BlenisFingamTORreview Copy 1

Oncogene (2004) 23, 31513171
& 2004 Nature Publishing Group All rights reserved 0950-9232/04 $25.00
www.nature.com/onc
Target of rapamycin (TOR): an integrator of nutrient and growth factor signals and coordinator of cell growth and cell cycle progression
Diane C Fingar1,2 and John Blenis*,1
1 Department of Cell Biology, Harvard Medical School, 240 Longwood Ave., Boston, MA 02115,USA; 2Department of Cell and Developmental Biology, University of Michigan Medical School, 1150 W. Medical Center Dr., Ann Arbor, MI 48109-0678, USA
Cell growth (an increase in cell mass and size through macromolecular biosynthesis) and cell cycle progression are generally tightly coupled, allowing cells to proliferate continuously while maintaining their size. The target of rapamycin (TOR) is an evolutionarily conserved kinase that integrates signals from nutrients (amino acids and energy) and growth factors (in higher eukaryotes) to regulate cell growth and cell cycle progression coordinately. In mammals, TOR is best known to regulate translation through the ribosomal protein S6 kinases (S6Ks) and the eukaryotic translation initiation factor 4Ebinding proteins. Consistent with the contribution of translation to growth, TOR regulates cell, organ, and organismal size. The identication of the tumor suppressor proteins tuberous sclerosis1 and 2 (TSC1 and 2) and Ras-homolog enriched in brain (Rheb) has biochemically linked the TOR and phosphatidylinositol 3-kinase (PI3K) pathways, providing a mechanism for the crosstalk that occurs between these pathways. TOR is emerging as a novel antitumor target, since the TOR inhibitor rapamycin appears to be effective against tumors resulting from aberrantly high PI3K signaling. Not only may inhibition of TOR be effective in cancer treatment, but rapamycin is an FDA-approved immunosuppressive and cardiology drug. We review here what is known (and not known) about the function of TOR in cellular and animal physiology. Oncogene (2004) 23, 31513171. doi:10.1038/sj.onc.1207542 Keywords: TOR; rapamycin; S6K1; 4E-BP1; cell growth
Introduction Tumorigenesis is a multistep process in which cells accumulate disparate complementing mutations that permit them to escape normal cellular and environmental constraints on proliferation (reviewed in Hanahan and Weinberg, 2000; Hahn and Weinberg, 2002). While only recently appreciated as an important process in tumorigenesis, an increased rate of cell growth (an increase in cell mass and size via macromolecular
*Correspondence: J Blenis, E-mail: john_blenis@hms.harvard.edu
biosynthesis) is required to support the rapid proliferation (an increase in cell number) central to tumorigenesis. A sufciency of nutrients and growth factors is required for cell growth and progression through the cell cycle. If the rate of cell growth is unable to keep up with a rapid rate of cell division, then cell proliferation cannot be sustained, since cells would progressively lose mass and size with each division cycle, resulting in inevitable cell death. Therefore, proliferating cells exhibit tight coordination between cell growth and cell cycle progression. In addition, such coordination ensures that individual cells, and indeed organs and whole organisms, maintain a characteristic size. The target of rapamycin, TOR (mTOR in mammals, also known as FRAP, RAFT, or RAPT), is an evolutionarily conserved serine/threonine kinase that regulates both cell growth and cell cycle progression through its ability to integrate signals from nutrients (amino acids and energy) and growth factors (reviewed in Shamji et al., 2003). In mammals and ies, the bestcharacterized downstream targets of TOR are two families of proteins that control protein translation, the ribosomal protein S6 kinases (S6Ks) and the eukaryotic initiation factor 4E (eIF4E)-binding proteins (4E-BPs). By virtue of its ability to sense environmental conditions, TOR functions as a rheostat to regulate the rate of cell growth and cell proliferation by, at least in part, regulating protein biosynthesis. It is thus not surprising that mTOR and its downstream targets have emerged as novel targets for cancer therapeutics (reviewed in Vivanco and Sawyers, 2002; Huang and Houghton, 2003). Inhibition of mTOR-dependent signaling in cancer cells should mimic nutrient and growth factor deprivation, limiting cell growth and conferring a signicant proliferative disadvantage. The bacterially derived drug rapamycin (also known as sirolimus) specically inhibits TOR, resulting in reduced cell growth, a reduced rate of cell cycle progression, and a reduced rate of proliferation (reviewed in Abraham and Wiederrecht, 1996; Schmelzle and Hall, 2000; Gingras et al., 2001b). As a result, rapamycin analogs, such as CCI-779 and RAD001, are currently being tested in clinical trials for efcacy against a variety of human tumors (reviewed in Elit, 2002; Hidalgo and Rowinsky, 2000; Huang and Houghton, 2003) In addition to being a promising
TOR coordinates cell growth and cell cycle progression DC Fingar and J Blenis
3152
anticancer agent, rapamycin is an FDA-approved immunosuppressive drug to inhibit kidney transplant rejection and cardiology drug to inhibit the restenosis that often occurs after angioplasty (Morice et al., 2002; Moses et al., 2003; and reviewed in Marks, 2003). Thus, in addition to the insights we will gain into how cell, organ, and organismal growth are controlled, understanding TOR action may have direct clinical application in human disease. In this review, we outline our emerging understanding of TOR regulation, TOR signaling, and the control and coordination of cell growth and cell cycle progression by TOR, with an outlook toward how disregulated TOR signaling may contribute to tumorigenesis. While this review will focus on mTOR and its implications for cancer biology, it is important to keep in mind that disregulated TOR function likely has pathophysiological signicance for other human disease states, which we will briey highlight.
Discovery of rapamycin and identication of TOR The identication of striking antifungal activity in extracts of the soil bacterium, Streptomyces hygroscopicus, found on Easter Island (also known as Rapa Nui), led to the purication of the active compound (rapamycin) in the early 1970s (reviewed in Sehgal, 2003). It was subsequently discovered that this drug also inhibited the proliferation of mammalian cells and suppressed the mammalian immune system by potently blocking the proliferation of B and T cells as well as exhibiting activity against solid tumors (reviewed in Sehgal, 2003). Early studies on rapamycin in the budding yeast Saccharomyces cerevisiae led to the identication of FKBP12 (FK506-binding protein of 12 kDa; encoded by the FPR1 gene) as a direct cellular receptor for the drug (Harding et al., 1989; Siekierka et al., 1989, 1990). FKBP12 is an abundant and ubiquitously expressed peptidyl-prolyl cis/trans isomerase that may function in protein folding (reviewed in Schreiber, 1991). While rapamycin inhibits the isomerase activity of FKBP12 (Heitman et al., 1991; Koltin et al., 1991; Wiederrecht et al., 1991), it appears that inhibition of this activity is not responsible for rapamycin sensitivity: Deletion of FPR1 (FKBP12) (or deletion of all four FKBP12 genes (FPR1-4)) in S. cerevisiae is not lethal; rather, the yeast are viable and exhibit resistance to the toxic effects of rapamycin (Heitman et al., 1991; Koltin et al., 1991; Dolinski et al., 1997). Therefore, it is the presence of FKBP12, not its activity, that is required for the toxic, antiproliferative action of rapamycin in yeast. Genetic screens for mutations that rescue the antiproliferative effects of rapamycin in S. cerevisiae identied three genes that when mutated conferred rapamycin resistance, the targets of rapamycin TOR1 and TOR2, and FPR1 (FKBP12) (Heitman et al., 1991; Cafferkey et al., 1993). Intriguingly, the TOR mutants were not recessive, like the FPR1 mutant, but rather dominantly acting point mutants. Wild-type Tor1p and Tor2p were found
Oncogene
to interact in a ternary manner with a gain-of-function rapamycin/FKBP12 complex, and thus were unable to bind either rapamycin or FKBP12 alone (Stan et al., 1994; Chen et al., 1995; and reviewed in Schmelzle and Hall, 2000; Gingras et al., 2001b). The rapamycinresistant point mutants (serine 1972 to arginine in Tor1p; serine 1975 to isoleucine in Tor2p) mapped to a region of TOR coined the FKBP12-rapamycin-binding (FRB) domain. In the presence of drug, these TOR mutants can neither bind to nor be inhibited by the rapamycinFKBP12 complex, thereby allowing proliferation in the presence of rapamycin (Stan et al., 1994; Chen et al., 1995). While both the yeast Tor1p and Tor2p proteins are 67% identical and are targeted by rapamycin, they do not share identical functions. Deletion of TOR2 is lethal due to arrested proliferation at all points in the cell cycle (Kunz et al., 1993), while deletion of TOR1 is not lethal, resulting only in a modest decrease in proliferation rate (Helliwell et al., 1994). Deletion of both TOR1 and TOR2 mimics rapamycin treatment the yeast arrest within one cell cycle in G1 phase (Kunz et al., 1993; Helliwell et al., 1994). Importantly, the lethality of the TOR2 deletion cannot be rescued upon overexpression of TOR1. These phenotypes, together with the data that yeast bearing a rapamycin-resistant point mutation in either Tor1p or Tor2p show resistance to rapamycininduced toxicity, suggest that Tor2p, in addition to sharing a rapamycin-sensitive function with Tor1p, shares an essential function that is resistant to rapamycin and that is not shared with Tor1p. This function is thought to be important for cell cycle-dependent organization of the actin cytoskeleton (Schmidt et al., 1996). Subsequent biochemical studies in mammalian cells led to the identication and cloning of the mammalian target of rapamycin, mTOR, from human, rat, and mouse. Since several groups cloned the gene at about the same time, TOR is also known as FRAP (FKBP12rapamcyin-associated protein), RAFT (rapamycin and FKBP12 target), RAPT (rapamycin target), or SEP (sirolimus effector protein) (Brown et al., 1994; Chen et al., 1994; Chiu et al., 1994; Sabatini et al., 1995; Sabers et al., 1995). TOR is a large (290 kDa) evolutionarily conserved member of the phosphatidylinositol 3-kinase (PI3K)-kinase-related kinase (PIKK) superfamily in which a COOH-terminal kinase domain with lipid kinase homology functions as a serine/ threonine protein kinase (reviewed in Keith and Schreiber, 1995). All mammalian PIKK family members (e.g. ATM, ATR, and DNA-PK) are similarly large and appear to function in cell cycle checkpoint control (reviewed in Keith and Schreiber, 1995), consistent with the function of TOR in a nutrient-sensitive cell cycle checkpoint. In contrast to yeast species, which contain two TOR homologs (Tor1 and Tor2) (Helliwell et al., 1994), only one TOR ortholog has been identied in higher eukaryotes, including mammals, ies, and worms (Brown et al., 1994; Chen et al., 1994; Chiu et al., 1994; Sabatini et al., 1995; Sabers et al., 1995; Oldham et al., 2000; Zhang et al., 2000; Long et al., 2002).
3153
TOR is 4060% identical at the amino-acid level among mammals, ies, worms, and yeast (Heitman et al., 1991; Oldham et al., 2000; Zhang et al., 2000; Long et al., 2002). Consistent with contacts predicted by the structure of the mTOR/rapamycin/FKBP complex (Chen et al., 1994; Stan et al., 1994; Choi et al., 1996) and with data from yeast, mutation of Ser2035 in the FRB domain of mTOR (analogous to serines 1972 and 1975 in Tor1 and Tor2, respectively) confers resistance to the known cellular and biochemical effects of rapamycin (Brown et al., 1995; Brunn et al., 1997). While rapamycin inhibits the rate of cell cycle progression and proliferation of mammalian cells via inhibition of mTOR, as it does in yeast, for reasons that are not currently understood, the potency of rapamycinmediated cell cycle inhibition in mammalian cells varies widely among cell types. Domain structure of mTOR The mTOR protein (2549 amino acids) is composed of numerous highly conserved yet poorly understood domains (Figure 1). It is found in a 2 mDa complex on gel ltration (Kim et al., 2002), suggesting associations with numerous cellular proteins, although only three mTOR-interacting proteins have been identied to date (Sabatini et al., 1999; Hara et al., 2002; Kim et al., 2002, 2003). The NH2-terminal half of TOR is composed of tandem HEAT repeats. Modeling of the tertiary structure of mTOR, however, suggests that most of the protein, except the kinase domain, consists largely of helical repeat units (Perry and Kleckner, 2003). These HEAT repeats may form an extended superhelical structure to create multiple large interfaces for pro-
teinprotein interactions (Perry and Kleckner, 2003; and reviewed in Gingras et al., 2001b). The COOH-terminus contains a highly conserved kinase domain; immediately NH2-terminal to the kinase domain lies the B100 amino-acid FRB domain, which binds the inhibitory rapamycinFKBP12 complex. The FRB/kinase domain pair is anked by an NH2-terminal B500 amino acid moderately conserved FAT domain and by a smaller, highly conserved FATC domain at the extreme COOHterminus of the protein. The FAT and FATC domains are found in all PIKK family members and may interact with each other to regulate kinase activity. Deletion analysis has mapped a repressor domain (amino acids 24302450) to the COOH-terminus of mTOR, just upstream of the FATC domain: deletion of these 20 amino acids results in increased TOR kinase activity toward downstream targets in vitro and in vivo (Sekulic et al., 2000). In general, mutagenesis of mTOR has been difcult, as subtle changes evoke signicant defects in function, suggesting that the extended tertiary structure of mTOR is essential for function, rendering conventional structure/function analysis of mTOR difcult (Brown et al., 1995; Peterson et al., 1999; Vilella-Bach et al., 1999; Kim et al., 2002). While mTOR localizes predominantly to the cytoplasm under steady-state conditions, cytoplasmic/nuclear shuttling is important for it to regulate its downstream targets S6 kinase 1 (S6K1) and 4E-BP1 (Kim and Chen, 2000). How does the rapamycin/FKBP12 complex inhibit TOR? While there is no question that rapamycin treatment of cells in vivo inhibits the ability of mTOR to send downstream signals and that the kinase activity of mTOR is absolutely required for the transmission of these signals (Brown et al., 1995; Brunn et al., 1997; Burnett et al., 1998), the mechanism by which rapamycin inhibits mTOR remains unclear. In one simple model of rapamycin function, the rapamycin/FKBP12 complex directly inhibits the kinase activity of mTOR. Unfortunately, several lines of evidence argue against this model. When mTOR is immunoprecipitated from cells treated with rapamycin in vivo, there is no inhibitory effect on mTORs ability to autophosphorylate in vitro (Peterson et al., 2000). Although it could be argued that the bound rapamycin/FKBP12 complex is lost during immunoprecipitation, rapamycin, in complex with recombinant FKBP12, inhibits mTOR autophosphorylation or exogenous substrate phosphorylation in vitro only at supraphysiological concentrations (0.510 mM); physiological rapamycin concentrations (1020 nM) have very little effect. In addition, nutrient deprivation, rapamycin treatment, or growth factor stimulation, which potently regulate the activity of mTOR-dependent downstream targets, alter neither the in vitro kinase activity nor the in vivo autophosphorylation (on Ser2481) of mTOR (Peterson et al., 2000). Collectively, these data suggest that in vivo, rapamycin does not directly inhibit the kinase activity of mTOR. Rather, rapamycin likely acts by altering the
