by the Autophagic/Lysosomal and Proteasomal
Pathways in Atrophying Muscle Cells
Jinghui Zhao,1Jeffrey J. Brault,1Andreas Schild,1Peirang Cao,2Marco Sandri,3,4,5Stefano Schiaffino,3,4,6
Stewart H. Lecker,2and Alfred L. Goldberg1,*
1Department of Cell Biology, Harvard Medical School, Boston, MA 02115, USA
2Renal Unit, Beth Israel Deaconess Medical Center and Harvard Medical School, Boston, MA 02115, USA
3Venetian Institute of Molecular Medicine, 35129 Padova, Italy
4Department of Biomedical Sciences, University of Padova, 35121 Padova, Italy
5Dulbecco Telethon Institute, 35129 Padova, Italy
6Institute of Neuroscience, Consiglio Nazionale delle Ricerche, 35121 Padova, Italy
Muscle atrophy occurs in many pathological
states and results primarily from accelerated
protein degradation and activation of the ubiq-
uitin-proteasome pathway. However, the im-
portance of lysosomes in muscle atrophy has
scription factors is essential for the atrophy in-
duced by denervation or fasting, and activated
FoxO3 by itself causes marked atrophy of mus-
cles and myotubes. Here, we report that FoxO3
does so by stimulating overall protein degrada-
tion and coordinately activating both lysosomal
and proteasomal pathways. Surprisingly, in
C2C12 myotubes, most of this increased pro-
teolysis is mediated by lysosomes. Activated
cle (and other cell types) by activating autoph-
agy. FoxO3 also induces the expression of
many autophagy-related genes, which are in-
duced similarly in mouse muscles atrophying
due to denervation or fasting. These studies in-
dicate that decreased IGF-1-PI3K-Akt signaling
activates autophagy not only through mTOR
but also more slowly by a transcription-depen-
dent mechanism involving FoxO3.
Muscle atrophyisadebilitating processthatleadstorapid
loss of strength and endurance. It occurs in specific mus-
cles with inactivity and denervation and systematically in
fasting and many diseases, including cancer, diabetes,
sepsis, and renal failure (Kandarian and Jackman, 2006;
Lecker et al., 2006). In these various conditions, the rapid
loss of muscle mass occurs primarily through an activa-
tion of protein breakdown. We have previously identified
a set of atrophy-specific genes or ‘‘atrogenes’’ that are
up- or downregulated similarly in muscles in these diverse
These observations indicate that common transcriptional
adaptations occur in various types of atrophy leading to
accelerated protein degradation.
Proteasomes and lysosomes comprise the two major
intracellular proteolytic systems in mammalian cells and
have generally been assumed to be regulated indepen-
dently and to serve distinct functions. The ubiquitin-pro-
teasome pathway degrades both cytosolic and nuclear
proteins (Glickman and Ciechanover, 2002), as well as
comprise most of the protein in adult skeletal muscle. The
acid hydrolases in lysosomes degrade most membrane
and extracellular proteins taken up by endocytosis, as
well as cytoplasmic proteins and organelles through au-
tophagy (Scott and Klionsky, 1998). In diverse types of
muscle wasting, the ubiquitin-proteasome pathway is ac-
tivated, as shown by increased sensitivity to proteasome
inhibitors; increased levels of ubiquitin conjugates; en-
hanced rates of ubiquitin conjugation; and induction of
genes for ubiquitin, several proteasomal subunits, and
two critical ubiquitin ligases (E3s), atrogin-1/MAFbx and
MuRF1 (Bodine et al., 2001; Lecker et al., 2004). Induction
of these muscle-specific E3s is essential for rapid atrophy
(Bodineetal., 2001). Because inhibitorsof lysosomalfunc-
in atrophying muscles (Furuno et al., 1990), the possible
contributions of lysosomes to atrophy have not attracted
much attention. However, an increased capacity for lyso-
somal proteolysis has been demonstrated in various types
of atrophy (Bechet et al., 2005; Furuno et al., 1990).
Our prior work showed that activation of FoxO tran-
scription factors is essential for fiber atrophy and atro-
gin-1 induction upon denervation, fasting, and glucocorti-
coid treatment (Sandri et al., 2004, 2006). Moreover,
expression of constitutively active FoxO3 (ca-FoxO3)
induces expression of multiple atrogenes, including the
critical E3 atrogin-1, and causes dramatic atrophy of
mouse muscles and myotubes (Sandri et al., 2004). The
472 Cell Metabolism 6, 472–483, December 2007 ª2007 Elsevier Inc.
to their inactivation through binding to 14-3-3 proteins in
the cytosol (Tran et al., 2003), but in ca-FoxO3, the three
Akt phosphorylation sites are mutated to alanines, allow-
ing free entry into the nucleus. Through activation of the
PI3K-Akt pathway, IGF-1 stimulates protein synthesis
and can induce hypertrophy of skeletal muscle (Glass,
2005). In addition, IGF-1/insulin inhibits overall protein
breakdown, degradation of myofibrillar proteins (Sacheck
et al., 2004), and expression of atrogin-1 and MuRF1
(Sacheck et al., 2004; Stitt et al., 2004). This reduced pro-
teolysis appears to contribute to muscle growth and to re-
sult from the inactivation of FoxO by Akt phosphorylation
(Sacheck et al., 2004; Stitt et al., 2004), while the acceler-
Sandri et al., 2004). The present studies were undertaken
to analyze how FoxO3 affects protein degradation and to
assess the contributions of lysosomal and proteasomal
pathways to the loss of muscle protein during atrophy.
We demonstrate here a marked ability of FoxO3 to stimu-
late lysosomal proteolysis by activating autophagy via
a transcriptional mechanism in muscle cells (as well as
is activated coordinately with the proteasomal pathway
and contributes importantly to atrophy.
FoxO3 Causes Atrophy by Stimulating Proteolysis
To determine the effects of FoxO3 on protein degrada-
tion, we infected C2C12 myotubes with an adenovirus
expressing ca-FoxO3, which within 2 days caused a dra-
matic reduction in myotube diameter (Figure 1A) and
a reduction in the protein:DNA ratio (data not shown) to
a level below that observed in control cells infected
with an adenovirus expressing GFP only. Thus, viral ex-
pression of ca-FoxO3 in myotubes mimics the effects
of FoxO3 overexpression in adult muscle (Sandri et al.,
To measure overall rates of protein degradation, the
majority of the myotube proteins were labeled by a 20 hr
exposure to [3H]tyrosine prior to viral infection. After infec-
tion, overall rates of protein degradation were determined
by measuring [3H]tyrosine release from the prelabeled
proteins into the medium, which contained a large excess
of nonradioactive tyrosine to prevent reincorporation
(Gronostajski et al., 1984). After 16 hr exposure to the
ca-FoxO3-expressing adenovirus, the overall rate of pro-
tein degradation in the myotubes was significantly in-
creased (Figure 1B, p < 0.01), and by 24 hr, the rate was
at least 50% greater than in cells infected with control
adenovirus (Figure 1B and Figure 2A). If maintained, this
enhancement of proteolysis is sufficient to cause marked
atrophy in 2 days. When expressed at a level similar to
of proteolysis by only 15% (Figure 1B, p < 0.01), presum-
ably because most of the wt-FoxO3 was phosphorylated
FoxO3-Induced Increase in Proteolysis
in Myotubes Is Mediated Mainly by Lysosomes
To determine to what extent proteasomes contribute to
FoxO3-induced proteolysis, we used two specific protea-
some inhibitors, Velcade (PS-341) and lactacystin (Kisse-
lev and Goldberg, 2001). To evaluate the contributions of
lysosomes, we used a specific inhibitor of the lysosomal
proton pump, concanamycin A (Woo et al., 1992), or chlo-
roquine and NH4Cl, which accumulate in lysosomes and
raise intralysosomal pH (Seglen, 1983). At concentrations
used, the inhibitors caused maximal inhibition of proteoly-
sis, and residual rates of degradation were linear for up to
5 hr with no detectable cell death. The actual contribution
of proteasomes or lysosomes was then determined as the
somal inhibitors, by subtracting the rates of proteolysis in
Figure 1. FoxO3 Induces Atrophy and
(A) Constitutively active FoxO3 (ca-FoxO3) in-
duces marked atrophy. Myotubes were infected
with control (top panel, expressing only GFP) or
ca-FoxO3 (bottom panel, expressing ca-FoxO3
and GFP) adenoviruses. Cells expressing GFP
are shown 48 hr after infection.
