REVIEW
When translation meets transformation: the mTOR story
J Averous
1and CG Proud
21Unite´ de Nutrition Humaine, INRA de Theix, Saint Gene`s Champanelle, France and2Department of Biochemistry & Molecular Biology, University of British Columbia, Life Sciences Centre, Vancouver, BC, Canada
There is currently a high level of interest in signalling through the mammalian target of rapamycin (mTOR). This reflects both its key role in many cell functions and its involvement in disease states such as cancers. The best understood targets for mTOR signalling are proteins involved in controlling the translational machinery, including the ribosomal protein S6 kinases and proteins that regulate the initiation and elongation phases of translation. Indeed, there is compelling evidence that at least one of these targets of mTOR (eukaryotic initiation factor eIF4E) plays a key role in tumorigenesis. It is regulated through the mTOR-dependent phosphorylation of inhibitory proteins such as eIF4E-binding protein 1. Thus, targeting mTOR signalling may be an effective anticancer strategy, in at least a significant subset of tumours. Not all effects of mTOR are sensitive to the classical anti-mTOR drug rapamycin, and this compound also interferes with other processes besides eIF4E function. Developing new approaches to targeting mTOR for cancer therapy requires more detailed knowledge of signalling downstream of mTOR. Such advances are likely to come from further work to understand the regulation of mTOR targets such as components of the translational apparatus.
Oncogene(2006) 25, 6423–6435. doi:10.1038/sj.onc.1209887 Keywords: apoptosis; mRNA translation; protein kinase; initiation; elongation; rapamycin
The signalling pathway involving the mammalian target of rapamycin (mTOR) is now the centre of substantial attention due to its key roles in regulating cell function and its links with human diseases, including certain types of cancer. The best-understood roles of mTOR in mammalian cells are related to the control of mRNA translation. The purpose of this article is, firstly, to review current knowledge about signalling downstream of mTOR to the translational machinery and, secondly, to described the links between mTOR signalling and the control of the eukaryotic translation initiation factor eIF4E. eIF4E plays important roles in mRNA translation and probably in mRNA metabolism, and is
strongly implicated in diverse human cancers. Rapamy-cin, which inhibits certain mTOR functions, inhibits the proliferation of certain tumour cells and can cooperate with other agents to induce apoptosis. We will also highlight related areas that are poorly understood and require further study.
Control of mTOR
As described in detail in the article by (Wullschleger et al., 2006), mTOR is a multidomain protein that forms complexes with other proteins (Figure 1a and b). Some of these partners appear to regulate the activity of mTOR (e.g., the small G-protein Rheb) while others, such as raptor and rictor, are involved in mediating downstream signalling. The interactions of raptor and rictor with mTOR seem to be mutually exclusive. The complexes in which they participate are termed mTOR-complex 1 (mTORC1) and mTORC2, respectively (Sarbassov et al., 2005) (Figure 1b). Raptor binds to short (five residue) motifs in certain targets of mTOR signalling whose control is sensitive to rapamycin, such as the ribosomal protein S6 (p70) kinases and eIF4E-binding proteins (4E-BPs). These motifs are termed TOR-signalling or TOS motifs (Schalm and Blenis, 2002). mTORC1 mediates effects of mTOR that show the ‘classical’ signature of rapamycin-sensitivity, while the known effects of mTORC2 are insensitive to this drug. Thus, by no means all of the biological functions of effects of mTOR are blocked by rapamycin.