Oncogene
Figure 1 Domain structure of TOR. The NH2-terminal half of TOR is composed of tandem HEAT repeats. Modeling of the tertiary structure of TOR, however, suggests that most of the protein (except the kinase domain) consists largely of helical repeat units (Perry and Kleckner, 2003). The COOH-terminal half of the protein contains the central FAT domain, followed by the FRB, the kinase, and the FATC domains. mTOR is regulated by nutrients (amino acids and energy) and by growth factors to modulate rates of cell growth and cell cycle progression, which coordinately control cell proliferation. Rapamycin, in complex with FKBP12, directly binds to the FRB domain to inhibit mTORdependent downstream signaling
3154
composition of multiprotein mTOR complexes to hinder the integration of critical upstream regulatory signals or the accessibility of the kinase to downstream substrates. Indeed, two novel mTOR-interacting proteins, raptor and GbL, were recently identied and shown to be important for the ability of mTOR to phosphorylate S6K1 and 4E-BP1 in vitro and in vivo (Hara et al., 2002; Kim et al., 2002, 2003) (to be discussed more below). The discovery that mTOR/ raptor complexes are easily disrupted with nonionic detergents may explain why measuring the in vitro kinase activity of mTOR has been so challenging and variable. Several pieces of data suggest that perhaps specic domains of TOR interact with critical regulators: Overexpression of the isolated FRB domain of mTOR in mammalian cells inhibits G1 phase cell cycle progression, and this dominant-negative effect is abrogated with a single point mutation (Vilella-Bach et al., 1999). In addition, a dominant-negative effect is produced when the TOR FAT domain, also known as the toxic effector domain, is overexpressed in yeast, resulting in the inhibition of proliferation (Alarcon et al., 1999). As a result, while the rapamycin/FKBP12 complex may not directly inhibit TOR kinase activity, it may block interactions with regulatory proteins through steric hindrance or conformational change.
TOR senses and integrates nutrient and growth factor signals Protein synthesis comprises a large share of cellular energetic expenditure. It is therefore not surprising that protein synthesis is initiated only when nutrients (amino acids and energy) and growth factors are present at sufcient levels. TOR is thought to sense and integrate environmental signals to control protein translation (and likely other cellular processes), which in turn is crucial for proper control of cell growth and cell cycle progression. Nutrient sensing of amino-acid and energy signals The nutrient sensing function of TOR is evolutionarily conserved (reviewed in Rohde et al., 2001; Crespo and Hall, 2002). In S. cerevisiae, TOR is sensitive to changes in amino-acid, nitrogen, and glucose levels, and inhibition of TOR with rapamycin or by TOR deletion triggers a response program similar to nutrient starvation. In the y, null mutations of dTOR impair larval growth and mimic the phenotype of amino-acid withdrawal (Oldham et al., 2000; Zhang et al., 2000). In mammals as well, phosphorylation of the best-characterized downstream targets of mTOR, the ribosomal S6K1 and 4E-BP1, are sensitive to changes in aminoacid levels and glucose concentration (Hara et al., 1998; Inoki et al., 2003b; Kim et al., 2002). Amino-acid deprivation and rapamycin likely act via a similar mTOR-dependent mechanism, since expression of an S6K1 mutant bearing both NH2- and COOH-terminal deletions (DNT/CT), which is resistant to the inhibitory
Oncogene
effect of rapamycin but sensitive to wortmannin, is also resistant to inhibition by amino-acid deprivation (Hara et al., 1998). While the mechanism by which mTOR senses amino-acid levels is unclear, several models have been proposed: mTOR may sense the charging of aminoacylated tRNA, since amino-acid alcohols, which are competitive inhibitors of aminoacyl-tRNA synthetases, inhibit mTOR regulation of S6K1 (Iiboshi et al., 1999). It has also been suggested, however, that mTOR may be regulated directly or indirectly by intracellular amino acids, by their metabolites, or by amino acidactivated second messengers (Beugnet et al., 2003). As part of its ability to sense levels of environmental nutrients, mTOR also senses energy sufciency. Reduction of cellular ATP levels with the glycolytic inhibitor 2-deoxyglucose specically inhibits mTOR-dependent phosphorylation of S6K1 and 4E-BP1 (Dennis et al., 2001). Interestingly, mTOR has an apparent Km for ATP greater than 1 mM, which is signicantly higher than the Km of most protein kinases (1020 mM) (Dennis et al., 2001). The requirement for such a high Km may allow mTOR to effect downstream signaling only when cellular energy is high. mTOR may sense cellular energy levels by another mechanism that is dependent on the AMP-activated protein kinase (AMPK). The drug AICAR, an activator of AMPK, inhibits the phosphorylation of S6K1 and 4E-BP1 in an mTOR-dependent manner in some, but not all, cell types (Kimura et al., 2003). AMPK is activated under states of low cellular energy (high AMP/ATP ratio) and acts to reduce ATP expenditure by inhibiting key enzymes of biosynthetic pathways and to increase the ATP supply by activating pathways producing ATP (reviewed in Hardie and Hawley, 2001). One mechanism by which AMPK may modulate mTOR function was described recently: activated AMPK phosphorylates the tumor suppressor protein tuberous sclerosis 2 (TSC2; also known as tuberin), enhancing the ability of the TSC complex (a heterodimer composed of TSC1 and TSC2) to inhibit mTOR (Inoki et al., 2003b) (see below for more discussion of TSC1 and TSC2 in regulation of mTOR). In the absence of TSC2, ATP depletion no longer results in dephosphorylation of S6K1 and 4E-BP1, and these cells are more susceptible to glucose-deprivation-induced apoptosis (Inoki et al., 2003b). Thus, energy sufciency is sensed by AMPK to modulate the tumor suppressor function of the TSC complex, which in turn regulates mTOR function (see Figure 2). Hence, mTOR may not directly sense ATP levels, as suggested previously (Dennis et al., 2001), but rather may be regulated by the AMP/ATP ratio via AMPK. Growth factor sensing The relationship of mTOR to regulation by growth factors, mitogens, and hormones continues to evolve. Certainly, the kinase activity of immunoprecipitated mTOR is not greatly altered following growth factor treatment. As noted above, it is unclear whether this result reects the true in vivo activity of mTOR and therefore indicates that mTOR kinase activity does not
3155
Figure 2 Integration of nutrient and growth factor signaling regulates mTOR-dependent downstream signaling. mTOR is regulated by PI3K/Akt-dependent signaling and by nutrients, such as amino acids and energy. PI3K converts the lipid PI-4,5-P2 (PIP2) into PI-3,4,5-P3 (PIP3), which localizes Akt to the membrane where it can be phosphorylated and activated by PDK1. The lipid phosphatase PTEN antagonizes PI3K action by converting PIP3 back to PIP2. Activated Akt phosphorylates TSC2, resulting in the inhibition of the tumor suppressor function of the TSC1/2 complex. Rheb, a small GTPase that is inactivated by TSC2s GAP activity, positively modulates mTOR function. mTOR phosphorylates both S6K1 and 4E-BP1 via independent pathways, resulting in the activation of S6K1 and inactivation of 4E-BP1 as a repressor of the translation initiation factor eIF4E. Increased S6K1 and eIF4E action independently promote cell growth and cell cycle progression. S6K1 and 4E-BP1 are also phosphorylated by PI3Kdependent but mTOR-independent mechanisms. Amino-acid sufciency is somehow sensed by TSC1/2, Rheb, and/or mTOR. Under low-energy conditions, mTOR is inhibited: the AMP/ATP ratio rises and activates AMPK. AMPK then phosphorylates TSC2 to enhance the tumor suppressor function of the TSC1/2 complex. Arrows depict activation, bars depict inhibition, and dotted lines depict unknown pathways. The Drosophila TOR pathway is very similar to the mammalian TOR pathway, as shown
more clearly dened the mechanism by which PA leads to mTOR activation: Cdc42, a small Rho family Gprotein that was previously shown to be an upstream activator of S6K1 (Chou and Blenis, 1996), activates PLD1 to increase cellular PA levels, which leads to the activation of S6K1 in a mTOR-dependent manner (Fang et al., 2003). Consistent with a role for PLD in the regulation of mTOR action, high-level PLD expression and PA production correlates with rapamycin resistance in breast cancer cell lines, while inhibition of PLD activity restores rapamycin sensitivity (Chen et al., 2003). The recent identication of the PI3K/Akt/TSC/ Ras-homolog enriched in brain (Rheb) pathway as an upstream, nutrient- and growth factor-responsive regulator of Drosophila and mammalian TOR adds substantially more weight to the idea that mTOR function is controlled by growth factors (discussed below).
Coordinate control of the translational effectors S6K1 and 4E-BP1 by TOR- and PI3K-dependent signals Both mTOR- and PI3K-dependent signals coordinately control the translational effectors S6K1 and 4E-BP1. In response to a wide variety of extracellular stimuli, PI3K phosphorylates the lipid phosphatidylinositol-4, 5- bisphosphate (PIP2) at the 30 -OH position to generate the lipid second messenger phosphatidylinositol-3, 4, 5triphosphate (PIP3). PIP3 lipids then recruit proteins containing pleckstrin homology (PH) domains, such as phosphatidylinositol-dependent kinase-1 (PDK1) and the serine/threonine kinase Akt (also known as protein kinase B, PKB), to the plasma membrane where they can be fully activated (see, Rameh and Cantley, 1999; Cantley, 2002, for more complete reviews of PI3Kdependent signaling). PI3K function is important for a wide range of cellular processes including cell growth, cell cycle progression and proliferation, cell survival, and cell migration. Aberrantly high PI3K-dependent signaling has been implicated in a wide variety of human cancers, and importantly, PI3K-stimulated oncogenesis is dependent on mTOR (reviewed in Vogt, 2001; Vivanco and Sawyers, 2002). As phosphorylation of S6K1 and 4E-BP1 is sensitive to inhibition by both rapamycin and wortmannin, researchers often place PI3K and mTOR in a linear signaling pathway. Such a linear wiring diagram is too simplistic, however, as a mutant of S6K1 (DNT/CT) that is resistant to inhibition by rapamycin is still sensitive to inhibition by wortmannin (Cheatham et al., 1995; Weng et al., 1995); thus, the mTOR and PI3K signals to S6K1 can be dissociated. The relationship of mTOR to PI3K has been further complicated by the fact that wortmannin, when used at 100-fold higher concentrations than is required to inhibit PI3K inhibits mTOR and other PIKK family members, including mTOR (Brunn et al., 1996); more alarmingly, LY294002 may inhibit both PI3K and mTOR at similar concentrations (Brunn et al., 1996). Thus, one must be
Oncogene
respond signicantly to growth factors, or whether the conditions that most researchers employ for measuring mTOR in vitro kinase activity are prone to artifact. While some growth factors may modestly increase phosphorylation of serine 2448, in a wortmanninsensitive and Akt-dependent manner (Nave et al., 1999), the signicance of serine 2448 phosphorylation remains unclear, as mutation of this site has no effect on the ability of mTOR to signal to the downstream targets S6K1 and 4E-BP1 (Sekulic et al., 2000). The observation that phosphatidic acid (PA) stimulates the phosphorylation of S6K1 and 4E-BP1 in an amino-aciddependent and rapamycin-sensitive manner supports the idea that mTOR may indeed be regulated by growth factors and mitogens, in addition to its regulation by nutrients and energy (Fang et al., 2001). PA is generated by mitogen-activated phospholipase D (PLD) and has been found to bind directly to the FRB domain of mTOR in vitro. In addition, a rapamycin-resistant mutant of S6K1 (DNT/CT) was insensitive to 1- and 2-butanol-induced inactivation (1- and 2-butanol block PA production by PLD), indicating that PA signals to S6K1 via mTOR (Fang et al., 2001). Recent work has
3156
extremely cautious when interpreting experiments employing pharmacological inhibition of PI3K with wortmannin and LY294002 so as not to ascribe a function to PI3K that is truly controlled by mTOR. The consensus in the eld has thus been, until recently, that mTOR and PI3K lie in parallel signaling pathways that converge on common downstream targets, with aminoacid and energy sufciency being mediated primarily via mTOR and growth factor, mitogen, and hormone sufciency being mediated primarily via PI3K. The recent discovery that the PI3K/Akt/TSC/Rheb pathway (reviewed in Manning and Cantley, 2003a) (to be discussed below) lies upstream of mTOR highlights the complexity of mTOR regulation and indicates that PI3K-regulated growth factor signaling pathways indeed crosstalk with the nutrient sensing function of mTOR. Thus, a current model would suggest that PI3K lies both upstream of and parallel to mTOR, with both branches converging on common downstream targets (Figure 2). Signaling to S6K1: regulation of ribosome biogenesis? While this review mainly focuses on the mTOR target S6K1, it is important to note that mammals possess two S6 kinase genes, S6K1 and S6K2, each of which are found as two alternatively spliced isoforms (reviewed in Martin and Blenis, 2002). The B70 kDa isoform of S6K1 (aII) is the best-studied S6 kinase and localizes predominantly to the cytoplasm, while the B85 kDa isoform (aI) localizes to the nucleus due to a nuclear localization sequence (NLS) contained within an NH2terminal 23 amino-acid extension. Both S6K2 isoforms localize to the nucleus due to an NLS in the COOHterminus. Similar to S6K1, two isoforms of S6K2 have been described; splice variant bI contains a 13 aminoacid extension NH2-terminal extension compared to the bII isoform. S6K1 is ubiquitously expressed and activated by a wide variety of extracellular signals. When activated by the convergence of mTOR- and PI3K-dependent signals, S6K1 functions as an in vivo kinase toward the 40S ribosomal protein S6 (Jeno et al., 1988), which has been suggested to increase the translational efciency of a class of mRNA transcripts that contain a terminal oligopyrimidine (TOP) tract at their 50 -end (Jefferies et al., 1994, 1997; Terada et al., 1994). Many of these 50 TOP mRNAs encode components of the translation machinery, such as ribosomal proteins and elongation factors (reviewed in Sonenberg and Gingras, 1998). S6K1 is therefore thought to upregulate the general translational capacity of the cell by enhancing the translation of components required for protein synthesis, a process known as ribosome biogenesis. It is important to note, however, that the role of S6K1 and ribosomal protein S6 phosphorylation in 50 -TOP translation has recently been questioned (Tang et al., 2001; Stolovich et al., 2002). For example, through a variety of approaches, amino-acid- and growth factor-induced S6K1 activation and S6 phosphorylation can be dissociated from 50 -TOP translation (Tang et al., 2001;
