LYSOSOMAL DEGRADATION OF α-SYNUCLEIN IN VIVO Sally K. Mak1, Alison L. McCormack2 ...

22 Pages·2010·1.88 MB·English

JBC Papers in Press. Published on March 3, 2010 as Manuscript M109.074617
The latest version is at http://www.jbc.org/cgi/doi/10.1074/jbc.M109.074617
1
LYSOSOMAL DEGRADATION OF -SYNUCLEIN IN VIVO
Sally K. Mak1, Alison L. McCormack2, Amy B. Manning-Boğ2, Ana Maria Cuervo3
and Donato A. Di Monte2
The Parkinson’s Institute, Sunnyvale, California 940851, Center for Health Sciences, SRI International,
Menlo Park, California 940252 and Albert Einstein College of Medicine, Bronx, New York 104613
Running head: CMA and -synuclein in vivo
Address correspondence to: Donato A. Di Monte: 333 Ravenswood Avenue, Menlo Park, CA 94025. Tel:
[phone]; E-mail: [email protected]

Pathologic accumulation of -synuclein is a neurodegenerative disorders. The precise
feature of human parkinsonism and other mechanisms by which this endogenously
neurodegenerative diseases. This accumulation expressed protein becomes involved in pathologic
may be counteracted by mechanisms of protein processes have yet to be fully elucidated.
degradation that have been investigated in vitro However, data from both clinical and laboratory
but remain to be elucidated in animal models. studies indicate that enhanced -synuclein
In this study, lysosomal clearance of - expression is itself capable of triggering a
synuclein in vivo was indicated by the detection parkinsonian syndrome in humans and PD-like
D
of -synuclein in the lumen of lysosomes pathology in animal models. Indeed, ow
n
isolated from the mouse midbrain. When multiplication mutations of the -synuclein gene lo
a
neuronal -synuclein expression was enhanced that result in increased expression of the wild-type de
d
as a result of toxic injury (i.e. treatment of mice protein are causally associated with autosomal fro
m
with the herbicide paraquat) or transgenic dominant parkinsonism (1,2). From the h
protein overexpression, the intralysosomal experimental standpoint, evidence of a gain of ttp://w
content of -synuclein was also significantly toxic function of elevated -synuclein includes the w
w
increased. This effect was paralleled by a observation of neurodegeneration and neuronal .jb
c
marked elevation of the lysosome-associated inclusions in the substantia nigra of rats and .o
rg
membrane protein type 2A (LAMP-2A) and the monkeys that overexpress -synuclein after viral- b/
y
lysosomal heat shock cognate protein of 70 kDa mediated neuronal transduction (3,4). g
u
e
(hsc70), two essential components of chaperone- An important corollary to the concept that st o
mediated autophagy (CMA). enhanced -synuclein may have deleterious n J
a
Immunofluorescence microscopy revealed an consequences is that intraneuronal mechanisms of nu
a
increase in punctate (lysosomal) LAMP-2A protein homeostasis, such as degradation ry
1
0
staining that co-localized with -synuclein pathways, could well play a key role in , 2
0
within nigral dopaminergic neurons of maintaining “non-toxic” levels of -synuclein. 19
paraquat-treated and -synuclein Although -synuclein clearance is likely to occur
overexpressing animals. The data provide in through different mechanisms, recent reports have
vivo evidence of lysosomal degradation of - underscored the important contribution of
synuclein under normal conditions and, quite lysosomal pathways of protein degradation and, in
importantly, under conditions of enhanced particular, chaperone-mediated autophagy (CMA)
protein burden. In the latter, increased (5-8). CMA targets specific cytosolic proteins that
lysosomal clearance of-synuclein was are recognized by the heat shock cognate protein
mediated, at least in part, by CMA induction. It of 70 kDa (hsc70) and, after interacting with the
is conceivable that these neuronal mechanisms lysosome-associated membrane protein type 2A
of protein clearance play an important role in (LAMP-2A), are translocated into the lysosomal
neurodegenerative processes characterized by lumen for rapid degradation (9). The amino acid
abnormal -synuclein buildup. sequence of -synuclein contains a pentapeptide
succession consistent with a CMA recognition
Several lines of clinical and experimental evidence motif. Experiments in which purified -synuclein
support a pathogenetic role of -synuclein in was added to intact lysosomes revealed that this
Parkinson’s disease (PD) and other CMA motif is essential for the internalization of
Copyright 2010 by The American Society for Biochemistry and Molecular Biology, Inc.
2
-synuclein into the lysosomal lumen and were sacrificed at day 3 post treatment.
degradation by lysosomal proteases (6). Experimental protocols were in accordance with
Furthermore, inhibition of lysosomal proteolysis the NIH guidelines for use of live animals and
and CMA activity has been found to decrease - were approved by the Animal Care and Use
synuclein clearance in a variety of in vitro Committee at the Parkinson’s Institute.
systems, including primary neuronal cultures (6- Isolation of lysosomes and lysosomal fractions -
8). For each preparation, midbrain tissues from 10
Despite these compelling in vitro data, the mice were pooled and homogenized. Lysosomes
relationship between -synuclein, lysosomes and were isolated from a light mitochondrial-
CMA remains unexplored in the brain in vivo. It is lysosomal fraction in a discontinuous metrizamide
density gradient (12,13). This protocol is
unknown if, in this setting, -synuclein is
optimized to isolate a fraction of lysosomes with
translocated into lysosomes and if enhanced -
high activity for CMA, i.e. lysosomes enriched in
synuclein expression is associated with CMA
the lysosomal chaperone hsc70. Preparations with
induction and increased lysosomal levels of the
more than 10% broken lysosomes, as assessed by
protein. In this study, lysosome--synuclein
-hexosaminidase latency, were discarded.
interactions were investigated in normal mice as
Following isolation, some lysosomal preparations
well as in mice in which increased -synuclein
expression was a consequence of either toxic were treated with proteinase K for 15 min at 0°C, Dow
with subsequent addition of 2 mM n
injury or genetic manipulation. Lysosomes were lo
phenylmethylsulfonyl fluoride (14). Lysosomal ad
isolated from midbrain tissue since neuronal cells e
d
within this region (i.e. dopaminergic neurons in matrices and membranes were isolated by fro
ultracentrifugation after hypotonic shock (15). m
the substantia nigra) are highly vulnerable to h
Lysosomes disrupted by hypotonic shock were ttp
neurodegeneration and -synuclein pathology in also used to test -synuclein proteolysis after ://w
PD. In a first set of experiments, levels of - w
incubation for 5 min at 37 ºC (7). w
synuclein and CMA markers were compared in .jb
Immunoblotting - Analyses were carried out using c
midbrain lysosomes from control mice vs. animals .o
tissue homogenates (post-nuclear fraction), rg
injected with the herbicide paraquat. Paraquat b/
lysosomal preparations (see above) or eluates from y
administration was used as a model of enhanced g
u
immunoprecipitation (see below). Equal amounts e
-synuclein burden because it has previously been of protein were loaded to each gel lane, separated st on
shown to induce a marked upregulation of the J
by SDS-PAGE and transferred to nitrocellulose an
protein within nigrostriatal dopaminergic neurons u
(10,11). In a second set of experiments, the mdreym mbirlakn ein. 1B0lo mtsM w eprheo sbplhoactkee db uwffietrhe d5 %sa lifnaet-,f preHe ary 10
association between elevated -synuclein and 7.4 (PBS) for 1 h and incubated with primary , 20
CMA induction was assessed in transgenic mice 19
antibodies against LAMP-2A (1:3000; Invitrogen),
overexpressing -synuclein. Taken together,
hsc70 (1:4000; Abcam), -synuclein (1:3000; BD
results of these studies provide in vivo evidence of
Bioscience), LAMP-1 (1:20; Developmental
lysosomal clearance of -synuclein that is
Studies Hybridoma Bank), cathepsin B (1:500;
enhanced under conditions of increased protein
Santa Cruz Biotechnology), cytochrome C
burden and is mediated, at least in part, by CMA
(1:1000; Cell Signaling), -synuclein (1:1000;
activation.
Chemicon) or -actin (1:7,500; Sigma) at 4°C