The basis of the rapamycin-sensitivity of mTORC1-mediated effects is unclear – rapamycin does not, for example, affect the overall integrity of mTORC1, although it may alter the stability of the interactions between some of the partner proteins (Jacinto et al., 2004; Sarbassov et al., 2006). The latter paper also showed that long-term treatment of certain types of cells results in the loss of mTORC2 complexes, in addition to the short-term inhibitory effects of rapamycin on mTORC1. It is suggested that this effect on mTORC2 reflects the ability of the rapamycin-FKBP12 complex to prevent the binding of rictor to newly-synthesized mTOR. It is also unclear why short-term control by mTORC2 of protein kinase B (PKB, also termed Akt) and the cytoskeleton are insensitive to rapamycin. More work is required, for example, to understand how rapamycin interferes with mTORC1 signalling and
Correspondence: Dr CG Proud, Department of Biochemistry & Molecular Biology, University of British Columbia, Life Sciences Centre, 2350 Health Sciences Mall, Vancouver, BC, Canada V6T 1Z3. E-mail: [email protected]
how the mTORC2 complex functions to regulate its targets. By analogy with mTORC1 and raptor, rictor may also bind specific features of its targets, but so far there is no information on this. Furthermore, there is at least one effect of mTOR that is not sensitive to rapamycin, but appears not to be mediated by mTORC2, that is, the phosphorylation of certain regulatory sites in 4E-BP1/2 (Wang et al., 2005). This hints at the existence of further complexity in signalling downstream of mTOR.
The upstream control of mTOR is described else-where in this issue of Oncogene. mTOR (or at least mTORC1) is regulated by signalling pathways linked to several oncoproteins or tumour suppressors including phosphatidyl inositol (PI) 3-kinase, the lipid phospha-tase PTEN, the protein kinase LKB1, Ras and Raf. These effects involve signalling through PKB, the AMP-activated protein kinase (AMPK) or the ‘classical’ MAP kinase ERK (Guertin and Sabatini, 2005; Ma et al., 2005) (see article by Corradetti and Guan, in this volume). mTOR signalling is activated in cells with loss-of-function mutations in, for example, TSC1/2 or PTEN. As assessed by the phosphorylation of 4E-BP1, it is also hyperactivated in Src-transformed cells (Tuhackova et al., 1999). Several aspects of the links between mTOR signalling and cancer have recently been reviewed (Guertin and Sabatini, 2005; Wullschleger et al., 2006).
mTOR provides a link between the availability of amino acids and the the regulation of translation. The
function of mTORC1 requires a supply of amino acids, with the essential branched-chain amino acid leucine being most effective (Kimball and Jefferson, 2006). It is not yet clear how leucine activates mTOR, but it does so by acting within the cell, and can do so by mechanisms that are independent of the TSC1/2 complex that mediates the effects of anabolic and mitogenic signals (Byfield et al., 2005; Nobukuni et al., 2005; Smith et al., 2005). Recent evidence suggests a role for the PI 3-kinase-related kinase vps34 in the control of mTOR by amino acids (Byfield et al., 2005; Nobukuni et al., 2005). vps34 functions in the regulation of macroauto-phagy, a process that is negatively controlled by mTOR (Klionsky and Emr, 2000). Since autophagy likely supplies intracellular amino acids to maintain mTOR signalling, there is some logic in connecting these processes, although the mechanisms by which vps34 may link to mTOR signalling and to the control of autophagy remain to be established. The role of the TSC1/2 complex in controlling mTOR is described in detail elsewhere in this volume by Corradetti and Guan. To understand how inhibition of mTOR by rapamy-cin may exert antiproliferative effects, it is crucial to consider the target proteins and processes that are known to be controlled through mTOR in a rapamycin-sensitive manner. There is now substantial evidence that dysregulation of the translational machinery can contribute to cell transformation (for a reviews see, e.g., Bjornsti and Houghton, 2004; Clemens, 2004; Mamane et al., 2004; Guertin and Sabatini, 2005).
Signalling downstream of mTOR The S6 kinases
Ribosomal protein S6 has been known for more than 30 years to be a phosphoprotein (Gressner and Wool, 1974): at least five sites of phosphorylation exist in its C-terminus. In the following years, two classes of protein kinases were identified as being able to phosphorylate S6 in vitro, the 70 and 90 kDa S6 kinases, which are now known as the p70 S6 kinases (S6Ks) and the p90 ribosomal protein S6 kinases, or RSKs (Avruch et al., 2001; Roux and Blenis, 2004). The subsequent discovery that the phosphorylation of S6 is sensitive to, and in many cases completely blocked by, rapamycin pointed to the former enzymes as the main physiological S6 kinases, as their activation is mediated by mTOR (Avruch et al., 2001). The RSKs, in contrast, are not affected by rapamycin as they are activated via the classical MAP kinase (ERK) pathway (Roux and Blenis, 2004). Recent data from S6 kinase knockout mice have revealed an input from ERK signalling to S6 phospho-rylation, at least in response to certain stimuli (Pende et al., 2004). Thus, RSKs or other kinases downstream of ERK may contribute to S6 phosphorylation under specific conditions.
