lactacystin, proteasome function,
and cell fate. J Biol Chem 273:8545 8548.
Garcia-Mata R, Bebok Z, Sorscher EJ, Sztul ES. 1999. Characterization and dynamics of aggresome formation by a cytosolic GFPchimera. J Cell Biol 146:1239 1254.

Gotoh H, Takenaka T, Horie H, Hiramoto Y. 1985. Organelle motility

in rat pituitary clonal cells. I. Dynamic movements of intracellular organelles. Cell Struct Funct 10:233243.
Holleran EA, Karki S, Holzbaur EL. 1998. The role of the dynactin
complex in intracellular motility. Int Rev Cytol 182:69 109.
Jensen TJ, Loo MA, Pind S, Williams DB, Goldberg AL, Riordan JR.
1995. Multiple proteolytic systems, including the proteasome,
contribute to CFTR processing. Cell 83:129 135.
Johnston JA, Ward CL, Kopito RR. 1998. Aggresomes: a cellular
response to misfolded proteins. J Cell Biol143:18831898.
Karki S, Holzbaur EL. 1999. Cytoplasmic dynein and dynactin in cell
division and intracellular transport. Curr Opin Cell Biol 11:4553.
Kim KB. 1999. Proteasome inhibition by the natural products epoxomicin and dihydroeponemycin: insights into specificity and
potency. Bioorg Med Chem Lett 9:33353340.
King SJ, Schroer TA. 2000. Dynactin increases the processivity of the
cytoplasmic dynein motor. Nature Cell Biol 2:20 24.
Lippincott-Schwartz J. 1998. Cytoskeletal proteins and Golgi dynamics. Curr Opin Cell Biol 10:5259.
Matteoni R, Kreis TE. 1987. Translocation and clustering of endosomes and lysosomes depends on microtubules. J Cell Biol
105:12531265.
Mayer RJ, Lowe J, Lennox G, Doherty F, Landon M. 1989. Intermediate filaments and ubiquitin: a new thread in the understanding
of chronic neurodegenerative diseases. Prog Clin Biol Res
317:809 818.
Presley JF, Cole NB, Schroer TA, Hirschberg K, Zaal KJ, LippincottSchwartz J. 1997. ER-to-Golgi transport visualized in living
cells [see comments]. Nature 389:81 85.
Sin N. 1999. Total synthesis of the potent proteasome inhibitor epoxomicin: a useful tool for understanding proteasome biology.
Bioorg Med Chem Lett 9:22832288.
Sodeik B, Ebersold MW, Helenius A. 1997. Microtubule-mediated
transport of incoming herpes simplex virus 1 capsids to the
nucleus. J Cell Biol 136:10071021.
Terasaki M, Chen LB, Fujiwara K. 1986. Microtubules and the endoplasmic reticulum are highly interdependent structures. J Cell
Biol 103:15571568.
Vaughan K, Vallee RB. 1995. Cytoplasmic dynein binds dynactin
through a direct interaction between the intermediate chains
and p150Glued. J Cell Biol 131:15071516.
Velasco A, Hendricks L, Moremen KW, Tulsiani DR, Touster O,
Farquhar MG. 1993. Cell type-dependent variations in the
subcellular distribution of alpha- mannosidase I and II. J Cell
Biol 122:39 51.
Vinitsky A, Michaud C, Powers JC, Orlowski M. 1992. Inhibition of
the chymotrypsin-like activity of the pituitary multicatalytic
proteinase complex. Biochemistry 31:94219428.
Ward CL, Kopito RR. 1994. Intracellular turnover of cystic fibrosis
transmembrane conductance regulator. Inefficient processing
and rapid degradation of wild-type and mutant proteins. J Biol
Chem 269:25710 25718.
Ward CL, Omura S, Kopito RR. 1995. Degradation of CFTR by the
ubiquitin-proteasome pathway. Cell 83:121127.
Waterman-Storer CM, Karki SB, Kuznetsov SA, Tabb JS, Weiss DG,
Langford GM, Holzbaur EL. 1997. The interaction between
cytoplasmic dynein and dynactin is required for fast axonal
transport. Proc Natl Acad Sci USA 94:12180 12185.
Zhang B, Tu P, Abtahian F, Trojanowski JQ, Lee VM. 1997. Neurofilaments and orthograde transport are reduced in ventral root
axons of transgenic mice that express human SOD1 with a
G93A mutation. J Cell Biol 139:13071315.