(B) ca-FoxO3 stimulates proteolysis. Myotubes
were incubated with [3H]tyrosine for 20 hr and
then washed with chase medium for 2 hr. New
chase media containing control, ca-FoxO3, or
were added, and media samples were collected
over 36 hr. The released radioactivity (indicating
proteins degraded) was plotted as a percentage
of total [3H]tyrosine incorporated into cell pro-
teins. The rates of proteolysis (calculated from
the linear slopes between 20 and 32 hr) are
also shown. Levels of FoxO3 protein at different
times after infection were analyzed by western
blot. Error bars represent SEM.
FoxO3 Stimulates Autophagy in Atrophying Muscle
Cell Metabolism 6, 472–483, December 2007 ª2007 Elsevier Inc. 473
cells treated with inhibitors from that in untreated cells
(Figure 2A, middle panel). Similar rates were found using
the two proteasome inhibitors or the three inhibitors of
lysosomal acidification (Figure 2A). In control myotubes,
breakdown of long-lived proteins appeared to be ?50%
proteasomal and ?40% lysosomal (Figure 2A, top panel).
This large contribution of lysosomes to total proteolysis is
much higher than that estimated previously in well-nour-
ished HeLa cells (Rock et al., 1994) and nutrient-deprived
rodent muscles (Furuno et al., 1990).
By 24 hr after infection in cells expressing ca-FoxO3,
overall rates of degradation increased by 66%. Surpris-
ingly, this increase was reduced only modestly after treat-
ment with proteasome inhibitors but was markedly in-
hibited by the lysosomal inhibitors (Figure 2A, top panel).
In fact, approximately 70% of the increase in proteolysis
required lysosomal function, while only 20% required pro-
teasomal activity (Figure 2A, bottom panel). Thus, FoxO3
stimulates both degradative pathways coordinately but
has a substantially greater effect on lysosomal proteolysis
in C2C12 myotubes. This increase in lysosomal proteoly-
sis is independent of proteasomal activity, sincesimilar in-
creases were seen in cells pretreated with Velcade for 4 hr
(seeFigure S1 inthe Supplemental Data available withthis
article online), and is also independent of atrogin-1. In pri-
mary muscle cultures derived from wild-type or atrogin-1-
deficient mice, ca-FoxO3 expression caused similar large
increases in overall proteolysis, and most of this increase
was lysosomal (Figure S2).
Evidence has been presented that caspases (Du et al.,
2004) and/or calpains (Kramerova et al., 2005) might
alsoplayanessential roleinenhancingmuscle proteolysis
during atrophy by promoting conversion of myofibrillar
proteins to forms easily digested by the ubiquitin-protea-
some pathway. However, addition of both proteasomal
proteolysis (data not shown). Thus, if caspases or cal-
pains, which are not affected by these inhibitors, do con-
tribute to overall proteolysis in myotubes, their contribu-
tions must be minor. Furthermore, treatment with the
general caspase inhibitors DEVD-CHO or cpm-VAD-
CHO did not significantly reduce overall proteolysis or
proteasomal proteolysis in either control or ca-FoxO3-
expressing myotubes (Figure S3). Thus, the great majority
of caspases in these cells.
In Other Cell Types, FoxO3 Enhances Lysosomal
but Not Proteasomal Proteolysis
Additional experiments tested whether the stimulation of
lysosomal and proteasomal proteolysis by FoxO3 is spe-
cific to muscle cells. ca-FoxO3 expression using the
same adenovirus in H4IIE rat hepatoma cells and PC12
rat pheochromocytoma cells also caused an increase in
Figure 2. Effects of FoxO3 on Proteaso-
mal and Lysosomal Proteolysis
(A) ca-FoxO3 stimulates both lysosomal and
proteasomal proteolysis in myotubes. Top
panel: overall rates of proteolysis in control or
ca-FoxO3-expressing cells after proteasomal
or lysosomal inhibition.Myotubes werelabeled
and infected with ca-FoxO3 as in Figure 1B.
Twenty-four hours after infection, new medium
containing proteasomal or lysosomal inhib-
itors (1 mM Velcade, 8 mM lactacystin [Lactac],
0.1 mM concanamycin A [ConcA], 50 mM chlo-
roquine [Chloroq], or 10 mM NH4Cl) was
added, and the rates of proteolysis were
determined. Middle panel: proteasomal or
lysosomal proteolysis was calculated as in Ex-
perimental Procedures. Bottom panel: FoxO3-
induced increase in proteolysis is mediated
mainly by lysosomes. The ca-FoxO3-induced
increasesin proteasomal and lysosomal prote-
olysis were plotted as a percentage of the in-
crease in total proteolysis.
(B) ca-FoxO3 increases lysosomal proteolysis,
but not proteasomal proteolysis, in H4IIE and
PC12 cells. Both types of cells were labeled
and infected with ca-FoxO3 adenovirus as in
(A). *p < 0.05, **p < 0.01 by two-tailed t test.
Error bars represent SEM.
474 Cell Metabolism 6, 472–483, December 2007 ª2007 Elsevier Inc.
FoxO3 Stimulates Autophagy in Atrophying Muscle
lysosomal proteolysis (Figure 2B). However, in these cells,
the rates of proteasomal degradation were unchanged,
presumably because these nonmuscle cells do not ex-
press the FoxO-induced ubiquitin ligases and do not con-
tain a pool of myofibrillar proteins. These results also indi-
cate that the activation of these two proteolytic processes
by FoxO3 is not necessarily linked.
FoxO3 Activates Autophagy
The major route for delivery of cytoplasmic proteins or
organelles into lysosomes is by macroautophagy, which
involves de novo formation of autophagosomes. These
vesicles sequester cytoplasmic constituents and then
fuse with lysosomes (Levine and Klionsky, 2004; Ohsumi,
2001). Changes in macroautophagy have long been
known to correlate with rates of proteolysis in liver (Morti-
more et al., 1983) and cultured cells (Knecht et al., 1984),
and rapid changes in macroautophagy occur in response
to changes in nutrient supply, mTOR, and insulin (Meijer
and Codogno, 2006). To test whether FoxO3 increases
lysosomal proteolysis by activating macroautophagy, we
used the macroautophagy inhibitor 3-methyladenine
(Seglen and Gordon, 1982). This agent reduced basal ly-
sosomal proteolysis by 0.14% ± 0.03%/hr in control cells
but caused a much larger decrease (0.50% ± 0.08%/hr) in
the lysosomal proteolysis in ca-FoxO3-expressing cells
(Figure 3A), strongly suggesting a marked activation of
macroautophagy. This inhibition by 3-methyladenine ap-
pears to result from its capacity to block class III PI3-
kinase (Petiot et al., 2000) since LY294002, an inhibitor
of both class I and class III PI3-kinases, failed to inhibit
and actually enhanced lysosomal and total proteolysis in
these cells (Sacheck et al., 2004). Another mode of deliv-
ery of cytosolic proteins to lysosomes is chaperone-medi-
ated autophagy (Dice, 2007), which is not sensitive to
3-methyladenine. ca-FoxO3 does not seem to activate
chaperone-mediated autophagy under these conditions
since it did not affect the expression of Lamp2a
(Figure S4), a marker for chaperone-mediated autophagy
(Cuervo and Dice, 2000), although ca-FoxO3 induced
many autophagy-related genes (see below). Another pos-
sible source of substrates for lysosomal degradation
could be endocytosis of membrane components, but
they cannot account for the large fraction of cell proteins
findings suggest that macroautophagy (hereafter referred
to as autophagy) is required for most of the increased
lysosomal proteolysis by ca-FoxO3.