Two p70S6 kinase genes exist in mouse and man (S6K1 and S6K2), each of which can give rise to two distinct proteins due to alternative splicing of the mRNAs (Avruch et al., 2001; Figure 2a). The longer
HEAT domains FRB FAT FATC kinase mTOR GβL PKCα? mTORC2 mTOR raptor GβL 4E-BPs S6K’s Via TOS motif
mTORC1 Others? Rheb.GTP TSC1 TSC2 PKB/Akt Erk/p90 RSK AMKP
Growth factors, etc. Growth factors, insulin, etc.
Amino acids ? LKB1 Ras/Raf a b PTEN PI 3-kinase PDK1 rictor
Figure 1 Summary of mTOR signalling, showing (a) the basic features of mTOR and (b) its upstream control, and its complexes mTORC1 and mTORC2. In (b), components in red boxes are tumour suppressors, while those in green boxes are proto-oncogenic. The function of GbL is unclear, but it is found in mTORC1 and mTORC2. As indicated, mTORC2 effectively acts upstream of mTORC1, by phosphorylating and activating PKB, a positive regulator of mTOR signalling. Phosphorylation of TSC2 by PKB and Erk/p90RSKappears to inhibit its function as a Rheb.
GTPase-activator protein (GAP) while AMPK-mediated phos-phorylation is thought to activate this function.
version of each contains, in its N-terminal region (which is not present in the shorter version), a potential nuclear-localization signal. S6K2 also possesses a functional nuclear localization signal in the C-terminus (Koh et al., 1999).
The S6 kinases are activated by phosphorylation at multiple sites (Figure 2). Several lie within the C-terminus, while two others lie immediately C-terminal to the catalytic domain. One of these, Thr389in the shorter form of S6K1, is directly phosphorylated by mTOR as part of the mTORC1 complex (Alessi et al., 1998; Pullen et al., 1998). Phosphorylation here is required for the phosphorylation of S6K1 by PDK1 (phosphoinositide-dependent kinase 1) at a threonine in the activation loop of the catalytic domain, Thr229. Phosphorylation at Thr229allows complete activation of S6K1 (Avruch et al., 2001). S6K2 is likely regulated in a similar manner. The N-termini of S6K1 and 2 each contain a TOS motif that interacts with the mTOR partner, raptor, to facilitate phosphorylation of S6Ks by mTORC1 (Schalm and Blenis, 2002). The phosphoryla-tion of the sites in the C-terminal phosphoregulatory region is believed to allow access to the sites (Thr389/ Ser229) whose phosphorylation leads to full activation (Avruch et al., 2001). It is not clear how the phosphorylation of the C-terminal sites is brought about or which kinase catalyses this: mTORC1 prima-rily phosphorylates Thr389, which as discussed is controlled by the C-terminal phosphorylation sites, which are sensitive to rapamycin. This implies that mTOR also, probably indirectly, brings about the phosphorylation of the C-terminal sites. A motif RSPRR present in this region appears to plays a crucial role in the inhibitory effect of the C-terminal region. It is proposed that a negative regulator of S6K1 binds to this motif and that mTOR is able to disrupt its binding (Schalm et al., 2005). The C-terminal region of S6K1 also governs whether S6K1 can be phosphorylated by mTORC2. A deletion mutant lacking this region is a substrate for mTORC2 (Ali and Sabatini, 2005).
