SIRT1 negatively regulates the mammalian target of rapamycin.

Abstract

The IGF/mTOR pathway, which is modulated by nutrients, growth factors, energy status and cellular stress regulates aging in various organisms. SIRT1 is a NAD+ dependent deacetylase that is known to regulate caloric restriction mediated longevity in model organisms, and has also been linked to the insulin/IGF signaling pathway. Here we investigated the potential regulation of mTOR signaling by SIRT1 in response to nutrients and cellular stress. We demonstrate that SIRT1 deficiency results in elevated mTOR signaling, which is not abolished by stress conditions. The SIRT1 activator resveratrol reduces, whereas SIRT1 inhibitor nicotinamide enhances mTOR activity in a SIRT1 dependent manner. Furthermore, we demonstrate that SIRT1 interacts with TSC2, a component of the mTOR inhibitory-complex upstream to mTORC1, and regulates mTOR signaling in a TSC2 dependent manner. These results demonstrate that SIRT1 negatively regulates mTOR signaling potentially through the TSC1/2 complex.

Full text

Page 1

SIRT1 Negatively Regulates the Mammalian Target of
Rapamycin
Hiyaa Singhee Ghosh1¤, Michael McBurney2, Paul D. Robbins1*
1Department of Microbiology and Molecular Genetics, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania, United States of America, 2Ottawa Health
Research Institute and Departments of Medicine and Biochemistry, Microbiology and Immunology, University of Ottawa, Ottawa, Canada
Abstract
The IGF/mTOR pathway, which is modulated by nutrients, growth factors, energy status and cellular stress regulates aging
in various organisms. SIRT1 is a NAD+ dependent deacetylase that is known to regulate caloric restriction mediated
longevity in model organisms, and has also been linked to the insulin/IGF signaling pathway. Here we investigated the
potential regulation of mTOR signaling by SIRT1 in response to nutrients and cellular stress. We demonstrate that SIRT1
deficiency results in elevated mTOR signaling, which is not abolished by stress conditions. The SIRT1 activator resveratrol
reduces, whereas SIRT1 inhibitor nicotinamide enhances mTOR activity in a SIRT1 dependent manner. Furthermore, we
demonstrate that SIRT1 interacts with TSC2, a component of the mTOR inhibitory-complex upstream to mTORC1, and
regulates mTOR signaling in a TSC2 dependent manner. These results demonstrate that SIRT1 negatively regulates mTOR
signaling potentially through the TSC1/2 complex.
Citation: Ghosh HS, McBurney M, Robbins PD (2010) SIRT1 Negatively Regulates the Mammalian Target of Rapamycin. PLoS ONE 5(2): e9199. doi:10.1371/
journal.pone.0009199
Editor: Mikhail V. Blagosklonny, Roswell Park Cancer Institute, United States of America
Received August 29, 2009; Accepted January 19, 2010; Published February 15, 2010
Copyright: ? 2010 Ghosh et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits
unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This work was supported in part by Department of Defense (DOD) grants 17-03-1-0488 and 17-03-0412 to P.D.R. The funders had no role in study
design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing Interests: The authors have declared that no competing interests exist.
* E-mail: probb@pitt.edu
¤ Current address: Columbia University Medical Center, New York, New York, United States of America
Introduction
The NAD+dependent deacetylase, SIRT1 (Sir2) has been
shown to regulate a wide variety of cellular processes including
aging and lifespan extension [1,2,3,4]. Transgenic mice overex-
pressing SIRT1 have a beneficial Calorie Restriction (CR)-like
phenotype, whereas downregulation of SIRT1 accelerates the
aging phenotype in mice [5,6]. Interestingly, SIRT1 orthologs are
linked to the insulin/IGF signaling pathway in C. elegans, drosophila
and mice through its ability to deacetylate the FOXO proteins
[7,8,9]. For example, the longevity phenotypes in C. elegans are
suppressed by mutations in daf-16, a forkhead family transcription
factor, which is regulated by SIRT1 [10]. Notably, CR induces
SIRT1 expression, which can be attenuated by IGF-1. Further-
more, treatment of cells with either insulin or IGF-1 lowers SIRT1
levels, suggesting an inverse relationship between SIRT1 and the
insulin/IGF pathway [8]. However, the role of SIRT1 in CR
induced longevity remains controversial because in yeast, the
cyclic-AMP-dependent kinase (PKA) signaling pathway has been
implicated in CR induced longevity, independent of Sir2 [11]. In
addition, severe CR has been shown to involve the ‘‘target of
rapamycin’’ (TOR) pathway for lifespan extension in yeast
[12,13].
The mammalian target of rapamycin (mTOR) is a serine/
threonine protein kinase that regulates cell growth and prolifer-
ation by modulating protein synthesis and transcription. mTOR
acts as nutrient, energy and redox sensor by integrating signals
from multiple upstream signaling pathways, including insulin,
growth factors (IGF1/2), and mitogens. The mTOR complex 1
(mTORC1) consists of mTOR, regulatory associated protein of
mTOR(Raptor), LST8/G-protein
(mLST8/GbL) and PRAS40. mTORC1 is stimulated by growth
promoting conditions and inhibited by low nutrient levels, growth
factor deprivation, reductive stress and the specific inhibitor of
mTORC1, Rapamycin. Upstream to mTORC1 is the TSC1-
TSC2 inhibitory complex, which functions as a GTPase activating
protein (GAP) for the GTPase Rheb, an upstream activator of
mTOR. The TSC1-TSC2 complex inactivates Rheb to inhibit
mTOR signaling [14,15]. Diverse growth and stress signals
converge at the TSC1-TSC2 complex to regulate mTORC1
signaling.
The mTOR pathway has been implicated in longevity in model
organisms such as yeast, worms and flies. Over-expression of the
Drosophila homologs dTSC1 or dTSC2 or mutation in dTOR or its
downstream target dS6K, leads to longevity phenotype in
Drosophila [16]. In yeast, 6 out of 10 gene mutations that are
known to increase replicative life span correspond to components
of the TOR pathway including TOR and S6K1 (Sch9) [13].
Furthermore, TOR inhibition has been shown to extend lifespan
in yeast by increasing Sir2p activity [17]. Resveratrol, a known
activator of SIRT1, has been demonstrated to inhibit mTOR
activity and cellular senescence [18,19,20]. In a recent extensive
study, rapamycin, the inhibitor of mTOR, was shown to extend
the median and maximal lifespan of mice [21].
The two best characterized substrates of mTORC1 are p70-S6
Kinase 1 (S6K1) and 4E-BP1, the eukaryotic initiation factor 4E
(eIF4E) binding protein 1. Activation of mTOR results in
phosphorylation of S6K1 and 4EBP1, which increases protein
b-subunitlike protein
PLoS ONE | www.plosone.org1 February 2010 | Volume 5 | Issue 2 | e9199