Figure 3. FoxO3 Activates Autophagy in Myotubes
(A) Most of the ca-FoxO3-induced increase in lysosomal proteolysis is sensitive to 3-methyladenine. After labeling cell proteins and adenoviral infec-
tion,myotubes weretreated with10mM3-methyladeninefor1hrandconcanamycinAforanother1hrbeforemeasurementoflysosomalproteolysis.
Error bars represent SEM.
(B) ca-FoxO3 induces lipidation of LC3 and increases Gabarapl1 protein levels. LC3, Gabarapl1, and beclin1 protein levels were determined by west-
ern blot over 36 hr after infection. b-actin was used as a loading control.
(C) ca-FoxO3 enhances lysosomal degradation of LC3 and Gabarapl1 proteins. Myotubes were infected with control or ca-FoxO3 viruses for 22 hr
and then treated with concanamycin A or Velcade for an additional 6 hr before lysis and western blotting. Quantified protein levels of LC3 and
Gabarapl1 (relative to b-actin) were plotted. Values shown are the means of duplicates.
Cell Metabolism 6, 472–483, December 2007 ª2007 Elsevier Inc. 475
FoxO3 Stimulates Autophagy in Atrophying Muscle
To confirm the activation of autophagy, we performed
western blots for LC3, the homolog of yeast Atg8, which
is cleaved and conjugated to phosphatidylethanolamine
of ca-FoxO3 was associated with a shift of some LC3-I to
time that protein degradation increased (Figure 1B). In ad-
dition, the level of Gabarapl1 protein (another homolog of
yeast Atg8) was increased by ca-FoxO3, but without any
change in its migration on SDS-PAGE (Figure 3B).
To confirm that these changes in lysosomal proteolysis
are due largely to changes in autophagy, we used mouse
embryonic fibroblasts (MEFs) that lack the Atg5 gene and
are therefore defective in autophagy (Mizushima et al.,
2001). When we applied a variety of agents or conditions
known to activate autophagy, including rapamycin (an
inhibitor of mTOR), API-2 (an inhibitor of Akt), LY294002
(an inhibitor of PI3K), or serum deprivation, all stimulated
lysosomal proteolysis in wild-type MEFs, as in myotubes
(see below). However, in Atg5-deficient MEFs, these
treatments had little effect on lysosomal proteolysis
(Figure S5). Thus, these measurements of change in lyso-
somal proteolysis provide a sensitive and quantitative
measure of change in autophagic activity.
FoxO3 Increases Expression of Many
Autophagy-Related Genes in Myotubes
Studies in yeast have identified many autophagy-related
(Atg) genes,mostof which have counterparts inhigher eu-
karyotes (Klionsky et al., 2003), but regulation of their ex-
pression has not been demonstrated in either system. To
clarify how FoxO3 might stimulate autophagy, we tested
whether FoxO3 alters the expression in myotubes of the
Atg genes LC3b, Atg12l, Atg4b, and Beclin1 (homolog of
yeast Atg6) as well as Ulk2 (possible homolog of yeast
Atg1), Gabarapl1 (homolog of yeast Atg8), and PI3KIII(ho-
molog of yeast Vps34, which is required for autophagy).
Previously, we found that LC3 and Gabarapl1 are induced
in muscles in diverse types ofatrophy (Lecker etal.,2004).
Using quantitative real-time PCR, we demonstrated that
these seven genes are induced by the expression of
ca-FoxO3 (Figure 4A). mRNAs for LC3b, Gabarapl1,
Vps34, Ulk2, and Atg12l were clearly increased by 20 hr
after infection with the FoxO3 adenovirus. Those for
Atg4b and Beclin1increased less,and did so moreslowly.
FoxO transcription factors share a characteristic DNA-
binding domain that recognizes the specific consensus
sequence (C/G)(A/T)AAA(C/A)A in the promoters of target
genes (Barthel et al., 2005). To determine whether FoxO3
binds directly to the promoters of Atg genes or whether it
we performed chromatin immunoprecipitation (ChIP) to
test whether ca-FoxO3 associates with the putative
FoxO-binding sites in the proximal promoter regions of
the three most induced genes, LC3b, Gabarapl1, and
Atg12l. We analyzed five consensus sequences within
the 5 kb genomic region upstream of the LC3b coding se-
quence, six for Gabarapl1, and four for Atg12l. From the
ca-FoxO3 immunoprecipitates, we were able to enrich
DNA fragments that cover three of these consensus se-
quences in the promoter of LC3b (L1, L2, and L4), three
for Gabarapl1 (G2, G3, and G4), and two for Atg12l (A2
and A4) (Figure 4B). These results thus demonstrate that
FoxO3 binds directly to these promoters and also allow
us to identify its specific binding sites.
After expression of ca-FoxO3, no increase in the levels
of beclin1 and LC3 proteins was evident on western blots,
infection (Figure 3B). Interestingly, in ca-FoxO3-express-
ing cells, LC3 and Gabarapl1 proteins declined 30 hr after
infection (Figure 3B). Most likely, FoxO3 enhances the
production and also degradation of these proteins, since
when these cells were exposed to concanamycin A to
block lysosomal proteolysis, both proteins accumulated
to higher levels in FoxO3-expressing myotubes than in
controls (Figure 3C), indicating enhanced lysosomal deg-
radation. Thus, FoxO3, through direct binding to the LC3b
and Gabarapl1 promoters, increases the production of
these Atg8 homologs, which are then consumed by lyso-
FoxO3 Increases Autophagosome Formation
in Isolated Adult Mouse Muscle Fibers
Several findings indicate that this FoxO-dependent induc-
covered or shown to be necessary for muscle wasting, in-
creased autophagy (Schiaffino and Hanzlikova, 1972) and
a greater capacity for lysosomal proteolysis (Furuno et al.,
formation of autophagosomes labeled byGFP-LC3 has re-
cently been demonstrated in mouse muscles upon fasting
(Mizushima et al., 2004) and denervation (Mammucari
et al., 2007 [this issue of Cell Metabolism]). To determine
ined whether overexpression of ca-FoxO3 activates au-
formation, we delivered a GFP-LC3 plasmid together with
a control or ca-FoxO3 construct into isolated mouse flexor
digitorum brevis (FDB) fibers by electroporation. Two days
later, we measured the number of GFP-positive puncta.
The fibers expressing ca-FoxO3 had 6-fold more puncta
onstrated that the activation of autophagy upon starvation
of the muscle requires FoxO3. When the isolated GFP-
LC3-expressing fibers were deprived of nutrients and se-
rum, they showed a dramatic increase in the formation of
GFP-positive puncta. However, in fibers where FoxO3
was knocked down by electroporation of RNAi for FoxO3,
there was a much smaller increase in these GFP-tagged
autophagosomes (Figure 5B).