Oncogene
Stolovich et al., 2002). The authors of these studies concluded that amino-acid and growth factor regulation of 50 -TOP mRNAs requires PI3K-dependent signals, partially requires mTOR-dependent signals, and does not require S6K1 activity or phosphorylation of ribosomal protein S6 (Tang et al., 2001; Stolovich et al., 2002). In light of these ndings, it is important to keep in mind that a major function of S6K may not involve the driving of 50 -TOP translation. Regulation of S6K1 is complex, involving at least eight phosphorylation sites regulated by mTOR- and growth factor-dependent signaling pathways. PI3Kdependent activation of S6K1 is mediated by a variety of effectors, including PDK1, Akt, PKCz and l, and the small G proteins Cdc42 and Rac1 (Chou and Blenis, 1996; Alessi et al., 1998; Pullen et al., 1998; Dufner et al., 1999; Romanelli et al., 1999; and reviewed in Martin and Blenis, 2002). The best understood phosphorylation event on S6K1 is the direct phosphorylation of threonine 229, which lies in the activation loop of the kinase domain, by PDK1 (Alessi et al., 1998; Pullen et al., 1998). Phosphorylation of all the remaining sites is less well understood. The C-terminus of S6K1 contains four proline-directed sites (serines 411, 418, 421, and threonine 424). It is important to note that the in vivo kinase(s) responsible for these C-terminal phosphorylation events is not denitively known. Two phosphorylation sites in the linker region, serine 371 and threonine 389, are essential for S6K1 activity (Pearson et al., 1995; Moser et al., 1997). While the kinase responsible for phosphorylation of 371 is not known, mTOR can directly phosphorylate threonine 389 in vitro (Burnett et al., 1998; Isotani et al., 1999). Indeed, threonine 389 is the major rapamycin-sensitive site, although rapamycin also leads to the rapid dephosphorylation of threonine 229 and serine 404 (Pearson et al., 1995; Weng et al., 1998), indicating that mTOR somehow inuences the phosphorylation of these sites as well. The domain structure of S6K1 can be divided into four regions: an acidic NH2-terminus, followed by the catalytic domain, a linker region, and a basic COOHterminus containing an autoinhibitory pseudosubstrate domain. S6K1 is thought to exist in an inactive conformation when the acidic NH2-terminus interacts with the basic COOH-terminal region, thereby stabilizing the interaction of the pseudosubstrate domain with the catalytic site (reviewed in Martin and Blenis, 2002). Phosphorylation of the four C-terminal sites is thought to be an initiating event in S6K1 activation, relaxing the conformation of the kinase to allow phosphorylation of threonine 389, which is proposed to create a docking site for PDK1, resulting in the nal phosphorylation of threonine 229 by PDK1 (Biondi et al., 2001). Of course, this model is likely too simplistic to explain S6K1 activation. For example, while phosphorylation of threonine 229 is enhanced by prior phosphorylation at threonine 389, phosphorylation of threonine 389 is also enhanced by phosphorylation at threonine 229 (Weng et al., 1998; Williams et al., 2000). With regard to threonine 389 phosphorylation, the major rapamycinsensitive site, rapamycin treatment induces its rapid
3157
dephosphorylation (Pearson et al., 1995), suggesting that an mTOR-inhibited phosphatase may dephosphorylate threonine 389. Furthermore, the kinase activity and threonine 389 phosphorylation status in the DNT/ CT S6K1 mutant is resistant to rapamycin, yet still stimulated with insulin in a wortmannin-sensitive manner, suggesting that a PI3K-dependent signal leads to phosphorylation of threonine 389 as well (Cheatham et al., 1995; Weng et al., 1995). Lastly, autophosphorylation may also play a signicant role in regulating the phosphorylation of threonine 389 (Romanelli et al., 2002). While the ribosomal protein S6 is the best-characterized substrate of S6K1, other targets have been reported but not extensively studied, including the transcription factor CREMt (de Groot et al., 1994), the RNA splicing and export factor CBP80 (Wilson et al., 2000), the apoptotic protein BAD (Harada et al., 2001), and the eukaryotic elongation factor 2 kinase (eEF2K) (Wang et al., 2001). In addition, we have recently identied a novel S6K1-interacting protein and in vivo substrate, SKAR, which bears homology to the ALY/REF family of proteins that couple transcription, splicing and RNA export (Richardson et al., submitted), suggesting that perhaps S6K1 may have a role in RNA processing. At this point in time, it is fair to say that our understanding of the biochemical function of S6K1 is poor, although S6K1 clearly plays an important role in the regulation of cell growth, cell cycle progression, and cell proliferation (Kawasome et al., 1998; Shima et al., 1998; Brennan et al., 1999; Montagne et al., 1999; Vinals et al., 1999; Fingar et al., 2002, 2004) (see below). Signaling to 4E-BP1: regulation of cap-dependent translation Treatment of cells with growth factors, mitogens, and hormones increases rates of protein translation, while nutrient deprivation or environmental stresses have the opposite effect (reviewed in Gingras et al., 2001b). Similar to S6K1, 4E-BP1 (also known as PHAS-I), a repressor of the translation initiation factor eIF4E and thus an inhibitor or protein biosynthesis, is regulated by the above cellular conditions via the coordinate action of mTOR- and PI3K-dependent signals. S6K1 and 4EBP1 function in parallel pathways that bifurcate downstream of mTOR (von Manteuffel et al., 1997). While this review will focus on 4E-BP1, it is important to note that mammals possess three 4E-BPs isoforms, 4E-BP1, 2, and 3. eIF4E is the rate-limiting translation initiation factor that binds to the Cap structure (m7GpppN) at the 50 -end of mRNA transcripts to initiate cap-dependent translation (reviewed in Gingras et al., 2001b). In the absence of nutrients or growth factors, 4E-BP1 is hypophosphorylated and associates tightly with eIF4E to inhibit eIF4E function (reviewed in Gingras et al., 2001b). Growth factor stimulation in the presence of sufcient nutrients leads to the phosphorylation of 4E-BP1 at multiple sites and induces its dissociation from eIF4E. Free eIF4E then binds to the scaffolding protein eIF4G,
which assembles the initiation complex by recruiting other initiation factors including eIF4A, which is a helicase that unwinds RNA secondary structure. eIF4G further interacts with eIF3, which recruits the 40S ribosome to the 50 -end of the mRNA. As 4E-BP1 and eIF4G share an overlapping binding site on eIF4E, their binding is competitive and mutually exclusive (Marcotrigiano et al., 1999). Phosphorylation of 4E-BP1 is also complex, occurring on multiple sites in an ordered manner, and is sensitive to rapamycin and PI3K inhibitors (reviewed in Gingras et al., 2001b). In the presence of sufcient growth factors and nutrients, major sites of 4E-BP1 phosphorylation include the proline-directed sites threonines 37 and 46, serine 65, and threonine 70 (reviewed in Gingras et al., 2001b). Phosphorylation at both threonines 37 and 46 is thought to be a priming event that is absolutely required for mitogen-induced serine 65 and threonine 70 phosphorylation, which results in the release of eIF4E (Gingras et al., 1999a, 2001a; Mothe-Satney et al., 2000a). Phosphorylation of threonine 70 appears to be required for the subsequent phosphorylation of serine 65 (Gingras et al., 1999a, 2001a; Mothe-Satney et al., 2000a, 2000b). Threonines 37 and 46 are basally phosphorylated in the absence of serum and stimulated only slightly by serum (Gingras et al., 1999a). Under serum-starved but amino-acid-rich conditions, rapamycin potently inhibits threonines 37 and 46 phosphorylation, indicating that mTOR is required for the phosphorylation of these sites (Gingras et al., 2001a). Indeed, immunoprecipitated mTOR phosphorylates 4EBP1 on threonines 37 and 46 in vitro, suggesting that mTOR can directly phosphorylate these sites (Brunn et al., 1997; Burnett et al., 1998; Gingras et al., 1999a). The modest mitogen-stimulated phosphorylation of threonines 37 and 46 that occurs upon growth factor stimulation is mostly resistant to rapamycin, implying that mTOR-independent kinases also phosphorylate these sites upon mitogen stimulation (Gingras et al., 2001a). Serine 65 and threonine 70 are potently phosphorylated upon mitogen stimulation. In the presence of serum, rapamycin induces a rapid dephosphorylation of serine 65 and threonine 70 (Gingras et al., 2001a), indicating that mTOR plays a critical role in maintaining serine 65 and threonine 70 phosphorylation. It is unlikely that mTOR phosphorylates serine 65 and threonine 70 directly, given that mTOR does not phosphorylate these sites in vitro (Burnett et al., 1998). Therefore, phosphorylation of serine 65 and threonine 70 has been proposed to be regulated by an mTORdependent kinase or mTOR-inhibited phosphatase (Brunn et al., 1997; Heesom and Denton, 1999; Mothe-Satney et al., 2000b). Treatment of cells with calyculin A, which inhibits type 1 and 2 protein phosphatases, prevents rapamycin-induced dephosphorylation of 4E-BP1 (Peterson et al., 1999), underscoring the possibility that serine 65 and threonine 70 are dephosphorylated by an mTOR inhibited phosphatase. 4E-BP1 phosphorylation is also regulated by the PI3K/Akt pathway (von Manteuffel et al., 1996; Gingras et al., 1998; Takata et al., 1999). 4E-BP1
Oncogene
3158
phosphorylation is stimulated by the expression of an activated PI3K catalytic subunit or an activated Akt mutant and inhibited by a dominant-negative Akt mutant or the PI3K inhibitors wortmannin and LY294002 (Gingras et al., 1998; Dufner et al., 1999; Takata et al., 1999). Akt does not directly phosphorylate 4E-BP1 in vitro, however, suggesting that an unidentied Akt-regulated kinase(s) phosphorylates 4E-BP1 (Gingras et al., 1998). It is also reported that mitogenactivated protein kinase (MAPK) may phosphorylate and regulate 4E-BP1 (Haystead et al., 1994; Lin et al., 1994; Herbert et al., 2002). TOR-dependent regulation of phosphatases? In budding yeast, protein phosphatases are critical downstream targets of TOR. In a nutrient-dependent manner, yeast TOR proteins stimulate the association of the phosphatases Sit4 (PP6-like) and Pph21/22 (PP2Alike) with the phosphatase-interacting protein, Tap42, an interaction that results in the inhibition of phosphatase function by competing out other regulatory subunits (Di Como and Arndt, 1996; Jiang and Broach, 1999). Consistent with a role for TOR in regulating the Tap42/phosphatase interaction, nutrient deprivation or rapamycin induces dissociation of the Tap42/phosphatase complex, and mutations in Tap42 confer rapamycin resistance (Di Como and Arndt, 1996; Jiang and Broach, 1999). Recently, the TAP42-interacting protein, TIP41, was identied as a novel regulator of Tap42 and Sit4 (Jacinto et al., 2001). TIP41 interacts with TAP42 during conditions of nutrient deciency and therefore promotes PP2A activity. TOR directly phosphorylates TIP41 leading to the release of TAP42, which can then bind to and inhibit PP2A or Sit4 (Jacinto et al., 2001). This role of TOR in the regulation of phosphatase function in yeast is important for the transcriptional and translational events required for ribosome biogenesis (reviewed in Crespo and Hall, 2002). In mammalian cells, a human ortholog of TAP42 has been identied, the B-cell receptor-binding protein alpha 4 (a4), that interacts with the catalytic subunit of PP2A (Murata et al., 1997). As the rapamycin sensitivity of this interaction is controversial, it is not clear whether the mTOR pathway regulates formation of a phosphatase/Tap42-like complex in mammals, as in yeast (Murata et al., 1997; Nanahoshi et al., 1998). Consistent with a role for a phosphatase in mediating mTOR signaling, TAP42 and a4 have been shown to interfere with PP2A-induced dephosphorylation of 4E-BP1 in vitro (Nanahoshi et al., 1998), and rapamycin- or aminoacid deprivation-induced dephosphorylation of S6K1 can be blocked with the phosphatase calyculin A (Peterson et al., 1999). In addition, PP2A has been shown to interact with wild-type S6K1 but not with a rapamycin-resistant mutant (DNT/CT) (Peterson et al., 1999; Westphal et al., 1999), adding support to the idea that mTOR may regulate the activity of a phosphatase. More work will be required before it is determined whether mTOR-dependent signaling denitively regulates phosphatase function in mammals, as in yeast.
Oncogene
TOR-dependent regulation of cellular processes other than translation While the role of mTOR in control of protein synthesis in mammals is well documented, analogies to TOR function in S. cerevisiae suggest that mTOR will be shown to play important roles in a diverse array of cellular processes in addition to protein synthesis (reviewed in Rohde et al., 2001; Crespo and Hall, 2002). Mutation of TOR in budding yeast or rapamycin treatment elicits changes in gene expression, amino-acid permease function, autophagy, and the organization of the actin cytoskeleton, as well as inhibition of protein synthesis. TOR regulates, both positively and negatively, a diverse array of nutrient-regulated metabolic genes. TOR controls ribosome biogenesis by controlling the expression of ribosomal protein mRNAs as well as the transcription of rRNA and tRNA (reviewed in Rohde et al., 2001; Crespo and Hall, 2002). Gene expression is also regulated by TOR by controlling the subcellular localization of transcription factors; for example, TOR-dependent phosphorylation of the GATA transcription factor GLN3 allows it to bind to and be sequestered in the cytoplasm by the repressor URE2 (Beck and Hall, 1999). TOR also regulates, both positively and negatively, the stability of various aminoacid permeases, thus allowing the yeast cell to ne tune the intracellular amino-acid environment (reviewed in Rohde et al., 2001; Crespo and Hall, 2002). TORdependent signaling also inhibits autophagy, the process whereby cytoplasmic proteins are degraded into constituent amino acids; thus, when TOR is inactivated or when yeast are grown on rapamycin, poor carbon sources, or poor nutrient media, TOR is downregulated and autophagy is activated as a survival response (reviewed in Rohde et al., 2001; Crespo and Hall, 2002). In this way, TOR functions as a rheostat that responds sensitively to nutritional cues to help yeast survive extreme conditions. As basic cellular processes in lower and higher eukaryotes tend to be conserved, it is likely that mTOR in mammals controls cellular processes other than protein synthesis. Indeed, regulation of cellular functions in addition to protein synthesis is beginning to be ascribed to mTOR. While the role of TOR in controlling gene expression in budding yeast is well documented, a role for mTOR in controlling mammalian gene expression is just beginning to be appreciated. mTOR has been reported to directly phosphorylate STAT3 (signal transducer and activator of transcription) in a rapamycin-sensitive manner (Yokogami et al., 2000), and S6K1 has been reportedtophosphorylatethetranscriptionfactorCREMt (de Groot et al., 1994). In addition, microarray analysis of RNA isolated from cells deprived of nutrients or treated with rapamycin has demonstrated an important role for mTOR in controlling the expression of genes involved in many metabolic and biosynthetic pathways (Peng et al., 2002). mTOR, via S6K1, has also been recently shown to activate rDNA transcription, a major rate-limiting step in ribosome biogenesis, through the phosphorylation of the transcription factor UBF