overnight. Blots were then incubated in HRP-
Experimental Procedures
conjugated secondary antibody (Pierce) for 1 h

and proteins were detected using SuperSignal
Animals - Experiments were carried out in 10- to
West Pico Substrate (Pierce). Membranes were
12-week old male C57BL/6 mice (Charles River)
exposed to Hyperfilm ECL, and densitometric
or 10-week old male transgenic mice
quantification was performed with ImageQuant
overexpressing wild-type mouse -synuclein
(GE Healthcare).
(Jackson Laboratory). C57BL/6 mice received a
Immunoprecipitation - For each preparation,
single intraperitoneal injection of either saline or
ventral mesencephalic tissues from 5 mice were
paraquat (Sigma) at the dose of 10 mg/kg and
3
pooled and sonicated in RIPA buffer with antibody conjugated to Cy3 (1:200; Jackson
Complete Mini (Roche). After centrifugation ImmunoResearch), rinsed and coverslipped. For
(100,000 x g for 30 min), supernatants were triple immunolabeling, a separate set of midbrain
collected and pre-cleared in 1% BSA (Sigma) sections were washed and blocked in 5% normal
containing protein A-sepharose beads (Pierce). donkey serum and 5% normal goat serum prior to
Equal protein aliquots were incubated with an immersion in sheep anti--synuclein (1:500;
antibody against hsc70 (1:4000; Abcam). To Millipore) and mouse anti-TH (1:500; Chemicon)
prevent non-specific binding, immune complexes overnight at 4oC. After washing, sections were
(pellets) were washed three times with RIPA incubated in donkey anti-sheep-FITC (1:200;
buffer. Then, immune complexes were eluted and Jackson ImmunoResearch) and goat anti-mouse
separated by SDS gel electrophoresis. Proteins conjugated to Alexa 350 (1:200; Invitrogen).
were transferred to nitrocellulose membrane, and Tissues were washed in PBS and incubated in
blots were blocked and incubated with antibodies rabbit anti-LAMP-2A (1:200; Invitrogen) for 2
against hsc70 or -synuclein. hours. Following rinsing, sections were immersed
Quantitative PCR (qPCR) - Total RNA (0.5 g) in donkey anti-rabbit-Cy3 (1:200; Jackson
was extracted from ventral mesencephalic tissues ImmunoResearch), washed and coverslipped.
using the RNeasy lipid tissue mini kit (Qiagen). Tissues were observed using a Nikon light
D
cDNA was synthesized by reverse transcriptase microscope equipped for epifluorescence. Control o
w
(SuperScript III, Invitrogen) with oligo(dT) sections were incubated in normal IgG in lieu of n
lo
primer. The DNA products of lamp-2a, lamp-2b, primary antibody and were devoid of staining. ad
e
d
lamp-2c, -synuclein and -actin (as an internal Statistical analysis – Data analysis was conducted fro
control) were amplified using the SYBR Green on at least three samples prepared separately from m
h
PCR Master Mix (Applied Biosystems). Primers different sets of animals. Differences among ttp
for LAMP-2A, LAMP-2B, LAMP-2C and -actin means were analyzed using one-way ANOVA. ://w
w
have been previously reported (16). The primers Newman-Keuls post-hoc analysis was used when w.jb
for -synuclein were 5’- differences were observed in ANOVA testing (p c.o
GATCCTGGCAGTGAGGCTTA-3’ and 5’- <0.05). brg/
y
GCTTCAGGCTCATAGTCTTGG -3’. g
Amplification of the DNA products was measured RESULTS ues
t o
in real time by the 7000 Sequence Detection n
J
System (Applied Biosystems). For all genes, the Enhanced lysosomal content of -synuclein an
u
presence of a single amplified product was after paraquat administration to mice. A single ary
verified by agarose gel electrophoresis. intraperitoneal injection of 10 mg/kg paraquat 10
, 2
Immunohistochemistry - Midbrain blocks were causes an increase in -synuclein within 01
9
immersion fixed in 4% paraformaldehyde, dopaminergic neurons of the mouse substantia
cryoprotected in sucrose and frozen. Cryostat-cut nigra that reaches its maximum at 3 days post
coronal sections containing the substantia nigra exposure (10). To assess lysosome--synuclein
(40 µm thick) were processed for fluorescent associations, lysosomes were isolated from the
microscopy. For dual immunolabeling, sections midbrain of mice sacrificed 3 days after a single
were blocked in 5% normal donkey serum intraperitoneal injection of either saline or
followed by incubation in sheep anti--synuclein paraquat. The purity of the lysosomal fractions
(1:600; Millipore) overnight at 4oC. Tissues were isolated from both groups of animals, their
washed in PBS prior to incubation in donkey anti- enrichment in lysosomal markers (LAMP-1 and
sheep IgG conjugated to FITC (1:200; Jackson cathepsin B) and the absence of contaminant
ImmunoResearch). Sections were then washed in mitochondria (as assessed by the cytochrome C
PBS, blocked in 5% normal donkey serum and marker) in the fractions is shown in the
immersed in either rabbit anti-LAMP-2A (1:200; Supplementary Material (Fig. S1). Immunoblot
Invitrogen) or rabbit anti-hsc70 (1:200; Stressgen) analysis revealed the presence of -synuclein
overnight at 4oC. Following washes, tissues were protein in isolated lysosomes from both control
incubated in donkey anti-rabbit secondary and paraquat-treated animals (Fig. 1A). In the
4
latter, denser -synuclein-positive bands indicated consequence of paraquat-induced -synuclein
a significant increase (2 fold) in -synuclein upregulation, levels of the protein were assayed in
content, thus providing in vivo evidence that, as a samples from midbrain homogenates (post-nuclear
consequence of paraquat-induced cytosolic fraction) and lysosomal fractions in parallel. Data
elevation of -synuclein (10), a greater amount of in Fig. 1C show that paraquat administration
the protein is associated with lysosomes. increased -synuclein by 1.8 fold in homogenate
Increased association of -synuclein with samples. Under the same conditions, a much
lysosomes could result from three different greater proportion of the protein was retrieved in