Induction of Autophagy-Related Genes
in Atrophying Mouse Muscles
In prior work, we found that LC3, Gabarapl1, and the
lysosomal protease cathepsin L are among the set of
476 Cell Metabolism 6, 472–483, December 2007 ª2007 Elsevier Inc.
FoxO3 Stimulates Autophagy in Atrophying Muscle
atrogenes that are induced similarly in muscles atrophy-
ing due to fasting, renal failure, diabetes, or cancer
(Lecker et al., 2004). Therefore, we examined by real-
time PCR whether all of the autophagy-related genes
induced by ca-FoxO3 in myotubes (Figure 4A) are also
induced in mouse muscles after denervation or food dep-
rivation, where FoxO3 appears to be activated. Indeed,
these genes were induced in both types of atrophy (ex-
cept Ulk2 and Beclin1, which were not induced in fasting)
(Figures 6A and 6B). LC3b and Gabarapl1 showed the
greatest induction (6- to 8-fold 1 week after denervation),
while Beclin1, which has been viewed as a key regulator
of autophagy (Liang et al., 1999), was induced least. Ad-
ditionally, the Atg mRNAs induced most by FoxO3 in the
myotubes also changed most in atrophying muscles. Sur-
prisingly, although autophagy is activated in muscle upon
food deprivation (Mizushima et al., 2004), the induction of
these genes in muscle of animals fasted for 1 day was
consistently less than following denervation. Presumably,
the enhancement of autophagy in muscle of fasted ani-
mals occurs largely via a nontranscriptional mechanism
(perhaps decreased mTOR activity), unlike that in disuse
atrophy, where there is no nutrient deficiency. Thus, acti-
vation of autophagy and induction of many autophagy-
related genes occur in multiple types of atrophy, although
it is unclear from these experiments whether these
FoxO3-dependent transcriptional changes drive the en-
hanced lysosomal proteolysis or only support the mainte-
nance of this process by replacing components con-
sumed during autophagy (e.g., LC3 and Gabarapl1).
Figure 4. FoxO3 Induces Autophagy-Related Genes in Myotubes
(A) ca-FoxO3 increases mRNA levels for many autophagy-related genes in myotubes (as shown by real-time PCR). Error bars represent SEM.
(B) ca-FoxO3 directly binds to the promoters of LC3b, Gabarapl1, and Atg12l. Five consensus (C/G)(A/T)AAA(C/A)A sequences for FoxO3 were iden-
tified in the 5 kb promoter region of LC3b, six for Gabarapl1, and four for Atg12l. Myotubes were infected with control or ca-FoxO3 adenoviruses for
12 hr, chromatin immunoprecipitation was performed using an anti-HA antibody (ca-FoxO3 is HA tagged), and pairs of primers that cover these sites
were used to amplify the related DNA fragments from the immunoprecipitates by PCR. We were unable to amplify the DNA fragments that cover
the two tentative sites between G1 and G2 in the promoter of Gabarapl1. Asterisks indicate tentative sites that are positive for binding of
ca-FoxO3 in ChIP assays.
Cell Metabolism 6, 472–483, December 2007 ª2007 Elsevier Inc. 477
FoxO3 Stimulates Autophagy in Atrophying Muscle
Two Distinct Mechanisms for Regulating
Autophagy via the PI3K-Akt Pathway
IGF-1 or insulin can rapidly inhibit proteolysis in muscle
and other cells, and this effect is generally assumed to re-
pathway. Accordingly, treatment of C2C12 myotubes with
IGF-1 for 2 hr decreased overall proteolysis, while treat-
ment with the Akt inhibitor API-2 (Yang et al., 2004) stimu-
lated this process, and these rapid responses reflected
mainly changes in the lysosomal pathway (Figure 7B).
mTOR is a kinase downstream of Akt that can inhibit au-
mTOR has been reported to phosphorylate Atg13, thus
inactivating the Atg1 complex (Levine and Klionsky,
2004). The mTOR inhibitor rapamycin stimulates autoph-
effective concentrations caused only a small but consis-
tent increase (10%–12%, p < 0.05) in lysosomal proteoly-
sis. This response was much smaller than the 50% in-
crease (p < 0.01) in this process induced by API-2
(Figure 7C). Therefore, the inhibition of Akt must stimulate
autophagy by an additional mTOR-independent mecha-
nism. Since most of the effect of API-2 is not through
mTOR, it most likely occurs through FoxO-dependent
activated FoxO transcription factors in the myotubes, as
shown by dephosphorylation of FoxO3 and induction of
atrogin-1 expression (Figure 7A).
To test the importance of transcription for this enhance-
ment of autophagy, we blocked mRNA synthesis with ac-
tinomycin D. This treatment did not reduce, but rather
seemed to enhance, the increase in lysosomal proteolysis
caused by rapamycin or its rapid decrease caused by
IGF-1, but inhibiting RNA synthesis blocked most of the
increase in this process caused by API-2 (Figure 7C).
Thus, activation of lysosomal proteolysis by API-2 re-
quires gene transcription either for the autophagy-related
genes described here or perhaps for some unidentified
activators of this process. Another clear difference in the
Figure 5. FoxO3 Induces Autophagosome Formation in Adult
(A) Flexor digitorum brevis (FDB) muscle fibers were transfected by
electroporation with a GFP-LC3 construct together with a control or
ca-FoxO3 construct. Two days later, autophagosomes (i.e., GFP-pos-
itive puncta) were examined by confocal microscopy and counted.
(B) Induction of autophagosome formation by nutrient and serum dep-
rivation is blocked by RNAi for FoxO3. FDB fibers were cotransfected
with GFP-LC3 and pSUPER vectors expressing shRNA for FoxO3 or
control vector. Two days later, the fibers were incubated in PBS for
24 hr and analyzed for autophagosome formation by confocal micros-
copy. Quantification was expressed as the percentage of fibers
containing more than 13 3 104GFP-positive vesicles/area, the level
of autophagosomes in fed control fibers.
Error bars represent SEM.
Muscles after Denervation or Fasting
Autophagy-related genes that are induced by ca-FoxO3 in cultured my-
otubes are also upregulated in muscles from denervated and fasted
adult CD-1 mice. For denervation, a 2–4 mm section of the sciatic nerve
was removed from one hindlimb. After 3 or 7 days, tibialis anterior mus-
cles from both hindlimbs were dissected. For food deprivation, all food
was removed in the late afternoon, but water was provided. Muscles
were collected 24 hr later. mRNA levels of these seven genes in tibialis
anterior muscles following denervation (A) or food deprivation (B) were
analyzed by real-time PCR. Error bars represent SEM.
478 Cell Metabolism 6, 472–483, December 2007 ª2007 Elsevier Inc.
FoxO3 Stimulates Autophagy in Atrophying Muscle
Figure 7. Increase in Lysosomal Proteolysis by Akt Inhibition Is Transcription Dependent
(A) API-2 treatment results in rapid dephosphorylation of Akt and FoxO3 proteins (by western blot) and an increase in mRNA levels of atrogin-1 (by
real-time PCR). Myotubes were treated with 1 mM API-2 for the indicated times.
(B) API-2 stimulates, while IGF-1 suppresses, overall proteolysis in myotubes mainly by affecting lysosomal proteolysis. Myotubes were incubated
with [3H]tyrosine for 20 hr and switched to chase medium for 2 hr. These cells were then treated with 1 mM API-2, 10 ng/ml IGF-1, or vehicle for an
additional 2 hr before proteolysis measurement.
(C) Increase in lysosomal proteolysis induced by API-2 requires transcription, unlike that induced by rapamycin. Experiments were performed as
above, but myotubes were pretreated with actinomycin D (10 mg/ml) for 1 hr before the addition of 1 mM API-2, 0.3 mM rapamycin, or 10 ng/ml IGF.