3159
(Hannan et al., 2003). In addition, the phosphorylation of the protein URI, which indirectly binds to RNA polymerases I, II, and III, and is required for nutrientsensitive transcriptional regulation, has been shown to be mTOR-dependent and thus rapamycin-sensitive (Gstaiger et al., 2003). mTOR-dependent signaling may also regulate post-transcriptional processes important for mRNA biogenesis. For example, the RNA splicing and export factor CBP80 was reported to be a substrate of S6K1 (Wilson et al., 2000), and we have identied a novel substrate specic for S6K1 (i.e. not phosphorylated by S6K2), named SKAR, that bears homology to the ALY/REF family of proteins that couple transcription, splicing, and RNA processing (Richardson et al., submitted), suggesting a posttranscription role for S6K1. The fact that three of four isoforms of S6K (p85-S6K1 aI, S6K2 bI, and p54-S6K2 bII) localize to the nucleus suggests that the mTOR to S6K signaling branch regulates nuclear targets and processes, an area that awaits future research. A role for mTOR in cellular processes other than translation and gene expression is also beginning to be identied. For example, mTOR has been reported to regulate autophagy (Blommaart et al., 1995; Shigemitsu et al., 1999). mTOR may also provide antiapoptotic function: Akt-mediated protection from apoptosis is mediated, at least in part, by mTOR-dependent stabilization of cell surface amino-acid transporters (Edinger and Thompson, 2002), and the proapoptotic protein BAD has been reported to be a substrate of S6K1 (Harada et al., 2001). Clearly, we are just beginning to dene the role of mTOR in the regulation of a diverse array of cellular functions.
Mechanism of TOS to S6K1 and 4E-BP1: identication of TOR-interacting proteins and the TOS motif Although mTOR is found in a high molecular weight complex and is composed of multiple domains that may function in proteinprotein interactions, the identication of mTOR-interacting proteins has proven elusive, until recently. Raptor (150 kDa) and GbL (36 kDa) (also known as mLST8), both WD40-repeat-containing proteins, were found to associate with mTOR (Hara et al., 2002; Kim et al., 2002, 2003). The association of raptor and GbL with mTOR is evolutionarily conserved, as budding yeast TOR1 and TOR2 were found to associate with the raptor ortholog KOG1 and the GbL ortholog LST8 (Loewith et al., 2002). Raptor and mTOR are believed to function as part of a nutrient-sensitive complex, translating this information into an appropriate downstream response (reviewed in Abraham, 2002). Kim et al. (2002) suggest that raptor both positively and negatively modulates mTOR function: a constitutive mTOR/raptor interaction is required for mTOR-dependent downstream signaling, while a second, high-afnity interaction negatively regulates mTOR during nutrient insufciency. Hara et al. (2002) suggest only a positive role for raptor in regulation of
mTOR by functioning as a scaffold that brings the mTOR kinase in close proximity to its substrates. Consistent with this model, the binding of raptor to mTOR signicantly enhances mTOR-mediated phosphorylation of 4E-BP1 and S6K1, and raptor interacts with both S6K1 and 4E-BP1 (Hara et al., 2002). The interaction of mTOR with GbL appears to mediate positively mTOR function, as cotransfection of GbL with mTOR signicantly stimulates the in vitro kinase activity of mTOR toward S6K1 and 4E-BP1, and stimulates the ability of mTOR to autophosphorylate (Kim et al., 2003). Since reduction of raptor or GbL expression with siRNA reduces the phosphorylation of S6K1 on the rapamycin-sensitive site, threonine 389, it seems clear that these proteins play an important role in mTOR action (Kim et al., 2003). In budding yeast, TOR1 or TOR2 associates with KOG1 (raptor) and LST8 (GbL), a complex coined TORC1, while TOR2 is also found in a distinct complex with the proteins AVO1, 2, and 3, and with LST8, a complex coined TORC2 (Loewith et al., 2002). Interestingly, while the TORC1 complex binds rapamycin/FKBP12, and disruption of the TORC1 complex mimics rapamycin treatment, the TORC2 complex does not bind rapamycin/FKBP12 and its disruption produces an actin defect (Loewith et al., 2002). These data argue that the TORC1 complex mediates rapamycin-sensitive signaling functions, while the TORC2 complex mediates TOR2specic, rapamycin-insensitive functions. It is important to note that currently it is not known whether mTOR is able to signal in a rapamycin-insensitive manner in mammals as in yeast. As mentioned earlier, the sensitivity of mTOR/raptor complex to disruption by detergents likely explains why mTOR-interacting proteins have proven so difcult to identify in the past and may explain why the kinase activity of mTOR measured in vitro correlates poorly with the activation state of downstream targets in vivo. Indeed, when endogenous mTOR is immunoprecipitated under conditions that preserve raptor association, signicant regulation of mTOR kinase activity toward exogenous substrates is observed (e.g. amino-acid withdrawal or glucose deprivation result in decreased in vitro kinase activity of mTOR toward S6K1 and 4EBP1 that is restored upon their readdition) (Kim et al., 2002). As raptor association plays a positive role in mTOR function, the effect of rapamycin on the mTOR/raptor complex was studied: Rapamycin appears to destabilize the complex but not to disrupt it completely (Hara et al., 2002; Kim et al., 2002). The mechanism by which raptor positively mediates mTOR-dependent downstream signaling was recently claried upon the discovery of the TOS motif in the NH2-terminus of the S6 kinases and the COOHterminus of the 4E-BPs (Schalm and Blenis, 2002). A functional TOS motif is required for the mTORmediated phosphorylation of S6K1 and 4E-BP1 in vivo, as mutation of the TOS motif mimics the effect of rapamycin treatment on phosphorylation of these substrates (Schalm and Blenis, 2002; Schalm et al., 2003). The TOS motif is required for raptor to interact
Oncogene
3160
with 4E-BP1 (either directly or indirectly), thus functioning to assemble an mTOR/raptor/4E-BP1 complex (Choi et al., 2003; Nojima et al., 2003; Schalm et al., 2003). In addition, a functional TOS motif is required for the mTOR/raptor complex to phosphorylate efciently 4E-BP1 in vitro, and overexpression of a 4E-BP1 mutant bearing an inactive TOS motif (F114A) that cannot bind raptor dominantly inhibits cell growth and reduces cell size, indicating that the TOS motif on 4EBP1 is important physiologically for control of cell growth (Choi et al., 2003; Nojima et al., 2003; Schalm et al., 2003). These data support the scaffold model whereby raptor presents the mTOR kinase to its downstream substrates.
TOR regulation by the PI3K/Akt/TSC/Rheb pathway: integration of nutrient and growth factor signals An exciting new area of research in the mTOR signaling eld has been the discovery that the proteins TSC1 and TSC2 and Rheb provide an evolutionarily conserved link between the growth factor-regulated PI3K/Akt pathway and the nutrient-sensitive mTOR pathway. TSC1/2, a novel negative regulator of TOR In both mammals and Drosophila, TSC1 (also known as hamartin) and TSC2 (also known as tuberin) associate with each other to form a heterodimer that provides tumor suppressor function by inhibiting cell cycle progression and cell proliferation (reviewed in Kwiatkowski, 2003; Manning and Cantley, 2003b). Consistent with the requirement for heterodimer formation in tumor suppressor function, mutation of either TSC1 or TSC2 results in the pediatric disease tuberous sclerosis, TSC, which is an autosomal dominant disorder affecting B1 in 6000 children (reviewed in Kwiatkowski, 2003). In TSC, benign tumors known as hamartomas form in many organs including the brain, heart, kidney, lungs, and skin, and aficted individuals often have neurological problems including seizures, mental retardation, and autism, heart, kidney, and lung failure, and severe skin rashes. While the hamartomas can grow in size to cause serious medical problems, they rarely become malignant; for example, only B2% of TSC patients develop malignant tumors, which tend to occur in the kidney (reviewed in Kwiatkowski, 2003). In mice, homozygous inactivation of either TSC1 or TSC2 results in embryonic lethality, while heterozygous inactivation predisposes the animals to tumorigenesis, consistent with these proteins providing tumor suppressor function in vivo (Onda et al., 1999; Kobayashi et al., 1999, 2001; and reviewed in Kwiatkowski, 2003). The molecular function of the TSC1/2 complex has proven elusive until recently. Complementary data from genetics and biochemistry in both mammalian systems and Drosophila have provided novel insight into the molecular function of TSC1 and TSC2 and their role in organismal physiology. Genetic screens in the Drosophila compound eye geared toward identifying negative
Oncogene
regulators of cell growth and proliferation identied TSC1 as a gene, which when mutated, produced an overgrowth phenotype large eyes containing large ommatidia as a result of a cell autonomous increase in cell size (Gao and Pan, 2001; Potter et al., 2001; Tapon et al., 2001a). Indeed, inactivation of either TSC1 or TSC2 (also known as gigas) produces large eyes as a result of increased cell size, while overexpression of TSC1 and TSC2 together but not alone reduces overall organ size as a result of a reduction in both cell size and cell number (Gao and Pan, 2001; Potter et al., 2001; Tapon et al., 2001a). Thus, a primary role for TSC1 and TSC2 appears to be in restricting cell and organ size. The role of TSC1/2 in restricting cell growth is evolutionarily conserved, as overexpression of TSC1 or TSC2 in cultured mammalian cells also reduces cell size (Inoki et al., 2003b; Rosner et al., 2003). Genetic epistasis experiments placed TSC1/2 in the Drosophila insulin signaling pathway: TSC1 and TSC2 were found to be epistatic (dominant) to the insulin receptor, PTEN, and dAkt, but dS6K was found to be epistatic to TSC1/ 2 (Gao and Pan, 2001; Potter et al., 2001; Tapon et al., 2001a; Radimerski et al., 2002). These interactions suggested that TSC1/2 functioned genetically downstream of PI3K and Akt but upstream of dS6K. Interestingly, manipulation of both PTEN and TSC1/2 produce a synergistic phenotype (e.g. overexpression of both more strongly reduces eye size than overexpression of either one individually, or inactivation of both more strongly increases eye size than inactivation of either one individually), indicating that PI3K also signals independent of TSC1/2 to regulate cell and organ size (Gao and Pan, 2001; Tapon et al., 2001a). Consistent with data from mice, homozygous inactivation of either TSC1 or TSC2 in ies results in embryonic lethality (Gao and Pan, 2001; Potter et al., 2001; Tapon et al., 2001a). The observation that a reduction of dTOR or dS6K signaling partially rescues the lethality and increased cell size phenotype of ies lacking TSC1/2 function conrms the placement of dTOR downstream of (or parallel to) TSC1/2 in a biochemical signaling pathway that controls cell growth (Gao and Pan, 2001; Radimerski et al., 2002). On the heels of genetic analyses in ies identifying TSC1/2 in the insulin signaling pathway as a regulator of cell growth came biochemical experiments conrming the placement of TSC1/2 upstream of dTOR, functioning to regulate negatively it and its downstream targets. In cultured cells from both Drosophila and mammals, overexpression of TSC1 and TSC2 together inhibits the phosphorylation of S6K and 4E-BP1 (Inoki et al., 2002; Tee et al., 2003b), while depletion of TSC1 or TSC2 with siRNA leads to increased phosphorylation of S6K and 4E-BP1 that is rapamycin-sensitive (Gao et al., 2002). Importantly, TSC2 alleles derived from TSC patients fail to inhibit phosphorylation of S6K1 and 4E-BP1 (Inoki et al., 2002; Tee et al., 2003b). Mammalian cells lacking TSC1 or TSC2 also exhibit increased S6K1 phosphorylation and activity that is rapamycin sensitive (Gao et al., 2002; Goncharova et al., 2002). Moreover, mutants of S6K1 that are insensitive to an mTOR input,
3161
and thus resistant to the inhibitory effect of rapamycin, are not inhibited by an overexpressed TSC1/2 complex (Inoki et al., 2002; Tee et al., 2002). These data collectively demonstrate TOR to be an obligate intermediate in the TSC-dependent regulation of S6K1. Several pieces of data suggest that the TSC1/2 complex is somehow required for mTOR to sense nutritional sufciency. Overexpression of TSC1/2 blocks the ability of amino acids to activate S6K1 in nutrientdeprived cells (Inoki et al., 2002; Tee et al., 2002). Moreover, amino-acid deprivation fails to reduce the increased phosphorylation of S6K observed in TSC1 or TSC2 null mammalian cells, or in Drosophila cells with siRNA-mediated reduction of TSC1 or TSC2 expression; importantly, the S6K activation that results from PTEN deletion in ies is still sensitive to amino-acid withdrawal (Gao et al., 2002). These data argue that a functional TSC1/2 complex is required for the nutrientsensing ability of TOR. It is important to note, however, that the mechanism for such a regulation is unknown; furthermore, it is also not known whether the nutrient responsiveness of mTOR is sensed solely via the upstream TSC complex or whether nutrient signals can be sensed by both the TSC complex and TOR itself (Figure 2). Akt phosphorylates and inhibits TSC2 While genetic analysis placed dTSC1/2 downstream of dPI3K and dAkt, the biochemical explanation for such a relationship was unknown. Several groups showed that PI3K-dependent activation of Akt results in the phosphorylation of TSC2 in both mammalian and Drosophila cells (Dan et al., 2002; Inoki et al., 2002; Manning et al., 2002; Potter et al., 2002). Human TSC2 is phosphorylated in vitro on both serine 939 and threonine 1462, and growth factors that activate PI3K phosphorylate these sites in a wortmannin-sensitive manner (Dan et al., 2002; Inoki et al., 2002; Manning et al., 2002). In PTEN/ cells that exhibit constitutive PI3K-dependent signaling, phosphorylation of threonine 1462 is constitutive (Manning et al., 2002). Moreover, overexpression of a mutant of TSC2 in which serine 969 and threonine 1462 are mutated to alanine more strongly represses S6K1 phosphorylation and activity than overexpression of wild type (Inoki et al., 2002; Manning et al., 2002). The mechanism by which phosphorylation of TSC2 by Akt inhibits its tumor suppressive function remains controversial. Some groups have demonstrated that phosphorylation of TSC2 destabilizes it, therefore disrupting the formation of functional TSC1/2 complexes: Coexpression of TSC2 with constitutively active Akt or expression of a phosphomimetic TSC2 mutant results in shorter TSC2 half-life; conversely, expression of a TSC2 mutant bearing alanine substitutions at the Akt sites results in longer half-life (Dan et al., 2002; Inoki et al., 2002). Phosphomimetic TSC2 mutants were found to be ubiquitinated, and proteosome inhibitors were found to stabilize TSC2 (Dan et al., 2002; Inoki et al., 2002). Lastly, TSC1/2 complexes were found to be less