scenarios: (i) enhanced binding and translocation the lysosomal matrix fraction; indeed, the ratio of
of the protein into lysosomes, (ii) decreased lysosomal matrix/homogenate -synuclein was 5.5
proteolysis of the protein once it reaches the times higher in paraquat-treated as compared to
lysosomal lumen, and (iii) reduced ability of the saline-injected mice (Fig. 1C). Taken together,
protein to cross the lysosomal membrane with these findings demonstrate that the association of
consequent accumulation at the membrane. The -synuclein with lysosomes reflects uptake and
next set of experiments was designed to evaluate internalization of the protein into lysosomes.
these possibilities. Two separate approaches were Furthermore, they suggest that -synuclein
used to discriminate between -synuclein bound degradation via lysosomal pathways is markedly
D
to the external surface of lysosomal membranes vs. enhanced when neuronal expression of the protein o
w
n
-synuclein internalized into lysosomes: first, - is increased. lo
a
synuclein levels were assayed in lysosomal To assess whether the increase in lysosomal de
d
preparations treated with proteinase K and then -synuclein caused by paraquat was due to fro
m
levels of the protein were compared in lysosomal accumulation of protease-resistant forms of the h
membrane vs. matrix fractions. Incubation of protein in the lysosomal lumen, the rate of - ttp://w
intact lysosomes with proteinase K digests synuclein degradation was assessed in incubations w
w
proteins bound at the cytosolic side of the of disrupted lysosomes from the midbrain of .jb
c
membrane while preserving proteins contained saline- and paraquat-injected mice (7). .o
rg
within the lysosomal lumen (14). To ensure that Immunoblot analysis compared -synuclein levels b/
y
proteinase K treatment did not digest luminal at time 0 and after a 5-min incubation. At time 0, g
u
proteins, levels of cathepsin B were assayed (Fig. lysosomal -synuclein content was greater in est o
1A). Proteinase K treatment reduced but did not preparations from paraquat- as compared to saline- n J
a
completely eliminate lysosome-associated - treated animals, and at 5 min, protein reduction nu
a
synuclein in preparations from saline- and was 47% in the former vs. 51% in the latter (Fig. ry 1
paraquat-injected mice (Fig. 1A). The extent of 2). Data indicate therefore that paraquat-induced 0, 2
this reduction was 58% and 36% in the former and lysosomal accumulation of -synuclein is not a 019
latter, respectively (data not shown). Of note, consequence of impaired degradation of the
levels of -synuclein recovered after proteinase K protein once it reaches the lysosomal lumen.
digestion were significantly higher (4 fold) in the Because the difference in -synuclein levels at
lysosomes isolated from paraquat-treated animals time 0 vs 5 min was more pronounced in samples
(Fig. 1A). from paraquat-injected mice (Fig. 2), data are also
In the second set of experiments, midbrain consistent with enhanced lysosomal clearance of
lysosomes were subjected to hypotonic shock and, the protein after toxicant administration.
after centrifugation, membrane and matrix LAMP-2A upregulation in the midbrain of
fractions were evaluated for -synuclein content. paraquat-treated mice. CMA is a pathway of
A comparison of protein levels in lysosomal lysosomal protein degradation that has been shown
fractions isolated from saline- vs. paraquat-treated to contribute to -synuclein clearance in a variety
mice revealed a marked increase (> 6 fold) in - of in vitro cell systems (6-8). We therefore tested
synuclein in the matrix of lysosomes from animals if paraquat-induced elevation of lysosomal -
injected with the herbicide (Fig. 1B). To further synuclein was accompanied by an increase in
evaluate how the relative amounts of extra- and levels of the lysosomal receptor LAMP-2A.
intralysosomal -synuclein changed as a Binding of cytosolic substrates to LAMP-2A is the
5
rate limiting step in CMA, and a direct correlation Paraquat-induced increase in lysosomal hsc70
between levels of LAMP-2A at the lysosomal and hsc70/-synuclein interactions. LAMP-2A
membrane and CMA activity has been previously upregulation, enhanced binding of substrates to
established (17,18). LAMP-2A is one of the three this receptor and increased protein translocation
spliced variants of a single gene, lamp-2; the other across the lysosomal membrane are often
two protein variants coded by this gene, LAMP- associated with increased levels of the lysosomal
2B and C, are not directly implicated in CMA chaperone hsc70 (12,21,22). In the next set of
(17,19). For this reason, our experiments were experiments, the content of hsc70 was compared
carried out using a specific antibody that in intact lysosomes isolated from the midbrain of
recognizes LAMP-2A but not LAMP-2B and C control vs. paraquat-injected mice. Data revealed
(20). When LAMP-2A levels were compared in that levels of this chaperone were 2.5-fold higher
the membrane fraction of lysosomes from the as a consequence of toxicant administration (Fig.
midbrain of saline- vs. paraquat-injected mice, 6A). This finding, together with the observation of
they were found to be significantly enhanced (4 LAMP-2A elevation, demonstrates that CMA
fold) as a consequence of toxicant exposure (Fig. activity is upregulated in the midbrain of mice
3A). To determine if this effect was due to challenged with paraquat.
transcriptional upregulation, mRNA levels of If CMA upregulation plays a role in the
LAMP-2A, LAMP-2B and LAMP-2C were clearance of excessive -synuclein after paraquat Do
w
assayed by qPCR in homogenates from the mouse exposure, one would expect augmented n
lo
ventral mesencephalon. The results showed no interactions between -synuclein and cytosolic ad
e
d