(D) API-2, but not rapamycin, induces mRNAs for LC3b and Gabarapl1 within 4 hr in myotubes. *p < 0.05, **p < 0.01 by two-tailed t test.
(E) Effects of API-2 and rapamycin on lysosomal proteolysis in control and ca-FoxO3-expressing myotubes. Experiments were performed as in
Figure 2A, but 22 hr after infection, cells were treated with API-2 or rapamycin for 2 hr before lysosomal proteolysis was determined.
(F) Expression of constitutively active Akt (ca-Akt) reduces both proteasomal and lysosomal proteolysis in myotubes. Myotubes were infected with
adenovirus expressing ca-Akt for 24 hr before proteolysis measurement.
(G) Rapamycin treatment prevents the specific phosphorylation of S6K and 4E-BP1, while expression of ca-FoxO3 has no apparent effects on their
phosphorylation or that of mTOR. Antibodies indicated were used for western blotting.
(H) Schematic of two different mechanisms of regulating autophagy downstream of PI3K-Akt signaling.
Error bars represent SEM.
Cell Metabolism 6, 472–483, December 2007 ª2007 Elsevier Inc. 479
FoxO3 Stimulates Autophagy in Atrophying Muscle
activation of autophagy by ca-FoxO3 is that it had no ap-
parent effects on the specific phosphorylation of mTOR or
that of mTORC1’s substrates, S6K and 4E-BP1. In con-
trast, rapamycin rapidly and completely abolished the
specific phosphorylation of S6K and 4E-BP1 (Figure 7G).
pathway can activate autophagy by two mechanisms: (1)
a rapid transcription-independent mechanism through
mTOR, and (2) a slower mechanism independent of
mTOR and requiring gene expression, apparently medi-
ated by FoxO3 (Figure 7H). Accordingly, API-2 increased
the transcription of Gabarapl1 and to a lesser extent
LC3b within 4 hr, but rapamycin did not (Figure 7D),
even though both rapamycin and API-2 increased the lip-
idation ofLC3(FigureS6).Another indication thatblocking
Akt in myotubes activates autophagy via FoxO3 is that
sosomal proteolysis in control cells (0.41% ± 0.03%/hr)
than in ca-FoxO3-expressing cells (0.13% ± 0.04%/hr),
where activation of autophagy by FoxO3 is maximal and
independent of Akt. By contrast, rapamycin, which does
not act through FoxOs, caused similar effects in both
types of cells (0.11% ± 0.02%/hr versus 0.17% ±
0.03%/hr) (Figure 7E). These findings indicate that IGF-1
or insulin can reduce protein degradation rapidly by sup-
also suppress this process by inactivating FoxOs, which
also inhibits proteasomal degradation through the reduc-
tion of atrogin-1 and MuRF1 transcription (Sacheck et al.,
2004). Accordingly, overproduction of Akt by adenoviral
expression of constitutively active Akt (ca-Akt) in C2C12
myotubes for 24 hr led to a reduction in overall proteolysis
(as was noted previously in HT29 cells [Arico et al., 2001]),
and this effect involved coordinate inhibition of both lyso-
somal and proteasomal processes (Figure 7F).
The present studies demonstrate that FoxO3-dependent
transcription enhances the cell’s capacity for autophagy
and thus that in muscle, the lysosomal and proteasomal
pathways for protein degradation are coordinately regu-
lated. These results can account for our earlier findings
(Furuno et al., 1990) that atrophying muscle shows an in-
creased capacity for lysosomal proteolysis together with
a large increase in nonlysosomal ATP-dependent proteol-
ysis (i.e., the ubiquitin-proteasome pathway). While serv-
ing complementary functions in promoting protein loss,
activation of these two processes seems to result from
the coordinate transcription of key genes rather than
crosstalk between the two systems since inhibition of ei-
ther process for many hours (or loss of atrogin-1) did not
alter the rate of the remaining degradative process. It
has been reported that inhibition of proteasomes leads
to a compensatory increase in autophagy (Iwata et al.,
2005), but no such change was observed in our studies.
Interestingly, this simultaneous activation of the two
proteolytic processes was restricted to muscle cells. In
hepatic (H4IIE) cells or neuronal (PC12) cultures, FoxO3
stimulated lysosomal proteolysis but not the proteasomal
process, presumably because these cells lack the mus-
cle-specific ubiquitin ligases atrogin-1 and MuRF1 that
are induced during atrophy. In muscle, protein degrada-
tion serves a physiological role distinct from proteolysis
in other tissues since muscle components, especially
myofibrillar proteins, are the major amino acid reservoir
in the organism. The stimulation of muscle proteolysis is
therefore critical in fasting and disease to mobilize amino
these two pathways catalyze the degradation of different
cellular components. While proteasomes degrade myofi-
brillar and most soluble proteins (Solomon and Goldberg,
1996), organelles (especially mitochondria) are degraded
primarily in lysosomes (Scott and Klionsky, 1998). The si-
multaneous activation of these two proteolytic pathways
by FoxO3 presumably ensures that the loss of different
cell components is coordinated upon fasting or disuse
and leaves the muscle with relatively normal composition,
ponents and in endurance due to loss of mitochondria.
Because of the importance of the ubiquitin-proteasome
pathway in degrading myofibrillar proteins, there has been
appreciable interest in developing inhibitors of the atro-
phy-specific ubiquitin ligases or the proteasome to retard
muscle atrophy and cachexia (Beehler et al., 2006; Glass,
2005). Our findings argue that such approaches would not
prevent the loss of those muscle components digested by
lysosomes. Unlike in adult muscle, in which they comprise
60%–70% of cell protein, in cultured myotubes, myofibrils
comprise only a small fraction of total protein. This differ-
ence probably explains why proteasome-dependent pro-
teolysis comprises only a small fraction of the FoxO3-
induced proteolysis in myotubes, while this system
accounts for most of the accelerated proteolysis in normal
and atrophying adult muscles (Lecker et al., 2006).
The presentfindings thatFoxO3 inducestheexpression
of many autophagy-related genes in culture and that sim-
ilar changes in their expression occur upon denervation
and fasting in mice are in agreement with our previous
finding that FoxO activation is essential for muscle wast-
ing upon denervation and fasting (Sandri et al., 2004). In
adult muscle, in which FoxO3 causes profound atrophy,
it also induces these same genes and causes a large
as also occurs in mouse muscles atrophying due to food
deprivation (Mizushima et al., 2004) or denervation (Mam-
mucari et al., 2007). These in vivo responses appear to
depend on FoxO3 since RNAi against FoxO3 blocked
the enhanced autophagy induced bystarvation of isolated
muscle fibers (Figure 5B).
Elucidation of the pathway for autophagy in yeast has
greatly advanced our knowledge of this process in mam-
malian tissues, but how autophagy in specific tissues is
regulated is still poorly understood. For example, mTOR,
the major nutrient sensor in eukaryotic cells, is an impor-
tant regulator of autophagy, and its inhibitor rapamycin
is widely used to activate this process, although how it
stimulates autophagy remains unclear. In myotubes,
480 Cell Metabolism 6, 472–483, December 2007 ª2007 Elsevier Inc.
FoxO3 Stimulates Autophagy in Atrophying Muscle
rapamycin induces onlya small increase in lysosomal pro-
teolysis (<15%),less than thatinduced in myotubes bythe
Akt inhibitor API-2. Thus, most of Akt’s effects on lyso-
mechanism, as has been suggested previously (Sarkar
et al., 2007; Tassa et al., 2003; Yamamoto et al., 2006).