abundant in the presence of active PI3K/Akt-dependent signaling ((Dan et al., 2002; Inoki et al., 2002; Potter et al., 2002). It is important to note, however, that other groups have not observed PI3K/Akt-dependent regulation of TSC2 stability or abundance of TSC1/2 complexes (Manning et al., 2002; Tee et al., 2002). The recent discovery that the PI3K/Akt/TSC pathway is an upstream regulator of mTOR and S6K1 has helped to clarify the controversy as to whether Akt is a physiological upstream activator of S6K1. In mammalian cells, while membrane-targeted Myr-Akt has been shown to increase S6K1 activity, activated Akt mutants that remains cytoplasmically localized fail to augment S6K1 activity (Dufner et al., 1999). It was suggested that Akt must exert its action at the membrane to effect S6K1 regulation. Dominant-negative Akt, however, inhibits 4E-BP1 phosphorylation while having no inhibitory effect on S6K1 activity (Dufner et al., 1999). It is possible that Myr-Akt augments S6K1 activity by phosphorylating and inactivating a larger pool of TSC2, and perhaps this event must occur at the membrane. Dominant-negative Akt (PKB) may not inhibit S6K1 if the TSC1/2 tumor suppressor complex can be phosphorylated and inactivated by other signals independent of PI3K/Akt. Indeed, the phosphorylation and inactivation of TSC2 has been shown to also occur via a PI3K/ Akt-independent mechanism, one that possibly involves PKC- and MEK-regulated pathways (Tee et al., 2003a). Thus, it appears that both nutrients, via an unknown pathway, and growth factors, via PI3K/Akt-dependent and -independent phosphorylation of TSC2, inhibit the tumor suppressor function of the TSC1/2 complex, allowing the phosphorylation of the downstream TOR targets S6K1 and 4E-BP1. In this way, the TSC complex functions to sense and integrate signals from both nutrients and growth factors to regulate TOR function. How the TSC complex inhibits TOR was a mystery until recently. Rheb, an activator of TOR that is inhibited by TSC2 Once again, complementary data derived from genetics and biochemistry in both Drosophila and mammalian systems have combined to reveal a mechanism by which the TSC1/2 complex inhibits TOR function: TSC2 acts as a GTPase-activating protein (GAP) toward the Rasrelated small GTPase Rheb , which is a positively acting upstream regulator of TOR. GAPs are a class of proteins that convert GTPases from an active GTPbound state to an inactive GDP-bound state. Therefore, TSC2 inactivates Rheb, which results in the downregulation of TOR. Genetic screens in the Drosophila compound eye identied Rheb as a gene that when mutated decreased eye and cell size or when overexpressed increased eye and cell size (Saucedo et al., 2003; Stocker et al., 2003). Genetic epistasis experiments demonstrated Rheb to be epistatic to PI3K, dAkt, and TSC1/2, while dTOR and dS6K were found to be epistatic to Rheb (Saucedo et al., 2003; Stocker et al., 2003). Thus, Rheb functions as a positive regulator of cell and organ growth, localizing genetically
Oncogene
3162
downstream of PI3K, Akt, and TSC1/2, but genetically upstream of dTOR and dS6K. While TSC2 was known to contain a Rap1-like GAP domain in its COOH-terminus, the physiologic in vivo target was unknown. As a result, at the same time other groups sought the GTPase target of TSC2s GAP domain. A clue came from the phenotypes resulting from the inactivation of the Schizosaccharomyces pombe Rheb ortholog, Rhb1, or the TSC1/2 orthologs, tsc1/2: loss of Rhb1 or tsc1/2 results in a G0/G1 arrest that is similar to that induced by nitrogen starvation (Mach et al., 2000; Matsumoto et al., 2002). Furthermore, loss of Rheb in S. cerevisiae results in increased uptake of arginine and lysine (Urano et al., 2000). Using the fact that Rheb in yeast functions in a nutrient sensing pathway, biochemical approaches demonstrated that TSC2 indeed functions as a GAP toward Rheb: TSC2 catalyses the conversion of Rheb-GTP into Rheb-GDP in vitro (Tee et al., 2003a; Zhang et al., 2003), and expression of TSC2 or the TSC1/2 complex in vivo increases the ratio of GDP:GTP-bound Rheb (Castro et al., 2003; Garami et al., 2003; Inoki et al., 2003a; Zhang et al., 2003). Importantly, TSC2-bearing mutations in the GAP domain fails to affect the GTP-bound state of Rheb, and cells lacking TSC2, either via RNAimediated reduction of TSC2 or via gene knockout, display increased Rheb-GTP loading (Garami et al., 2003; Zhang et al., 2003). Consistent with the genetic results, Rheb overexpression in mammalian cells increases S6K1 and 4E-BP1 phosphorylation in a nutrient- and growth factor-independent manner that is sensitive to rapamycin but insensitive to wortmannin, indicating that Rheb lies upstream of mTOR but downstream of PI3K (Castro et al., 2003; Garami et al., 2003; Inoki et al., 2003a; Tee et al., 2003b). Morever, reduction of Rheb expression with RNAi leads to decreased phosphorylation of S6K1 in response to insulin (Garami et al., 2003). Overexpression of wildtype TSC2, but not a GAP-domain mutant, reduces the ability of Rheb to activate the mTOR pathway, showing TSC2 to be a negative regulator of Rheb in mammalian cells (Castro et al., 2003; Garami et al., 2003; Tee et al., 2003b). Many important questions remain regarding how Akt, TSC1/2, and Rheb modulate TOR function. The molecular mechanism by which Akt inhibits the tumor suppressor function of TSC1/2 remains controversial, as different groups have reported different effects of Aktmediated phosphorylation on the stability of TSC2 itself and on expression of a TSC1/2 complex. How nutrients modulate TOR function remains a major unresolved question. Since inactivation of TSC1/2 or Rheb overexpression leads to constitutive TOR-dependent signaling in the absence of nutrients, the TSC1/2 complex and Rheb may function as nutrient sensors. How TSC1/2 or Rheb might respond to nutrient signals is completely unclear at this time. The identity of the guanine nucleotide exchange factor (GEF) that loads GTP on to Rheb is also unknown, although it has been suggested that Rheb may not require a dedicated GEF as it exists in a highly active GTP-bound state and possesses low
Oncogene
intrinsic GTPase activity. Finally, the molecular mechanism by which Rheb activates TOR, either directly or indirectly, is unknown and is a question that is likely being vigorously pursued at this time. The nding that TSC1/2 exert their tumor suppressive function by downregulating mTOR-dependent signaling suggests that rapamycin and analogs may be clinically efcacious for the treatment of tuberous sclerosis (reviewed in Manning and Cantley, 2003a).
Coordinate control of cell growth and cell cycle progression by the TOR and PI3K pathways While the proliferation of lower eukaryotes like yeast is driven simply by nutrient availability, the development of more complex, multicellular eukaryotes requires that signals from nutrients and growth factors be tightly coordinated. The crosstalk that occurs between the nutrient-regulated TOR pathway and the growth factorregulated PI3K pathway ensures that cell proliferation only occurs when amino acids, energy, and growth factors are in sufcient supply. While TOR and PI3K have been known to regulate cell division and proliferation for some time (rapamycin and PI3K inhibitory drugs inhibit cell cycle progression and proliferation by inducing an accumulation of cells in G1 phase; reviewed in Abraham and Wiederrecht, 1996), only more recently has a role for TOR and PI3K in regulation of cell growth been appreciated. The growth of an organ or whole organism is mediated by increases in both cell size and cell number through the coordinated action of cell growth and cell cycle progression. Cell growth refers to the increase in mass and size that occurs as a cell progresses through the cell division cycle (cdc) in preparation for division in mitosis, whereby a large mother cell symmetrically splits into two smaller daughter cells. It is fairly intuitive that a cell must double in mass through increased macromolecular biosynthesis and grow to increased size during each cell division cycle to maintain its size. Thus, the relative rates of cell growth vs cell division determine the nal size of the cell. By analysing temperaturesensitive cdc mutants in S. cerevisiae, Hartwell and coworkers gained important insight into the relationship between cell growth and cell division. They noted that when cell division is blocked upon the inactivation of cdc genes, cell growth continues and the yeast arrest at a large cell size (Johnston et al., 1977). In contrast, when deprived of nutrients or when cdc genes encoding biosynthetic proteins are inactivated, cell division and cell growth are coordinately blocked, suggesting that sufcient cell growth is required for cell cycle progression, but not vice versa (Johnston et al., 1977). Similarly, disruption of cell cycle regulatory genes in Drosophila (e.g. dE2F or cdc2) results in cell cycle arrest at a large cell size (Weigmann et al., 1997; Neufeld et al., 1998), and induction of a G1 phase cell cycle block by overexpressing the cyclin D/cdk4 inhibitor p16 in mammalian broblasts induces a shift to increased cell
3163
size (Fingar et al., 2002). These data from evolutionarily diverse organisms demonstrate that while normally tightly coupled, cell growth and cell division can be experimentally dissociated, indicating that distinct pathways mediate these fundamental processes. Moreover, since cell division is dependent on a sufcient level of cell growth, cell growth pathways must crosstalk with pathways that control cell cycle progression. It is important to note that the molecular mechanism(s) by which cell growth is coupled to cell division is poorly understood. The TOR and PI3K pathways are highly conserved between Drosophila and higher eukaryotes, making Drosophila an attractive model genetic system to study the role of these pathways in cell growth, proliferation, and development. Manipulation of virtually all components of the Drosophila insulin signaling pathway has been shown to alter organ and body size as a result of cell autonomous changes in cell size and number (reviewed in Stocker and Hafen, 2000; Weinkove and Leevers, 2000; Oldham and Hafen, 2003). For example, inactivation of positively acting molecules such as the insulin receptor, PI3K, chico (the insulin receptor substrate (IRS) ortholog), PDK1, Akt, TOR, Rheb, or S6K decreases organ and cell size concomitant with a slowing down of the cell cycle, while overexpression of many of these molecules increases organ and cell size (Bohni et al., 1999; Chen et al., 1996; Leevers et al., 1996; Montagne et al., 1999; Verdu et al., 1999; Oldham et al., 2000; Zhang et al., 2000; Rintelen et al., 2001; Saucedo et al., 2003; Stocker et al., 2003). Consistently, inactivation of negatively acting molecules such as PTEN or TSC1/2 increases organ and cell size, while their overexpression decreases organ and cell size (Gao et al., 2000; Goberdhan et al., 1999; Huang et al., 1999; Gao and Pan, 2001; Tapon et al., 2001a; Potter et al., 2002). In addition, overexpression of a dominant mutant of 4E-BP1 with high afnity for eIF4E reduces cell and organ size (Miron et al., 2001). These genetic studies clearly demonstrate the TOR and PI3K pathways to be important regulators of cell and organ growth. It is important to bear in mind that these results were initially somewhat unexpected, as many of these proteins were originally considered to be proto-oncogenes with a role in regulating cell cycle progression and proliferation. Studies in mice and mammalian cell culture have conrmed the evolutionarily conserved function of the TOR and PI3K pathways in the regulation of cell, organ, and organismal growth. Homozygous knock out of IRS-1, Akt1 (PKBa), or S6K1 in mice results in a small animal phenotype, although whether the mice are small as a result of reduced cell number, reduced cell size, or a combination of both mechanisms was not reported (Araki et al., 1994; Shima et al., 1998; Cho et al., 2001b; Lawlor et al., 2002; Mora et al., 2003). Modulation of PI3K, PDK1, or Akt expression in the heart affects heart size as a result of changes in cardiomyocyte size but not cell number (Shioi et al., 2000, 2002; Mora et al., 2003). PI3K/Akt and TOR have also been implicated in the control of skeletal muscle