The physiological role of S6 phosphorylation is still quite unclear. Recent studies used ‘knock-in’ mice in
which all five of the serines that are phosphorylated by the S6 kinases were mutated to alanine. This work indicated a role for S6 phosphorylation in cell growth (Ruvinsky et al., 2005). Despite this decrease in cell size, the knock-in cells actually showed faster rates of protein synthesis. This could reflect increased access of the S6 kinases to other substrates involved in translation, such as eIF4B and eEF2 kinase (Figure 2). The phosphoryla-tion of these proteins by S6K1 is likely to lead to more efficient initiation or elongation, respectively (Wang et al., 2001; Raught et al., 2004). An earlier idea that the role of the S6 kinases, and probably S6 phosphoryla-tion, was to control the translation of the mRNAs for ribosomal proteins (so-called 50-TOP mRNAs) was not substantiated by later data from the S6K knockout mice or from the animals in which the phosphorylation sites in S6 were converted to alanines (Pende et al., 2004; Ruvinsky et al., 2005).
In many investigations that focused on the role of mTOR in transformation, the levels of phosphorylation of S6K1 and S6 have been studied as a read-out for mTOR activity. This reflects the fact that phosphoryla-tion of S6K and S6 is very sensitive to rapamycin treatment. The antiproliferative/apoptotic effects of rapamycin were often attributed to its effect on S6K and the presumed role of S6K in controlling the translational machinery, although there is little evidence that S6 phosphorylation/the S6Ks play a major role in promoting protein synthesis. However, it is clear that S6K activity is enhanced in several cancer cell lines or tumours. In numerous cases, this is due to the increase of mTOR activity that occurs when the tumour suppressors LKB1, PTEN or TSC are mutated with loss of function. It has also been observed that the level of S6K1 could be also increased. It is established that the amplification of chromosomal locus 17q23, which contains the S6K1 gene, is a feature of numerous cancers such as meningioma (Surace et al., 2004) or breast cancer (Barlund et al., 2000). This suggests that S6K1 may participate in carcinogenesis independently of other targets of mTOR, but this idea requires further investigation.
Few studies have addressed the role of S6K in the processes leading to carcinogenesis. Single and double knockout (S6K1/, S6K2/ and S6K1/ S6K2/)
mice have been generated. The major observations are that the S6K1/ animals exhibit a smaller size and
reduced growth and the S6K1/S6K2/, which exhibit
a pronounced perinatal lethality (Pende et al., 2004). However, there are no data concerning tumorigenesis in such cells or animals. Addressing the role of S6Ks in transformation presents a number of technical difficul-ties. For example, (i) there is no specific inhibitor of S6K activity; (ii) the elimination of S6K1 could potentially be rescued by S6K2; and (iii) the elimination of both S6K1 and S6K2 may profoundly alter the physiology of the cell. The overexpression of a rapamycin-resistant form of S6K1 increased cell size (Fingar et al., 2002) and the proportion of cells in S phase in the presence of serum (Fingar et al., 2004). Conversely, depletion of S6K1 by RNAi decreases the percentage of cell in S phase Catalytic domain NLS (only in long splice variants) P P P P P P P PDK1 mTOR Phosphoregulatory region P Upstream kinases? rpS6 eEF2 kinase SKAR eIF4B IRS1 mTOR CREMτ BAD
Figure 2 Summary of the 70 kDa S6 kinases (based on S6K1), showing the main features of their structures and control (e.g., phosphorylation sites) as well as known substrates. As indicated, phosphorylation at the mTOR site (Thr389in the short form of S6K1) allows PDK1 to phosphorylate and fully turn on the S6 kinases. NLS¼ nuclear localisation signal; the black box denotes the kinase catalytic domain. Other phosphorylation sites are mentioned in the text.
(Fingar et al., 2004). This effect on cell cycle could be the consequence of a direct regulation of the activity of one or more cell cycle regulators. This effect could also be an indirect consequence of the regulation of cell growth by S6K1 via mechanisms that are so far obscure. More recently, it has been demonstrated that the overexpression of a rapamycin-resistant S6K1 mutant in ovarian cancer cell line increased the expression of metalloproteinase 9, which favours the invasiveness of the cells (Zhou and Wong, 2006). S6K1 has recently been shown to phosphorylate eIF4B (Raught et al., 2004) and facilitate its recruitment into initiation complexes (Holz et al., 2005), and this could play an important role in the regulation of the translation by mTOR.