Page 2

synthesis and ribosome biogenesis. Thus activation of mTOR
results in an increase in cell size and mass [22].
Clearly mTOR and SIRT1 regulate many common effectors
critical to the longevity signaling pathways in lower organisms and
mice. However, no direct link has yet been established between
these two important regulators. Here we investigated the potential
functional interrelationship between these two proteins in
regulating the stress response. Our results demonstrate that SIRT1
indeed regulates mTOR signaling, potentially through TSC2.
Results
SIRT1 Regulates mTOR Signaling in Human and Mouse
Cells
We investigated the activity of mTOR pathway in SIRT1
deficient mouse embryonic fibroblasts (MEFs) by analyzing the
phosphorylation of mTOR and its substrates S6K1 and 4EBP1.
Phosphorylation of S6, the downstream target of S6K1 was also
analyzed. As shown in Figure 1A, absence of SIRT1 resulted in
higher phosphorylation of mTOR, S6K1, 4EBP1 and S6,
suggesting a role for SIRT1 in mTORC1 regulation. Since
mTOR pathway is responsive to nutrient and cellular stress and is
downregulated in response to stress signals, we examined if stress
induced by amino acid (leucine) starvation inhibited the
upregulated mTOR signaling in SIRT1 deficient cells. We
observed that in contrast to WT cells, the upregulated mTOR
activity in SIRT1 null cells was not fully abrogated even under
leucine starvation.
To verify these results in human cells, HeLa cells were depleted
for SIRT1 using stable retroviral delivery of an shRNAi specific
for SIRT1 [23]. The SIRT1 depleted and matched-control HeLa
cells were then treated with a number of different stress conditions
and phosphorylation of mTOR, S6K1 and S6 examined.
Consistent with our data from murine cells, the SIRT1-depleted
HeLa cells showed higher mTOR signaling regardless of the stress
conditions (Figure 1B). To further confirm the role of SIRT1 in
regulating mTOR pathway in animals, we analyzed the mTOR
pathway in SIRT12/2and WT mouse tissue. As indicated by
higher levels of phospho-S6 and phospho-4EBP1 (Figure 1C),
tissues from SIRT1 knock-out mice showed elevated mTOR
activity.
Consistent with a role for mTOR in regulating cell size by
modulating protein synthesis, we also observed a larger morphol-
ogy for the SIRT1 deficient MEFs compared to WT MEFs
(Figure 1D). Furthermore, on a single cell level, the SIRT1
deficient cells showed a higher protein content per cell when
compared to wild-type MEFs (Figure 1E). SIRT1 is predominantly
known as a nuclear protein, although several recent reports suggest
the cytoplasmic presence of SIRT1 [24,25]. Since mTOR
signaling is a cytoplasmic process, we examined the intracellular
localization of SIRT1 by immunofluroscence of HeLa cells and
subcellular fractionation of several human cell lines and mouse
embryonic fibroblasts. Our results demonstrate that SIRT1 is
present in the cytoplasm of HeLa, 293T, Jurkat cells and mouse
embryonic fibroblasts, consistent with a cytoplasmic role for
SIRT1 (Figure 1F and G).
Deacetylase Activity of SIRT1 Plays a Role in mTOR
Signaling
SIRT1 is a NAD+ dependent deacetylase whose catalytic
activity is important for most of its known functions. To determine
if the catalytic activity of SIRT1 is important for regulation of
mTOR signaling, HeLa cells were treated with the SIRT1
activator resveratrol (RES), under stress (-leucine) or growth
(insulin) conditions and mTOR signaling measured by examining
