ARTICLE
Sestrin prevents atrophy of disused and aging
muscles by integrating anabolic and catabolic
signals
Jessica Segalés1,2, Eusebio Perdiguero 1,12, Antonio L. Serrano1,12, Pedro Sousa-Victor1,10,12, Laura Ortet1,
Mercè Jardí1, Andrei V. Budanov3,4, Laura Garcia-Prat1,2,11, Marco Sandri5, David M. Thomson6, Michael Karin7,
Jun Hee Lee 8 & Pura Muñoz-Cánoves1,2,9*
A unique property of skeletal muscle is its ability to adapt its mass to changes in activity.
Inactivity, as in disuse or aging, causes atrophy, the loss of muscle mass and strength, leading
to physical incapacity and poor quality of life. Here, through a combination of transcriptomics
and transgenesis, we identify sestrins, a family of stress-inducible metabolic regulators, as
protective factors against muscle wasting. Sestrin expression decreases during inactivity and
its genetic deficiency exacerbates muscle wasting; conversely, sestrin overexpression suffices
to prevent atrophy. This protection occurs through mTORC1 inhibition, which upregulates
autophagy, and AKT activation, which in turn inhibits FoxO-regulated ubiquitin–proteasome-
mediated proteolysis. This study reveals sestrin as a central integrator of anabolic and
degradative pathways preventing muscle wasting. Since sestrin also protected muscles
against aging-induced atrophy, our findings have implications for sarcopenia.
https://doi.org/10.1038/s41467-019-13832-9 OPEN
1 Department of Experimental & Health Sciences, University Pompeu Fabra, CIBERNED, 08003 Barcelona, Spain. 2 Centro Nacional de Investigaciones
Cardiovasculares, 28019 Madrid, Spain. 3 School of Biochemistry and Immunology, Trinity Biomedical Sciences Institute, Trinity College Dublin, Dublin D02
R590, Ireland. 4 Engelhardt Institute of Molecular Biology, Center for Precision Genome Editing and Genetic Technologies for Biomedicine, 119991 Moscow,
Russia. 5 Department of Biomedical Science, University of Padova, 35100 Padova, Italy. 6 Department of Physiology and Developmental Biology, Brigham
Young University, Provo, UT 84602, USA. 7 Department of Pharmacology, University of California San Diego, La Jolla, CA 92093, USA. 8Department of
Molecular and Integrative Physiology, University of Michigan, Ann Arbor, MI 48109-2200, USA. 9 ICREA, 08003 Barcelona, Spain. 10Present address:
Instituto de Medicina Molecular (iMM), Faculdade de Medicina, Universidade de Lisboa, 1649 Lisbon, Portugal. 11Present address: Princess Margaret Cancer
Centre, University Health Network, Toronto M5G 1L7 ON, Canada. 12These authors contributed equally: Eusebio Perdiguero, Antonio L. Serrano, Pedro
Sousa-Victor. *email: pura.munoz@upf.edu
NATURE COMMUNICATIONS |          (2020) 11:189 | https://doi.org/10.1038/s41467-019-13832-9 |www.nature.com/naturecommunications 1
12
34
56
78
9
0
()
:,;
The control of mass in adult skeletal muscle is determinedby a dynamic balance between anabolic and catabolicprocesses triggered by changes in activity or pathological
conditions1. Muscle hypertrophy is associated with increased
protein synthesis induced by activated AKT and mammalian
target of rapamycin complex 1 (mTORC1) pathways1. Unlike
muscle hypertrophy, muscle atrophy always involves a proteo-
static shift in favor of catabolic versus anabolic processes1,2. Chief
among these implicated catabolic changes is the activation of the
autophagy-lysosome and ubiquitin–proteasome pathways, in
particular the induction of muscle-specific ubiquitin ligases of the
atrophy-related gene family, also known as atrogenes. Atrogenes,
which are regulated by FoxO transcription factors (TF), remove
proteins and organelles in atrophying fibers, whereas anabolic
myofiber growth depends on AKT and mTOR activation3–7. AKT
phosphorylates FoxO TF and impedes their nuclear activity, thus
suppressing FoxO-dependent atrogene expression8. A complex
scenario is thus emerging in which catabolic signaling connects
with biosynthetic pathways during muscle atrophy, although little
is known about how these connections are established.
Skeletal muscle atrophy is a major health problem and is the
consequence of a wide variety of pathological conditions,
including inactivity (immobilization or nerve injury), chronic
diseases, and neuromuscular disorders. Muscle atrophy also
complicates many aging-associated diseases, lowers life quality,
and increases mortality. Regardless of the driving origin, muscle
atrophy always involves loss of muscle mass, strength, and
function1,2. Preventing or reversing muscle atrophy is therefore of
the utmost importance, yet the mechanisms driving muscle
atrophy are largely unknown.
Through a transcriptomic/bioinformatic screen for potential
atrophy regulators, we identify the sestrin genes, particularly
sestrin1 (but also sestrin 2), as genes downregulated rapidly in
several models of muscle atrophy in vivo, including disuse,
denervation, and aging (sarcopenia). Sestrins are a family of
stress-inducible metabolic regulators that are conserved
throughout metazoans9. Cell-based studies showed that sestrins
have an antioxidant function that suppresses reactive oxygen
species (ROS)9,10. Genetic studies of Drosophila sestrin (dSesn)
revealed that, by activating AMPK, dSesn also functions as
negative regulator of dTORC1, leading to several age-related
pathologies11. Similar age-associated metabolic defects are also
observed in cSesn-mutated Caenorhabditis elegans12. Recent
studies indicate that mouse sestrins attenuate obesity-associated
metabolic liver diseases, such as insulin resistance and steatohe-
patitis by suppressing oxidative stress or modulating AMPK/
mTORC1 activity10,13–15. However, the role of mammalian ses-
trins in the regulation of muscle mass control is unknown, despite
skeletal muscle being the site of maximal sestrin 1 (Sesn1)
expression in humans and mice16.
In this study, we have made the surprising observation that not
all catabolic activities are enhanced during muscle atrophy;
rather, while proteasome activity is induced in response to
inactivity, autophagy is blunted. Using sestrin gain-of-function
and ablation approaches in mice, we find that sestrin preserves
muscle mass and force in atrophying conditions by coordinating
anabolic and catabolic pathways. This protective effect extends to
age-induced muscle atrophy. Our results reveal that sestrin is a
key instructor of skeletal muscle mass.
Results
Sestrins protect from disuse-induced muscle atrophy. Through
a bioinformatic analysis of the atrophy-associated transcriptome,
we searched for potential growth and atrophy regulators. Sesn1 was
identified as one of six genes dysregulated in distinct models of
muscle wasting in vivo, such as disuse and denervation (Fig. 1a and
Supplementary Data 1). Sesn1 belongs to a conserved stress-
inducible family of proteins with antioxidant and metabolic func-
tions, which in vertebrates are encoded by the Sesn1, Sesn2, and
Sesn3 loci9. Sesn1, among sestrins, is highly expressed in mam-
malian skeletal muscle16 (Supplementary Fig. 1a). Considering that
sestrins control mTOR and AKT pathways9,11,14,17,18, critical
regulators of muscle function1,2, we examined the influence of
Sesn1 downregulation on muscle atrophy.
Sesn1 RNA and protein expression decreased in mouse skeletal
muscle after immobilization-induced limb inactivity (Fig. 1b),
correlating with muscle atrophy and loss of force (Fig. 1c–f, white
and gray charts). To investigate the involvement of Sesn1 in disuse
atrophy, we generated transgenic mice overexpressing human
Sesn1 (96% homologous to murine Sesn1) in skeletal muscle
under the MCK promoter (Sesn1SkM-Tg mice). Sesn1SkM-Tg mice
expressed Sesn1 in both fast and slow muscles, without changes in
Sesn2 and Sesn3 expression (Supplementary Fig. 1b, c), were of
normal weight, and showed no obvious muscle abnormalities
under normal conditions when compared to littermate Sesn1WT
(wild-type, WT) mice (Supplementary Fig. 1d, e). However,
Sesn1SkM-Tg mouse muscle, but not the control Sesn1WT muscle,
was strongly protected against all measures of disuse atrophy,
including muscle weight, myofiber size, fibrosis index, and force
production by the extensor digitorum longus (EDL) (Fig. 1c, d,
navy and blue charts, and Supplementary Fig. 1f) and soleus
muscles (Supplementary Fig. 1g, navy and blue charts). No
changes in myonuclear number were observed (Supplementary
Fig. 1h). Importantly, Sesn1 also protected from atrophy induced
by denervation, as shown by the preservation of myofiber size and
muscle force in Sesn1SkM-Tg mice compared to Sesn1WT mice
upon sciatic denervation (Supplementary Fig. 1i, j).
Given the high sequence homology of mammalian sestrins
(Sesn1–3), and to discriminate whether the atrophy-preventing
effect was specific for Sesn1 or could also be exerted by another
sestrin family member, we tested muscle atrophy protection with
a transgenic mouse line overexpressing human Sesn2 (92%
homologous to murine Sesn2) in skeletal muscle (Sesn2SkM-Tg
mice), in comparison to littermate control Sesn2WT mice
(Supplementary Fig. 2a–c). Sesn2SkM-Tg overexpression provided
similar protection from disuse atrophy and force loss (Fig. 1e, f
and Supplementary Fig. 2d), indicating that, upon overexpres-
sion, Sesn1 and Sesn2 are equally effective at preventing
immobilization-induced muscle atrophy.
Transgenic Sesn1/2 expression throughout development has
the potential to produce adaptive effects unrelated to sestrin
activity that may mask sestrin functions in the adult. To exclude
this, we overexpressed Sesn1 and Sesn2 in skeletal muscle of 4-
month-old WT mice via adeno-associated virus (AAV) transduc-
tion (Supplementary Fig. 2e, f). Consistent with the results from
transgenic mice, AAV-based Sesn1 or Sesn2 overexpression in
WT muscle also prevented disuse-induced muscle atrophy and
weakness (Fig. 1g, h). Sestrins may function as antioxidants
through their oxidoreductase activity10,19, prompting us to test
the effect of an oxidoreductase-disrupting C130 mutation in
Sesn10,19. Overexpression of Sesn1-C130S still impeded disuse
muscle atrophy (Supplementary Fig. 2g), indicating that the
protective action of sestrin is independent of its antioxidant
function.
Loss of Sesn 1 aggravates disuse-induced muscle atrophy. Of
the three sestrins, Sesn1 is the main form expressed in
muscle10,16,19 (Supplementary Fig. 1a). We therefore analyzed
mice deficient in Sesn1 (Sesn1KO mice)20 in comparison to their
WT controls. Although these mice showed no detectable
ARTICLE NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-13832-9
2 NATURE COMMUNICATIONS |          (2020) 11:189 | https://doi.org/10.1038/s41467-019-13832-9 | www.nature.com/naturecommunications
alterations in muscle weight, myofiber size, or force under basal
conditions (Supplementary Fig. 3a–d), they showed more pro-
nounced myofiber atrophy and force loss in EDL and soleus
muscles in response to inactivity (Fig. 2a–d). No changes in
Sesn2/Sesn3 levels were found in muscles of Sesn1KO mice
(Supplementary Fig. 3e); Sesn1 is thus critical for preventing
muscle wasting.