size: Overexpression of activated Akt or S6K1 induces hypertrophy of cultured myotubes, while inhibition of TOR or PI3K with rapamycin or LY294002, respectively, blocks IGF-1-induced hypertrophy (Rommel et al., 2001). In cultured cells that are asynchronously cycling, rapamycin and LY294002 both induce a shift to reduced cell size, indicating mTOR and PI3K to be important mediators of cell growth (Fingar et al., 2002). Consistently, reduction of mTOR, raptor, or GbL expression with RNAi also reduces cell size (Kim et al., 2002; Kim et al., 2003). Overexpression of PI3K or Akt increases cell size in a rapamycin-sensitive manner, indicating that the ability of the PI3K/Akt pathway to drive cell growth is mTOR dependent (Alvarez et al., 2003; Faridi et al., 2003). Overexpression of S6K1 or eIF4E also increases cell size, while overexpression of a dominantly acting mutant of 4E-BP1 decreases cell size (Fingar et al., 2002). Moreover, overexpression of rapamycin-resistant mutants of S6K1 or overexpression of eIF4E partially rescues the small cell size phenotype induced by rapamycin, indicating S6K1 and eIF4E to be downstream mediators of mTOR-dependent cell growth (Fingar et al., 2002). mTOR is best known as a regulator of cell cycle progression and cell proliferation (reviewed in Abraham and Wiederrecht, 1996), although the biochemical mechanisms by which mTOR mediates these events have been poorly dened until recently. In B and T cells, rapamycin induces a strong G1phase arrest and is therefore a potent antiproliferative drug. In most other cell types, however, rapamycin merely delays cell cycle progression and reduces proliferation rate. These features of rapamycin make it a great immunosuppressive drug, effecting strong immunosuppression with limited toxicity. When cycling nonimmune cells are treated with rapamycin, they proliferate slowly at a smaller than normal size (Fingar et al., 2002). These data clearly show that a given cell type does not divide at a xed size. Rather, cell size is dynamic and responsive to the extracellular milieu. Indeed, budding yeast grown on poor carbon sources proliferate at reduced cell size (Flick et al., 1998), and rat Schwann cells divide at a size that varies depending on the concentration of extracellular growth factors in the media (Conlon et al., 2001). It is intriguing that the same set of mTOR-dependent downstream effectors that regulate cell growth and cell size, S6K1, 4E-BP1, and eIF4E, also regulate the rate of cell cycle progression: When quiescent U2OS cells are stimulated with serum to enter G1 phase from G0, overexpression of S6K1 and eIF4E accelerates S phase entry, while reduced expression of S6K1 with RNAi or overexpression of a dominant mutant of 4E-BP1 inhibits the rate of S phase entry (Fingar et al., 2004). Moreover, overexpression of rapamycin-resistant mutants of S6K1 or overexpression of eIF4E partially rescues the rapamycin-induced delay in G1 progression to S phase, indicating S6K1 and eIF4E to be downstream mediators of mTOR-dependent cell division (Fingar et al., 2004). These data are consistent with the ability of rapamycin-resistant mutants of S6K1 to partially restore rapamycin-suppressed
Oncogene
3164
E2F-dependent transcription and to partially rescue rapamycin-inhibited proliferation of vascular smooth muscle cells (Brennan et al., 1999; Vinals et al., 1999); furthermore, embryonic stem cells lacking S6K1 proliferate in culture at a reduced rate (Kawasome et al., 1998). Consistent with RNAi against S6K1 delaying G1 phase progression, cells containing reduced S6K1 expression exhibit reduced cell size (Fingar et al., 2004; Richardson et al., submitted). These data demonstrate mTOR to be a central coordinator of cell growth and cell cycle progression, but what is the mechanism? We favor a model whereby the primary role of mTOR is to regulate the rate of cell growth, with mTOR-dependent regulation of cell cycle progression a secondary consequence (Figure 3). It is also possible, however, that in addition to driving cell growth, mTOR may also directly inuence the cell cycle machinery (e.g. phosphorylation), similar to PI3K, although there is currently no strong evidence for such a model. mTOR likely restricts the rate of cell cycle progression, and thus cell division and proliferation rates, when environmental conditions are inappropriate and thus growth rate is low. When conditions are optimal, however, it is likely that overactivation of the mTOR pathway, while able to accelerate G1 phase progression modestly , is not sufcient to accelerate cell division and cell proliferation. In support of this idea, overexpression of S6K1 confers a proliferative advantage in low serum-containing media, a hallmark of transformation, but does not increase proliferation rate in full serum-containing media (Fingar et al., 2004). Moreover, overexpression of a constitutively active mutant of S6K1 is not sufcient to induce oncogenic transformation (Mahalingam and Templeton, 1996). It is interesting to note, however, that overexpression of eIF4E in NIH-3T3 and Rat2
Figure 3 Coordination of cell growth and cell cycle progression by mTOR. mTOR may coordinate cell growth and cell cycle progression my making cell cycle progression dependent on a sufcient level of cell growth. In this model, which we favor, mTOR primarily regulates cell growth, and as a secondary consequence regulates the rate of cell cycle progression. In another possible model, mTOR may directly regulate the activity of cell cycle machinery components (e.g. phosphorylation), in parallel to its role in control of cell growth. Cell growth and cell cycle progression would therefore be coordinated by virtue of sharing a common upstream regulator, mTOR
Oncogene
broblasts is sufcient to induce transformation (Lazaris-Karatzas et al., 1990), although there are no other reports of eIF4E-induced transformation in other cell types. These data support the idea that signaling along the mTOR/S6K1 and mTOR/4E-BP1/ eIF4E branches are required but not sufcient for oncogenesis. The role of mTOR in coordinating cell growth and cell cycle progression differs from the proposed role of PI3K as such a coordinator (Alvarez et al., 2003). While the PI3K pathway promotes cell growth in an mTORdependent manner, it also is able to accelerate cell cycle progression, cell division, and cell proliferation under optimal conditions, likely via mTOR-independent pathways (Alvarez et al., 2003). More mechanistically speaking, the PI3K/Akt pathway is reported to control directly the phosphorylation and stability of cyclin D1 (Diehl et al., 1998), and the phosphorylation and function of the cdk inhibitory protein p21 (Rossig et al., 2001; Zhou et al., 2001). Thus, PI3K-dependent signals likely branch into those controlling cell growth via mTOR and those that more directly inuence the cell cycle machinery. This role of PI3K is consistent with the observation that aberrantly high PI3K-dependent signaling is found in many tumors, while disregulated mTOR-mediated signaling is not commonly associated with tumorigenesis. Indeed, the hamartomas that arise in tuberous sclerosis as a result of aberrantly high mTOR-dependent signaling are benign and rarely progress to malignancy. How would an increased rate of TOR-dependent cell growth inuence the rate of cell cycle progression? If the synthesis of cell cycle driving proteins were sensitive to the protein biosynthetic capacity of the cell, then the rate of cell growth could be effectively coupled to cell cycle progression (reviewed in Tapon et al., 2001b). Consistent with this model, the S6K1 and 4E-BP1/ eIF4E pathways function in translational control. Indeed, in budding yeast, the synthesis of the G1-cyclin CLN3 is nutrient- and rapamycin-sensitive and controlled by TOR and Cdc33 (an eIF4E ortholog) at the level of translation initiation (Barbet et al., 1996). A short upstream open reading frame in the 50 leader of the CLN3 transcript functions as a translational control element to repress CLN3 expression when protein synthesis and cell growth rate are low, as during nutrient deprivation (Polymenis and Schmidt, 1997). When growth conditions are favorable, CLN3 protein is efciently translated and accumulates to drive passage through G1 phase. Data from Drosophila support the idea that expression of a cyclin may be a mechanism by which cell growth is coupled to cell cycle progression: In dTOR null mutant tissue, levels of cyclin E protein are low, and ectopic expression of cyclin E rescues S phase entry in dTOR null cells (Zhang et al., 2000). By analogy, perhaps the translation of cyclins in mammalian cells is regulated by an mTOR-regulated, cell growth-dependent mechanism. By such a mechanism, such cyclins could be thought of as cell growth sensors rather than cell growth drivers. It is important to note that while there are examples in the literature of
3165
rapamycin inhibiting the expression of various cell cycle control proteins (Braun-Dullaeus et al., 2001), it not clear whether the observed effect is a direct effect of mTOR inhibition or whether it is an indirect secondary consequence of a delay in cell cycle progression. Certain cyclins may also be able to function as cell growth drivers, as it is reported that in Drosophila, cyclin D/ cdk4 complexes promote cell growth in addition to their role in driving cell cycle progression (Datar et al., 2000; Meyer et al., 2000). Whether cyclin/cdk complexes also have a growth-promoting function in mammals is not known.
Disregulation of coordinate TOR and PI3K signaling contributes to tumorigenesis The signicance of the connection between the PI3K pathway and the mTOR pathway is highlighted by the observation that the growth and proliferation of tumors displaying aberrantly high PI3K-dependent signaling, either through constitutive activation of PI3K or Akt or via inactivation of the PI3K antagonist PTEN, display enhanced sensitivity to rapamycin and analogs (Aoki et al., 2001; Neshat et al., 2001; Podsypanina et al., 2001). As this topic has been the subject of many comprehensive reviews, including one in this issue, we will not extensively discuss it here (reviewed in Vogt, 2001; Huang and Houghton, 2003; Luo et al., 2003; Shamji et al., 2003). This differential sensitivity of PI3K/ Akt-dependent tumors to rapamycin makes mTOR and its downstream targets S6K1 and 4E-BP1/eIF4E exciting novel targets for cancer therapeutics. As a result, rapamycin analogs, such as CCI-779 and RAD001, are currently being tested in clinical trials against a variety of human tumors (Elit, 2002; and reviewed in Vivanco and Sawyers, 2002; Huang and Houghton, 2003; Luo et al., 2003). Since inhibition of mTOR produces only modest effects on rates of cell proliferation in most cell types, it is likely that treating patients with a combination of mTOR inhibitors plus other chemotherapy drugs may be the best approach for treating cancer in the clinic. Indeed, a combination of rapamycin plus the ABL tyrosine kinase inhibitor Imatinib (also known as Gleevac or STI571) synergistically inhibits the proliferation and survival of BCR/ABL-transformed myeloid and lymphoid cell lines, models of chronic myelogenous leukemia (Ly et al., 2003; Mohi et al., 2004). Use of a combination of drugs that target distinct signaling pathways or that target different points within a signal transduction pathway may allow clinicians to better treat tumors and more effectively battle the pervasive problem of drug resistance. That eIF4E induces oncogenic transformation in certain cell types suggests that aberrantly high capdependent translation may predispose a cell to tumorigenesis. Indeed, eIF4E is overexpressed in many human cancers, although it is not known whether this is a cause or consequence of oncogenesis (reviewed in De Benedetti and Harris, 1999). Moreover, aberrantly high rates
of protein biosynthesis are observed in tumors (reviewed in Ruggero and Pandol, 2003), and tumorigenesis has been reported to require a higher degree of capdependent translation to inhibit apoptosis (Li et al., 2002). In accordance with these ideas, a recent paper reports that the proto-oncogenes Ras and Akt act rapidly to increase the association of mRNA transcripts with polyribosomes, and therefore may acutely inuence protein translation; changes in global transcription occur with a slower time course (Rajasekhar et al., 2003). Thus, increased translation may be an initiating event in tumorigenesis that is followed later by global transcriptional changes (Rajasekhar et al., 2003). Consistent with the idea that oncogenes may act to alter protein translation state, cyclin D1 and c-Myc, two proteins important for driving cell proliferation, have been identied as targets of mTOR whose transcription and mRNA association with polysomes (and presumably their translation) is decreased by rapamycin in cells with aberrantly high Akt-dependent signaling (Gera et al., 2003). Lastly, rapamycin is reported to enhance proteolytic degradation of hypoxia-inducible factor 1a (HIF-1a), a transcription factor that drives hypoxiainducible gene expression (Hudson et al., 2002). As a result, inhibition of the cellular response to hypoxic stress, an event crucial for the survival of cancer cells under hypoxic conditions and for tumor progression, may underlie, at least in part, the apparent efcacy of mTOR inhibitors as antineoplastic drugs. The discovery that the TSC1/2 complex lies upstream of mTOR and inhibits mTOR-dependent signaling makes mTOR and its downstream effectors exciting novel targets for the treatment of the benign, but devastating tumors that arise in tuberous sclerosis (reviewed in Manning and Cantley, 2003a). In addition, mTOR inhibitors may prove useful for treating the tumors associated with PeutzJeghers syndrome, a disease caused by inactivating mutations in the tumor suppressor protein LKB1 (Hemminki et al., 1998; Jenne et al., 1998). LKB1 has recently been shown to phosphorylate and activate AMPK, thus enhancing the inhibitory effect of the TSC1/2 complex toward mTOR (Hawley et al., 2003; Hong et al., 2003; Woods et al., 2003). Thus, LKB1 inuences the ability of energy status to modulate mTOR function, and as a result, its inactivation results in aberrrantly high mTOR-dependent signaling even under low-energy conditions (Shaw et al., 2004). To begin to develop efcacious cancer therapeutics that target the mTOR pathway, it will be important in the future to more completely identify the downstream targets that mediate mTORs effects on cell growth, cell cycle progression, cell proliferation, and cell survival. It is likely that mTOR will differentially control the expression of such targets depending on the cell or tumor type examined, thus making the identication of mTOR-dependent targets a complicated endeavor. It appears as though the mTOR pathway exerts control of cellular proliferation, and thus oncogenesis, through a diverse spectrum of biochemical mechansims.
Oncogene
3166