change between saline- and paraquat-injected mice hsc70, the chaperone responsible for recognizing fro
in LAMP-2C, a small (not statistically significant) CMA substrates and delivering them to the m
h
increase in LAMP-2B and a robust elevation (6 lysosomal membrane (20,23). To assess hsc70/- ttp
fold) of LAMP-2A (Fig. 3B). This significant synuclein interactions, cytosolic fractions of the ://w
w
eallseov atoibosne rovfe dL AwMhePn- 2tAhe claaumsepd- 2bay trpaanrsacqruipatt wwaass mimomusuen oprecvipeinttartaeld withm aense hnscce7p0h aalonnti body,w aenrde w.jbc.o
amplified by PCR and separated by agarose gel eluates were then assayed for both hsc70 and - brg/
electrophoresis (Fig. 3C). y
synuclein. Immunoprecipitation conditions were g
Further evidence of a relationship between adjusted to yield similar levels of hsc70 in tissue ues
paraquat exposure, -synuclein upregulation and t o
eluates from control and paraquat-exposed mice n
increased clearance of this protein via CMA was Ja
(Fig. 6B). In agreement with previous findings in n
u
provided by experiments in which midbrain cultured cells (6), a portion of cytosolic - ary
sweecrteio dnos ufbrloem-im cmonutnroosl taainnded pwariathq ua-ts-ytrneuacteledi nm aincde siny ntuiscslueein ewluaast erse cforvoemre dc obnotruonld atnoi mcyatlos so(Fliicg .h s6cB7)0. 10, 201
9
LAMP-2A antibodies. Immunoreactivity for both
To evaluate specificity of this interaction, eluates
-synuclein and LAMP-2A was enhanced within
were also assayed with an antibody against -
nigral neurons in sections from paraquat-injected
synuclein. Immunoblots probed with this antibody
animals (Figs. 4A vs. 4D and 4B vs. 4E), with
showed no immunoreactivity (data not shown).
substantial co-localization of the two proteins (Fig.
As compared to control mice, levels of -
4F). -Synuclein and LAMP-2A antibodies
synuclein were higher in immunoprecipitates from
stained neurons in a punctate pattern, most likely
animals injected with paraquat (Fig. 6B),
reflecting the lysosomal accumulation of -
consistent with the interpretation that increased -
synuclein that co-localized with the CMA marker
synuclein expression resulted in enhanced binding
(Figs. 4G-I). Additional midbrain sections were
to hsc70 and augmented clearance of the protein
triple-immunostained with antibodies against TH,
via CMA.
LAMP-2A and -synuclein. In sections from CMA activation in transgenic mice
paraquat-treated mice, increased immunoreactivity
overexpressing -synuclein. The relationship
for LAMP-2A and -synuclein was observed
between increased expression of -synuclein and
within neurons that also stained for TH, indicating
CMA activation was further investigated in
their dopaminergic phenotype (Fig. 5).
transgenic mice overexpressing mouse -
6
synuclein under the regulatory control of the Thy- labeling for hsc70 in transgenic mice (Figs. 10D
1 promoter. Earlier work has shown that these vs. 10A and 10E vs. 10B). Significant co-
transgenics do not develop overt localization of the two proteins was also observed
neurodegenerative changes (24). Levels of - within nigral neurons (Fig. 10F).
synuclein mRNA were determined in the ventral
mesencephalon, hippocampus and frontal cortex of
DISCUSSION
these animals and found to be elevated by 23, 16
and 8 times, respectively, as compared to values in
the same brain regions of control littermates (Figs. Results of this study provide evidence of
7A-C). -Synuclein transgenics also displayed a lysosomal degradation of -synuclein in vivo by
marked increase in LAMP-2A but not LAMP-2B showing internalization of the protein into
mRNA (Figs. 7A-C). Interestingly, the extent of lysosomes from the midbrain of control mice and
LAMP-2A upregulation was correlated with the mice injected with paraquat. In the latter, selective
degree of -synuclein overexpression throughout injury of nigrostriatal dopaminergic neurons is
the brain; LAMP-2A levels were increased by 18 accompanied by an increased expression of -
fold in the ventral mesencephalon, 10 fold in the synuclein (10,25). Of note, levels of
hippocampus and 3 fold in the frontal cortex (Fig. intralysosomal -synuclein were also markedly
D
7D). enhanced after paraquat exposure, suggesting a o
w
Midbrain sections from transgenic and control relationship between cytosolic upregulation and nlo
a
animals were immunostained with both - lysosomal translocation of the protein. One de
d
synuclein and LAMP-2A antibodies. As expected, possible explanation for the increased lysosomal fro
m
-synuclein immunoreactivity was more robust -synuclein after paraquat exposure may be the h
within nigral neurons in samples from formation of post-translationally modified forms ttp
overexpressing mice as compared to non- of the protein resistant to degradation by ://w
w
transgenic animals (Figs. 8D vs. 8A). In the same lysosomal proteases. In particular, formation of w.jb
tissue sections, labeling for LAMP-2A was also oxidized -synuclein could be hypothesized under c.o
more pronounced (Figs. 8E vs. 8B) and co- our experimental conditions given the pro-oxidant brg/
localized with -synuclein immunoreactivity (Fig. effects of paraquat on nigral dopaminergic neurons y g
u
8F), supporting a relationship between - (26,27). A recent report, however, has es
t o
synuclein overexpression and CMA activation. In demonstrated that the rate of lysosomal clearance n J
a separate set of tissue sections, robust LAMP-2A is not significantly different between unmodified anu
a
and -synuclein staining within dopaminergic and oxidized -synuclein, making it unlikely that ry
1
neurons of transgenic animals was confirmed by paraquat would cause a lysosomal accumulation of 0, 2
fluorescence microscopy using antibodies against protease-resistant forms of the protein (7). 01
9
both proteins and TH (Figs. 8G-J). Furthermore, we found that approximately 50% of