We next evaluated whether sestrin provided equal protection to
distinct types of muscle fibers. Disuse muscle atrophy was
associated with muscle fiber-type switching from IIA/IIX to IIB
fibers in WT mice and in Sesn1SkM-Tg mice (Supplementary
Fig. 3f). Preservation of muscle mass upon Sesn1 overexpression
and increased atrophy after Sesn1 deletion was observed in type
IIA/IIX and type IIB fibers (Supplementary Fig. 3f, g), indicating
3d Den
(This study)
Growth & atrophy
regulators
14d Imm
(Renaud 2013)
3000
1549 2245
20
1901 522 3
1265 4
1225 6 0
2 8
5
a b
–2 30.5
14
d D
en
3d
 D
en
14
d I
mm
Cilp
Igfbp5
Mtor
Itfg2
Rnf152
Sesn1
0
0.5
1.0
1.5
Basal 10d
Imm
Basal 3d 10d
0
0.5
1.0
1.5
* *
Imm
Sesn1
Tubulin
R
el
at
iv
e
pr
ot
ei
n 
le
ve
ls
Basal 10d Imm
*
c
*
d
0
1000
2000
3000
0
1000
2000
3000
*
0
500
1000
1500
2000
2500
Freq (Hz)
0 50 100 150
Freq (Hz)
Freq (Hz)
0 50 100 150
0
50
100
150
200
250
0
50
100
150
200
250
Ab
so
lu
te
 fo
rc
e 
(m
N)
*
R
el
at
iv
e 
Se
sn
1
m
R
N
A 
ex
pr
es
sio
n
*
**
0
500
1000
1500
2000
2500
0
50
100
150
200
250
e f
0 50 100 150 200
Freq (Hz)
0 50 100 150 200
0
100
200
300
0
50
100
150
200
250
*
50
150
250
g
AAV-Control AAV-Sesn1
Immobilization
AAV-Control basal
AAV-Control Imm
AAV-Sesn1 Imm
AAV-Control basal
AAV-Control Imm
AAV-Sesn2 Imm
0
1000
2000
3000
*
0
100
200
300
*
*
h
*
0
50
100
150
200
250
* *
14d Den
(Sartori 2013)
CS
A 
(µm
2 )
CS
A 
(µm
2 )
CS
A 
(µm
2 )
CS
A 
(µm
2 )
CS
A 
(µm
2 )
CS
A 
(µm
2 )
W
ei
gh
t (m
g)
W
ei
gh
t (m
g)
W
ei
gh
t (m
g)
W
ei
gh
t (m
g)
Ab
so
lu
te
 fo
rc
e 
(m
N)
Ab
so
lu
te
 fo
rc
e 
(m
N)
Ab
so
lu
te
 fo
rc
e 
(m
N)
M
ax
 s
pe
cif
ic 
fo
rc
e
(m
N 
mm
–
2 )
Sesn1WT
Sesn1SkM-Tg
Sesn1WT Imm
Sesn2WT
Sesn2WT Imm
Sesn2WT
Sesn2WT Imm
Sesn2WT
Sesn2WT Imm
Sesn1WT
Sesn1WT Imm
Sesn1WT
Sesn1WT Imm
Sesn1SkM-Tg Imm
Sesn2SkM-Tg
Sesn2SkM-Tg Imm
Sesn2SkM-Tg
Sesn2SkM-Tg Imm
Sesn2SkM-Tg
Sesn2SkM-Tg Imm
Sesn1SkM-Tg
Sesn1SkM-Tg Imm
Sesn1SkM-Tg
Sesn1SkM-Tg Imm
M
ax
im
um
 fo
rc
e
(m
N)
M
ax
im
um
 fo
rc
e
(m
N)
Sp
ec
ific
 fo
rc
e
(m
N 
mm
–
2 )
Sp
ec
ific
 fo
rc
e
(m
N 
mm
–
2 )
AAV-Control
basal AAV-Control AAV-Sesn2
Immobilization
AAV-Control
basal
*
0
10
20
30
40
50
60
70
0
10
20
30
40
50
60
0
50
100
150
200
250
300
M
ax
 s
pe
cif
ic 
fo
rc
e
(m
N 
mm
–
2 )
0
50
100
150
200
250
300
M
ax
 s
pe
cif
ic 
fo
rc
e
(m
N 
mm
–
2 )
M
ax
 s
pe
cif
ic 
fo
rc
e
(m
N 
mm
–
2 )
0
50
100
150
200
250
300
0
10
20
30
40
50
60
70
0
10
20
30
40
50
60
0
100
200
300
0
1000
2000
3000
0
50
100
150
200
kDa
55
55
NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-13832-9 ARTICLE
NATURE COMMUNICATIONS |          (2020) 11:189 | https://doi.org/10.1038/s41467-019-13832-9 |www.nature.com/naturecommunications 3
that sestrin protects against muscle atrophy independently of
fiber-type changes.
Sestrin blunts FoxO-dependent atrogenes in disused muscle.
To interrogate the mechanistic basis of sestrin-mediated protec-
tion against muscle atrophy, we performed RNAseq analysis on
control and immobilized TA muscles of all available mouse
genotypes: Sesn1SkM-Tg, Sesn2SkM-Tg, and Sesn1KO mice, and
their respective WT counterparts. Like other atrophy models21,
muscle immobilization enriched gene expression signatures for
apoptosis, inflammation, cell-cycle inhibition, and anabolic reg-
ulation in all mouse groups (Fig. 3a and Supplementary Data 2).
Interestingly, the association between immobilization and ana-
bolic signaling hallmarks (PI3K/AKT/mTORC1 regulation) was
strengthened by Sesn1/2 overexpression, whereas these hallmarks
correlated negatively with Sesn1 deficiency (Fig. 3a).
We also performed a hierarchical clustering analysis to identify
genes regulated by both immobilization and sestrins (Supplemen-
tary Fig. 4a). We focused on gene clusters whose immobilization-
dependent induction was increased or decreased by Sesn over-
expression (Clusters A and B) but were behaving contrarily in
Sesn1-deficient muscles (Clusters C and D) (Supplementary Fig. 4a
and Supplementary Data 3). Intersection between clusters A/C and
clusters B/D yielded a gene set (defined as Sestrin-regulated genes)
that was highly enriched in canonical pathways implicated in
ubiquitin–proteasome-mediated proteolysis, including proteasome
subunits and the E3 ubiquitin-ligase atrogenes MuRF1 (Trim63)
and Atrogin1 (Fbxo32)6,22–24 (Supplementary Fig. 4a and Supple-
mentary Data 3). Other proteostasis pathways, such as
macroautophagy-lysosome degradation, were also enriched (Sup-
plementary Fig. 4b). Interestingly, genes regulated by immobiliza-
tion and sestrins were often found to contain binding sites for
FoxO, Myc/Maz, and Sp1 TF (Fig. 3b). FoxO TF-binding sites were
highly enriched in atrogenes and autophagy genes present in the
sestrin-regulated gene set (Supplementary Fig. 4c, left and right). Of
note, a significant overlap was found between sestrin-regulated
atrogenes and a previously described list of FoxO-regulated
atrogenes24 (Supplementary Fig. 4c, middle). These results are
consistent with the former findings that upregulation of atrogenes,
particularly Atrogin1 and MuRF1, is directed by FoxO3a and
required for muscle atrophy3,6,7,24.
Based on these findings, we examined the status of AKT-FoxO
signaling in immobilized muscle of Sesn1/2 transgenic mice.
Immobilized muscles of Sesn1SkM-Tg and Sesn2SkM-Tg mice had
higher levels of AKT and FoxO3 phosphorylation than WT
counterparts (Fig. 3c, d; Supplementary Fig. 5a). Notably, sestrin
overexpression in muscle upregulated both PDK1-dependent AKT
phosphorylation at Thr308 and mTORC2-dependent phosphor-
ylation at Ser473 (Fig. 3d). Phosphorylation at both sites activates
AKT kinase activity, while AKT-mediated phosphorylation of
FoxO inhibits its transcriptional activity by retaining it in the
cytoplasm8. Consequently, immobilized Sesn1/2SkM-Tg muscle
exhibited lower expression of well-known FoxO targets24 such as
Atrogin1 and MuRF1 (Fig. 3e and Supplementary Fig. 5b),
consistent with reduced FoxO transcriptional activity and with the
transcriptome studies (Supplementary Figs. 4b, c and 5c). In
contrast, disused muscle of Sesn1KO mice expressed higher levels
of these atrogenes (Supplementary Fig. 5d). In line with atrogene-
mediated upregulation of ubiquitin–proteasome activity, the
immobilization-induced increase in chymotrypsin-like protea-
some activity in muscles of WT mice was blunted by sestrin
overexpression (Fig. 3f). Conversely, blocking proteasome activity
with bortezomib prevented the disuse atrophy in WT mice
(Supplementary Fig. 5e). These results suggest that sestrins protect
disused muscles from wasting, at least in part, by repressing the
induction of FoxO-regulated atrogenes encoding muscle proteo-
lytic enzymes.
Involvement of FoxO in sestrin-mediated muscle protection
was confirmed through genetic modulation of FoxO. Muscle-
specific overexpression of constitutively active FoxO3 (FoxO3
TM7) abolished the protective effect of Sesn1 against muscle
atrophy in Sesn1SkM-Tg mice (Fig. 3g). Contrarily, muscle-specific
genetic deletion of all three FoxO genes (FoxO1,3,4SkM-KO triple-
knockout mice) (Supplementary Fig. 5f) produced strong
protection against inactivity-induced muscle atrophy-like Sesn1/
2 overexpression (Fig. 3h, see Fig. 1 for comparison). AKT
activation by sestrin thus blocks ubiquitin–proteasome-depen-
dent proteolysis via FoxO-signaling inhibition, subsequently
attenuating muscle wasting during disuse.
