to disrupt the interaction between GATOR1 and
GATOR2 (fig. S3, C and D).
Given that SAMTOR is an inhibitor of m TORC1
signaling and methionine starvation promotes
the interaction between SAMTOR and GATOR1-
KICSTOR, we hypothesized that methionine starvation would also inhibit m TORC1 signaling.
Indeed, in multiple cell types, methionine starvation
inhibited mTORC1 signaling in a SAMTOR-dependent fashion, as measured by the phosphorylation of the m TORC1 substrates S6K1 at
Thr389 and 4E-BP1 at Ser65 (Fig. 4C and fig. S4,
A to C). In contrast, loss of SAMTOR did not prevent the inhibition of m TORC1 signaling caused
by withdrawal of leucine, arginine (Fig. 4D), or
growth factors (fig. S4D).
Consistent with the effects of SAMTOR overexpression (Fig. 2B), methionine starvation also
reduced the colocalization of m TOR with lysosomes in wild-type but not SAMTOR-null cells
(Fig. 4E). Furthermore, reexpression of wild-type
SAMTOR, but not a SAM binding–deficient mutant, restored the capacity of the m TORC1 pathway
to sense methionine in the SAMTOR-null HEK-
293T cells (Fig. 4F). Methionine starvation partially
reduced SAMTOR levels in a proteasome-dependent
manner (Fig. 4, B, C, and F, and fig. S4D), but this
degradation was not required for mTORC1 to
respond to methionine starvation (fig. S4E).
As in mammalian cells, dTOR signaling in
Drosophila S2R+ cells also responds to environmental methionine and leucine levels, as detected
by the phosphorylation of dS6K at residue Thr398
(Fig. 4G). Using double-stranded RNA (dsRNA)–
induced RNA interference, we found that knockdown of dSamtor (encoded by the gene CG3570),
but not of dSesn, prevented inhibition of d TOR
signaling by methionine starvation (Fig. 4G and
fig. S4F). However, the dsRNA targeting dSesn
did prevent inhibition of d TOR by leucine starvation. Thus, the fly orthologs of SAMTOR and
Sestrin2 have conserved roles in methionine and
leucine sensing, respectively.
Our results show that SAMTOR is required
for the mTORC1 pathway to detect changes in
methionine levels and that this function requires
its capacity to bind SAM. Moreover, the addition
of SAM to methionine-starved cells reactivated
m TORC1 signaling (Fig. 4H), indicating that it
is the drop in SAM levels that mediates the inhibitory effects of methionine restriction on
mTORC1. Given these findings, we predicted
that the loss of methionine adenosyltransferase
(MAT2A) would prevent m TORC1 from sensing
methionine by blocking its conversion to SAM.
Because MAT2A is essential in human cells (27, 28),
we generated a doxycycline-repressible system in
order to acutely suppress MAT2A expression (29).
Consistent with SAMTOR sensing SAM rather
than methionine directly, the loss of MAT2A
greatly attenuated the capacity of m TORC1 to
sense methionine while leaving its activation
by SAM largely intact (Fig. 4H).
Several properties of SAMTOR suggest that it
functions as a SAM sensor that signals methionine
sufficiency to m TORC1 (Fig. 4I): (i) SAMTOR binds
SAM with an affinity that is compatible with the
drop in intracellular SAM concentrations caused
by methionine starvation, (ii) SAMTOR is required
for methionine starvation to inhibit mTORC1
signaling, and (iii) SAMTOR mutants that do
not bind SAM cannot signal methionine suf-
ficiency to m TORC1. Because SAM levels can be
affected by the availability of folate, betaine, and
vitamin B12, SAMTOR may also link m TORC1 sig-
naling to the availability of these metabolites (30).
The Rag GTPase pathway senses and integrates
the presence of multiple amino acids upstream
of m TORC1 (4, 8). Sestrin1 and Sestrin2 detect
leucine, whereas CASTOR1 and SLC38A9 sense cy-
tosolic and lysosomal arginine, respectively (16, 19).
In contrast to the Sestrins and CASTOR1, which
bind to GATOR2, SAMTOR interacts with GATOR1-
KICSTOR. Our genetic data suggest that SAMTOR
potentiates GATOR1 function through an un-
known mechanism that may involve disruption
of the binding of GATOR1 to GATOR2. The in-
teraction between SAMTOR and GATOR1 re-
quires KICSTOR, which may reflect either a
composite binding site or the requirement for
KICSTOR to localize GATOR1 to the lysosomal
surface. In addition, structural information will
be needed if we are to understand how the
binding of SAM to SAMTOR disrupts its inter-
action with GATOR1 and KICSTOR.
Unlike leucine and arginine, which directly
bind sensors upstream of m TORC1, methionine
is sensed indirectly through SAM. SAM is a
central metabolite required for most methylation
reactions, including that of DNA (31), histones
(25, 30), and phospholipids (32), and our work
highlights its additional role as a signaling mol-
ecule. Whereas S. cerevisiae does not have a
SAMTOR homolog, the yeast TOR pathway does
sense methionine through the regulated methyl-
ation of the PP2A family of phosphatases (33).
In metazoans, the mTORC1 pathway senses
multiple amino acids, which suggests that these
nutrients were, at times, scarce during their evolution. Two inferences can be drawn from the
existence of SAMTOR: (i) SAM can become limiting in certain nutritional states, and (ii) modulation of mTORC1 under these conditions is
beneficial for maintaining organismal homeostasis. Indeed, diets low in methionine reduce tissue
SAM levels, improve insulin sensitivity, and extend lifespan in mice and rats (34–38). It is intriguing to speculate that these benefits might
be mediated in part via the SAMTOR-dependent
inhibition of m TORC1, which is well appreciated
for its impact on glucose metabolism and the
aging process (1). Given that SAMTOR has a SAM-binding pocket, it may be possible to modulate
SAMTOR function pharmacologically.