LAMP-2A changes as a consequence of - -synuclein was degraded during a 5 min-
synuclein overexpression were also assessed at the incubation of lysosomal matrix extracts, regardless
protein level by western blot analysis. Because of of whether they were obtained from control or
limited tissue availability, measurements were not paraquat-treated mice. Thus, lysosomal
carried out in the substantia nigra. Rather, LAMP- accumulation of -synuclein following treatment
2A protein levels were compared in homogenates with the herbicide cannot simply be explained by
from the frontal cortex of control vs. transgenic reduced clearance capability. Rather, it reflects an
mice. In the latter, in which cortical LAMP-2A increase in lysosomal uptake triggered by the
mRNA was enhanced by 3 fold (see above), cytosolic buildup of the protein.
protein levels were also 2.8-fold higher (Fig 9). To elucidate mechanisms involved in the
To evaluate changes in hsc70 that may arise enhanced translocation of -synuclein into
from -synuclein overexpression, nigral tissues lysosomes after paraquat administration, we
were dual-labeled with antibodies against these assessed changes in lysosomal hsc70 and LAMP-
two proteins. Increased -synuclein 2A as markers of CMA induction. The rationale
immunoreactivity was paralleled by enhanced for investigating the relationship between -
7
synuclein clearance and CMA in the paraquat lysosomal -synuclein and CMA activation in the
model was twofold. First, earlier in vitro data are paraquat model are compatible with but do not
consistent with -synuclein being a CMA necessarily indicate a direct link between these
substrate and with CMA playing an important role events. In particular, they do not address the issue
in preventing neuronal accumulation of the protein of whether -synuclein accumulation is itself a
(6-8). Second, paraquat administration has condition leading to CMA induction. To further
previously been shown to upregulate CMA investigate this possibility, a final set of
activity in rat liver and cultured fibroblasts experiments was carried out in mice in which
(16,22). Our current findings provide evidence in increased -synuclein levels resulted from
favor of a contribution of CMA to -synuclein transgenic overexpression of the protein. Data
degradation in the midbrain of paraquat-exposed revealed a robust upregulation of LAMP-2A in the
mice. In these animals, higher levels of lysosomal brain of transgenic animals and supported a direct
hsc70 and LAMP-2A paralleled the enhanced correlation between levels of -synuclein
cytosolic expression and lysosomal internalization overexpression and CMA activation. Indeed, the
of -synuclein. Moreover, a punctate -synuclein increase in LAMP-2A was greatest in the brain
and LAMP-2A co-immunoreactivity most likely region with the highest -synuclein level (ventral
reflected the lysosomal localization of both these mesencephalon) and smallest in the area with the
D
proteins within nigral neurons. Finally, binding of lowest protein expression (frontal cortex). CMA o
w
-synuclein to cytosolic hsc70 was augmented in induction as a consequence of transgene nlo
a
the midbrain of paraquat-treated mice, further expression was also indicated by enhanced hsc70 de
d
supporting a relationship between CMA induction immunoreactivity within nigral neurons laden with fro
m
and increased lysosomal clearance of-synuclein. -synuclein. h
upreQguuliateti on ofi nLteAreMstPin-g2lAy , was apcahriaeqvueadt -itnhdrouucgehd of wAilsd -styupgeg eosrt emdu btayt eeda rli-esry nsutucdleieins, mthaey cinlevaorlavnec ea ttp://ww
w
de novo synthesis, as indicated by the dramatic variety of mechanisms (5,6,8,28-31). The present .jb
elevation of LAMP-2A mRNA in the ventral findings do not rule out this possibility but clearly c.o
mesencephalon of mice injected with the reveal the importance of lysosomal pathways and, brg/
y
herbicide. At least two mechanisms have been in particular, CMA in the neuronal degradation of g
u
described to increase LAMP-2A under conditions wild-type -synuclein in vivo. The observation of est o
associated with CMA induction. During nutritional CMA induction and enhanced lysosomal n J
a
stress, the most extensively studied inducer of internalization of -synuclein under conditions of nu
a
CMA, enhanced LAMP-2A results from a increased protein expression (i.e. paraquat-induced ry 1
decrease in its degradation as well as a relocation 0
upregulation and transgenic overexpression of - , 2
of the receptor protein from the lysosomal lumen synuclein) bears significant implications. Both 019
to the lysosomal membrane (18). In contrast,
clinical and experimental data suggest that higher
LAMP-2A changes associated with CMA
levels of neuronal -synuclein may trigger or
activation during oxidative stress were found to be
predispose to pathologic consequences, including
due to an actual increase in LAMP-2A synthesis
aggregate formation and neurodegeneration
(22). This latter mechanism could explain our
(2,4,32). If so, induction of lysosomal degradation
present data: the oxidative damage caused by
pathways may represent an important mechanism
paraquat administration would induce new
that counteracts -synuclein-related pathology.
synthesis of LAMP-2A as a critical CMA factor
Failure of this mechanism, which may arise from
involved in the lysosomal degradation of -
normal aging (19), could play a role in the
synuclein and other toxic/damaged proteins.
pathogenesis of PD and other synucleinopathies,
A causal relationship between intraneuronal
whereas enhancement of these pathways could
-synuclein levels, CMA induction and enhanced
become a valuable strategy for therapeutic
translocation of -synuclein into lysosomes may
intervention in these human diseases.
be difficult to demonstrate in the in vivo setting. It
could be argued that the findings of increased
8
REFERENCES