Sestrin coaxes autophagy by blunting mTORC1 to protect
muscle. Gene set enrichment analysis (GSEA) also identified
mTORC1-signaling regulation as another hallmark of sestrin
modulation in muscle (Fig. 4a). mTORC1 is a major anabolic
factor, classically associated with promotion of muscle growth
and hypertrophy. Although mTORC1 was originally thought to
decline during muscle atrophy1,2, we found sustained—and even
increased—mTORC1 activity in immobilized atrophic muscles of
non-transgenic mice, as revealed by phosphorylation of the
mTORC1 downstream targets S6 and ULK (Fig. 4b). Over-
expression of Sesn1 or Sesn2, which restores muscle mass in
the disuse condition, strongly inhibited mTORC1 signaling
Fig. 1 Sestrins prevent disuse-induced skeletal muscle atrophy. a Left Venn diagram showing overlap between a gene set of growth and atrophy
regulators (see the “Methods” section) and dysregulated genes in immobilized (Imm) or denervated (Den) muscles reported in published muscle atrophy
models or identified in our RNAseq comparison. Right Heat map for the six genes dysregulated in all gene sets analyzed. b Analysis of Sesn1 mRNA (left)
and protein (right) in tibialis anterior (TA) muscles from non-immobilized (basal) and immobilized (Imm) limbs of WT mice for the indicated number of
days (d). c TA muscle weight of and mean TA fiber cross-sectional area (CSA) in Sesn1SkM-Tg mice (overexpressing human Sesn1) and corresponding wild
type (WT) mice (Sesn1WT) in basal conditions and after 10 days of limb immobilization. d Force measurements in extensor digitorum longus (EDL) muscle
of Sesn1WT and Sesn1SkM-Tg mice in basal conditions and after 10 days of limb immobilization. Charts show force–frequency curves (left) and maximum
specific force (maximum force normalized by muscle area) (right). e Weight of TA muscles and mean TA fiber CSA in Sesn2SkM-Tg mice (overexpressing
human Sesn2) and corresponding WT mice (Sesn2WT) in basal conditions and after 10 days of limb immobilization. f Force measurements in EDL muscles
of Sesn2WT and Sesn2SkM-Tg mice in basal conditions and after 10 days of limb immobilization. Charts show force–frequency curves and maximum specific
force. g Histology and muscle force in EDL muscles of young mice transduced with AAV-Sesn1 or AAV-Control followed 4 days later by limb
immobilization for 10 additional days. The upper panels show representative images of hematoxylin/eosin (H/E) staining in basal and immobilized
muscles. Scale bar= 50 µm. The charts show fiber size (CSA), maximum force, and specific force. h Histology and muscle force in EDL muscles transduced
with AAV-Sesn2 or AAV-Control and treated as described in g. Scale bar= 50 µm. All data are shown as mean with SEM. Statistical comparisons by
unpaired two-tailed Student's t-test (*p < 0.05 vs. basal conditions). Sample numbers were n= 4–6 mice per group for c–f and n= 4 mice per condition for
g, h. Source data are provided as a Source Data file.
ARTICLE NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-13832-9
4 NATURE COMMUNICATIONS |          (2020) 11:189 | https://doi.org/10.1038/s41467-019-13832-9 | www.nature.com/naturecommunications
(i.e. reduced S6 and ULK1 phosphorylation) after immobilization
(Fig. 4b), whereas Sesn1KO muscle showed constitutive
mTORC1-signaling activation even in resting muscle (Supple-
mentary Fig. 6a). Therefore, muscle immobilization unexpectedly
conveys mTORC1 activation, whereas sestrins strongly down-
regulate it.
mTORC1-mediated ULK1 phosphorylation inhibits
autophagy13,25, another major protein and organelle degradation
pathway. The autophagy pathway was also found in the sestrin-
regulated gene set, which included many autophagy-related genes
enriched in FoxO TF-binding sites (Supplementary Fig. 4b, c and
Supplementary Data 3). Skeletal muscle autophagy flux was
determined in vivo using colchicine, which blocks autophagosome
degradation26. Colchicine-triggered marked accumulation of the
autophagosome marker LC3-II in WT and sestrin-overexpressing
muscles (Fig. 4c, left panel, and Supplementary Fig. 6b), indicating
that basal autophagy flux is active in all these tissues, being barely
affected by Sesn1/2 overexpression in non-disuse conditions.
However, muscle immobilization in WT mice strongly decreased
colchicine-dependent LC3-II accumulation (Fig. 4c, right panel,
and Supplementary Fig. 6c), indicating that autophagy flux was
attenuated. Importantly, autophagy flux was preserved in
a b
c d
*
*
0
1000
2000
3000
*
0
20
40
60
*
*
WT
WT Imm
*
*
*
0
100
200
300
0
100
200
300
EDL Soleus  
* *
0
50
100
150
200
250
0
50
100
150
200
250
**
*
250
Freq (Hz)Freq (Hz)
0 50 100 150 200
Freq (Hz)
0 50 100 150 200
0
50
100
150
200
0
100
200
300
0 50 100 150 200
Freq (Hz)
0 50 100 150 200
0
50
100
150
200
0
50
100
150
200
250
*
*
*
*
50
150
250
CS
A 
(µm
2 )
W
ei
gh
t (m
g)
M
ax
 s
pe
cif
ic 
fo
rc
e
(m
N 
mm
–
2 )
M
ax
 s
pe
cif
ic 
fo
rc
e
(m
N 
mm
–
2 )
Sesn1KO 
Sesn1KOImm 
WT
WT Imm
WT
WT Imm
WT
WT Imm
Sesn1KO 
Sesn1KOImm 
Sesn1KO 
Sesn1KOImm 
Sesn1KO 
Sesn1KOImm 
WT
WT Imm
Sesn1KO 
Sesn1KOImm 
WT
WT Imm
Sesn1KO 
Sesn1KOImm 
M
ax
im
um
 fo
rc
e 
(m
N)
M
ax
im
um
 fo
rc
e 
(m
N)
Ab
so
lu
te
 fo
rc
e 
(m
N)
Ab
so
lu
te
 fo
rc
e 
(m
N)
Ab
so
lu
te
 fo
rc
e 
(m
N)
Ab
so
lu
te
 fo
rc
e 
(m
N)
*
*
Fig. 2 Loss of sestrin1 exacerbates disuse-induced muscle atrophy. a Mean CSA of TA fibers from WT and Sesn1KO mice in basal conditions and after
10 days of limb immobilization. b TA muscle weight in WT and Sesn1KO mice in basal conditions and after 10 days of limb immobilization. c, d Force
measurements in EDL c and soleus dmuscles of WT and Sesn1KO mice in basal conditions and after 10 days of limb immobilization. Charts show maximum
and specific force (top) and force–frequency curves (bottom). All data are shown as mean with SEM. Comparisons by unpaired two-tailed Student's t-test
(*p < 0.05). N= 3–7 mice per genotype and condition. Source data are provided as a Source Data file.
NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-13832-9 ARTICLE
NATURE COMMUNICATIONS |          (2020) 11:189 | https://doi.org/10.1038/s41467-019-13832-9 |www.nature.com/naturecommunications 5
immobilized muscles overexpressing Sesn1 or Sesn2 (Fig. 4c, right
panel, and Supplementary Fig. 6c). This observation is consistent
with downregulation of mTORC1-dependent ULK1 activation by
overexpression of sestrins during immobilization conditions
(Fig. 4b). These findings were reinforced by experiments with a
tandem fluorescent autophagy flux reporter (mRFP-GFP-LC327)
(Fig. 4d). Although mature autolysosomes (red LC3 puncta) were
abundant in basal WT muscle, the red LC3 signal was eliminated
by muscle immobilization, leaving only yellow puncta (undegraded
autophagosomes). Sesn1/2 overexpression substantially preserved
Enrichment plot: AKT_MTORC1_SIGNALING
Enrichment plot: AKT_MTORC1_SIGNALING
En
ric
hm
en
t s
co
re
 (E
S)
R
an
ke
d 
lis
t m
et
ric
 (S
ign
al2
No
ise
)
En
ric
hm
en
t s
co
re
 (E
S)
R
an
ke
d 
lis
t m
et
ric
 (S
ign
al2
No
ise
)
'Sesn1-2Skm Tg' (positively correlated)
'WT' (negatively correlated)
Zero cross at 7350
'Sesn1KO' (positively correlated)
'WT' (negatively correlated)
Zero cross at 6642
0
0.20
0.15
0.10
0.05
0.00
–0.05
–0.5
1.5
1.0
0.5
0.0
–1.0
0.10
0.05
0.00
–0.05
–0.10
–0.15
–0.20
–2
–1
0
1
2
3
50002500 7500 10,000
Rank in Ordered Dataset
Rank in Ordered Dataset
Enrichment profile Hits Ranking metric scores
12,500
0 50002500 7500 10,000
Enrichment profile Hits Ranking metric scores
12,500
c
pFoxO3
Basal 3d Imm
Basal 3d Imm
Basal 3d Imm Basal 3d Imm
Basal 3d Imm Basal 3d Imm
Basal 3d Imm
Basal 3d Imm
*
pF
ox
O
3/
Fo
xO
3
pF
ox
O
1/
Fo
xO
1
pA
KT
T3
08
/A
KT
pAKTThr308pAKTS473
pA
KT
S4
73
/A
KT
Basal
pFoxO3
3d Imm
Tubulin
pAKT S473
pAKTT308
Tubulin
pFoxO1
d
pFoxO1
Tubulin
pAKT S473
pAKT T308
Tubulin pA
KT
T3
08
/A
KT
pFoxO3
pFoxO1
pF
ox
O
1/
Fo
xO
1
pAKT S473
pA
KT
S4
73
/A
KT
pAKT T308
pFoxO3
*
Atrogin1
Basal 3d Imm Basal 3d Imm Basal 3d Imm
MurF1
0
1
2
3
4 Cathepsin L
0
0.5
1.0
1.5
2.0
*
*
*
g h
*
ns
* *
Basal
Imm
C.A. FoxO3 Imm
0
500
1000
1500
2000
2500
FoxO1,3,4WT
FoxO1,3,4SkM-KO
FoxO1,3,4SkM-KOImm
FoxO1,3,4WTImm
*
e
0
25
50
75
100
125 ns
Genes
–
lo
g1
0 
p 
va
lu
e
–
lo
g1
0 
p 
va
lu
e
–
lo
g1
0 
p 
va
lu
e
TNFA SIGNALING VIA NFKB
KRAS SIGNALING UP
E2F TARGETS
ALLOGRAFT REJECTION
APOPTOSIS
P53 PATHWAY
INTERFERON GAMMA
RESPONSE
INFLAMMATORY RESPONSE
G2M CHECKPOINT
PI3K AKT MTOR SIGNALING
versus
WT Imm
5
10
15
20
10 20 30
Immobilization hallmarksa b
10
20
30
40
50 100 150
Genes
E12
Maz
Maz
Myc
AP4 SP1
SP1
NFAT
ETS2
ELK1
NRF1AP1
TATA
FOXO
FOXOMEF2 Pax4
Unknown
LEF1
LEF1
Imm Sesn-regulated genes
Genes
25
50
75
100 200 300
f
Sesn1SkM-TgImm 
Sesn2SkM-TgImm 
versus
WT Imm
Sesn1KOImm 
MTORC1_SIGNALING
Sesn1WT Sesn1SkM-Tg Sesn1WT Sesn1SkM-Tg
Basal 3d Imm
Sesn2WT Sesn2SkM-Tg Sesn2WT Sesn2SkM-Tg
Sesn1WT
Sesn1SKM-Tg
Sesn1WT
Sesn1WT
Sesn1WTImm
Sesn1SKM-TgImm
Sesn1SKM-Tg
Sesn1SKM-Tg
Sesn2WT
Sesn2SKM-Tg
Re
la
tiv
e 
m
RN
A
ex
pr
es
sio
n
CS
A 
(µm
2 )
CS
A 
(%
 of
 ba
sa
l)
Sesn1WT Sesn1SKm-Tg
Pr
ot
ea
so
m
e 
ac
tiv
ity
(F
U 
µg
–
1 
pr
ot
) 
*
AKT
AKT
FoxO1
FoxO3
FoxO1
FoxO3
pFoxO1
*
*
*
*
*
*
*
0
1
2
3
0
1
2
3
4
pF
ox
O
3/
Fo
xO
3
0
1
2
3
4
0
1
2
3
4
0
1
2
3
4
*
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
0
5
10
15
0
1
2
3
4
5
0
1
2
3
4
5
0
20
40
60
80
kDa
70
70
55
55
55
55
55
70
100
70
70
70
kDa
70
70
55
55
55
55
55
70
100
70
70
70
100
ARTICLE NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-13832-9
6 NATURE COMMUNICATIONS |          (2020) 11:189 | https://doi.org/10.1038/s41467-019-13832-9 | www.nature.com/naturecommunications
the red LC3 puncta during immobilization (Fig. 4d). Conversely,
Sesn1KO mice showed impaired autophagy flux already in basal
conditions, with reduced colchicine-dependent LC3-II accumula-
tion (Supplementary Fig. 6d) and loss of red LC3 puncta in the
reporter assay (Fig. 4e). The defective autophagy in non-
immobilized muscle of Sesn1KO mice could be explained by the
constitutive mTORC1-dependent ULK1 activation in the absence
of sestrin (Supplementary Fig. 6a), since stress-induced sestrin
expression inhibits mTORC1 signaling via activation of AMPK28.