The activation of autophagy by FoxO3 provides an addi-
tional mechanism by which the cell’s capacity for autoph-
tion of a large number of autophagy-related genes, seven
of which were found to be increased in these studies.
FoxO3 seems to directly activate their transcription since
ca-FoxO3 was shown to bind directly to the promoters of
LC3b, Gabarapl1, and Atg12l (Figure 4B).
These studies, however, have not resolved precisely
how FoxO3 activates autophagy or the consequences
of enhanced expression of Atg genes. Possibly the
increased levels of multiple Atg proteins per se lead to
greater autophagy if the levels of one or more of them
are rate limiting for formation of autophagic vacuoles.
Alternatively, FoxO3 may also induce one or more other
proteins to enhance this process. Surprisingly, expression
of Beclin1, which had been proposed to be critical in the
activation of autophagy in other cells, changes much
less than expression of other autophagy-related genes
in response to ca-FoxO3 and denervation in mouse mus-
cle. Possibly, the FoxO3-dependent expression of many
autophagy-related genes may be important not for the
enhancement of autophagy but to enable high levels of
autophagy to be maintained for extended periods. The
need forsuchincreasedproduction ofLC3andGabarapl1
was clearly evident in these experiments, where both pro-
teins were found to be destroyed by lysosomes during
FoxO-induced autophagy and their increased production
could only be demonstrated when lysosomal proteolysis
was blocked (Figure 3C). It is also noteworthy that al-
though rapamycin stimulated autophagy rapidly without
altering transcription, after 7 hr of treatment, an increase
in the expression of some Atg genes was seen (data not
shown), perhaps to allow maintenance of this response.
These transcriptional and nontranscriptional mecha-
nisms for regulation of autophagy by the IGF-1/insulin-
PI3K-Akt pathway must function together in the suppres-
sion of proteolysis during growth (asshown here when Akt
was overexpressed) and also in enhancing protein loss in
nutrient- or insulin-deficient conditions (as shown here
when Akt was inhibited). It is noteworthy that in myotubes,
the increase in lysosomal proteolysis seen with the Akt
inhibitor was severalfold larger than that seen with rapa-
mycin. Presumably, in atrophying muscle, both modes
of activating autophagy function together with the en-
hancement of the ubiquitin-proteasome pathway to cause
cellsislikely tobeimportantinotherphysiological orpath-
ological processes and other cell types (e.g., hepatic
H4IIE and neuronal PC12 cells). Accordingly, it was re-
cently reported that Drosophila FoxO is required for star-
vation-induced autophagy in fat bodies (Juhasz et al.,
2007). In many cells, autophagy is activated during apo-
ptosis, which is often triggered by FoxOs (Tran et al.,
2003). Additionally, FoxO-induced autophagy may be im-
portant in the extension of life span induced by the reduc-
tion in the PI3K-Akt signaling pathway in worms and flies.
In such mutants (daf-2) in C. elegans, the prolonged life
span requires the FoxO ortholog (daf-16) or an Atg gene
(Beclin1), and cells exhibit enhanced autophagy (Melen-
dez etal.,2003).Autophagy alsohelps protect cells byen-
viruses, or intracellular bacteria. Therefore, activation of
FoxO3, like treatment with rapamycin, may offer promise
as a strategy to stimulate autophagy and help cells with-
stand such threats to their viability.
Cell Culture and Materials
C2C12 myoblasts were maintained and differentiated to myotubes as
described previously (Sacheck et al., 2004). Adenoviral infection was
conducted on the sixth or seventh day of differentiation. Adenoviruses
expressing ca-FoxO3, wt-FoxO3, and ca-Akt were described previ-
ously (Sandri et al., 2004). Wild-type and Atg5-deficient MEFs were
provided by Y. Ohsumi (National Institute for Basic Biology, Okazaki,
Japan). Velcade (bortezomib) was provided by Millennium Pharma-
ceuticals. Lactacystin,API-2, rapamycin,and actinomycin D were pur-
chased from Calbiochem. Concanamycin A, 3-methyladenine, IGF-1,
and chloroquine were from Sigma-Aldrich.
Determination of Proteasomal and Lysosomal Proteolysis
C2C12 myotubes were incubated with [3H]tyrosine (5 mCi/ml) for 20 hr
to label cell proteins and then switched to chase medium (containing
2 mM unlabeled tyrosine to prevent reincorporation of [3H]tyrosine)
for 2 hr. Fresh chase medium containing proteasomal or lysosomal
inhibitors (1 mM Velcade, 8 mM lactacystin, 0.1 mM concanamycin A,
50 mM chloroquine, or 10 mM NH4Cl) was added. Starting 1 hr after
addition of these inhibitors, medium samples were collected for
3–4 hr and combined with 10% TCA (final concentration) to precipitate
proteins. The acid-soluble radioactivity reflects the amount of prela-
beled, long-lived proteins degraded at different times and was ex-
pressed relative to the total radioactivity initially incorporated. Plotting
these values versus time gave the total rates of proteolysis. Proteaso-
mal or lysosomal proteolysis was determined by subtracting the rates
of proteolysis in cells treated with proteasomal or lysosomal inhibitors
from that inuntreatedcells. Whenthe inhibitors used are not indicated,
proteasomal and lysosomal proteolysis were determined with Velcade
and concanamycin A, respectively. All measurements were performed
RNA Extraction, Reverse Transcription, and Quantitative
Total RNA was isolated with TRIzol (Invitrogen). Reverse transcription
was performed using TaqMan reverse transcription reagents (Applied
Biosystems). Mouse gene-specific primers were selected with Primer
3 software. PCR reactions were performed using a DyNAmo HS SYBR
Green qPCR kit (Finnzymes) and an ABI 7900HT Fast Real-Time PCR
system (Applied Biosystems). Genes were quantified as described
previously (Sacheck et al., 2004), using GAPDH as the internal control.
Primers for Lamp2a were provided by A.M. Cuervo (Albert Einstein
College of Medicine, New York). Sequences of primers used for real-
time PCR are listed in Table S1.
Cells were solubilized in lysis buffer (1% Triton X-100, 10 mM Tris [pH
7.6], 50 mM NaCl, 30 mM sodium pyrophosphate, 50 mM NaF, 5 mM
Cell Metabolism 6, 472–483, December 2007 ª2007 Elsevier Inc. 481
FoxO3 Stimulates Autophagy in Atrophying Muscle
EDTA, 0.1 mM Na3VO4, and protease inhibitor cocktail [Roche]). Thirty
micrograms of total proteins was separated by SDS-PAGE, trans-
ferred to PVDF membranes, and analyzed by western blot using the
ECL method (Amersham). We used polyclonal antibodies against
FoxO3, p-FoxO3(Thr32) (Upstate), 4E-BP1, S6K, p-S6K(Thr389),
mTOR, p-mTOR(Ser2448) (Cell Signaling), b-actin (Abcam), LC3, be-
clin1 (Santa Cruz), Gabarapl1 (Proteintech Group, Inc.). Protein levels
were quantified by using Quantity One software.
Chromatin Immunoprecipitation Assay
Myotubes were infected with control or ca-FoxO3 adenoviruses for
12 hr before analysis using the chromatin immunoprecipitation (ChIP)
assay kit (Upstate). Cellular chromatin was crosslinked by formalde-
hyde and then sheared by sonication. The soluble chromatin was im-
munoprecipitated with an anti-HA polyclonal antibody (Santa Cruz).
After decrosslinking, the immunoprecipitates were subjected to
PCR. Primers used are listed in Table S2.