phosphorylation levels of S6 and 4EBP1 (Figure 2A). Resveratrol
suppressed mTOR signaling regardless of stress or growth
conditions, suggesting that inducing the catalytic activity of SIRT1
negatively regulates mTOR signaling. To further confirm this
observation, matched-control and SIRT1-depleted HeLa cells
were treated with the SIRT1 inhibitor nicotinamide (NAM).
Consistent with the resveratrol results, the NAM treated control
cells showed upregulation of S6 and 4EBP1 phophorylation
(Figure 2B). This result further demonstrates that the catalytic
activity of SIRT1 is important for mTOR regulation and
inhibition of the catalytic activity results in elevated mTOR
signaling in normal cells.
Rapamycin Inhibits Upregulated mTOR in SIRT1 Deficient
Cells
To determine if SIRT1 modulates the expression of the critical
proteins involved in regulation of the mTOR pathway, protein
expression analysis of TSC1, TSC2, Raptor and Rheb in SIRT1
depleted cells under various stress conditions was performed. As
shown in figure 3A, SIRT1 deficiency does not affect the
expression of the mTOR regulatory proteins.
To determine if SIRT1 regulates the mTOR signaling pathway
upstream or downstream from mTORC1, we examined whether
rapamycin, a specific mTOR inhibitor that inhibits mTOR
complex 1 (mTORC1) activity, could inhibit the upregulated
mTOR activity in the SIRT1 depleted cells. Treatment of the
matched-control or SIRT1 depleted HeLa cells with rapamycin
significantly abrogated the phophorylation of S6K1, S6 and
4EBP1 (Figure 3B), suggesting that SIRT1 regulates mTOR
upstream of the mTORC1.
mTOR is downregulated by the upstream TSC1/TSC2
complex. TSC2 null cells show upregulated mTOR signaling
whereas over-expression of TSC1 and TSC2 result in mTOR
inhibition [26,27]. Thus we examined if rapamycin could inhibit
upregulated mTOR signaling in SIRT1 deficient MEFs compared
to TSC2 deficient MEFs. Treatment of both TSC22/2 and
SIRT12/2 MEFs with rapamycin resulted in abrogation of the
elevated mTOR activity, demonstrating that SIRT1 null MEFs
are similarly sensitive to rapamycin as the TSC2 null MEFs
(Figure 3C). Interestingly, in contrast with the SIRT1 null MEFs,
the TSC2 null cells did not show a significant size difference from
the TSC2+/+ MEFs (Figure S1).
SIRT1 Inhibits mTOR Signaling through TSC2
Since the above results suggested that SIRT1 acts upstream of
mTORC1 to downregulate mTOR signaling similar to TSC2, we
investigated if the SIRT1-mediated down-regulation of mTOR
signaling was TSC2 dependent. We treated WT and TSC2 null
MEFs with the SIRT1 activator resveratrol (RES), and SIRT1
inhibitor nicotinamide (NAM). As expected, only the TSC2 null
MEFs showed an increased S6 phosphorylation. In contrast, NAM
treatment induced S6 phosphorylation in WT MEFs (Figure 4A),
suggesting a potential role for SIRT1 in TSC2-mediated inhibition
of mTORC1 activity. Interestingly, the SIRT1 activator, RES,
could not inhibit S6 phosphorylation in absence of TSC2,
indicating that SIRT1 may be dependent on TSC2 for inhibiting
mTORC1 activity. In addition, no further decrease in S6
phosphorylation in response to RES treatment was observed in
WT cells, possibly due to the lower basal levels of S6
phosphorylation in WT cells which are both TSC2 and SIRT1
positive.
Resveratrol has been shown to activate the AMP-activated
kinase (AMPK) pathway independent of SIRT1 [28,29]. AMPK, a
SIRT1 Regulates mTOR
PLoS ONE | www.plosone.org2February 2010 | Volume 5 | Issue 2 | e9199