Consistent with this idea, AMPK activation was readily observed in
non-immobilized Sesn1-overexpressing muscle, providing a
mechanistic explanation for how sestrin inhibits mTORC1 and
activates autophagy (Supplementary Fig. 6e).
We further tested whether sestrin-induced autophagy is
important for attenuating disuse atrophy through genetic and
pharmacological approaches. Sesn1-mediated protection against
disuse atrophy was abolished upon constitutive mTORC1
activation by silencing TSC2, the negative regulator of
mTORC129–31 (Fig. 4f and Supplementary Fig. 6f). Conversely,
rapamycin, which inhibits mTORC1 and induces autophagy32,
increased WT muscle autophagy flux (Supplementary Fig. 7a),
and substantially attenuated muscle atrophy (Fig. 4g), like
sestrins. A similar protective effect against muscle atrophy was
exerted by the autophagy inducer spermidine32,33 (Fig. 4h).
Autophagy activation through Atg7 overexpression32 also
protected myofibers from disuse-induced atrophy (Supplemen-
tary Fig. 7b, c), whereas muscle-specific genetic deletion of Atg7
(Atg7SkM-KO) (Supplementary Fig. 7d) nullified Sesn1-mediated
protection against disuse-induced muscle atrophy (Fig. 4i). No
significant crosstalk between both catabolic activities was
observed during muscle immobilization (not shown). Taken
together, these data demonstrate that muscle inactivity produces
detrimental effects by differentially acting on catabolic mechan-
isms (i.e. decreasing autophagic while inducing proteasomal
pathways). By reversing this concert, sestrins preserve autophagy
that is essential for muscle homeostasis, while preventing
proteasome overactivation that wastes muscle by majorly driving
loss of muscle mass during disuse condition.
Sestrins protect muscle against aging-associated atrophy. We
finally investigated whether sestrins can also protect against aging-
associated muscle atrophy (sarcopenia). Compared with young
mice (4 months old), aged mice (24 months old) showed lower
expression of Sesn1 protein in skeletal muscle (Fig. 5a) accom-
panied by pronounced loss of skeletal muscle force and mass and
myofiber atrophy (Fig. 5b–e). All these muscle parameters were
improved by transduction of AAV-Sesn1 for one month, including
a slight reduction in muscle fibrosis (Fig. 5c–e and Supplementary
Fig. 7e), suggesting that sestrin activation presents a promising
strategy for protecting against age-associated muscle atrophy and
related conditions. No changes in myonuclear number were
observed (Supplementary Fig. 7f). Mechanistically, we found that
mTORC1 activity is increased in muscles of aged mice and is
reduced by Sestrin overexpression (Fig. 5f), correlating with the
protection exerted by Sestrin on age-related muscle atrophy.
Finally, we compared publicly available transcriptomic data of
skeletal muscles of young and old mice (GEO: GSE53959)34 to
assess for potential gene expression similarities between aged and
immobilized muscles. Interestingly, we found that aged-muscle
transcriptome was enriched in gene expression signatures for
apoptosis, inflammation, cell-cycle inhibition, UV response, Myc
targets, and anabolic signaling (PI3K/AKT/mTORC1 regulation)
(Fig. 5g). Supporting the existence of commonalities between
genes induced by muscle immobilization and aging, the aged
muscle transcriptome was found to be also enriched in atrogenes
(Fig. 5h), while the upregulated genes contained binding sites for
E12, Myc/Maz, Sp1, and FoxO TFs (Fig. 5i). In fact, the
transcriptome of old mice was enriched in FoxO and Sestrin-
regulated atrogenes24, including the E3 ubiquitin-ligases Atrogin1
and Musa1 (Fbxo30)23 and autophagy-related genes Bnip3 and
p62 (Sqstm1) (Fig. 5i). Taken together, these results suggest that,
similar to disuse-induce muscle atrophy, sestrins may protect
muscle against aging-associated muscle atrophy by coordinating
anabolic and catabolic pathways.
Discussion
Autophagy and ubiquitin–proteasome proteolytic activities have
been independently linked to muscle wasting6,31,35. Our analysis
combining unbiased transcriptomics and targeted transgenesis
approaches has identified an important mechanism for protecting
muscle from wasting during disuse and aging. Sestrins, strongly
downregulated by disuse, preserve muscle mass by coordinately
inhibiting atrophic proteolysis and activating homeostatic
autophagy. Sestrins may achieve disuse-induced protection by
regulating the balance between distinct growth-associated mTOR
complexes, strongly inhibiting mTORC128 while supporting
mTORC2 activity14. Through upregulation of AKT, sestrins
inhibit FoxO-dependent transcription of atrogenes, which nor-
mally promote muscle wasting by accelerating protein degrada-
tion. At the same time, through the activation of AMPK and
inhibition of mTORC128, sestrins upregulate autophagy, thus
maintaining proteostasis and organelle quality for homeostatic
preservation of muscle mass and force. Of interest, the protective
effect of sestrin was also extended to denervation-induced muscle
Fig. 3 Sestrin blunts FoxO-dependent upregulation of muscle atrogenes. a Gene set enrichment analysis (GSEA) of immobilization-related genes. Bubble
plot of enriched GSEA hallmarks in the dysregulated genes upon 3-day muscle immobilization in WT mice (left). Enrichment plots of the combined PI3K-
AKT-MTOR signaling and mTORC1-signaling hallmarks (AKT-MTORC1 signaling) in comparisons of Sesn1/2SkM-Tg vs. WT mice and Sesn1KO vs. WT mice
(right). b Bubble plot of enriched transcription factor-binding sites (GSEA) among immobilization-dysregulated genes (left) and the sestrin-regulated gene
set defined in Supplementary Fig. 4a (right). cWestern blot analysis showing phosphorylation levels of FoxO1, FoxO3, and AKT in muscles of Sesn1WT and
Sesn1SkM-Tg mice in basal conditions and after 3 days of immobilization. Representative blots are shown (left) with corresponding quantification. For each
experimental condition (basal and immobilization) values of Sesn1SkM-Tg are referred to averaged values for Sesn1WT samples, which were set to one.
dWestern blot analysis of the phosphorylation levels of FoxO1, FoxO3, and AKT in muscles of Sesn2WT and Sesn2SkM-Tg mice in basal conditions and after
3 days of immobilization. Representative blots are shown (left) with corresponding quantification (right), relatively to Sesn2WT values, as in c. e Atrogin1,
MurF1, and Cathepsin L mRNA levels in skeletal muscle from Sesn1WT and Sesn1SkM-Tg mice in basal conditions and after 3 days of immobilization.
f Proteasome activity in total homogenates of TA muscles from Sesn1WT and Sesn1SkM-Tg mice in basal conditions and after 3 days of immobilization.
g Mean myofiber CSA in TA muscle from Sesn1WT and Sesn1SkM-Tg mice electrotransferred with control vector or with a plasmid encoding constitutively
active FoxO3 (C.A. FoxO3) and then immobilized for 10 days. Values are relative to basal conditions. h Mean myofiber CSA in TA muscle from
FoxO1,3,4WT and FoxO1,3,4SkM-KO mice in basal conditions and after 10 days of immobilization. All data are shown as mean with SEM. Comparisons by
Student's t-test (*p < 0.05). Sample numbers were n= 3–4 mice per group for c, d, n= 3–6 animals for e–g and n= 3–5 mice per genotype and condition
for h. Source data are provided as a Source Data file.
NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-13832-9 ARTICLE
NATURE COMMUNICATIONS |          (2020) 11:189 | https://doi.org/10.1038/s41467-019-13832-9 |www.nature.com/naturecommunications 7
wasting and age-associated muscle atrophy. The relevance of our
data is further supported by recent findings demonstrating
reduction of sestrins’ protein levels in muscles of aging human
individuals36 including muscles from the frail elderly popula-
tion37. Future studies should explore the effect of Sestrin/mTOR
modulators such as NV-513838 on these conditions. Loss of
muscle mass and force significantly affects health, life quality,
and even survival. Given their beneficial effects on muscle wasting
during disuse and aging, sestrins should be regarded as
nodal regulators of mammalian muscle growth with potentially
broad applications in the treatment of common catabolic
conditions.