S6K1 has also been shown to be responsible for a negative feedback control of IRS1 and 2. This point should be taken carefully into consideration. For example, in S6K1/ mice, it appears that
insulin-sensitivity is increased and this apparently decreases the onset of obesity and diabetes with age (Pende et al., 2000). So it appears that in cancer cells where S6K1 is specifically inhibited, IRS 1 and 2 could be sensitized to insulin (and other stimuli) or even constitutively activated. This effect on the PI3K/PKB axis could actually improve the survival of cancer cell lines by inhibiting apoptotic pathway components such as FOXO. Thus, inhibiting S6K could actually promote antiapoptotic signalling. Finally, the evidence that S6K1 is a good target in cancer therapy is not so clear (for a review see Manning, 2004). Although additional functions have been ascribed to S6K1 (see, e.g., Holz et al., 2005; for a review see Ruvinsky and Meyuhas, 2006), it is not clear how they could contribute to transformation.
Regulation of eEF2 and eEF2 kinase
Eukaryotic elongation factor 2 (eEF2) mediates the translocation step of translation elongation in which the ribosome moves relative to the mRNA by one codon, and the peptidyl-tRNA shifts from the A- to the P-site (the spent tRNA moving to the exit site). eEF2 binds GTP and this nucleotide is hydrolysed during the translocation process. eEF2 undergoes phosphorylation at Thr56 within the GTP-binding domain and this modification interferes with its ability to bind the ribosome, thus inhibiting its function (Carlberg et al., 1990; Browne and Proud, 2002).
The kinase acting on eEF2 is eEF2 kinase, a calcium/ calmodulin-dependent enzyme that has a very unusual catalytic domain, unrelated in sequence to almost all other protein kinases (Ryazanov et al., 1997) (Figure 3). Insulin and other agents that activate protein synthesis cause the rapid dephosphorylation of eEF2. The effect of insulin is accompanied by accelerated rates of elongation, and by decreased activity of eEF2 kinase: all three effects are blocked by rapamycin (Redpath et al., 1996). mTOR thus negatively regulates eEF2 kinase and thereby activates eEF2.
eEF2 kinase is a phosphoprotein. At least three phosphorylation sites are regulated by mTOR, as judged by their sensitivities to rapamycin and/or amino-acid starvation (which lead to their dephosphorylation) (Browne and Proud, 2002; Browne and Proud, 2004; Figure 3). However, eEF2 kinase does not possess a TOS motif, does not bind to raptor, and is not a direct substrate for phosphorylation by mTOR in vitro (Smith and Proud, unpublished). Ser366 (in human eEF2 kinase) is phosphorylated by S6K1 (and by RSKs), and phosphorylation here causes the inactivation of eEF2 kinase at low calcium concentrations (i.e., lowers its sensi-tivity to activation by Ca2þ) (Wang et al., 2001).
Phospho-rylation at Ser359also inactivates eEF2 kinase (Knebel et al., 2001), but the kinase acting here has yet to be identi-fied. Phosphorylation at the third site, Ser78, also causes inactivation of eEF2 kinase, in this case by inhibiting the binding of CaM, which binds immediately C-terminal to Ser78 (Browne and Proud, 2004). The mTOR-con-trolled kinase acting at this site is also unknown.
