Signal transduction

Linking nutrients to growth

How cells sense nutrients to control growth is largely unknown. One missing link involved in conveying the nutrient signal to the TOR protein, which regulates growth, seems to be the Rag proteins.

In mammalian cells, nutrients (such as amino acids), growth factors and cellular energy together trigger a molecular signalling pathway, mediated by the protein TOR, that controls cell growth. Deregulation of this pathway is implicated in cancer, and TOR inhibition by the anticancer drug rapamycin prevents unruly cell growth. Extensive research has led to characterization of many components of this signalling pathway. A remaining question is how amino acids activate TOR. Two teams (Kim et al.1 reporting in Nature Cell Biology and Sancak et al.2 writing in Science) now provide an important clue.

TOR ('target of rapamycin') is an evolutionarily highly conserved protein kinase that is found in two functionally and structurally distinct multiprotein complexes: TORC1 and TORC2 (ref. 3). TORC1 controls many cellular processes that ultimately determine cell growth, including protein synthesis, ribosome formation, nutrient transport and autophagy (a survival mechanism that kicks in in response to starvation).

Activation of TORC1 requires simultaneous availability of amino acids, growth factors and energy. Although inputs from growth factors (such as insulin) and energy are relatively well understood, determining the way that amino acids trigger TORC1 activation has been elusive. Amino-acid depletion results in rapid dephosphorylation of two molecules downstream of TORC1, S6K and 4E-BP, whereas addition of amino acids leads to rapid, TORC1-dependent phosphorylation of these molecules. But what is the molecular link between the amino-acid signal and TORC1 activation?

To answer this question, Kim et al. and Sancak et al. used complementary approaches. Using the technique of RNA interference (RNAi), Kim and colleagues1 performed a screen with the S2 cell line of the fruitfly Drosophila to search for GTPase proteins that regulate S6K phosphorylation in response to amino acids. Sancak et al.2 performed a proteomic analysis in mammalian cells to identify new binding partners of TORC1. Both teams independently identified Rag GTPase proteins as mediators of the amino-acid signal to TORC1. Sancak and colleagues further showed that these regulatory molecules interact with raptor, a component of TORC1. Meanwhile, Kim et al. concentrated on the physiological aspects of the Rag GTPases' function, demonstrating their role in TORC1-mediated regulation of autophagy and cell size in Drosophila.

Regulatory GTPases come in different varieties and are commonly components of signalling pathways. These proteins are active when bound to the nucleotide GTP and inactive when bound to GDP — the product of GTP hydrolysis. The intrinsic enzymatic activity of the GTPases converts the GTP to GDP. Specific GTPase-activating proteins (GAPs) stimulate this intrinsic enzymatic activity of the GTPases, and another group of proteins called GTPase exchange factors (GEFs) mediate dissociation of GDP from GTPases so that the GTPases can bind to a new GTP molecule and resume activity. Thus, GAPs inhibit and GEFs stimulate the signalling function of GTPases.

There are four Rag proteins (RagA–D), with high sequence similarity existing between RagA and RagB, and between RagC and RagD. These proteins function as heterodimers — RagA or B binding to RagC or D (ref. 4). Rag heterodimers form independently of GTP/GDP binding status. Nonetheless, GTP binding to the RagA/B subunit is crucial for the activation and localization of the heterodimer. The two studies1,2 show that a RagA/B that cannot bind to GTP also fails to stimulate TORC1, and a constitutively active RagA/B supports TORC1 activity even in the absence of amino acids.

But what is the consequence of Rag binding to TORC1? Sancak et al. propose that amino-acid-activated Rags are molecular matchmakers for TORC1 and another GTPase called Rheb. When bound to Rag, TORC1 is somehow delivered to Rheb, which is localized on the membrane of a so far ill-defined organelle. Rheb then directly stimulates the kinase activity of TORC1, which leads to S6K and 4E-BP phosphorylation. Amino acids impinge on this mechanism at the level of GTP loading of RagA/B (Fig. 1).

Figure 1: Passing on the message.