a b
pS6
S6
Tubulin
pULK1
ULK1
pS6
*p
S6
/S
6
AAV-Control
AAV-Sesn1
– – – ++ + – – – ++ +
c d
*
*
0
20
40
60
80
100 WT
WT Imm
*
*
shC shTSC2 shC shTSC2
AAV-Control AAV-Sesn1
Basal
Imm
0
1000
2000
3000
4000
* * *
ns
Basal
10d Imm
Basal rapamycin
10d Imm rapamycin
**
*
AAV-Control
AAV-Control Imm
AAV-Sesn1
AAV-Sesn1Imm
Atg7WT
Atg7SkM-KO
*
0
500
1000
1500
2000
*
*
0
1000
2000
3000
nsi
e
WT
0
20
40
60
80
*
*
f g
h
*
*
CS
A 
(µm
2 )
CS
A 
(µm
2 )
– – – – –– + ++ +Colchicine:
GAPDH
AAV-Control AAV-Sesn1 AAV-Control AAV-Sesn1
LC3-I
LC3-II
Colchicine:
GAPDH
LC3-I
LC3-II
LC
3-
II/
G
AP
DH
LC
3-
II/
G
AP
DH
– + – +Colchicine: – + – +Colchicine:
AAV-
Sesn1
AAV-
Control
AAV-
Sesn1
AAV-
Control
AAV-
Sesn1
AAV-
Control
AAV-
Sesn1
AAV-
Control
**
Basal 3d Imm Sesn2WT Sesn2SkM-Tg Imm Imm
 
AAV-Control
AAV-Control Imm
AAV-Sesn1
AAV-Control Imm
 
R
FP
+G
FP
+
 
pu
nc
ta
 (%
)
 
R
FP
+G
FP
+
 
pu
nc
ta
 (%
)
AAV-
Sesn1
  AAV-
Control
 AAV-
Sesn1
  AAV-
Control
Sesn1KO 
Sesn1KO 
Sesn1KOImm 
 
R
FP
+G
FP
+
 
pu
nc
ta
 (%
)
LC
3-
II/
G
AP
DH
(fo
ld 
inc
rea
se
co
lc
hi
ci
ne
 v
s 
ve
hi
cl
e)
LC
3-
II/
G
AP
DH
(fo
ld 
inc
rea
se
co
lc
hi
ci
ne
 v
s 
ve
hi
cl
e) *
pS
6/
S6
*
pS6
Basal 3d Imm Basal 3d Imm Basal 3d Imm Basal 3d Imm
*
pULK1
0.0
0.5
1.0
1.5
* *
pU
LK
1/
UL
K1
Basal
Sesn
2
WT
Sesn
2
SkM
-Tg
Sesn
2
WT
Sesn
2
SkM
-Tg
3d Imm Basal 3d Imm
*
Sesn2SkM-Tg
Sesn2WT
Sesn2SkM-Tg
Sesn2SkM-TgImm
Sesn2WT
Sesn2WTImm
0.0
0.5
1.0
1.5
2.0
0.0
0.5
1.0
1.5
2.0
*
pU
LK
1/
UL
K1
pULK1
5
10
15
20
25
10 20 30 40
mTORC1 signaling
EMT
UV response
Myc targets V1
p53 targets
Sesn-regulated genes
hallmarks
Genes
–
lo
g1
0 
p 
va
lu
e
0
1000
2000
3000
CS
A 
(µm
2 )
0
1000
2000
3000
CS
A 
(µm
2 )
CS
A 
(µm
2 )
CS
A 
(µm
2 )
0
1000
2000
3000
Basal
Basal spermidine
10d imm
10d imm spermidine
*
ns
0
1
2
3
0
1
2
3
4
0
1
2
3
4
0
1
2
3
0.0
0.5
1.0
1.5
2.0
2.5
0
20
40
60
80
*
*
kDa
35
35
55
130
130
kDa
35
35
55
130
130
kDa
15
35
kDa
15
35
ARTICLE NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-13832-9
8 NATURE COMMUNICATIONS |          (2020) 11:189 | https://doi.org/10.1038/s41467-019-13832-9 | www.nature.com/naturecommunications
Methods
Animals. The different mouse models used in this study were generated as follows:
The skeletal muscle-specific Sesn1 transgenic mouse model (Sesn1SkM-Tg) was
generated in C57BL/6 background and carry a transgene coding for human Sesn1
under the control of MCK promoter. The human Sesn1 cDNA sequence (from
LV-Sesn119) was subcloned into the pBluescript-MCK plasmid (a kind gift from
Markus Rüegg). A 4 kb PacI digestion fragment was excised, microdialyzed, and
microinjected into the pronuclei of fertilized mouse eggs (C57BL/6J × C57BL/6J) at
the Mouse Mutant Core Facility, Institute for Research in Biomedicine (Barcelona,
Spain). Embryos were implanted into pseudo-pregnant foster females (ICR), and
transgenic pups were identified. DNA samples from tail clips of subsequent litters
were screened by PCR with primers spanning different sequences of MCK promoter
and Sesn1 cDNA (Primer set 1 forward CTGCCCCCGGGTCACCACC reverse
TCGATTCAGGTCATATAGCGGGT; Primer set 2 forward CAGGGCTTATAC
GTGCCTGGGACTC reverse TGTGGGTGGAAAACCATCACTAACG; Primer set
3 forward GCCCCCGGGTCACCACCAAG reverse GGCCAAGCGCATGGATCC
TTTTA) that amplified a 1550, 600, and 76 bp fragments, respectively. The transgene
was maintained on the C57BL/6J background throughout the study.
The skeletal muscle-specific Sesn2 transgenic mouse model (Sesn2SkM-Tg) was
generated by crossing mice that express Sesn2 from a Tet-regulated promoter
(provided by Dr. M. Karin, UCSD, San Diego, USA) with mice harboring the
skeletal muscle-specific MCK-tTA construct (obtained from Dr. M. Ruegg,
Biozentrum, Basel, Switzerland). Non-transgenic littermates (that were WT for
Sesn1 and Sesn2 expressions) were used as controls for the two transgenic mouse
lines, and named Sesn1WT and Sesn2WT mice, respectively. Sesn1 KO mice were
provided by Dr. J.H. Lee (University of Michigan, USA) and analyzed in
comparison with normal, control WT mice. Atg7SkM-KO mice were generated as
previously described32, and FoxO1,3,4SkM-KO mice were obtained by crossing
FoxO1/3/4-floxed mice (a kind gift from Dr. M. Sandri, University of Padova, Italy)
with the Pax7Cre line (provided by Dr. M. Capecchi, University of Utah, USA).
Non-transgenic littermates (that were WT for FoxO1,3,4 and for Atg7 expression)
were used as controls for Atg7SkM-KO mice and for FoxO1,3,4SkM-KO mice, and
named Atg7WT and FoxO1,3,4WT mice, respectively. Therefore, Sesn1WT,
Sesn2WT, WT, Atg7WT, and FoxO1,3,4WT mice are all equally WT, non-transgenic
mice, but each one is used as littermate control of each individual genetically
modified mouse line.
Mice were housed in standard cages under 12 h light/dark cycles with ad
libitum access to food and water. All experiments were performed on 4–5-month-
old male mice. All animal experiments were approved by the Ethics Committee of
the Barcelona Biomedical Research Park (PRBB) and performed according to
Catalan and European legislation.
Induction of muscle atrophy. The immobilization protocol was performed uni-
laterally on anesthetized animals as formerly described39. Briefly, the right hin-
dlimb was immobilized with rigid plastic sticks fixed with a medical adhesive
bandage. This procedure prevented movement of the immobilized leg alone.
Denervation was performed as previously described40. In brief, a 5 mm segment of
the sciatic nerve was surgically removed down to the gluteus maximum from the
right leg. Mice not subjected to atrophy-promoting conditions were used as con-
trols (Basal). Muscles were removed at 3 or 10 days after atrophy induction and
frozen in liquid nitrogen for subsequent analyses.
In vivo gene electrotransfer and AAV injection. Expression plasmids used for
electrotransfer studies were purified using a Endofree plasmid kit (Qiagen) and
dissolved in 0.9% NaCl. 45 min before electrotransfer, muscles were pretreated with
hyaluronidase (10 U/muscle). Afterwards, naked plasmids (60 μg DNA) were
injected into the tibialis anterior muscle and 10 pulses of 20 ms each were applied
to each hindlimb at 175 V/cm and 1 Hz using an electroporator (ECM 830; BTX).
Empty vector was used as control.
AAV for in vivo expression of Sesn1 and Sesn2 were generated and provided by
the Virus Production Unit (UPV, UAB, Barcelona). AAVs were diluted in 0.9%
NaCl at 0.25*1013 gc/ml and directly injected into muscles (40 µl/TA and 10 µl/EDL
and soleus muscles). AAV-GFP was used as a control. Four days after transduction
mice were subjected to the immobilization protocol.
Pharmacological treatments. Mice were injected with rapamycin (4 mg/kg body
weight) (which induces autophagy) or vehicle (DMSO) intraperitoneally (i.p.)
every other day for 2 weeks. Colchicine (which inhibits autophagy) was injected i.p.
(0.4 mg/kg*day) 2 days before sacrifice. The proteasome inhibitor Bortezomib
(0.1 mg/kg) or vehicle (DMSO) was injected i.p. every other day for 2 weeks. Mice
were treated with 3 mM spermidine in drinking water for 2 weeks.
Muscle force measurement. Ex vivo force measurements of EDL and soleus
muscles was assessed as previously described41. Briefly, mice were sacrificed, and
muscles were immediately excised and placed into a dish containing oxygenated
Krebs–Henseleit solution. Muscles were mounted vertically in a temperature
controlled (30 °C) chamber and immersed in the Krebs–Ringer bicarbonate buffer
solution, with 10 mM glucose, also continuously oxygenated. One end of the
muscle was linked to a fixed clamp, while the other end was connected to the lever-
arm of an Aurora Scientific Instruments 300B actuator/transducer system, using a
nylon thread. The optimum muscle length (Lo) was determined from micro-
manipulations of muscle length to produce the maximum isometric twitch force.
Maximum isometric-specific tetanic force was determined from the plateau of the
curve of the relationship between specific isometric force with a stimulation fre-
quency ranging from 1 to 200 Hz. Force was normalized per muscle area (deter-
mined by dividing the muscle mass by the product of longitude and the density of
muscle (1.06 mg/mm3)) to calculate the specific force (mN/mm2).
Muscle histology and immunohistochemistry. Muscles were embedded in OCT
solution (TissueTek), frozen in isopentane cooled with liquid nitrogen and stored
at −80 °C until analysis. 10 μm muscle cryosections were collected and stained for
hematoxylin/eosin (H/E) or Sirius red (Sigma-Aldrich).
For immunohistochemistry assays, muscle cryosections were examined by
standard immunohistochemical procedures for the expression of myosin heavy
chain (MHC) isoforms. The primary monoclonal antibodies employed were anti-
myosin I (A4.840), anti-myosin IIA (A4.74), and anti-myosin IIB (BF-F3)
(Developmental Studies Hybridoma Bank).
Digital images were acquired using the Leica DMR600B microscope equipped
with a DFC300FX camera. Fiber type distribution, CSA, and percentage of muscle
area positive for Sirius red staining were quantified using Image J software, as
previously reported42,43.
For myonuclei quantification, muscle sections were immunostained for
dystrophin (1/400), the secondary antibody was coupled to Alexa-488 and nuclei
were stained with DAPI (Invitrogen). Images were acquired using a Leica TCS SP5
confocal scanning microscope system.