Generation of atrogin-1-Deficient Mice and Primary Muscle
atrogin-1knockoutmice(Bodine etal.,2001)wereprovided byRegen-
eron. Myoblasts were isolated using the procedure of Rando and Blau
(1994).Muscles were removed from the hindlimbs of 2-week-old mice.
After treatment with 0.1% collagenase D and dispase II (Roche), the
isolated cells were plated on collagen-coated dishes. Myoblasts
were cultured in F-10 nutrient medium with 20% fetal calf serum,
2.5 ng/ml basic fibroblast growth factor (Invitrogen). Myotubes
were induced by switching to differentiation medium. All media
contained Primocin (InvivoGen). Cultures were used on the third day
in differentiation medium, when myotubes formed.
Single-Fiber Analyses of Adult Mouse Muscle
Flexor digitorum brevis muscles from adult mice were digested in type
I collagenase at 4?C for 1 hr, at 37?C for 2 hr, and dissociated into
single fibers. The fibers were electroporated using a BTX porator
(50 volts/4 mm, 3 pulses, 200 ms intervals) to transfer plasmid DNA
and then plated on glass coverslips coated with laminin and cultured
in Tyrode’s salt solution (pH 7.3) containing 10% fetal bovine serum,
50 U/ml penicillin, 50 mg/ml streptomycin, and 5% CO2(37?C). The
shRNA construct used for knocking down FoxO3 was described pre-
viously (Sandri et al., 2004). The specific sequence used against
FoxO3 was 50-GGATAAGGGCGACAGCAAC-30.
with this article online at http://www.cellmetabolism.org/cgi/content/
This work was supported by grants to A.L.G. from the Muscular Dys-
trophy Association, the National Space Biomedical Research Institute,
the Fund for Innovation from Elan Pharmaceuticals, and the Ellison
Foundation. J.J.B. was supported by a Ruth L. Kirschstein National
ship from the Swiss National Science Foundation. M.S. wassupported
by grants from the Agenzia Spaziale Italiana (OSMA project), Telethon
Italy (TCP04009), and Compagnia San Paolo. S.S. was supported by
grants from the Agenzia Spaziale Italiana (OSMA project) and the EU
2004-005272). S.H.L. was supported by grant DK062307 from the
NIH. We thank Y. Ohsumi for providing Atg5-deficient MEFs and
A.M. Cuervo for providing Lamp2a primers.
Received: July 10, 2007
Revised: September 30, 2007
Accepted: November 6, 2007
Published: December 4, 2007
Arico, S., Petiot, A., Bauvy, C., Dubbelhuis, P.F., Meijer, A.J.,
Codogno, P., and Ogier-Denis, E. (2001). The tumor suppressor
PTEN positively regulates macroautophagy by inhibiting the phospha-
tidylinositol 3-kinase/protein kinase B pathway. J. Biol. Chem. 276,
Barthel, A., Schmoll, D., and Unterman, T.G. (2005). FoxO proteins
in insulin action and metabolism. Trends Endocrinol. Metab. 16,
Bechet, D., Tassa, A., Taillandier, D., Combaret, L., and Attaix, D.
(2005). Lysosomal proteolysis in skeletal muscle. Int. J. Biochem.
Cell Biol. 37, 2098–2114.
Beehler, B.C., Sleph, P.G., Benmassaoud, L., and Grover, G.J. (2006).
Reduction of skeletal muscle atrophy by a proteasome inhibitor in a rat
model of denervation. Exp. Biol. Med. (Maywood) 231, 335–341.
Bodine, S.C., Latres, E., Baumhueter, S., Lai, V.K., Nunez, L., Clarke,
B.A., Poueymirou, W.T., Panaro, F.J., Na, E., Dharmarajan, K., et al.
(2001). Identification of ubiquitin ligases required for skeletal muscle
atrophy. Science 294, 1704–1708.
Brunet, A., Bonni, A., Zigmond, M.J., Lin, M.Z., Juo, P., Hu, L.S.,
Anderson, M.J., Arden, K.C., Blenis, J., and Greenberg, M.E. (1999).
Akt promotes cell survival by phosphorylating and inhibiting
a Forkhead transcription factor. Cell 96, 857–868.
Cuervo, A.M., and Dice, J.F. (2000). Unique properties of lamp2a
compared to other lamp2 isoforms. J. Cell Sci. 113, 4441–4450.
Dice, J.F. (2007). Chaperone-mediated autophagy. Autophagy 3, 295–
Du, J., Wang, X., Miereles, C., Bailey, J.L., Debigare, R., Zheng, B.,
Price, S.R., and Mitch, W.E. (2004). Activation of caspase-3 is an initial
step triggering accelerated muscle proteolysis in catabolic conditions.
J. Clin. Invest. 113, 115–123.
Furuno, K., Goodman, M.N., and Goldberg, A.L. (1990). Role of differ-
ent proteolytic systems in the degradation of muscle proteins during
denervation atrophy. J. Biol. Chem. 265, 8550–8557.
Glass, D.J. (2005). Skeletal muscle hypertrophy and atrophy signaling
pathways. Int. J. Biochem. Cell Biol. 37, 1974–1984.
Glickman, M.H., and Ciechanover, A. (2002). The ubiquitin-protea-
some proteolytic pathway: destruction for the sake of construction.
Physiol. Rev. 82, 373–428.
Gronostajski, R.M., Goldberg, A.L., and Pardee, A.B. (1984). The role
of increased proteolysis in the atrophy and arrest of proliferation in
serum-deprived fibroblasts. J. Cell. Physiol. 121, 189–198.
Iwata, A., Riley, B.E., Johnston, J.A., and Kopito, R.R. (2005). HDAC6
and microtubules are required for autophagic degradation of aggre-
gated huntingtin. J. Biol. Chem. 280, 40282–40292.
Juhasz, G., Puskas, L.G., Komonyi, O., Erdi, B., Maroy, P., Neufeld,
T.P., and Sass, M. (2007). Gene expression profiling identifies
FKBP39 as an inhibitor of autophagy in larval Drosophila fat body.
Cell Death Differ. 14, 1181–1190.
Kandarian, S.C., and Jackman, R.W. (2006). Intracellular signaling
during skeletal muscle atrophy. Muscle Nerve 33, 155–165.
Kisselev, A.F., and Goldberg, A.L. (2001). Proteasome inhibitors: from
research tools to drug candidates. Chem. Biol. 8, 739–758.
Klionsky, D.J., Cregg, J.M., Dunn, W.A., Jr., Emr, S.D., Sakai, Y.,
Sandoval, I.V., Sibirny, A., Subramani, S., Thumm, M., Veenhuis, M.,
and Ohsumi, Y. (2003). A unified nomenclature for yeast autophagy-
related genes. Dev. Cell 5, 539–545.
Knecht, E., Hernandez-Yago, J., and Grisolia, S. (1984). Regulation of
lysosomal autophagy in transformed and non-transformed mouse
fibroblasts under several growth conditions. Exp. Cell Res. 154, 224–
Kramerova, I., Kudryashova, E., Venkatraman, G., and Spencer, M.J.
(2005). Calpain 3 participates in sarcomere remodeling by acting
482 Cell Metabolism 6, 472–483, December 2007 ª2007 Elsevier Inc.
FoxO3 Stimulates Autophagy in Atrophying Muscle
upstream of the ubiquitin-proteasome pathway. Hum. Mol. Genet. 14,
Lecker, S.H., Jagoe, R.T., Gilbert, A., Gomes, M., Baracos, V., Bailey,
skeletal muscle atrophy involve a common program of changes in
gene expression. FASEB J. 18, 39–51.
Lecker, S.H., Goldberg, A.L., and Mitch, W.E. (2006). Protein degrada-
tion by the ubiquitin-proteasome pathway in normal and disease
states. J. Am. Soc. Nephrol. 17, 1807–1819.