1. Singleton, A.B., Farrer, M., Johnson, J., Singleton, A., Hague, S., Kachergus, J., Hulihan, M.,
Peuralinna, T., Dutra, A., Nussbaum, R., Lincoln, S., Crawley, A., Hanson, M., Maraganore, D.,
Adler, C., Cookson, M.R., Muenter, M., Baptista, M., Miller, D., Blancato, J., Hardy, J., and Gwinn-
Hardy, K. (2003) Science 302, 841
2. Farrer, M., Kachergus, J., Forno, L., Lincoln, S., Wang, D.S., Hulihan, M., Maraganore, D., Gwinn-
Hardy, K., Wszolek, Z., Dickson, D., and Langston, J.W. (2004) Ann. Neurol. 55, 174-179
3. Lo Bianco, C., Ridet, J.L., Schneider, B.L., Deglon, N., and Aebischer, P. (2002) Proc. Natl. Acad.
Sci. USA 99, 10813-10818
4. Eslamboli, A., Romero-Ramos, M., Burger, C., Bjorklund, T., Muzyczka, N., Mandel, R.J., Baker,
H., Ridley, R.M., and Kirik, D. (2007) Brain 130, 799-815
5. Webb, J.L., Ravikumar, B., Atkins, J., Skepper, J.N., and Rubinsztein, D.C. (2003) J.Biol.Chem.
278, 25009-25013
6. Cuervo, A.M., Stefanis, L., Fredenburg, R., Lansbury, P.T., and Sulzer, D. (2004) Science 305,
1292-1295
7. Martinez-Vicente, M., Talloczy, Z., Kaushik, S., Massey, A.C., Mazzulli, J., Mosharov, E.V.,
D
Hodara, R., Fredenburg, R., Wu, D.C., Follenzi, A., Dauer, W., Przedborski, S., Ischiropoulos, H., ow
n
Lansbury, P.T., Sulzer, D., and Cuervo A.M. (2008) J. Clin. Invest. 118, 777-788 lo
a
d
8. Vogiatzi, T., Xilouri, M., Vekrellis, K., and Stefanis, L. (2008) J. Biol. Chem. 283, 23542-23556 e
d
9. Massey, A.C., Zhang, C., and Cuervo, A.M. (2006) Curr. Top. Dev. Biol. 73, 205-235 fro
m
10. Manning-Boğ, A.B., McCormack, A.L., Li, J., Uversky, V.N., Fink, A.L., and Di Monte, D.A. h
(2002) J. Biol. Chem. 277, 1641-1644 ttp://w
11. Manning-Boğ, A.B., McCormack, A.L., Purisai, M.G., Bolin, L.M., and Di Monte, D.A. (2003) J. w
w
Neurosci. 23, 3095-3099 .jb
c
12. Cuervo, A.M., Dice, J.F., and Knecht, E. (1997) J. Biol. Chem. 272, 5606-5615 .o
rg
13. Wattiaux, R., Wattiaux-De Coninck, S., Ronveaux-Dupal, M.F., and Dubois, F. (1978) J. Cell Biol. b/
y
78, 349-368 g
u
e
14. Aniento, F., Roche, E., Cuervo, A.M., and Knecht, E. (1993) J. Biol. Chem. 268, 10463-10470 st o
15. Ohsumi, Y., Ishikawa, T., and Kato, K. (1983) J. Biochem. 93, 547-556 n J
a
16. Massey, A.C., Kaushik, S., Sovak, G., Kiffin, R., and Cuervo, A.M. (2006) Proc. Natl. Acad. Sci. nu
a
USA 103, 5805-5810 ry
1
17. Cuervo, A.M., and Dice, J.F. (2000) J. Cell Sci. 113, 4441-4450 0, 2
0
18. Cuervo, A.M., and Dice, J.F. (2000) Traffic 1, 570-583 1
9
19. Zhang, C., and Cuervo, A.M. (2008) Nat. Med. 14, 959-965
20. Cuervo, A.M., and Dice, J.F. (1996) Science 273, 501-503
21. Cuervo, A.M., Knecht, E., Terlecky, S.R., and Dice, J.F. (1995) Am. J. Physiol. 269, C1200-1208
22. Kiffin, R., Christian, C., Knecht, E., and Cuervo, A.M. (2004) Mol. Biol. Cell 15, 4829-4840
23. Chiang, H.L., Terlecky, S.R., Plant, C.P., and Dice, J.F. (1989) Science 246, 382-385
24. Chandra, S., Gallardo, G., Fernandez-Chacon, R., Schluter, O.M., and Sudof, T.C. (2005) Cell 123,
383-396
25. McCormack, A.L., Thiruchelvam, M., Manning-Boğ, A.B., Thiffault, C., Langston, J.W., Cory-
Slechta, D.A., and Di Monte, D.A. (2002) Neurobiol. Dis. 10, 119-127
26. McCormack, A.L., Atienza, J.G., Johnston, L.C., Andersen, J.K., Vu, S., and Di Monte, D.A. (2005)
J. Neurochem. 93, 1030-1037
27. McCormack, A.L., Atienza, J.G., Langston, J.W., and Di Monte, D.A. (2006) Neuroscience 141,
929-937
28. Bennett, M.C., Bishop, J.F., Leng, Y., Chock, P.B., Chase, T.N., and Mouradian, M.M. (1999) J.
Biol. Chem. 274, 33855-33858
29. Lee, H.J., Khoshaghideh, F., Patel, S., and Lee, S.J. (2004) J. Neurosci. 24, 1888-1896
9
30. Sarkar, S., Davies, J.E., Huang, Z., Tunnacliffe, A., and Rubinsztein, D.C. (2007) J. Biol. Chem.
282, 5641-5652
31. Yu, W.H., Dorado, B., Figueroa, H.Y., Wang, L., Planel, E., Cookson, M.R., Clark, L.N., and Duff,
K.E. (2009) Am. J. Pathol. 175, 736-747
32. McCormack, A.L., Mak, S.K., Shenasa, M., Langston, W.J., Forno, L.S., and Di Monte, D.A. (2008)
J. Neuropathol. Exp. Neurol. 67, 793-802