Fluorescence microscopy analysis of muscle sections. TA muscles were
removed, fixed in PFA 2% for 4 h at 4 °C and incubated with 15% sucrose overnight
at 4 °C. Then, muscles were embedded in OCT solution (TissueTek), immediately
frozen in liquid nitrogen-cooled isopentane and stored at −80 °C. 10 μm cryo-
sections of TA muscle (which has been transfected with mRFP-GFP-LC344) were
analyzed using a Leica TCS SP5 confocal scanning microscope system. Colocali-
zation of RFP-LC3 and GFP-LC3 puncta was determined on the maximum
Fig. 4 Autophagy induction via sestrin-mediated mTORC1 blockade prevents atrophy. a Bubble plot showing the main molecular hallmarks (GSEA)
enriched in the sestrin-regulated gene set defined in Supplementary Fig. 4a. b Western blot analysis and quantification of S6 and ULK1 phosphorylation in
muscles from Sesn2WT and Sesn2SkM-Tg mice (left) and in muscles transduced with AAV-Sesn1 or AAV-Control (right) in basal conditions and at 3 days
post-immobilization. Values were normalized to basal control conditions. c Western blot analysis and quantification of LC3I and LC3II in TA muscles
transduced with AAV-Sesn1 or AAV-Control in basal conditions and at 3 days post-immobilization. Treatment of mice with colchicine or vehicle is
indicated. Lower chart shows the fold increase in LC3II content in colchicine-treated versus vehicle-treated mice. d Representative confocal images of
3-day-immobilized TA muscle electrotransferred with a tandem mRFP-GFP-LC3 reporter plasmid to enable detection of autophagosomes (yellow puncta)
and autolysosomes (red puncta). The chart shows double RFP+GFP+ puncta as a percentage of total puncta for the indicated genotypes. Scale bar= 5 µm.
e Representative confocal images of non-immobilized muscles in WT and Sesn1KO mice treated as in d. The chart shows RFP+GFP+ puncta as a
percentage of total puncta. Scale bar= 5 µm. f Mean TA myofiber CSA in basal conditions and at 10 days post-immobilization in muscles transduced with
AAV-Control or AAV-Sesn1 and electrotransferred with control short hairpin (sh) or sh targeting TSC2. g Mean TA myofiber CSA in basal conditions and
at 10 days post-immobilization in WT mice treated with vehicle (top) or rapamycin (bottom). h Mean TA myofiber CSA in basal conditions and at 10 days
post-immobilization in WT mice treated with vehicle or spermidine. i Mean TA myofiber CSA in basal conditions and at 10 days post-immobilization in
Atg7WT and Atg7SkM-KO mice transduced with AAV-Control or AAV-hSesn1. All data are shown as mean with SEM. Comparisons by Student's t-test (*p <
0.05 and **p < 0.01 vs. basal). Sample numbers were n= 3 mice per group for b, d, e, n= 3–5 animals for c, f, g, i and n= 4 for h. Source data are provided
as a Source Data file.
NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-13832-9 ARTICLE
NATURE COMMUNICATIONS |          (2020) 11:189 | https://doi.org/10.1038/s41467-019-13832-9 |www.nature.com/naturecommunications 9
projection of 10-z sections. Note: Measuring autophagy flux through this method is
based on the concept of lysosomal quenching of GFP. GFP is a stably folded
protein and relatively resistant to lysosomal proteases. However, the low pH inside
the lysosome quenches the fluorescent signal of GFP, which makes it difficult to
trace the delivery of GFP–LC3 to lysosomes. In contrast, RFP exhibits more stable
fluorescence in acidic compartments, and mRFP–LC3 can be readily detected in
autolysosomes. By exploiting the difference in the nature of these two fluorescent
proteins (that is, lysosomal quenching of GFP fluorescence versus lysosomal sta-
bility of RFP fluorescence), autophagic flux can be morphologically traced with an
mRFP–GFP–LC3 tandem construct. With this tandem construct, autophagosomes
and autolysosomes are labeled with yellow (mRFP and GFP) and red (mRFP only)
signals, respectively.
0 50 100 150 200
0
50
100
150
200
250
Young AAV-Control
Old AAV-Control
Old AAV-Sesn1
Freq (Hz)
Ab
so
lu
te
 fo
rc
e 
(m
N)**
CS
A 
(µm
2 )
*
a b
Tubulin
Sesn1
Young Old
0.0
0.5
1.0
1.5
Young Old
Se
sn
1/
tu
bu
lin
*
Young Old
AAV-
Sesn1
AAV-
Control
  AAV-
Control
Young Old
AAV-
Sesn1
AAV-
Control
  AAV-
Control
Young Old
AAV-
Sesn1
AAV-
Control
  AAV-
Control
Young Old
AAV-
Sesn1
AAV-
Control
  AAV-
Control
Young Old
AAV-
Sesn1
AAV-
Control
  AAV-
Control
0
20
40
60
W
ei
gh
t (m
g)
*
*
0 50 100 150 200
0
50
100
150
200
250
Young
Old
Freq (Hz)
Ab
so
lu
te
 fo
rc
e 
(m
N)
*
c
ed
g
h i
pS6
S6
Tubulin
pULK1
ULK1
pp70S6K
p70S6K
Young AAV-
Control
Old AAV-
Control
Old AAV-
Sesn1
0
1
2
3
4
5
pULK1
pU
LK
1/
UL
K1
0
2
4
6
pp70S6K
pp
70
S6
K/
p7
0S
6K
f
0.0
0.5
1.0
1.5
2.0
pS6
pS
6/
S6
* *
Protein secretion
Mitotic spindle
UV response DN
KRas signaling UP
Myc targets V1
TGFβ signaling
G2M checkpoint
Inflammatory response
Apoptosis
E2F targets
PI3K-AKT- MTOR signaling
MTORC1 signaling
Hedgehog signaling
0.8 0.9 1.0 1.1 1.2 1.3
NES
Genes in gene set (%)
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
–
lo
g1
0 
p 
va
lu
e
–
lo
g1
0 
p 
va
lu
eNES 1.358
5 10 15 20 25 30 35 40 45
4
6
8
10
12
14 E12
SP1
Unknown Unknown
MEF2
LEF1
LEF1
MAZ AP4
MYOD
NFAT
PAX4 FOXO
NES 1.270
Old Yo
un
g
Mt1
Fbxo30 (Musa1)
Psme4
Sqstm1 (p62)
Ezh1
Psdm11
Fbxo31
Bnip3
Ube4b
Eif4g3
Fbxo32 (Atrogin1)
Sesn1
Tubulin
 AAV-
Sesn1
AAV-
ControlAAV-Control
AAV-Sesn1
R
el
at
iv
e 
hS
es
n1
 
m
R
N
A 
ex
pr
es
sio
n *
AAV-Sesn1AAV-ControlAAV-Control
Young Old
0
400
800
1200
0
500
1000
1500
2000
2500
3000
3500
kDa
55
55
70
kDa
55
55
70
kDa
130
70
130
70
35
35
55
Enrichment plot: FOXO_ TARGETS
Enrichment plot: AKT_MTORC1_SIGNALINGEnrichment plot:
HALLMARK_PI3K_AKT_MTOR_SIGNALING
En
ric
hm
en
t s
co
re
 (E
S)
R
an
ke
d 
lis
t m
et
ric
 (S
ign
al2
No
ise
)
En
ric
hm
en
t s
co
re
 (E
S)
R
an
ke
d 
lis
t m
et
ric
 (S
ign
al2
No
ise
)
En
ric
hm
en
t s
co
re
 (E
S)
R
an
ke
d 
lis
t m
et
ric
 (S
ign
al2
No
ise
)
0.5
0.4
0.3
0.2
0.1
0.0
2.0 'Old' (positively correlated)
'Young' (negatively correlated)
1.5
1.0
0.5
Zero cross at 13754
'Old' (positively correlated)
'Old' (positively correlated)
Zero cross at 9122Zero cross at 13754
0.0
–0.5
–1.0
–1.5
0
0
50002500 7500 10,000
Rank in Ordered Dataset
Enrichment profile Hits Ranking metric scores
12,500 15,000 17,500 20,000 22,500 25,000
0
0
1
–1
2
50002500 7500 10,000
Rank in Ordered Dataset
12,500 15,000 17,500 20,000 22,500 25,000
'Young' (negatively correlated)
0
3
2
1
0
–1
–2
50002500 7500 10,000
Rank in Ordered Dataset
Enrichment profile Hits Ranking metric scoresEnrichment profile Hits Ranking metric scores
12,500 15,000 17,500 20,000
–0.1
0.4
0.25
0.20
0.15
0.10
0.05
0.00
–0.05
–0.10
–0.15
0.3
0.2
0.1
0.0
–0.1
Enrichment plot: ATROGENES
En
ric
hm
en
t s
co
re
 (E
S)
R
an
ke
d 
lis
t m
et
ric
 (S
ign
al2
No
ise
)
0.4
0.3
0.2
0.1
0.0
'Old' (positively correlated)
'Young' (negatively correlated)
Zero cross at 9122
3
2
1
0
–1
–2
50002500 7500 10,000
Rank in Ordered Dataset
Enrichment profile Hits Ranking metric scores
12,500 15,000 17,500 20,000
–0.2
–0.1
ARTICLE NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-13832-9
10 NATURE COMMUNICATIONS |          (2020) 11:189 | https://doi.org/10.1038/s41467-019-13832-9 | www.nature.com/naturecommunications
RNA isolation, reverse transcription, and quantitative PCR. Total RNA from
TA muscles was isolated with QIAzol Lysis Reagent (Qiagen) and quantified with
Nanodrop. M-MLV Reverse Transcriptase (Promega) was used to synthesize
cDNAs from 1 µg total RNA following the manufacturer’s instructions. RT-qPCR
reactions were performed with SYBR Green in 384-well plates using the Roche LC-
480 cycler (Roche Applied Science). All data were normalized to L7 expression.
Primer sequences are listed in Table 1.
Transcriptomic analysis. For RNAseq analysis, total RNA from TA muscles was
extracted using a protocol combining QIAzol Lysis Reagent and RNAeasy minikit
columns (Qiagen) following the manufacturer’s instructions. RNAseq services were
provided by the CNIC Genomics Unit, including quality control tests of total RNA
using Agilent Bioanalyzer and Nanodrop spectrophotometry. cDNA library pre-
paration and amplification were performed from 200 ng total RNA using NEBNext
Ultra RNA Library Prep Kit for Illumina. RNAseq analysis was performed with
3–4 samples per condition, using Illumina Hiseq 2500. Sequencing reads were pre-
processed by means of a pipeline that used FastQC, to asses read quality, and
Cutadapt 1.7.1 to trim sequencing reads, eliminating Illumina adaptor remains, and
to discard reads that were shorter than 30 bp. The resulting reads were mapped
against the mouse transcriptome (GRCm38, release 76; aug2014 archive) and
quantified using RSEM v1.2.20. Data were then processed with a differential
expression analysis pipeline that used Bioconductor package LIMMA for nor-
malization and differential expression testing. For differential expression analysis
we filtered for genes that showed at least Log2FC 0.25 (≥+0.25 for upregulation;
and ≤−0.25 for downregulation) and an adj. p value < 0.05.
FoxO promoter-reporter assay. TA muscles from Sesn1WT and Sesn1SkM-Tg mice
were electrotransferred with a reporter plasmid with three copies of forkhead
responsive element linked to the luciferase reporter gene (FHRE-luciferase;
Addgene). After 3 days of immobilization, muscles were collected and luciferase
activity was measured in muscle homogenates by using Dual-Luciferase Reporter
Assay Kit (Promega Corporation, USA). Values were normalized to non-
immobilized muscles.