Levine, B., and Klionsky, D.J. (2004). Development by self-digestion:
molecular mechanisms and biological functions of autophagy. Dev.
Cell 6, 463–477.
Liang, X.H., Jackson, S., Seaman, M., Brown, K., Kempkes, B.,
Hibshoosh, H., and Levine, B. (1999). Induction of autophagy and inhi-
bition of tumorigenesis by beclin 1. Nature 402, 672–676.
Mammucari, C., Milan, G., Romanello, V., Masiero, E., Rudolf, R., Del
Piccolo, P., Burden, S.J., Di Lisi, R., Sandri, C., Zhao, J., et al. (2007).
Meijer,A.J., and Codogno, P. (2006). Signalling and autophagy regula-
tion in health, aging and disease. Mol. Aspects Med. 27, 411–425.
Melendez, A.,Talloczy, Z., Seaman, M.,Eskelinen,E.L., Hall, D.H., and
Levine, B. (2003). Autophagy genes are essential for dauer develop-
ment and life-span extension in C. elegans. Science 301, 1387–1391.
Mizushima, N., Yamamoto, A., Hatano, M., Kobayashi, Y., Kabeya, Y.,
Suzuki, K., Tokuhisa, T., Ohsumi, Y., and Yoshimori, T. (2001). Dissec-
tion of autophagosome formation using Apg5-deficient mouse embry-
onic stem cells. J. Cell Biol. 152, 657–668.
Mizushima, N., Yamamoto, A., Matsui, M., Yoshimori, T., and Ohsumi,
Y. (2004). In vivo analysis of autophagy in response to nutrient starva-
tion using transgenic mice expressing a fluorescent autophagosome
marker. Mol. Biol. Cell 15, 1101–1111.
Mortimore, G.E., Hutson, N.J., and Surmacz, C.A. (1983). Quantitative
correlation between proteolysis and macro- and microautophagy in
mouse hepatocytes during starvation and refeeding. Proc. Natl.
Acad. Sci. USA 80, 2179–2183.
Ohsumi, Y. (2001). Molecular dissection of autophagy: two ubiquitin-
like systems. Nat. Rev. Mol. Cell Biol. 2, 211–216.
Petiot, A., Ogier-Denis, E., Blommaart, E.F., Meijer, A.J., and
Codogno,P.(2000). Distinctclassesof phosphatidylinositol30-kinases
are involved in signaling pathways that control macroautophagy in
HT-29 cells. J. Biol. Chem. 275, 992–998.
Rando, T.A., and Blau, H.M. (1994). Primary mouse myoblast purifica-
tion, characterization, and transplantation for cell-mediated gene ther-
apy. J. Cell Biol. 125, 1275–1287.
Rock, K.L., Gramm, C., Rothstein, L., Clark, K., Stein, R., Dick, L.,
Hwang, D., and Goldberg, A.L. (1994). Inhibitors of the proteasome
block the degradation of most cell proteins and the generation of pep-
tides presented on MHC class I molecules. Cell 78, 761–771.
Sacheck, J.M., Ohtsuka, A., McLary, S.C., and Goldberg, A.L. (2004).
IGF-I stimulates muscle growth by suppressing protein breakdown
and expression of atrophy-related ubiquitin ligases, atrogin-1 and
MuRF1. Am. J. Physiol. Endocrinol. Metab. 287, E591–E601.
Sacheck, J.M., Hyatt, J.P., Raffaello, A., Jagoe, R.T., Roy, R.R.,
Edgerton, V.R., Lecker, S.H., and Goldberg, A.L. (2007). Rapid disuse
and denervation atrophy involve transcriptional changes similar to
those of muscle wasting during systemic diseases. FASEB J. 21,
Sandri, M., Sandri, C., Gilbert, A., Skurk, C., Calabria, E., Picard, A.,
Walsh, K., Schiaffino, S., Lecker, S.H., and Goldberg, A.L. (2004).
Foxo transcription factors induce the atrophy-related ubiquitin ligase
atrogin-1 and cause skeletal muscle atrophy. Cell 117, 399–412.
Sandri, M., Lin, J., Handschin, C., Yang, W., Arany, Z.P., Lecker, S.H.,
Goldberg, A.L., and Spiegelman, B.M. (2006). PGC-1alpha protects
skeletal muscle from atrophy by suppressing FoxO3 action and atro-
phy-specific gene transcription. Proc. Natl. Acad. Sci. USA 103,
Sarkar, S., Davies, J.E., Huang, Z., Tunnacliffe, A., and Rubinsztein,
D.C. (2007). Trehalose, a novel mTOR-independent autophagy
enhancer, accelerates the clearance of mutant huntingtin and alpha-
synuclein. J. Biol. Chem. 282, 5641–5652.
Schiaffino, S., and Hanzlikova, V. (1972). Studies on the effect of
denervation in developing muscle. II. The lysosomal system. J. Ultra-
struct. Res. 39, 1–14.
Scott, S.V., and Klionsky, D.J. (1998). Delivery of proteins and organ-
elles to the vacuole from the cytoplasm. Curr. Opin. Cell Biol. 10,
Seglen, P.O. (1983). Inhibitors of lysosomal function. Methods
Enzymol. 96, 737–764.
Seglen, P.O., and Gordon, P.B. (1982). 3-Methyladenine: specific
inhibitor of autophagic/lysosomal protein degradation in isolated rat
hepatocytes. Proc. Natl. Acad. Sci. USA 79, 1889–1892.
Solomon, V., and Goldberg, A.L. (1996). Importance of the ATP-ubiq-
uitin-proteasome pathway in the degradation of soluble and myofibril-
lar proteins in rabbit muscle extracts. J. Biol. Chem. 271, 26690–
Stitt, T.N., Drujan, D., Clarke, B.A., Panaro, F., Timofeyva, Y., Kline,
W.O., Gonzalez, M., Yancopoulos, G.D., and Glass, D.J. (2004). The
IGF-1/PI3K/Akt pathway prevents expression of muscle atrophy-in-
duced ubiquitin ligases by inhibiting FOXO transcription factors. Mol.
Cell 14, 395–403.
Tassa, A., Roux, M.P., Attaix, D., and Bechet, D.M. (2003). Class III
phosphoinositide 3-kinase-Beclin1 complex mediates the amino
acid-dependent regulation of autophagy in C2C12 myotubes.
Biochem. J. 376, 577–586.
Tran, H., Brunet, A., Griffith, E.C., and Greenberg, M.E. (2003). The
many forks in FOXO’s road. Sci. STKE 2003, RE5.
Woo, J.T., Shinohara, C., Sakai, K., Hasumi, K., and Endo, A. (1992).
Isolation, characterization and biological activities of concanamycins
as inhibitors of lysosomal acidification. J. Antibiot. (Tokyo) 45, 1108–
Yamamoto, A., Cremona, M.L., and Rothman, J.E. (2006). Autophagy-
mediated clearance of huntingtin aggregates triggered by the insulin-
signaling pathway. J. Cell Biol. 172, 719–731.
Yang, L., Dan, H.C., Sun, M., Liu, Q., Sun, X.M., Feldman, R.I.,
Hamilton, A.D., Polokoff, M., Nicosia, S.V., Herlyn, M., et al. (2004).
Akt/protein kinase B signaling inhibitor-2, a selective small molecule
inhibitor of Akt signaling with antitumor activity in cancer cells overex-
pressing Akt. Cancer Res. 64, 4394–4399.
Cell Metabolism 6, 472–483, December 2007 ª2007 Elsevier Inc. 483
FoxO3 Stimulates Autophagy in Atrophying Muscle