FOOTNOTES

These studies were supported by the National Institutes of Health (ES12077), the Backus Foundation and
the Parkinson’s Disease Foundation.

The abbreviations used are: LAMP, lysosome-associated membrane protein; hsc70, heat shock cognate
protein of 70 kDa; PD, Parkinson’s disease; CMA, chaperone-mediated autophagy; PBS, phosphate-
buffered saline; qPCR, quantitative PCR; Thy-1, thymus cell antigen 1; PK, proteinase K; PQ, paraquat.

FIGURE LEGENDS
D
o
w
Figure 1. -Synuclein association with lysosomes is enhanced in the midbrain of paraquat-treated n
lo
mice. A, Lysosomes were isolated from the midbrain of animals sacrificed 3 days after a single ad
e
d
intraperitoneal injection of either saline or 10 mg/kg paraquat (PQ). After isolation, they were untreated fro
or incubated with proteinase K (PK). B, Lysosomes from the midbrain of saline- and paraquat-treated m
h
mice were subjected to hypotonic shock and ultracentrifugation for separation of lysosomal matrices and ttp
membranes. A & B, The -synuclein content was measured by western blot analysis. To ensure equal ://w
w
loading of the gels, immunoblots were probed with an antibody against LAMP-1 (a lysosomal w
.jb
transmembrane protein) or cathepsin B (a marker of lysosomal matrix). Results of the densitometric c
.o
quantification are the means + SEM (n=3) and are expressed as folds of the control (con) value in saline- rg
b/
treated mice. C, Homogenate (post-nuclear fraction) and lysosomal (membrane and matrix fractions) y
g
u
samples (10 g) from the midbrain of saline- and PQ-injected mice were assayed in parallel by western e
s
blotting using an anti--synuclein antibody. A representative blot is shown. Results of the densitometric t on
J
quantification are the means + SEM (n=3) and are expressed as the ratio of lysosomal matrix/homogenate an
u
a
-synuclein in control and PQ-treated animals, respectively. *p<0.01 versus the corresponding control ry
1
group. 0
, 2
01
9
Figure 2. Lysosomal degradation of -synuclein. Levels of -synuclein were measured by western blot
analysis at time 0 and after a 5-min incubation of lysosomes from the midbrain of saline- and paraquat
(PQ)-injected mice. After isolation, lysosomes were subjected to hypotonic shock. Results of the
densitometric quantification are the means + SEM of triplicate samples and are expressed as folds of the
control (con) value at time 0 in saline-treated mice. Differences in -synuclein levels (means + SEM)
were calculated by subtracting values at 5 min from the corresponding values at time 0 in preparations
from control (saline-injected) and paraquat-treated animals. *p<0.005 versus the corresponding control
group.