A role for TOR in other human disease states Besides contributing to cancer formation, disregulated mTOR and PI3K signaling likely contributes to the pathophysiology of other human disease states in nonproliferative tissue, including heart disease, diabetes, and muscular atrophy. As mentioned earlier, activation of the PI3K/Akt pathway increases the heart size in mice in an mTOR-dependent manner, while decreased signaling through this pathway, as observed in PDK1/ mice or through overexpression of dominant-negative alleles of PI3K or Akt, results in reduced heart size (Mora et al., 2003; Shioi et al., 2000; Shioi et al., 2002; Shiojima et al., 2002). These changes in heart size are mediated by changes in cardiomyocyte size but not cell number. The PDK1/ mice die at 511 weeks postbirth as a result of heart failure and increased sensitivity to hypoxia-induced cardiomyocyte death (Mora et al., 2003). Thus, disregulation of mTOR and PI3K signaling can promote cardiac hypertrophy or hypotrophy, both of which can lead to heart failure. In addition, the mTOR pathway positively mediates the intimal hyperplasia of vascular smooth muscle cells that often follows angioplasty or other vascular intervention and can lead to restenosis (reviewed in Marks, 2003). Intimal hyperplasia may be induced upon vessel injury, causing the vascular smooth muscle cells to change from a quiescent, differentiated, contractile phenotype to a more proliferative, dedifferentiated, and migratory phenotype (reviewed in Owens, 1995). Rapamycin inhibits the increased proliferation and migration of vascular smooth muscle cells that occurs after injury (Poon et al., 1996; Gallo et al., 1999; Roque et al., 2000) and also induces their differentiation back toward a more mature phenotype (Martin et al., 2003). The injury-induced expression of several cell cycle regulatory proteins is inhibited by the in vivo administration of rapamycin to rats, suggesting a mechanism for the antirestenotic effect of rapamycin on proliferation (Braun-Dullaeus et al., 2001). After successful clinical trials, rapamycin (sirolimus) is now FDA approved for use on drug-eluting stents to inhibit restenosis after coronary artery intervention (Morice et al., 2002; Moses et al., 2003). Disregulated TOR and PI3K function may also contribute to the development of type II diabetes. IRS2, Akt2 (PKBb)- and S6K1-decient mice develop type II diabetes, a disorder characterized by insulin resistance in peripheral tissues, impaired insulin suppression of hepatic glucose output, and impaired glucose-induced insulin production and secretion by pancreatic b cells (Shima et al., 1998; Withers et al., 1998; Cho et al., 2001a; and reviewed in Saltiel and Kahn, 2001; Kozma and Thomas, 2002). The diabetes in the IRS2/ mice results from impaired insulin action in liver and skeletal muscle and from an inability of b cells to compensate for the peripheral insulin resistance by upregulating insulin production (Withers et al., 1998). It is interesting that the Akt2/ and S6K1/ mice develop diabetes for different reasons than the IRS2/ mice: while both the Akt2/ and S6K1/ mice exhibit
Oncogene
hyperglycemia upon glucose injection, the Akt2-decient mice exhibit fasting hyperglycemia, while the S6K1-decient mice do not. Careful investigation of these phenotypes revealed that the Akt2-decient mice are defective in peripheral insulin action in skeletal muscle and liver while retaining normal b-cell compensation (Cho et al., 2001a); the S6K1-decient mice, however, are diabetic due to an impaired ability to produce insulin as a result of decreased pancreatic b-cell size (Pende et al., 2000). The overlapping but distinct diabetic phenotypes resulting from inactivation of IRS2, Akt2, and S6K1 suggest that while these signaling molecules operate in a shared signaling pathway, they likely also mediate distinguishable functions through distinct signaling pathways. As mentioned earlier, mTOR- and PI3K-dependent signaling have been shown to be important for controlling the size of skeletal muscle cells, as rapamcyin and LY294002 block IGF1-induced hypertrophy (Rommel et al., 2001). As a result, activation of these pathways could be therapeutically benecial for treating the muscular atrophy that is associated with many disease states.
Future directions Many issues remain with regard to the role of TOR in cellular and animal physiology. A new communication reports that TOR deciency in C. elegans (let-363) doubles lifespan compared to wild-type animals (Vellai et al., 2003), similar to the phenotype of animals decient in DAF-2/insulin-like growth factor signaling (Kenyon et al., 1993). Whether TOR controls lifespan in mammals is not known. In addition, while TOR2 in budding and ssion yeast clearly signals in a rapamycininsensitive manner, it is not known whether TOR can also signal in a rapamycin-insensitive manner in mammals. Historically, mTOR-regulated targets or processes have been identied based on their rapamycin sensitivity; thus, it is entirely possible that rapamycininsensitive targets or processes exist but have not yet been identied. Circumstantial data support the theory of rapamycin-insensitive signaling by TOR in mammals: While rapamycin reduces cell size and delays G1 phase progression in mammalian cells expressing wild-type mTOR, cells treated with rapamycin but expressing kinase-dead mTOR exhibit an even greater reduction in size and a more pronounced delay in G1 phase progression (Fingar et al., 2002, 2004). Such observations may simply indicate that rapamycin does not completely inhibit the rapamycin-sensitive function of mTOR, or alternatively, they may indicate that mTOR signals in both a rapamycin-sensitive and -insensitive manner to control cell growth and cell cycle progression. Kinasedead mTOR, by functioning as a dominant-negative allele, may eliminate the rapamycin-insensitive signaling that remains after the rapamycin-sensitive signal is eliminated with rapamycin treatment. Cells with targeted knockout of mTOR or those with complete
3167
knockdown of mTOR with RNAi technology should prove useful in determining whether mammalian TOR signals in a rapamycin-insensitive manner or not. Not only does TOR control cell growth in a cell autonomous manner, but a new publication suggests that dTOR in Drosophila may control cell growth by a non cell autonomous, humoral mechanism (Colombani et al., 2003): mutation of dTOR, mutation of an aminoacid transporter known as slimfast, or overexpression of TSC1/2 in an organ called the fat body affects the growth of other tissues. The authors propose a model whereby the fat body secretes a signal via a TOR- and S6K-dependent mechanism that is sensed by peripheral tissues. Thus, the TOR pathway may regulate growth in response to both local and whole-organism nutrient availability. As S6K1/ mice are perfectly proportioned but small (Shima et al., 1998), it will be important to determine whether the reduced body size is cell autonomous or whether it results from a humoral mechanism, perhaps due to decreased levels of circulating insulin. Whether TOR modulates growth in a non
References Abraham RT and Wiederrecht GJ. (1996). Annu. Rev. Immunol., 14, 483510. Abraham RT. (2002). Cell, 111, 912. Alarcon CM, Heitman J and Cardenas ME. (1999). Mol. Cell. Biol., 10, 25312546. Alessi DR, Kozlowski MT, Weng QP, Morrice N and Avruch J. (1998). Curr. Biol., 8, 6981. Alvarez B, Garrido E, Garcia-Sanz JA and Carrera AC. (2003). J. Biol. Chem., 278, 2646626473. Aoki M, Blazek E and Vogt PK. (2001). Proc. Natl. Acad. Sci. USA, 98, 136141. Araki E, Lipes MA, Patti ME, Bruning JC, Haag III B, Johnson RS and Kahn CR. (1994). Nature, 372, 186190. Barbet NC, Schneider U, Helliwell SB, Stanseld I, Tuite MF and Hall MN. (1996). Mol. Cell. Biol., 7, 2542. Beck T and Hall MN. (1999). Nature, 402, 689692. Beugnet A, Tee AR, Taylor PM and Proud CG. (2003). Biochem. J., 372, 555566. Biondi RM, Kieloch A, Currie RA, Deak M and Alessi DR. (2001). EMBO J., 20, 43804390. Blommaart EF, Luiken JJ, Blommaart PJ, van Woerkom GM and Meijer AJ. (1995). J. Biol. Chem., 270, 23202326. Bohni R, Riesgo-Escovar J, Oldham S, Brogiolo W, Stocker H, Andruss BF, Beckingham K and Hafen E. (1999). Cell, 97, 865875. Braun-Dullaeus RC, Mann MJ, Seay U, Zhang L, von Der Leyen HE, Morris RE and Dzau VJ. (2001). Arterioscler. Thromb. Vasc. Biol., 21, 11521158. Brennan P, Babbage JW, Thomas G and Cantrell D. (1999). Mol. Cell. Biol., 19, 47294738. Brown EJ, Albers MW, Shin TB, Ichikawa K, Keith CT, Lane WS and Schreiber SL. (1994). Nature, 369, 756758. Brown EJ, Beal PA, Keith CT, Chen J, Shin TB and Schreiber SL. (1995). Nature, 377, 441446. Brunn GJ, Hudson CC, Sekulic A, Williams JM, Hosoi H, Houghton PJ, Lawrence Jr JC and Abraham RT. (1997). Science, 277, 99101. Brunn GJ, Williams J, Sabers C, Wiederrecht G, Lawrence Jr JC and Abraham RT. (1996). EMBO J., 15, 52565267.
cell autonomous, humoral manner in mammals awaits future work. While signicant progress has been recently made in understanding TOR function at the biochemical, cellular, and physiological level, we have a great deal more to learn. In the future, it will be important to elucidate how Rheb positively modulates TOR function, how aminoacid sufciency is sensed by TOR, and how rapamycin inhibits TOR-dependent signaling. Moreover, it is imperative to identify the downstream targets and cellular processes that are regulated by TOR. TOR function is clearly vitally important in mammals, as homozygous mTOR inactivation in the mouse is embryonically lethal (Hentges et al., 2001).
Acknowledgements We thank Celeste J Richardson, Kathy A Martin, and Martin Myers Jr for critical reading of the paper. This work was supported by NIH Grant #GM051405. We are also thankful for the support of the Tuberous Sclerosis Alliance and The Rothberg Courage Fund.
Burnett PE, Barrow RK, Cohen NA, Snyder SH and Sabatini DM. (1998). Proc. Natl. Acad. Sci. USA, 95, 14321437. Cafferkey R, Young PR, McLaughlin MM, Bergsma DJ, Koltin Y, Sathe GM, Faucette L, Eng WK, Johnson RK and Livi GP. (1993). Mol. Cell. Biol., 13, 60126023. Cantley LC. (2002). Science, 296, 16551657. Castro AF, Rebhun JF, Clark GJ and Quilliam LA. (2003). J. Biol. Chem., 278, 3249332496. Cheatham L, Monfar M, Chou MM and Blenis J. (1995). Proc. Natl. Acad. Sci. USA, 92, 1169611700. Chen Y, Chen H, Rhoad AE, Warner L, Caggiano TJ, Failli A, Zhang H, Hsiao C-L, Nakanishi K and Molnar-Kimber KL. (1994). Biochem. Biophys. Res. Commun., 203, 17. Chen Y-R, Meyer CF and Tan T-H. (1996). J. Biol. Chem., 271, 631634. Chen J, Zheng XF, Brown EJ and Schreiber SL. (1995). Proc. Natl. Acad. Sci. USA, 92, 49474951. Chen Y, Zheng Y and Foster DA. (2003). Oncogene, 22, 39373942. Chiu MI, Katz H and Berlin V. (1994). Proc. Natl. Acad. Sci., 91, 1257412578. Cho H, Mu J, Kim JK, Thorvaldsen JL, Chu Q, Crenshaw III EB, Kaestner KH, Bartolomei MS, Shulman GI and Birnbaum MJ. (2001a). Science, 292, 17281731. Cho H, Thorvaldsen JL, Chu Q, Feng F and Birnbaum MJ. (2001b). J. Biol. Chem., 276, 3834938352. Choi J, Chen J, Schreiber SL and Clardy J. (1996). Science, 273, 239242. Choi KM, McMahon LP and Lawrence Jr JC. (2003). J. Biol. Chem., 278, 1966719673. Chou MM and Blenis J. (1996). Cell, 85, 573583. Colombani J, Raisin S, Pantalacci S, Radimerski T, Montagne J and Leopold P. (2003). Cell, 114, 739749. Conlon IJ, Dunn GA, Mudge AW and Raff MC. (2001). Nat. Cell. Biol., 3, 918921. Crespo JL and Hall MN. (2002). Microbiol. Mol. Biol. Rev., 66, 579591 table of contents. Dan HC, Sun M, Yang L, Feldman RI, Sui XM, Ou CC, Nellist M, Yeung RS, Halley DJ, Nicosia SV, Pledger WJ and Cheng JQ. (2002). J. Biol. Chem., 277, 3536435370.
Oncogene
3168 Datar SA, Jacobs HW, de la Cruz AF, Lehner CF and Edgar BA. (2000). EMBO J., 19, 45434554. De Benedetti A and Harris AL. (1999). Int. J. Biochem. Cell. Biol., 31, 5972. de Groot RP, Ballou LM and Sasssone-Corsi P. (1994). Cell, 79, 8191. Dennis PB, Jaeschke A, Saitoh M, Fowler B, Kozma SC and Thomas G. (2001). Science, 294, 11021105. Di Como CJ and Arndt KT. (1996). Genes Dev., 10, 19041916. Diehl JA, Cheng M, Roussel MF and Sherr CJ. (1998). Genes Dev., 12, 34993511. Dolinski K, Muir S, Cardenas M and Heitman J. (1997). Proc. Natl. Acad. Sci. USA, 94, 1309313098. Dufner A, Andjelkovic M, Burgering BM, Hemmings BA and Thomas G. (1999). Mol. Cell. Biol., 19, 45254534. Edinger AL and Thompson CB. (2002). Mol. Cell. Biol., 13, 22762288. Elit L. (2002). Curr. Opin. Invest. Drugs., 3, 12491253. Fang Y, Park IH, Wu AL, Du G, Huang P, Frohman MA, Walker SJ, Brown HA and Chen J. (2003). Curr. Biol., 13, 20372044. Fang Y, Vilella-Bach M, Bachmann R, Flanigan A and Chen J. (2001). Science, 294, 19421945. Faridi J, Fawcett J, Wang L and Roth RA. (2003). Am. J. Physiol. Endocrinol. Metab., 285, E96472. Fingar DC, Richardson CJ, Tee AR, Cheatham L, Tsou C and Blenis J. (2004). Mol. Biol. Cell., 24, 200216. Fingar DC, Salama S, Tsou C, Harlow E and Blenis J. (2002). Genes Dev., 16, 14721487. Flick K, Chapman-Shimshoni D, Stuart D, Guaderrama M and Wittenberg C. (1998). Mol. Cell. Biol., 18, 24922501. Gallo R, Padurean A, Jayaraman T, Marx S, Roque M, Adelman S, Chesebro J, Fallon J, Fuster V, Marks A and Badimon JJ. (1999). Circulation, 99, 21642170. Gao X and Pan D. (2001). Genes Dev., 15, 13831392. Gao X, Neufeld TP and Pan D. (2000). Dev. Biol., 221, 404418. Gao X, Zhang Y, Arrazola P, Hino O, Kobayashi T, Yeung RS, Ru B and Pan D. (2002). Nat. Cell. Biol., 4, 699704. Garami A, Zwartkruis FJT, Nobukuni T, Joaquin M, Roccio M, Stocker H, Kozma S, Hafen E, Bos JL and Thomas G. (2003). Mol. Cell, 11, 14571466. Gera JF, Mellinghoff IK, Shi Y, Rettig MB, Tran C, Hsu JH, Sawyers CL and Lichtenstein AK. (2003). J. Biol. Chem., 279, 27372746. Gingras AC, Gygi SP, Raught B, Polakiewicz RD, Abraham RT, Hoekstra MF, Aebersold R and Sonenberg N. (1999a). Genes Dev., 13, 14221437. Gingras AC, Kennedy SG, OLeary MA, Sonenberg N and Hay N. (1998). Genes Dev., 12, 502513. Gingras AC, Raught B and Sonenberg N. (2001b). Genes Dev., 15, 807826. Gingras AC, Raught B, Gygi SP, Niedzwiecka A, Miron M, Burley SK, Polakiewicz RD, Wyslouch-Cieszynska A, Aebersold R and Sonenberg N. (2001a). Genes Dev., 15, 28522864. Goberdhan DC, Paricio N, Goodman EC, Mlodzik M and Wilson C. (1999). Genes Dev., 13, 32443258. Goncharova EA, Goncharov DA, Eszterhas A, Hunter DS, Glassberg MK, Yeung RS, Walker CL, Noonan D, Kwiatkowski DJ, Chou MM, Panettieri Jr RA and Krymskaya VP. (2002). J. Biol. Chem., 277, 3095830967. Gstaiger M, Luke B, Hess D, Oakeley EJ, Wirbelauer C, Blondel M, Vigneron M, Peter M and Krek W. (2003). Science, 302, 12081212.
Oncogene
Hahn WC and Weinberg RA. (2002). N. Engl. J. Med., 347, 15931603. Hanahan D and Weinberg RA. (2000). Cell, 100, 5770. Hannan KM, Brandenburger Y, Jenkins A, Sharkey K, Cavanaugh A, Rothblum L, Moss T, Poortinga G, McArthur GA, Pearson RB and Hannan RD. (2003). Mol. Cell. Biol., 23, 88628877. Hara K, Maruki Y, Long X, Yoshino K, Oshiro N, Hidayat S, Tokunaga C, Avruch J and Yonezawa K. (2002). Cell, 110, 177189. Hara K, Yonezawa K, Weng QP, Kozlowski MT, Belham C and Avruch J. (1998). J. Biol. Chem., 273, 1448414494. Harada H, Andersen JS, Mann M, Terada N and Korsmeyer SJ. (2001). Proc. Natl. Acad. Sci. USA, 98, 96669670. Hardie DG and Hawley SA. (2001). BioEssays, 23, 11121119. Harding MW, Galat A, Uehling DE and Schreiber SL. (1989). Nature, 341, 758760. Hawley SA, Boudeau J, Reid JL, Mustard KJ, Udd L, Makela TP, Alessi DR and Hardie DG. (2003). J. Biol., 2, 28. Haystead TA, Haystead CM, Hu C, Lin TA and Lawrence Jr JC. (1994). J. Biol. Chem., 269, 2318523191. Heesom KJ and Denton RM. (1999). FEBS Lett., 457, 489493. Heitman J, Movva NR and Hall MN. (1991). Science, 253, 905909. Helliwell SB, Wagner P, Kunz J, Deuter-Reinhard M, Henriquez R and Hall MN. (1994). Mol. Cell. Biol., 5, 105118. Hemminki A, Markie D, Tomlinson I, Avizienyte E, Roth S, Loukola A, Bignell G, Warren W, Aminoff M, Hoglund P, Jarvinen H, Kristo P, Pelin K, Ridanpaa M, Salovaara R, Toro T, Bodmer W, Olschwang S, Olsen AS, Stratton MR, de la Chapelle A and Aaltonen LA. (1998). Nature, 391, 184187. Hentges KE, Sirry B, Gingeras AC, Sarbassov D, Sonenberg N, Sabatini D and Peterson AS. (2001). Proc. Natl. Acad. Sci. USA, 98, 1379613801. Herbert TP, Tee AR and Proud CG. (2002). J. Biol. Chem., 277, 1159111596. Hidalgo M and Rowinsky EK. (2000). Oncogene, 19, 66806686. Hong SP, Leiper FC, Woods A, Carling D and Carlson M. (2003). Proc. Natl. Acad. Sci. USA, 100, 88398843. Huang S and Houghton PJ. (2003). Curr. Opin. Pharmacol., 3, 371377. Huang H, Potter CJ, Tao W, Li DM, Brogiolo W, Hafen E, Sun H and Xu T. (1999). Development, 126, 53655372. Hudson CC, Liu M, Chiang GG, Otterness DM, Loomis DC, Kaper F, Giaccia AJ and Abraham RT. (2002). Mol. Cell. Biol., 22, 70047014. Iiboshi Y, Papst PJ, Kawasome H, Hosoi H, Abraham RT, Houghton PJ and Terada N. (1999). J. Biol. Chem., 274, 10921099. Inoki K, Li Y, Xu T and Guan KL. (2003a). Genes Dev., 17, 18291834. Inoki K, Li Y, Zhu T, Wu J and Guan KL. (2002). Nat. Cell. Biol., 4, 648657. Inoki K, Zhu T and Guan KL. (2003b). Cell, 115, 577590. Isotani S, Hara K, Tokunaga C, Inoue H, Avruch J and Yonezawa K. (1999). J. Biol. Chem., 274, 3449334498. Jacinto E, Guo B, Arndt KT, Schmelzle T and Hall MN. (2001). Mol. Cell, 8, 10171026. Jefferies HB, Fumagalli S, Dennis PB, Reinhard C, Pearson RB and Thomas G. (1997). EMBO J., 16, 36933704. Jefferies HBJ, Reinhard C, Kozma SC and Thomas G. (1994). Proc. Natl. Acad. Sci., 91, 44414445.