Muscle protein extraction and Western blotting. Total homogenates from
skeletal muscle were obtained in IP buffer (50mM Tris–HCl pH 7.5, 150mM NaCl,
1% NP-40, 5 mM EGTA, 5mM EDTA, 20mM NaF, 25mM β-glycerophosphate,
0.1 mM sodium vanadate, 1 mM PMSF) supplemented with protease and phos-
phatase inhibitors (Complete Mini, Roche Diagnostic Corporation; phosphatase
inhibitor cocktail, Sigma). Protein concentration was measured using the Bradford
method (Protein Assay, Bio-Rad). 40 µg of protein were resolved by SDS–PAGE and
transferred to PVDF membranes (Millipore). Membranes were blocked with 5%
milk in TBS-T for 1 h and incubated with primary antibodies overnight at 4 °C in
5% BSA in TBS-T (Tubulin 1:4000 and others 1:1000). Proteins were detected by the
ECL method and quantified by scanning densitometry. The antibodies used are
listed in Table 2. Uncropped versions of all blots shown in figures are supplied in the
Source Data File.
Proteasome activity in muscle. Proteasome activity in total homogenates from
TA muscles was determined by evaluating the cleavage of specific fluorogenic
substrates. Muscles were homogenized in lysis buffer (50 mM Tris–HCl pH 7.5,
250 mM Sucrose, 5 mM MgCl2, 0.5 mM EDTA, 2 mM ATP, and 1 mM DTT) and
centrifuged at 12,000 × g for 30 min at 4 °C. The supernatant was collected and
protein concentration determined by the method of Bradford. For chymotrypsin-
like activity, aliquots of 20 μg protein were incubated for 60 min at 37 °C in the
presence of 100 μM of the fluorogenic substrate succinyl-Leu-Leu-Val-Tyr-7-
amino-4-methylcoumarin (Suc LLVY-AMC). Each assay was conducted in the
absence and presence of the specific proteasomal inhibitor MG132 (Sigma-Aldrich)
at 20 μM. Fluorescence was read with a spectrofluorometer (390 nm excitation/
460 nm emission; Tecan Infinite M200). The activity was expressed as units of
fluorescence per microgram of protein, as a percentage of the control group. All
samples were assayed in triplicate using at least four animals.
Bioinformatic analysis. Hierarchical clustering of expression values (after filtering
expression values below one in all the samples) was carried out with Morpheus
(https://software.broadinstitute.org/morpheus/) using one minus Pearson correla-
tion with a complete linkage. GSEA of Sestrin-regulated genes was performed using
GSEA web interface with the Molecular Signatures Database “hallmarks” and
“transcription factor binding targets” genesets, to reveal pathways and cis-
regulatory motifs which can function as potential transcription factor-binding sites,
respectively45. The Java implementation code from the Broad Institute was used for
direct GSEA comparisons of the raw data from our RNA-seq experiments or with
GEO: GSE5395 data set with the “AKT-MTOR signaling” gene set (generated by
combining the “MTORC1 Signaling” and the “PI3K-AKT- MTOR signaling”
Fig. 5 Sestrins prevent aging-related muscle atrophy. aWestern blot and quantification of Sesn1 protein levels in skeletal muscle from young (4 months)
and old (24 months) mice. b Force–frequency curve of EDL muscles from young and old WT mice. c qPCR of human Sesn1 mRNA in skeletal muscle from
24-month-old mice transduced with AAV-Sesn1 for one month and Sesn1 protein expression by Western blotting in the same muscles (left). Weight of TA
muscles from young and old mice transduced with AAV-Sesn1 or AAV-Control for 1 month (right). d Representative H/E staining pictures of TA muscle
sections from 24-month-old mice transduced with AAV-Sesn1 or AAV-Control for 1 month and quantification of mean myofiber CSA. Scale bar= 50 µm.
e EDL force-frequency curve in 24-month-old mice transduced with AAV-Sesn1 or AAV-Control for 1 month. f Western blot analysis and quantification of
phosphorylation levels of p70S6K, S6, and ULK1 in muscles from old mice transduced with AAV-Sesn1 or AAV-Control for 1 month. Representative blots
are shown with corresponding quantification. Values were normalized to averaged values of young control mice. g Bubble plot of enriched GSEA hallmarks
in the dysregulated genes upon muscle aging (left). Enrichment plots of the PI3K-AKT-MTOR signaling hallmark and the combined PI3K-AKT-MTOR
signaling and mTORC1 signaling hallmarks (AKT-MTORC1 signaling) in comparisons of old and young mice (right). h As in g, enrichment plots of the
atrogenes gene set in comparisons of old and young mice. i Bubble plot of enriched transcription factor-binding sites (GSEA) in the dysregulated genes
upon muscle aging (left). Enrichment plots of the FoxO3a targets in skeletal muscle (as defined in Brocca et al.24) and heatmap illustrating genes with
higher enrichment in comparisons of old and young mice (right). All data are shown as mean with SEM. Comparisons by Student's t-test (*p < 0.05).
Sample numbers were n= 8 mice per group for a, n= 4–7 mice per condition for b and c, n= 3–5 mice per condition for d and e and n= 3 mice per group
for f. Source data are provided as a Source Data file.
Table 1 Primers used for qPCR.
Gene Species Forward primer Reverse primer
Sesn1 Mouse GTCTGGATAACATCACATTAG CCAGGTAGGAACACTGATGC
Human CAGCATTGGAAAACATTAGGCAA CCGAAGACTCGGTATTTGAAAGC
Sesn2 Mouse TAGCCTGCAGCCTCACCTAT TATCTGATGCCAAAGACGCA
Sesn3 Mouse CATGCGTTTCCTCACTCAGA GGCAAAGTCTTCGTACCCAA
Fbxo32 Mouse AAGGCTGTTGGAGCTGATAGCA CACCCACATGTTAATGTTGCCC
Trim63 Mouse TGCCTGGAGATGTTTACCAAGC AAACGACCTCCAGACATGGACA
Ctsl Mouse GTGGACTGTTCTCACGCTCAAG TCCGTCCTTCGCTTCATAGG
FoxO1 Mouse CTGGGTGTCAGGCTAAGAGT GGGGTGAAGGGCATCTTT
FoxO3 Mouse CTGGGGGAACCTGTCCTATG TCATTCTGAACGCGCATGAAG
FoxO4 Mouse CTTCCTCGACCAGACCTCG ACAGGATCGGTTCGGAGTGT
Atg7 Mouse TCTGGGAAGCCATAAAGTCAGG GCGAAGGTCAGGAGCAGAA
L7 Mouse GAAGCTCATCTATGAGAAGGC AAGACGAAGGAGCTGCAGAAC
NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-13832-9 ARTICLE
NATURE COMMUNICATIONS |          (2020) 11:189 | https://doi.org/10.1038/s41467-019-13832-9 |www.nature.com/naturecommunications 11
Molecular Signatures Database hallmarks), the growth and atrophy regulators gene
set (assembled by combining the following Gene Ontology categories: positive and
negative regulations of insulin-like growth factor receptor signaling pathway,
positive and negative regulations of TORC1 signaling and positive and negative
regulations of muscle atrophy) and the atrogenes gene set (that was manually
curated from selected bibliography1–4,6,7,22,24). Bubble plots were generated using
ggplot2 library in R or Seaborn in Python. Venn diagrams were generated using
The BEG Ugent tool (http://bioinformatics.psb.ugent.be/webtools/Venn/).
Statistical analysis. For mouse experiments, no specific blinding method was
used, but mice in each sample group were selected randomly. The sample size (n)
of each experimental group is described in each corresponding figure legend.
GraphPad Prism software was used for all statistical analyses. Quantitative data
displayed as histograms are expressed as means ± standard error of the mean
(represented as error bars). Results from each group were averaged and used to
calculate descriptive statistics.
Unpaired t-test (independent samples, two-sided) was used for pairwise
comparisons among groups at each time point, unless indicated in figure legends.
Statistical significance was set at a p < 0.05.
Data availability
The data that support the findings of this study are available from the corresponding
author upon reasonable request. Source data used to generate Figs. 1–5 and
Supplementary Figs. 1–7 are provided in the Source Data file. The RNA-sequencing data
have been deposited in the Gene Expression Omnibus (GEO) database under accession
code: GSE136866.
Received: 25 February 2019; Accepted: 6 November 2019;
References
1. Egerman, M. A. & Glass, D. J. Signaling pathways controlling skeletal muscle
mass. Crit. Rev. Biochem. Mol. Biol. 49, 59–68 (2014).
2. Schiaffino, S. & Mammucari, C. Regulation of skeletal muscle growth by the
IGF1-Akt/PKB pathway: insights from genetic models. Skelet. Muscle 1, 4 (2011).
3. Bodine, S. C. et al. Identification of ubiquitin ligases required for skeletal
muscle atrophy. Science 294, 1704–1708 (2001).
4. Gomes, M. D., Lecker, S. H., Jagoe, R. T., Navon, A. & Goldberg, A. L.
Atrogin-1, a muscle-specific F-box protein highly expressed during muscle
atrophy. Proc. Natl Acad. Sci. USA 98, 14440–14445 (2001).
5. Mammucari, C. et al. FoxO3 controls autophagy in skeletal muscle in vivo.
Cell Metab. 6, 458–471 (2007).
6. Milan, G. et al. Regulation of autophagy and the ubiquitin-proteasome system
by the FoxO transcriptional network during muscle atrophy. Nat. Commun. 6,
6670 (2015).
7. Sandri, M. et al. Foxo transcription factors induce the atrophy-related ubiquitin
ligase atrogin-1 and cause skeletal muscle atrophy. Cell 117, 399–412 (2004).
8. Stitt, T. N. et al. The IGF-1/PI3K/Akt pathway prevents expression of muscle
atrophy-induced ubiquitin ligases by inhibiting FOXO transcription factors.
Mol. Cell 14, 395–403 (2004).
9. Lee, J. H., Budanov, A. V. & Karin, M. Sestrins orchestrate cellular metabolism
to attenuate aging. Cell Metab. 18, 792–801 (2013).
10. Ho, A., Cho, C. S., Namkoong, S., Cho, U. S. & Lee, J. H. Biochemical basis of
sestrin physiological activities. Trends Biochem. Sci. 41, 621–632 (2016).
11. Lee, J. H. et al. Sestrin as a feedback inhibitor of TOR that prevents age-related
pathologies. Science 327, 1223–1228 (2010).
12. Yang, Y. L. et al. SESN-1 is a positive regulator of lifespan in Caenorhabditis
elegans. Exp. Gerontol. https://doi.org/10.1016/j.exger.2012.12.011 (2013).
13. Kim, J., Kundu, M., Viollet, B. & Guan, K. L. AMPK and mTOR regulate
autophagy through direct phosphorylation of Ulk1. Nat. Cell Biol. 13, 132–141
(2011).
14. Lee, J. H. et al. Maintenance of metabolic homeostasis by Sestrin2 and
Sestrin3. Cell Metab. 16, 311–321 (2012).
15. Tao, R., Xiong, X., Liangpunsakul, S. & Dong, X. C. Sestrin 3 protein enhances
hepatic insulin sensitivity by direct activation of the mTORC2-Akt signaling.
Diabetes 64, 1211–1223 (2015).
16. Peeters, H. et al. PA26 is a candidate gene for heterotaxia in humans:
identification of a novel PA26-related gene family in human and mouse. Hum.