Figure 3. Paraquat-induced LAMP-2A upregulation. A, Lysosomal membranes were isolated from the
midbrain of mice sacrificed 3 days after a single injection of either saline or paraquat (PQ). Levels of
LAMP-2A were measured by western blot analysis. Blots were probed with an antibody against LAMP-1
to ensure equal loading. B, mRNA levels of LAMP-2A, LAMP-2B and LAMP-2C were assayed by
qPCR in homogenates from the ventral mesencephalon of saline and paraquat-treated mice. A & B,
Results in the graphs are the means + SEM of triplicate samples and are expressed as folds of the
corresponding control value in saline-treated mice. *p<0.01 versus the corresponding control group. C,
10
The lamp-2a transcript was amplified by PCR and separated by agarose gel electrophoresis in samples
from the ventral mesencephalon of saline and paraquat-injected mice.

Figure 4. Enhanced LAMP-2A immunoreactivity and LAMP-2A/-synuclein co-localization within
nigral neurons from paraquat-treated mice. A-F, Midbrain sections from saline- (A-C) and paraquat-
(D-F) injected mice were immunostained with antibodies against -synuclein (A & D) and LAMP-2A (B
& E). Merged images are shown in C & F. Scale bar, 20 m. G-I, Higher magnification images of a
representative neuron from the substantia nigra of a paraquat-treated mouse show a punctate pattern of
immunostaining for both -synuclein (G) and LAMP-2A (H) with substantial co-localization of the two
proteins (I). Scale bar equals 10 m.

Figure 5. Paraquat-induced increase in LAMP-2A and -synuclein immunolabeling within
dopaminergic neurons. A-H, Midbrain sections from saline- (A-D) and paraquat- (E-H) injected mice
were triple-immunolabeled with antibodies against TH (A & E), -synuclein (B & F) and LAMP-2A (C
& G). Merged images of -synuclein and Lamp-2A immunoreactivity are shown in D & H.
Representative images show that, after paraquat treatment, labeling for -synuclein and Lamp-2A was
increased within dopaminergic (TH-positive) neurons. Scale bar equals 10 m. D
o
w
n
Figure 6. Paraquat-induced changes in hsc70. A, Lysosomes were isolated from the midbrain of loa
d
animals sacrificed 3 days after a single intraperitoneal injection of saline or paraquat (PQ). Levels of ed
hsc70 were measured by western blot analysis. Blots were probed with an antibody against LAMP-1 to fro
m
ensure equal loading. B, Cytosolic fractions of the mouse ventral mesencephalon were h
ttp
immunoprecipitated with an hsc70 antibody, and eluates were then assayed for hsc70 and -synuclein by ://w
western blot analysis. Immunoprecipitation conditions were adjusted to yield similar levels of hsc70 in w
w
tissue eluates from control and paraquat-exposed mice. Results of the densitometric quantification are the .jb
c
means + SEM of triplicate samples and are expressed as folds of the control (con) value in saline-treated .org
mice. *p<0.01 versus the corresponding control group. b/
y
gu
e
s
Figure 7. Increase in LAMP-2A mRNA in transgenic mice overexpressing -synuclein. A-C, mRNA t o
n
levels of -synuclein (-syn), LAMP-2A and LAMP-2B were assayed by qPCR in homogenates from the Ja
n
ventral mesencephalon (A), hippocampus (B) and frontal cortex (C) of control and transgenic mice. Data ua
ry
are the means + SEM of quadruplicate samples and are expressed as folds of the corresponding control 1
0
value. *p<0.0001 versus the corresponding control group. D, Correlation plot for -synuclein and LAMP- , 2
0
1
2A mRNAs in the three brain regions of the transgenic mice. R2=0.991. 9

Figure 8. Enhanced LAMP-2A labeling in the substantia nigra of -synuclein overexpressing mice.
A-F, Midbrain sections from control (A-C) and transgenic (D-F) mice were immunostained with
antibodies against -synuclein (A & D) and LAMP-2A (B & E). Representative images show increased
labeling for both -synuclein and LAMP-2A in transgenics and substantial co-localization of the two
proteins (C & F). G-J, A second set of midbrain sections from transgenic animals was triple-labeled with
antibodies recognizing TH (G), -synuclein (H), and LAMP-2A (I). Merged image of -synuclein and
Lamp-2A immunoreactivity is shown in J. Scale bars equal 15 m.

Figure 9. Increase in cortical LAMP-2A in -synuclein overexpressing mice. Levels of LAMP-2A
were measured by western blot analysis in homogenates from the frontal cortex of control and transgenic
mice. Blots were probed with an antibody against -actin to ensure equal loading. Results of the
densitometric quantification are the means + SEM of triplicate samples and are expressed as folds of the
value in control mice. *p<0.001 versus the control group.