3169 Jenne DE, Reimann H, Nezu J, Friedel W, Loff S, Jeschke R, Muller O, Back W and Zimmer M. (1998). Nat. Genet., 18, 3843. Jeno P, Ballou LM, Novak-Hofer I and Thomas G. (1988). Proc. Natl. Acad. Sci. USA, 85, 406410. Jiang Y and Broach JR. (1999). EMBO J., 18, 27822792. Johnston GC, Pringle JR and Hartwell LH. (1977). Exp. Cell. Res., 105, 7998. Kawasome H, Papst P, Webb S, Keller GM, Johnson GL, Gelfand EW and Terada N. (1998). Proc. Natl. Acad. Sci. USA, 95, 50335038. Keith CT and Schreiber SL. (1995). Science, 270, 5051. Kenyon C, Chang J, Gensch E, Rudner A and Tabtiang R. (1993). Nature, 366, 461464. Kim JE and Chen J. (2000). Proc. Natl. Acad. Sci. USA, 97, 1434014345. Kim DH, Sarbassov dos D, Ali SM, Latek RR, Guntur KV, Erdjument-Bromage H, Tempst P and Sabatini DM. (2003). Mol. Cell, 11, 895904. Kim DH, Sarbassov DD, Ali SM, King JE, Latek RR, Erdjument-Bromage H, Tempst P and Sabatini DM. (2002). Cell, 110, 163175. Kimura N, Tokunaga C, Dalal S, Richardson C, Yoshino K, Hara K, Kemp BE, Witters LA, Mimura O and Yonezawa K. (2003). Genes Cells, 8, 6579. Kobayashi T, Minowa O, Kuno J, Mitani H, Hino O and Noda T. (1999). Cancer Res., 59, 12061211. Kobayashi T, Minowa O, Sugitami Y, Takai S, Mitani H, Kobayashi E, Noda T and Hino O. (2001). PNAS, 98, 87628767. Koltin Y, Faucette L, Bergsma DJ, Levy MA, Cafferkey R, Koser PL, Johnson RK and Livi GP. (1991). Mol. Cell. Biol., 11, 17181723. Kozma SC and Thomas G. (2002). BioEssays, 24, 6571. Kunz J, Henriquez R, Schneider U, Deuter-Reinhard M, Movva NR and Hall MN. (1993). Cell, 73, 585596. Kwiatkowski DJ. (2003). Ann. Hum. Genet., 67, 8796. Lawlor MA, Mora A, Ashby PR, Williams MR, Murray-Tait V, Malone L, Prescott AR, Lucocq JM and Alessi DR. (2002). EMBO J., 21, 37283738. Lazaris-Karatzas A, Montine KS and Sonenberg N. (1990). Nature, 345, 544547. Leevers SJ, Weinkove D, MacDougall LK, Hafen E and Watereld MD. (1996). EMBO J., 15, 65846594. Li S, Sonenberg N, Gingras AC, Peterson M, Avdulov S, Polunovsky VA and Bitterman PB. (2002). Mol. Cell. Biol., 22, 28532861. Lin TA, Kong X, Haystead TA, Pause A, Belsham G, Sonenberg N and Lawrence Jr JC. (1994). Science, 266, 653656. Loewith R, Jacinto E, Wullschleger S, Lorberg A, Crespo JL, Bonenfant D, Oppliger W, Jenoe P and Hall MN. (2002). Mol. Cell, 10, 457468. Long X, Spycher C, Han ZS, Rose AM, Muller F and Avruch J. (2002). Curr. Biol., 12, 14481461. Luo J, Manning BD and Cantley LC. (2003). Cancer Cell, 4, 257262. Ly C, Arechiga AF, Melo JV, Walsh CM and Ong ST. (2003). Cancer Res., 63, 57165722. Mach KE, Furge KA and Albright CF. (2000). Genetics, 155, 611622. Mahalingam M and Templeton DJ. (1996). Mol. Cell. Biol., 16, 405413. Manning BD and Cantley LC. (2003a). Trends Biochem. Sci., 28, 573576. Manning BD and Cantley LC. (2003b). Biochem. Soc. Trans., 31, 573578. Manning BD, Tee AR, Logsdon MN, Blenis J and Cantley LC. (2002). Mol. Cell, 10, 151162. Marcotrigiano J, Gingras AC, Sonenberg N and Burley SK. (1999). Mol. Cell, 3, 707716. Marks AR. (2003). N. Engl. J. Med., 349, 13071309. Martin KA and Blenis J. (2002). Adv. Cancer Res., 86, 139. Martin KA, Rzucidlo EM, Merenick BL, Fingar DC, Brown DJ, Wagner RJ and Powell RJ. (2003). Am. J. Physiol. Cell. Physiol., 286, C507C517. Matsumoto S, Bandyopadhyay A, Kwiatkowski DJ, Maitra U and Matsumoto T. (2002). Genetics, 161, 10531063. Meyer CA, Jacobs HW, Datar SA, Du W, Edgar BA and Lehner CF. (2000). EMBO J., 19, 45334542. Miron M, Verdu J, Lachance PE, Birnbaum MJ, Lasko PF and Sonenberg N. (2001). Nat. Cell. Biol., 3, 596601. Mohi GM, Boulton C, Gu TL, Sternberg DW, Neuberg D, Grifn JD, Gilliland DG and Neel BG. (2004). Proc. Natl. Acad. Sci. USA, 101, 31303135. Montagne J, Stewart MJ, Stocker H, Hafen E, Kozma SC and Thomas G. (1999). Science, 285, 21262129. Mora A, Davies AM, Bertrand L, Sharif I, Budas GR, Jovanovic S, Mouton V, Kahn CR, Lucocq JM, Gray GA, Jovanovic A and Alessi DR. (2003). EMBO J., 22, 4666 4676. Morice MC, Serruys PW, Sousa JE, Fajadet J, Ban Hayashi E, Perin M, Colombo A, Schuler G, Barragan P, Guagliumi G, Molnar F and Falotico R. (2002). N. Engl. J. Med., 346, 17731780. Moser BA, Dennis PB, Pullen N, Pearson RB, Williamson NA, Wettenhall RE, Kozma SC and Thomas G. (1997). Mol. Cell. Biol., 17, 56485655. Moses JW, Leon MB, Popma JJ, Fitzgerald PJ, Holmes DR, OShaughnessy C, Caputo RP, Kereiakes DJ, Williams DO, Teirstein PS, Jaeger JL and Kuntz RE. (2003). N. Engl. J. Med., 349, 13151323. Mothe-Satney I, Brunn GJ, McMahon LP, Capaldo CT, Abraham RT and Lawrence Jr JC. (2000a). J. Biol. Chem., 275, 3383633843. Mothe-Satney I, Yang D, Fadden P, Haystead TA and Lawrence Jr JC. (2000b). Mol. Cell. Biol., 20, 35583567. Murata K, Wu J and Brautigan DL. (1997). Proc. Natl. Acad. Sci. USA, 94, 1062410629. Nanahoshi M, Nishiuma T, Tsujishita Y, Hara K, Inui S, Sakaguchi N and Yonezawa K. (1998). Biochem. Biophys. Res. Commun., 251, 520526. Nave BT, Ouwens M, Withers DJ, Alessi DR and Shepherd PR. (1999). Biochem. J., 344 (Part 2), 427431. Neshat MS, Mellinghoff IK, Tran C, Stiles B, Thomas G, Petersen R, Frost P, Gibbons JJ, Wu H and Sawyers CL. (2001). Proc. Natl. Acad. Sci. USA, 98, 1031410319. Neufeld TP, de la Cruz AF, Johnston LA and Edgar BA. (1998). Cell, 93, 11831193. Nojima H, Tokunaga C, Eguchi S, Oshiro N, Hidayat S, Yoshino K, Hara K, Tanaka N, Avruch J and Yonezawa K. (2003). J. Biol. Chem., 278, 1546115464. Oldham S and Hafen E. (2003). Trends Cell. Biol., 13, 7985. Oldham S, Montagne J, Radimerski T, Thomas G and Hafen E. (2000). Genes Dev., 14, 26892694. Onda H, Lueck A, Morks PW, Warren HB and Kwiatkowski DJ. (1999). J. Clin. invest., 104, 687695. Owens GK. (1995). Physiol. Rev., 75, 487517. Pearson RB, Dennis PB, Han JW, Williamson NA, Kozma SC, Wettenhall RE and Thomas G. (1995). EMBO J., 14, 52795287.
Oncogene
3170 Pende M, Kozma SC, Jaquet M, Oorschot V, Burcelin R, Le Marchand-Brustel Y, Klumperman J, Thorens B and Thomas G. (2000). Nature, 408, 994997. Peng T, Golub TR and Sabatini DM. (2002). Mol. Cell. Biol., 22, 55755584. Perry J and Kleckner N. (2003). Cell, 112, 151155. Peterson RT, Beal PA, Comb MJ and Schreiber SL. (2000). J. Biol. Chem., 275, 74167423. Peterson RT, Desai BN, Hardwick JS and Schreiber SL. (1999). Proc. Natl. Acad. Sci. USA, 96, 44384442. Podsypanina K, Lee RT, Politis C, Hennessy I, Crane A, Puc J, Neshat M, Wang H, Yang L, Gibbons J, Frost P, Dreisbach V, Blenis J, Gaciong Z, Fisher P, Sawyers C, Hedrick-Ellenson L and Parsons R. (2001). Proc. Natl. Acad. Sci. USA, 98, 1032010325. Polymenis M and Schmidt EV. (1997). Genes Dev., 11, 25222531. Poon M, Marx SO, Gallo R, Badimon JJ, Taubman MB and Marks AR. (1996). J. Clin. Invest., 98, 22772283. Potter CJ, Huang H and Xu T. (2001). Cell, 105, 357368. Potter CJ, Pedraza LG and Xu T. (2002). Nat. Cell. Biol., 4, 658665. Pullen N, Dennis PB, Andjelkovic M, Dufner A, Kozma SC, Hemmings BA and Thomas G. (1998). Science, 279, 707710. Radimerski T, Montagne J, Hemmings-Mieszczak M and Thomas G. (2002). Genes Dev., 16, 26272632. Rajasekhar VK, Viale A, Socci ND, Wiedmann M, Hu X and Holland EC. (2003). Mol. Cell, 12, 889901. Rameh LE and Cantley LC. (1999). J. Biol. Chem., 274, 83478350. Richardson CR, Broenstrup M, Fingar DC, Juelich K, Gygi S and Blenis J Submitted. Rintelen F, Stocker H, Thomas G and Hafen E. (2001). Proc. Natl. Acad. Sci. USA, 98, 1502015025. Rohde J, Heitman J and Cardenas ME. (2001). J. Biol. Chem., 276, 95839586. Romanelli A, Dreisbach VC and Blenis J. (2002). J. Biol. Chem., 277, 4028140289. Romanelli A, Martin KA, Toker A and Blenis J. (1999). Mol. Cell. Biol., 19, 29212928. Rommel C, Bodine SC, Clarke BA, Rossman R, Nunez L, Stitt TN, Yancopoulos GD and Glass DJ. (2001). Nat. Cell Biol., 3, 10091013. Roque M, Cordon-Cardo C, Fuster V, Reis ED, Drobnjak M and Badimon JJ. (2000). Atherosclerosis, 153, 315322. Rosner M, Hofer K, Kubista M and Hengstschlager M. (2003). Oncogene, 22, 47864798. Rossig L, Jadidi AS, Urbich C, Badorff C, Zeiher AM and Dimmeler S. (2001). Mol. Cell. Biol., 21, 56445657. Ruggero D and Pandol PP. (2003). Nat. Rev. Cancer, 3, 179192. Sabatini DM, Barrow RK, Blackshaw S, Burnett PE, Lai MM, Field ME, Bahr BA, Kirsch J, Betz H and Snyder SH. (1999). Science, 284, 11611164. Sabatini DM, Pierchala BA, Barrow RK, Schell MJ and Snyder SH. (1995). J. Biol. Chem., 270, 2087520878. Sabers CJ, Martin MM, Brunn GJ, Williams JM, Dumont FJ, Wiederrecht G and Abraham RT. (1995). J. Biol. Chem., 270, 815822. Saltiel AR and Kahn CR. (2001). Nature, 414, 799806. Saucedo LJ, Gao X, Chiarelli DA, Li L, Pan D and Edgar BA. (2003). Nat. Cell Biol., 5, 566571. Schalm SS and Blenis J. (2002). Curr. Biol., 12, 632639. Schalm SS, Fingar DC, Sabatini DM and Blenis J. (2003). Curr. Biol., 13, 797806.
Oncogene
Schmelzle T and Hall MN. (2000). Cell., 103, 253262. Schmidt A, Kunz J and Hall MN. (1996). Proc. Natl. Acad. Sci. USA., 93, 1378013785. Schreiber SL. (1991). Science, 251, 283287. Sehgal SN. (2003). Transplant Proc., 35, 7S14S. Sekulic A, Hudson CC, Homme JL, Yin P, Otterness DM, Karnitz LM and Abraham RT. (2000). Cancer Res., 60, 35043513. Shamji AF, Nghiem P and Schreiber SL. (2003). Mol. Cell., 12, 271280. Shaw R, Kosmatka M, Bardeesy N, Hurley RL, Witters LA, DePinho RA and Cantley LC.. (2004). Proc. Natl. Acad. Sci. USA., 101, 33293335. Shigemitsu K, Tsujishita Y, Hara K, Nanahoshi M, Avruch J and Yonezawa K. (1999). J. Biol. Chem., 274, 10581065. Shima H, Pende M, Chen Y, Fumagalli S, Thomas G and Kozma SC. (1998). Embo J., 17, 66496659. Shioi T, Kang PM, Douglas PS, Hampe J, Yballe CM, Lawitts J, Cantley LC and Izumo S. (2000). Embo J., 19, 25372548. Shioi T, McMullen JR, Kang PM, Douglas PS, Obata T, Franke TF, Cantley LC and Izumo S. (2002). Mol. Cell. Biol., 22, 27992809. Shiojima I, Yefremashvili M, Luo Z, Kureishi Y, Takahashi A, Tao J, Rosenzweig A, Kahn CR, Abel ED and Walsh K. (2002). J. Biol. Chem., 277, 3767037677. Siekierka JJ, Hung SH, Poe M, Lin CS and Sigal NH. (1989). Nature, 341, 755757. Siekierka JJ, Wiederrecht G, Greulich H, Boulton D, Hung SH, Cryan J, Hodges PJ and Sigal NH. (1990). J. Biol. Chem., 265, 2101121015. Sonenberg N and Gingras AC. (1998). Curr. Opin. Cell. Biol., 10, 268275. Stan R, McLaughlin MM, Cafferkey R, Johnson RK, Rosenberg M and Livi GP. (1994). J. Biol. Chem., 269, 3202732030. Stocker H and Hafen E. (2000). Curr. Opin. Genet. Dev., 10, 529535. Stocker H, Radimerski T, Schindelholz B, Wittwer F, Belawat P, Daram P, Breuer S, Thomas G and Hafen E. (2003). Nat. Cell. Biol., 5, 559565. Stolovich M, Tang H, Hornstein E, Levy G, Cohen R, Bae SS, Birnbaum MJ and Meyuhas O. (2002). Mol. Cell. Biol., 22, 81018113. Takata M, Ogawa W, Kitamura T, Hino Y, Kuroda S, Kotani K, Klip A, Gingras AC, Sonenberg N and Kasuga M. (1999). J. Biol. Chem., 274, 2061120618. Tang H, Hornstein E, Stolovich M, Levy G, Livingstone M, Templeton D, Avruch J and Meyuhas O. (2001). Mol. Cell. Biol., 21, 86718683. Tapon N, Ito N, Dickson BJ, Treisman JE and Hariharan IK. (2001a). Cell., 105, 345355. Tapon N, Moberg KH and Hariharan IK. (2001b). Curr. Opin. Cell. Biol., 13, 731737. Tee AR, Anjum R and Blenis J. (2003a). J. Biol. Chem., 278, 3728837296. Tee AR, Fingar DC, Manning BD, Kwiatkowski DJ, Cantley LC and Blenis J. (2002). Proc. Natl. Acad. Sci. USA., 99, 1357113576. Tee AR, Manning BD, Roux PP, Cantley LC and Blenis J. (2003b). Curr. Biol., 13, 12591268. Terada N, Patel HR, Takase K, Kohno K, Nairn AC and Gelfand EW. (1994). PNAS., 91, 1147711481. Urano J, Tabancay AP, Yang W and Tamanoi F. (2000). J. Biol. Chem., 275, 1119811206. Vellai T, Takacs-Vellai K, Zhang Y, Kovacs AL, Orosz L and Muller F. (2003). Nature, 426, 620.
3171 Verdu J, Buratovich MA, Wilder EL and Birnbaum MJ. (1999). Nat. Cell. Biol., 1, 500506. Vilella-Bach M, Nuzzi P, Fang Y and Chen J. (1999). J. Biol. Chem., 274, 42664272. Vinals F, Chambard JC and Pouyssegur J. (1999). J. Biol. Chem., 274, 2677626782. Vivanco I and Sawyers CL. (2002). Nat. Rev. Cancer, 2, 489 501. Vogt PK. (2001). Trends Mol. Med., 7, 482484. von Manteuffel SR, Dennis PB, Pullen N, Gingras AC, Sonenberg N and Thomas G. (1997). Mol. Cell. Biol., 17, 54265436. von Manteuffel SR, Gingras AC, Ming XF, Sonenberg N and Thomas G. (1996). Proc. Natl. Acad. Sci. USA., 93, 40764080. Wang X, Li W, Williams M, Terada N, Alessi DR and Proud CG. (2001). Embo J., 20, 43704379. Weigmann K, Cohen SM and Lehner CF. (1997). Development, 124, 35553563. Weinkove D and Leevers SJ. (2000). Curr. Opin. Genet. Dev., 10, 7580. Weng Q-p, Andrabi K, Kozlowski MT, Grove JR and Avruch J. (1995). M.C.B., 15, 23332340. Weng QP, Kozlowski M, Belham C, Zhang A, Comb MJ and Avruch J. (1998). J. Biol. Chem., 273, 1662116629. Westphal RS, Coffee RL, Jr, Marotta A, Pelech SL and Wadzinski BE. (1999). J. Biol. Chem., 274, 687692. Wiederrecht G, Brizuela L, Elliston K, Sigal NH and Siekierka JJ. (1991). Proc. Natl. Acad. Sci. USA., 88, 10291033. Williams MR, Arthur JS, Balendran A, van der Kaay J, Poli V, Cohen P and Alessi DR. (2000). Curr. Biol., 10, 439448. Wilson KF, Wu WJ and Cerione RA. (2000). J. Biol. Chem., 275, 3730737310. Withers DJ, Gutierrez JS, Towery H, Burks DJ, Ren JM, Previs S, Zhang Y, Bernal D, Pons S, Shulman GI, Bonner-Weir S and White MF. (1998). Nature, 391, 900904. Woods A, Johnstone SR, Dickerson K, Leiper FC, Fryer LG, Neumann D, Schlattner U, Wallimann T, Carlson M and Carling D. (2003). Curr. Biol., 13, 20042008. Yokogami K, Wakisaka S, Avruch J and Reeves SA. (2000). Curr. Biol., 10, 4750. Zhang H, Stallock JP, Ng JC, Reinhard C and Neufeld TP. (2000). Genes. Dev., 14, 27122724. Zhang Y, Gao X, Saucedo LJ, Ru B, Edgar BA and Pan D. (2003). Nat .Cell. Biol., 5, 578581. Zhou BP, Liao Y, Xia W, Spohn B, Lee MH and Hung MC. (2001). Nat. Cell. Biol., 3, 245252.
Oncogene

BlenisFingamTORreview Copy 1

Încărcat de

Informații document

Drepturi de autor

Formate disponibile

Partajați acest document

Partajați sau inserați document

Opțiuni de partajare

Vi se pare util acest document?

Este necorespunzător acest conținut?

Drepturi de autor:

Formate disponibile

BlenisFingamTORreview Copy 1

Încărcat de

Drepturi de autor:

Formate disponibile

Oncogene (2004) 23, 31513171

S-ar putea să vă placă și