PerspectiveCell Signaling

New mTOR Targets Grb Attention

See allHide authors and affiliations

Science  10 Jun 2011:
Vol. 332, Issue 6035, pp. 1270-1271
DOI: 10.1126/science.1208071

A cellular enzyme called mechanistic (or mammalian) target of rapamycin (mTOR) controls cell growth and division, and is an important drug target in cancer (1). Despite extensive study, a complete understanding of mTOR function has remained elusive. One reason is that rapamycin, the natural compound that led to the identification of mTOR, only partially inhibits the enzyme. In addition, mTOR functions in two distinct protein complexes (mTORC1 and mTORC2). Furthermore, only a few proteins have been identified as mTOR substrates, and these seem insufficient to explain its myriad functions. Two papers in this issue, by Hsu et al. on page 1317 (2) and Yu et al. on page 1322 (3), uncover dozens of new substrates and downstream targets of mTOR through proteomic screens. These results have major implications for research and drug development in cancer and metabolic disorders.

mTOR is a kinase that phosphorylates serine and threonine residues on target proteins. Both Hsu et al. and Yu et al. sought to identify all cellular proteins whose phosphorylation is directly or indirectly controlled by mTOR. Both studies used recently discovered mTOR kinase inhibitor (TORKI) compounds that are highly selective (4) and fully block the kinase activity of mTOR in mTORC1 and mTORC2. By contrast, rapamycin suppresses phosphorylation of only some mTORC1 substrates and is generally inactive toward mTORC2. Hsu et al. and Yu et al. each used two complementary approaches to stimulate mTOR activity in cells. One strategy involved the use of mammalian cells genetically modified to lack tuberous sclerosis complex 2, a negative regulator of mTORC1. In the second approach, cells were deprived of growth factors before treatment with insulin, a factor that rapidly activates mTORC1 and mTORC2. Phosphorylated proteins were isolated and analyzed by mass spectrometry—a “phosphoproteomics” approach. As a complementary strategy, Hsu et al. defined the consensus phosphorylation sequence for the mTOR enzyme in vitro.

Although the details of the phosphoproteomic methodology differ between the two studies, the conclusions are remarkably similar. Both studies identify known mTOR substrates as well as hundreds of previously unknown protein phosphorylation sites controlled by the kinase. A large fraction of these known and novel substrates is insensitive to rapamycin. Further, Hsu et al. show that most of the rapamycin-sensitive phosphorylation sites do not fit the consensus amino acid sequence for direct mTOR substrates, suggesting that they are downstream targets in the signaling network. Together, these results help to explain why experimentation with rapamycin has uncovered few direct mTOR substrates. Also, the incomplete effect of rapamycin on the mTOR signaling network helps to clarify why rapamycin analogs have shown limited efficacy as anticancer drugs. The new direct mTOR substrates inferred from the phosphoproteomic data include several that function in processes not previously linked closely to mTOR including mRNA processing, DNA replication, and vesicle-mediated transport.

mTOR targets.

A proteomic approach has expanded our knowledge of the processes controlled by mTOR. A large portion of the cellular phosphorylation response to insulin is controlled by mTOR.


An important finding of both studies is that mTOR controls a large portion of the phosphorylation response to insulin, as most of the phosphorylation events stimulated by this factor were suppressed in cells treated with TOR-KIs. A clinical implication of this result is that treatments using TOR-KIs are likely to cause some insulin resistance. This validates previous findings showing that mTORC1 and mTORC2 together regulate most members of the AGC superfamily, which includes kinases involved in insulin signaling such as Akt and S6 kinase 1 (S6K1) (1). Many of the TOR-KI–sensitive sites do not conform to the consensus motifs of either mTOR or AGC kinases, which suggests that these kinases initiate signaling cascades that activate other protein kinases.

Hsu et al. and Yu et al. illustrate the biological impact of their data by focusing on the same candidate mTOR substrate, Grb10. This cytoplasmic protein suppresses signaling by insulin and the related insulin-like growth factors (IGFs). Mice lacking Grb10 are larger than normal and exhibit enhanced insulin sensitivity (5, 6). Both studies show that Grb10 is directly phosphorylated by mTORC1, a modification that enhances Grb10 stability. Consistent with mouse genetic data, depletion of Grb10 from cells increased insulin sensitivity, whereas over-expression suppressed insulin signaling. Grb10 appears to participate in a negative feedback loop whereby mTORC1, activated downstream of insulin or IGFs, phosphorylates Grb10 and potentiates its ability to suppress continuing signals from insulin and IGF receptors (see the figure). This mechanism is complementary to a known feedback pathway in which S6K1 phosphorylates the insulin receptor substrate (IRS)—a protein that has a positive role in insulin and IGF s ignaling—targeting it for degradation.

It had long been suspected that the S6K1-IRS feedback loop was insufficient to explain the powerful negative control of insulin responsiveness by mTORC1; identification of the mTORC1-Grb10 mechanism thus fills an important gap in our knowledge. As Yu et al. describe, mining of transcriptome data pushes the implications of this finding further by suggesting that Grb10 has tumor-suppressor function in human cancer. A related point of relevance to cancer research and treatment is that TOR-KIs are likely to enhance signaling through IGF receptors, which can support survival of cancer cells. However, TOR-KIs profoundly suppress downstream events in insulin and IGF signaling, which might mitigate the loss of negative feedback control.

Grb10 is just one example of an exciting lead generated by phosphoproteomic screens. Follow-up studies will likely yield new insights about mTOR biology and will hopefully provide clinical investigators with new biomarkers to monitor mTOR inhibitor action, along with improved understanding of their anticancer activity as well as toxicities to normal tissues.

References and Notes

  1. S.S.Y. was supported by the Indang research grant of Inje University.

Navigate This Article