Page 3

Figure 1. mTOR activity in SIRT1 deficient mouse and human cells. (A) Western blot analysis for phosphorylation of the various mTOR target
protein substrates from mouse embryonic fibroblasts (MEFs) that were either un-treated or leucine starved. (B) Western blot analysis of extracts from
HeLa cells depleted for SIRT1 after stress treatments as indicated. (C) Western blot analysis of muscle extracts from WT or SIRT1 knock-out mice. Lane
1 and 2 are tissues from two WT mice, and Lane 3 and 4 are tissues from two SIRT1 null mice. (D) F-actin (red) staining of SIRT1 null and wild-type (WT)
MEFs. (E) Protein content of SIRT1+/+ and SIRT12/2 MEFs. (F) Immunofluroscence of HeLa cells using SIRT1 monoclonal antibody. (G) Subcellular
fractionation of HeLa, 293T, Jurkat cells and MEFs followed by Western blotting for SIRT1. Anti-Parp1 and Hsp90 shows the nuclear and cytoplasmic
fractionation. -leu: leucine deprived, -glc: glucose deprived, -ser: serum deprived, +TM: tunicamycin.
doi:10.1371/journal.pone.0009199.g001
SIRT1 Regulates mTOR
PLoS ONE | www.plosone.org3February 2010 | Volume 5 | Issue 2 | e9199