Amino acids are transported into the cell where they trigger a molecular signalling pathway that is mediated by the multiprotein complex TORC1. Two studies1,2 show that Rag GTPases — heterodimers of RagA or RagB and RagC or RagD — are involved in TORC1 activation in response to this nutrient signal. By binding to TORC1, Rag GTPases mediate its transfer to intracellular membranes that contain another GTPase, Rheb. There, other signals such as cellular energy and growth factors (insulin) integrate with the amino-acid signal, leading to Rheb-mediated activation of TORC1 and phosphorylation of its downstream effectors S6K and 4E-BP, which ultimately lead to protein synthesis and cell growth.

Previous work5 has shown that the Rheb–TORC1 interaction depends on amino-acid availability, and that TORC1 inhibition after amino-acid withdrawal can be overcome through increased Rheb expression. Intriguingly, inactivation of the TSC1–TSC2 complex, a GAP that inhibits Rheb, cannot overcome the effect of amino-acid withdrawal6, suggesting that amino acids probably mediate TORC1 interaction with activated Rheb, rather than activation (GTP loading) of Rheb. (Insulin and energy mediate the activation of Rheb through inhibition of TSC1–TSC2 GAP activity.)

The molecular mechanism by which Rheb activates TORC1 once Rheb has integrated amino acid, insulin and energy signals, and has bound to TORC1, remains to be determined. An earlier study suggested the involvement of a protein called FKBP38. In the absence of amino acids, FKBP38 binds to and inhibits TORC1, and dissociates only on amino-acid-stimulated binding of Rheb to TORC1 (ref. 7). It is difficult to imagine how FKBP38 might be involved in Rag-mediated activation of TORC1, as FKBP38 is itself membrane bound. Another protein, VPS34, also seems to mediate amino-acid stimulation of mammalian TORC1 activity8. But the relationship between VPS34 and the Rag proteins in activating mammalian TORC1 is unclear, as VPS34 is not involved in activating TORC1 in Drosophila9.

Of the other remaining questions, perhaps the most important is: what mediates amino acids' effect on the Rag proteins? So far, no GEF or GAP specific for regulation of the Rag family of proteins has been identified. This topic is likely to become the subject of considerable attention. The finding that Gtr1 and Gtr2, yeast proteins related to RagA/B and RagC/D, are implicated in nutrient sensing by TOR also cannot be overlooked10.

Previous work11 has shown that a high-protein diet (that is, increased availability of amino acids) affects glucose metabolism and insulin signalling through the TOR signalling pathway, leading to increased blood glucose levels and insulin resistance. Unravelling regulation of the Rag GTPases will help us to gain a better understanding of metabolic disorders such as obesity and type 2 diabetes.


  1. 1

    Kim, E., Goraksha-Hicks, P., Li, L., Neufeld, T. P. & Guan, K.-L. Nature Cell Biol. 10.1038/ncb1753 (2008).

  2. 2

    Sancak, Y. et al. Science 320, 1496–1501 (2008).

    ADS  Article  Google Scholar 

  3. 3

    Wullschleger, S., Loewith, R. & Hall, M. N. Cell 124, 471–484 (2006).

    Article  Google Scholar 

  4. 4

    Gao, M. & Kaiser, C. A. Nature Cell Biol. 8, 657–667 (2006).

    Article  Google Scholar 

  5. 5

    Long, X. et al. J. Biol. Chem. 280, 23433–23436 (2005).

    Article  Google Scholar 

  6. 6

    Smith, E. M. et al. J. Biol. Chem. 280, 18717–18727 (2005).

    Article  Google Scholar 

  7. 7

    Bai, X. et al. Science 318, 977–980 (2007).

    ADS  Article  Google Scholar 

  8. 8

    Gulati, P. et al. Cell Metab. 7, 456–465 (2008).

    Article  Google Scholar 

  9. 9

    Juhasz, G. et al. J. Cell Biol. 181, 655–666 (2008).

    Article  Google Scholar 

  10. 10

    Dubouloz, F. et al. Mol. Cell 19, 15–26 (2005).

    Article  Google Scholar 

  11. 11

    Patti, M.-E. & Kahn, B. B. Nature Med. 10, 1049–1050 (2004).

    Article  Google Scholar 

Download references

Author information



Rights and permissions

Reprints and Permissions

About this article

Cite this article

Zinzalla, V., Hall, M. Linking nutrients to growth. Nature 454, 287–288 (2008).

Download citation

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.