REFERENCES AND NOTES
1. R. A. Saxton, D. M. Sabatini, Cell 168, 960–976
2. C. C. Dibble, B. D. Manning, Nat. Cell Biol. 15, 555–564
3. J. L. Jewell, R. C. Russell, K. L. Guan, Nat. Rev. Mol. Cell Biol.
14, 133–139 (2013).
4. Y. Sancak et al., Science 320, 1496–1501 (2008).
5. E. Kim, P. Goraksha-Hicks, L. Li, T. P. Neufeld, K. L. Guan,
Nat. Cell Biol. 10, 935–945 (2008).
6. Z. Y. Tsun et al., Mol. Cell 52, 495–505 (2013).
7. L. Bar-Peled et al., Science 340, 1100–1106 (2013).
8. Y. Sancak et al., Cell 141, 290–303 (2010).
9. L. Bar-Peled, L. D. Schweitzer, R. Zoncu, D. M. Sabatini,
Cell 150, 1196–1208 (2012).
10. R. L. Wolfson et al., Nature 543, 438–442 (2017).
11. M. Peng, N. Yin, M. O. Li, Nature 543, 433–437 (2017).
12. S. Wang et al., Science 347, 188–194 (2015).
13. M. Rebsamen et al., Nature 519, 477–481 (2015).
14. J. Jung, H. M. Genau, C. Behrends, Mol. Cell. Biol. 35,
15. G. Wyant et al., Cell, in press.
16. R. L. Wolfson et al., Science 351, 43–48 (2016).
17. R. A. Saxton et al., Science 351, 53–58 (2016).
18. R. A. Saxton, L. Chantranupong, K. E. Knockenhauer,
T. U. Schwartz, D. M. Sabatini, Nature 536, 229–233
19. L. Chantranupong et al., Cell 165, 153–164 (2016).
20. E. L. Huttlin et al., Nature 545, 505–509 (2017).
21. L. Chantranupong et al., Cell Rep. 9, 1–8 (2014).
22. A. Parmigiani et al., Cell Rep. 9, 1281–1291 (2014).
23. A. Hildebrand, M. Remmert, A. Biegert, J. Söding, Proteins 77
(suppl. 9), 128–132 (2009).
24. P. Z. Kozbial, A. R. Mushegian, BMC Struct. Biol. 5, 19
25. S. J. Mentch et al., Cell Metab. 22, 861–873 (2015).
26. C. L. Quinlan et al., Nat. Chem. Biol. 13, 785–792 (2017).
27. T. Wang et al., Cell 168, 890–903.e15 (2017).
28. T. Wang et al., Science 350, 1096–1101 (2015).
29. M. Gossen, H. Bujard, Proc. Natl. Acad. Sci. U.S.A. 89,
30. J. W. Locasale, Nat. Rev. Cancer 13, 572–583 (2013).
31. J. A. Law, S. E. Jacobsen, Nat. Rev. Genet. 11, 204–220
32. C. Ye, B. M. Sutter, Y. Wang, Z. Kuang, B. P. Tu, Mol. Cell 66,
33. B. M. Sutter, X. Wu, S. Laxman, B. P. Tu, Cell 154, 403–415
34. N. Orentreich, J. R. Matias, A. DeFelice, J. A. Zimmerman,
J. Nutr. 123, 269–274 (1993).
35. R. A. Miller et al., Aging Cell 4, 119–125 (2005).
36. L. Sun, A. A. Sadighi Akha, R. A. Miller, J. M. Harper, J. Gerontol. A
64, 711–722 (2009).
37. H. M. Brown-Borg et al., Aging Cell 13, 1019–1027 (2014).
38. D. Wanders et al., Diabetes 65, 1499–1510 (2016).
We thank all members of the Sabatini lab for helpful insights
and suggestions, in particular N. Kory for helping validate
the efficacy of dsRNA knockdowns in Drosophila cells;
G. Wyant for experimental discussions and advice; C. Lewis,
B. Chan, and T. Kunchok for performing the LC/MS analysis;
P. Jouandin and N. Perrimon for generously providing Drosophila
S2R+ cells; and C. Thoreen for providing pCW57.1 vector.
D.M.S., X.G., and J.M.O. are inventors on patent application
submitted by Whitehead Institute for Biomedical Research
that relates to SAMTOR and its role in m TORC1 signaling.
Supported by NIH grants R01 CA103866 and R37 AI47389
and U.S. Department of Defense grant W81XWH-07-0448
(D.M.S.), NIH fellowships T32 GM007753 and F30 CA210373
(J.M.O.), NIH grant F31 GM121093-01A1 (R.A.S.), NSF grant
2016197106 (K.J.C.), Paul Gray UROP Fund grant 3143900
(S.M.S.), Michael J. Fox Foundation grants NS083524 and
AG011085 (J. W.H.), and NIH grant U41 HG006673 (S.P.G.
and J. W.H.). J. W.H. is a paid consultant for Takeda
Pharmaceuticals, SV Brahma Discovery, and the American
Society for Microbiology. D.M.S. is an investigator of the
Howard Hughes Medical Institute and a founding member
of the scientific advisory board, a paid consultant, and
a shareholder of Navitor Pharmaceuticals, which is targeting
for therapeutic benefit the amino acid sensing pathway
upstream of m TORC1.
Materials and Methods
Figs. S1 to S4
9 July 2017; accepted 29 September 2017
818 10 NOVEMBER 2017 • VOL 358 ISSUE 6364 sciencemag.org SCIENCE
RESEARCH | REPORTS