Genet. 112, 573–580 (2003).
17. Byun, J. K. et al. A positive feedback loop between Sestrin2 and mTORC2 Is
required for the survival of glutamine-depleted lung cancer cells. Cell Rep. 20,
586–599 (2017).
18. Wolfson, R. L. et al. Sestrin2 is a leucine sensor for the mTORC1 pathway.
Science 351, 43–48 (2016).
19. Budanov, A. V., Sablina, A. A., Feinstein, E., Koonin, E. V. & Chumakov, P.
M. Regeneration of peroxiredoxins by p53-regulated sestrins, homologs of
bacterial AhpD. Science 304, 596–600 (2004).
20. Lanna, A. et al. A sestrin-dependent Erk-Jnk-p38 MAPK activation complex
inhibits immunity during aging. Nat. Immunol. 18, 354–363 (2017).
21. Calura, E. et al. Meta-analysis of expression signatures of muscle atrophy: gene
interaction networks in early and late stages. BMC Genomics 9, 630 (2008).
22. Lecker, S. H. et al. Multiple types of skeletal muscle atrophy involve a common
program of changes in gene expression. FASEB J. 18, 39–51 (2004).
23. Sartori, R. et al. BMP signaling controls muscle mass. Nat. Genet. 45,
1309–1318 (2013).
24. Brocca, L. et al. FoxO-dependent atrogenes vary among catabolic conditions
and play a key role in muscle atrophy induced by hindlimb suspension. J.
Physiol. 595, 1143–1158 (2017).
25. Jung, C. H. et al. ULK-Atg13-FIP200 complexes mediate mTOR signaling to
the autophagy machinery. Mol. Biol. Cell 20, 1992–2003 (2009).
26. Ju, J. S., Varadhachary, A. S., Miller, S. E. & Weihl, C. C. Quantitation of
“autophagic flux” in mature skeletal muscle. Autophagy 6, 929–935 (2010).
27. Kimura, S., Noda, T. & Yoshimori, T. Dissection of the autophagosome
maturation process by a novel reporter protein, tandem fluorescent-tagged
LC3. Autophagy 3, 452–460 (2007).
28. Budanov, A. V. & Karin, M. p53 target genes sestrin1 and sestrin2 connect
genotoxic stress and mTOR signaling. Cell 134, 451–460 (2008).
29. Inoki, K., Li, Y., Xu, T. & Guan, K. L. Rheb GTPase is a direct target of TSC2
GAP activity and regulates mTOR signaling. Genes Dev. 17, 1829–1834
(2003).
30. Tee, A. R., Manning, B. D., Roux, P. P., Cantley, L. C. & Blenis, J. Tuberous
sclerosis complex gene products, Tuberin and Hamartin, control mTOR
signaling by acting as a GTPase-activating protein complex toward Rheb.
Curr. Biol. 13, 1259–1268 (2003).
Table 2 Antibodies.
Antibody Source Reference
Rabbit anti-Sesn1 Provided by Dr. Jun Hee Lee
Rabbit anti-Sesn2 Proteintech Cat#10795-1-AP
Rabbit anti-phospho-AKT (Ser473) Cell Signaling Technology Cat#9271
Rabbit anti-phospho-AKT (Trh308) Cell Signaling Technology Cat#4056
Mouse anti-AKT (pan) Cell Signaling Technology Cat#2920
Rabbit anti-phospho-FoxO1 (Thr24)/FoxO3a (Thr32) Cell Signaling Technology Cat#9464
Rabbit anti-phospho-S6 (Ser235/236) Cell Signaling Technology Cat#4858
Mouse anti-S6 Cell Signaling Technology Cat#2317
Anti-Phospho-AMPKα (Thr172) Cell Signaling Technology Cat#2531
Anti-AMPKα Cell Signaling Technology Cat#2532
Rabbit anti-phospho-ULK1 Cell Signaling Technology Cat#14202
Rabbit anti-ULK1 Cell Signaling Technology Cat#8054
Rabbit anti-LC3B Cell Signaling Technology Cat#2775
Mouse anti-Tubulin Sigma Aldrich Cat#T-6199
Mouse anti-GAPDH Santa Cruz Biotechnology Cat#sc-32233
ARTICLE NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-13832-9
12 NATURE COMMUNICATIONS |          (2020) 11:189 | https://doi.org/10.1038/s41467-019-13832-9 | www.nature.com/naturecommunications
31. Castets, P. et al. Sustained activation of mTORC1 in skeletal muscle inhibits
constitutive and starvation-induced autophagy and causes a severe, late-onset
myopathy. Cell Metab. 17, 731–744 (2013).
32. Garcia-Prat, L. et al. Autophagy maintains stemness by preventing senescence.
Nature 529, 37–42 (2016).
33. Morselli, E. et al. Spermidine and resveratrol induce autophagy by distinct
pathways converging on the acetylproteome. J. Cell Biol. 192, 615–629 (2011).
34. Barns, M. et al. Molecular analyses provide insight into mechanisms
underlying sarcopenia and myofibre denervation in old skeletal muscles of
mice. Int. J. Biochem. Cell Biol. 53, 174–185 (2014).
35. Masiero, E. et al. Autophagy is required to maintain muscle mass. Cell Metab.
10, 507–515 (2009).
36. Zeng, N., D’Souza, R. F., Mitchell, C. J. & Cameron-Smith, D. Sestrins are
differentially expressed with age in the skeletal muscle of men: a cross-
sectional analysis. Exp. Gerontol. 110, 23–34 (2018).
37. Rai, N. et al. Exploration of novel anti-oxidant protein sestrin in frailty
syndrome in elderly. Aging Dis. 9, 220–227 (2018).
38. Kato, T. et al. Sestrin modulator NV-5138 produces rapid antidepressant
effects via direct mTORC1 activation. J. Clin. Invest. 129, 2542–2554
(2019).
39. Madaro, L., Smeriglio, P., Molinaro, M. & Bouche, M. Unilateral
immobilization: a simple model of limb atrophy in mice. Basic Appl. Myol. 18,
149–153 (2008).
40. Serrano, A. L. et al. Calcineurin controls nerve activity-dependent
specification of slow skeletal muscle fibers but not muscle growth. Proc. Natl
Acad. Sci. USA 98, 13108–13113 (2001).
41. Del Prete, Z., Musaro, A. & Rizzuto, E. Measuring mechanical properties,
including isotonic fatigue, of fast and slow MLC/mIgf-1 transgenic skeletal
muscle. Ann. Biomed. Eng. 36, 1281–1290 (2008).
42. Vidal, B. et al. Fibrinogen drives dystrophic muscle fibrosis via a TGFbeta/
alternative macrophage activation pathway. Genes Dev. 22, 1747–1752 (2008).
43. Ardite, E. et al. PAI-1-regulated miR-21 defines a novel age-associated
fibrogenic pathway in muscular dystrophy. J. Cell Biol. 196, 163–175 (2012).
44. Zhu, J., Dagda, R. K. & Chu, C. T. Monitoring mitophagy in neuronal cell
cultures. Methods Mol. Biol. 793, 325–339 (2011).
45. Subramanian, A. et al. Gene set enrichment analysis: a knowledge-based
approach for interpreting genome-wide expression profiles. Proc. Natl Acad.
Sci. USA 102, 15545–15550 (2005).
Acknowledgements
We thank members of our laboratory for discussions, and specially V. Lukesova, M.
Raya, and J.M. Ballesteros for technical and logistic help, and J.M. López Martí for
bioinformatic analysis support. We are greatly indebted to C. Keller and M. Capecchi for
Pax–Cre mouse lines, E. Masliah and K. Kosberg for Atg7 lentivirus, A. Webb for FoxO
lentivirus; J. Martiń-Caballero (PRBB Animal Facility) and J.M. Ballestero for excellent
animal care, the CRG/UPF Microscopy Facility, A. Dopazo (CNIC Genomics Unit) and
Fátima Sánchez-Cabo (CNIC Bioinformatics Unit), M. Martínez-Vicente for advice on
macroautophagy, and S. Bartlett for editorial support. The authors acknowledge funding
from the Spanish Ministry of Science, Innovation and Universities, Spain (grants
SAF2015-67369-R, RTI2018-096068-B-I00, and SAF 2015-70270-REDT, a Mariá de
Maeztu Unit of Excellence award to UPF [MDM-2014-0370], and the UPF-CNIC
collaboration agreement, ERC-2016-AdG-741966, La Caixa-HEALTH (HR17-00040),
MDA, UPGRADE-H2020-825825, AFM and DPP-E. The CNIC is supported by the
Instituto de Salud Carlos III (ISCIII), the Ministerio de Ciencia, Innovación y Uni-
versidades (MCNU) and the Pro CNIC Foundation, and is a Severo Ochoa Center of
Excellence (SEV-2015-0505). Work was also supported by the Ellison Medical Foun-
dation (AG-SS-2440-10 to M.K. and AG-NS-0932-12 to J.H.L.), NIH (R01DK114131,
R01DK111465, and R01DK102850 to J.H.L.), CARIPARO and H2020-MSCA-RISE-2014
(to M.S.), and the Russian Science Foundation (Grant 17-14-01420 to A.V.B.). J.S.
acknowledges funding from a Juan de La Cierva Postdoctoral Fellowship.
Author contributions
J.S., E.P., A.L.S., P.S-V., L.O., M.J. and D.M.T. performed experiments. E.P. performed
bioinformatics analyses. M.K., M.S., L.G-P. and J.H.L. provided several mouse strains.
A.V.B. provided some sestrin-related reagents. J.S. and P.M-C., conceived and designed
the project. J.S., E.P., A.L.S. and P.M-C. designed experiments and analyzed and inter-
preted data. J.S. and P.M-C. wrote the manuscript. E.P. and A.L.S. extensively and
critically revised the manuscript. All authors commented on the manuscript and
approved it.
Competing interests
The authors declare no competing interests.
Additional information
Supplementary information is available for this paper at https://doi.org/10.1038/s41467-
019-13832-9.
Correspondence and requests for materials should be addressed to P.Mño-C.
Peer review information Nature Communications thanks the anonymous reviewers for
their contribution to the peer review of this work.
Reprints and permission information is available at http://www.nature.com/reprints
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in
published maps and institutional affiliations.
Open Access This article is licensed under a Creative Commons
Attribution 4.0 International License, which permits use, sharing,
adaptation, distribution and reproduction in any medium or format, as long as you give
appropriate credit to the original author(s) and the source, provide a link to the Creative
Commons license, and indicate if changes were made. The images or other third party
material in this article are included in the article’s Creative Commons license, unless
indicated otherwise in a credit line to the material. If material is not included in the
article’s Creative Commons license and your intended use is not permitted by statutory
regulation or exceeds the permitted use, you will need to obtain permission directly from
the copyright holder. To view a copy of this license, visit http://creativecommons.org/
licenses/by/4.0/.
© The Author(s) 2020
NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-13832-9 ARTICLE
NATURE COMMUNICATIONS |          (2020) 11:189 | https://doi.org/10.1038/s41467-019-13832-9 |www.nature.com/naturecommunications 13