Page 4

cellular energy sensor, also acts as a negative regulator of mTOR
pathway by activating TSC2 in response to energy stress.
However, consistent with another published report [30] we could
not detect any measurable difference in AMPK activity in the
absence of SIRT1 (Figure S2), but instead detected significant
differences in mTOR activity.
To confirm that the observed effects of resveratrol were indeed
mediated through SIRT1, we investigated the effect of resveratrol
on WT versus SIRT1 null MEFs. Resveratrol treatment strongly
inhibited S6 phosphorylation in WT MEFs, but with reduced
efficacy in SIRT1 null MEFs, suggesting that the effect of
resveratrol on mTOR signaling is mediated partly through SIRT1
(Figure 4B). We observed that the effect of resveratrol on mTOR
signaling was dose-dependent with lower concentrations (25 mM)
showing inhibition only in SIRT1-positive cells, while higher
concentrations (100 mM) inhibited mTOR both in WT as well as
in the SIRT1-deficient cells. This result suggests that resveratrol
regulates mTOR signaling through both SIRT1-dependent and -
independent pathways, consistent with reports showing SIRT1
independent effects of resveratrol.
SIRT1 Interacts with TSC2
Given that the effects of RES and NAM on mTOR regulation
are both SIRT1- and TSC2-dependent, we investigated if these
two proteins interact in vivo. Immunoprecipitation of endogenous
TSC2 resulted in coimmunoprecipitation of endogenous SIRT1
(Figure 5A) and conversely, TSC2 was coimmunoprecipitated with
SIRT1 (Figure 5B) with or without stress condition (-leucine),
indicating that SIRT1 associates with TSC2. We further examined
if TSC2 is an acetylated protein. Although we could not detect
acetylation of TSC2 by immunoblotting and by mass-spec analysis
(data not shown), it is still possible that TSC2 or another protein in
the complex is acetylated in response to stress or growth conditions
and can be regulated by the deacetylase activity of SIRT1.
Discussion
Both SIRT1 and mTOR have been linked to age-associated
diseases with SIRT1 activation having a protective effect, whereas
inhibition of mTOR conferring a beneficial effect. For example,
SIRT1 activation confers a therapeutic effect in type 2 diabetes,
obesity and neurodegenerative diseases such as Alzheimer’s and
amyotrophic lateral sclerosis, whereas inhibition of mTOR is
protective against cardiovascular and neurological diseases, diet-
induced obesity and cancer [31,32,33,34,35,36,37,38]. Autopha-
gy, a mechanism important in regulating stress response and aging
is negatively regulated by mTOR [39,40], whereas SIRT1 has
been reported to activate autophagy by deacetylating several
essential components of the autophagy machinery [41].
The inverse relationship between the roles of SIRT1 and
mTOR in aging-associated diseases and lifespan extension suggests
a functional interrelationship between these two proteins. Our
results demonstrate that SIRT1 and mTOR signaling pathways
are indeed interconnected in a way that promotes stress sensing
pro-survival signals, where the regulation of mTOR is mediated
potentially through an interaction of SIRT1 with the TSC1-TSC2
complex.
Stress conditions downregulate mTOR signaling thereby
reducing protein synthesis and cell growth. We found that this
mechanism is deregulated in the absence of SIRT1 in mouse and
human cells. SIRT1 deficiency caused upregulation of mTOR
signaling which could not be abolished even under cellular stress
caused by leucine starvation and other stress inducible stimuli.
Interestingly, SIRT1 has been suggested as a nutrient-sensitive
growth suppressor gene [42]. Although it was proposed that
SIRT1 functions through regulation of telomerase activity, our
results suggest that SIRT1 functions as a nutrient-responsive
growth suppressor also by regulating mTOR signaling.
SIRT1 has been shown to regulate many metabolic and stress
responsive pathways through the regulation of gene expression of
critical components. We observed that for mTOR regulation,
SIRT1 does not seem to function through regulating expression of
mTOR signaling proteins, instead SIRT1 potentially regulates
mTOR through an upstream inhibitory complex. Using SIRT1
deficient and TSC2 deficient cells, we observed that SIRT1’s
inhibitory effect on mTOR was similar to that of the mTOR
inhibitory protein TSC2. Further analysis using SIRT1 activator
and inhibitor indicated that the mTOR inhibitory effect of SIRT1
was at least partially dependent on TSC2. Resveratrol has been
reported to affect insulin signaling through SIRT1 independent
pathways. Consistent with these reports, our data demonstrated
that at lower doses, resvetratrol regulated the mTOR pathway in a
SIRT1 dependent manner. However, at higher doses, reveratrol
likely activated SIRT1 independent pathways in parallel, to inhibit
mTOR activity.
Figure 2. mTOR signaling in response to SIRT1 activator (resveratrol) and inhibitor (nicotinamide). (A) HeLa cells were either mock-
treated (vehicle alone) or treated with –leucine media or 200nM insulin, with or without 50 mM resveratrol as indicated. (B) SIRT1-depleted and
control HeLa cells were either mock-treated or treated with 10mM nicotinamide.
doi:10.1371/journal.pone.0009199.g002
SIRT1 Regulates mTOR
PLoS ONE | www.plosone.org4February 2010 | Volume 5 | Issue 2 | e9199

Page 5

Based on our data, we propose a model where negative
regulation of mTOR signaling by SIRT1 is mediated through its
association with TSC2 (Figure 5C). The TSC1-TSC2 complex is
the most prominent upstream inhibitor of mTOR signaling,
integrating several upstream signals such as growth factor,
energy, stress and possibly amino acids. In response to specific
stimuli, such as hormones, low energy, low nutrient or hypoxia,
specific kinases and regulatory proteins activate or inhibit the
TSC2 protein of the TSC1-TSC2 complex, thereby regulating
mTOR signaling. By acting through the main mTOR inhibitory
complex (TSC1/TSC2 complex), SIRT1 potentially responds to
more than one form of stress or growth signal to regulate mTOR
signaling.
Importantly, our results demonstrating a role for SIRT1 in
mTOR signaling is the first evidence for SIRT1 to directly
modulate translation-regulation. Previously, SIRT1’s role in
regulating cellular stress response was shown to involve various
transcription factors such as NF-kB, p53 and the FOXO proteins,
Figure 3. Expression and signaling in mTOR pathway in response to stress stimuli and rapamycin. (A) Western blot analysis of extracts
from matched-control and SIRT1 depleted HeLa cells after stress treatments as indicated. (B) SIRT1-depleted and matched-control HeLa cells, and (C)
Wild-type, SIRT1 null and TSC2 null MEFs were either mock-treated (vehicle alone) or treated with 25nM rapamycin for 1 hr followed by Western blot
analysis for mTOR activity. S+: SIRT1+/+, S-: SIRT12/2, T+: TSC2+/+, T-: TSC22/2.
doi:10.1371/journal.pone.0009199.g003
SIRT1 Regulates mTOR
PLoS ONE | www.plosone.org5 February 2010 | Volume 5 | Issue 2 | e9199