There are data indicating that eEF2 kinase activity is elevated in certain cancer cells (e.g., breast cancer; Parmer et al., 1999) and that its activity is enhanced in proliferating cells (Bagaglio et al., 1993; Bagaglio and Hait, 1994; Cheng et al., 1995; Parmer et al., 1997). This appears surprising since eEF2 kinase is a negative regulator of protein synthesis, and thus presumably of cell growth/proliferation. Recent studies show that a compound that acts as a selective inhibitor of eEF2 kinase (NH125) decreases the viability of a range of cancer cell lines, and this effect is countered by artificial overexpression of eEF2 kinase consistent with the idea that eEF2 kinase plays a role in cell viability (Arora et al., 2003). Earlier data obtained with another compound that inhibits eEF2 kinase, rottlerin, indicated loss of viability and growth, as well as changes in cell morphology (Parmer et al., 1997). However, it is now clear that rottlerin can inhibit other kinases in addition to eEF2 kinase (Davies et al., 2000) and caution must therefore be exercised in interpreting data obtained with such reagents.
Recent data suggest a role for eEF2 kinase in the activation of autophagy (Wu et al., 2006), which as noted above is negatively regulated by mTOR signalling. Analysis of cells, tissues or animals that lack eEF2
Catalytic domain CaM-binding eEF2-binding? 78 359 S6K kinase? kinase? mTOR eEF2(P) (inactivation) P P P 366
Figure 3 Overview of eEF2 and eEF2 kinase, including regulatory phosphorylation sites in eEF2 kinase that are linked to mTOR signaling. The (mTOR-linked) kinases that act at Ser78 and Ser359 remain to be identified.
kinase will be required to establish the contribution made by phosphorylation of eEF2 to the control of protein synthesis and to cell physiology. There is currently a high level of interest in autophagy based on recent data that autophagy may suppress tumour development, although the situation is not clearcut (Ng and Huang, 2005; Hait et al., 2006).
Control of eIF4E
Eukaryotic initiation factor 4E (eIF4E) is subject to two main types of rapid regulation: phosphorylation at
a site in its C-terminus (Ser209; Flynn and Proud, 1995; Joshi et al., 1995) and sequestration by small, heat-stable phosphoproteins termed 4E-binding proteins, or 4E-BPs (Gingras et al., 1999; Figure 4a). Very recent data also show that its intracellular localization is also regulated by mTOR.
The effect of phosphorylation on the function of eIF4E has been the subject of mixed reports, and its physiological role remains unclear (Scheper and Proud, 2002). Recent data suggest that phosphorylation of eIF4E weakens its binding to 7-methylguanosine triphosphate and to capped RNA (Scheper et al., 2002; Slepenkov et al., 2006). Detailed biophysical studies are
4E-BP1
PP PP
eIF4E P
Translation initiation eIF4E nuclear import
mRNA degradation (P-bodies)
P P PP ? mTOR? MAPK (Erk, p38 MAPK αβ) eIF4G 4E-T MNKs ? ? P P P raptor 4E-BP1 PP P P RAIP TOS mTOR GβL ? P Inhibition of translation initiation
mTOR,…? a b T37 T46 S65 T70 S101 1 2 3 4 P S112 P
Figure 4 (a) Summary of eIF4E and its binding partners (and their roles, i.e., eIF4G/F, 4E-BPs, 4E-T. Mnks); binding of 4E-BP1 to eIF4E prevents the formation of eIF4F complexes by competing with eIF4G. The phosphorylation of 4E-BP1 prevents its binding to eIF4E. The Mnks are activated by phosphorylation and need to bind to eIF4G in order to phosphorylate eIF4E. The biological role of eIF4E phosphorylation is not yet well-defined (see text). The role of mTOR-dependent phosphorylation of eIF4G is quite unclear. 4E-T is a phosphoprotein but the kinase(s) involved and the effects of the phosphorylation are unknown. (b) Outline of 4E-BPs (based on 4E-BP1), showing motifs and phosphorylation sites. It is not clear how the RAIP motif functions in the mTOR-dependent control of 4E-BP1. It is also not clear whether mTOR directly phosphorylates all four sites (Thr37/46/70 and Ser65) in vivo or recruits another kinase(s) to do so. Other kinases that are able to phosphorylate 4E-BP1 in vivo or in vitro independently of mTOR are not shown as their relevance for the physiological control of 4E-BP1 is unclear. The 101 and 112 sites are constitutively phosphorylated in HEK-293 cells. The numbered arrows indicate the apparent order of the hierarchical phosphorylation of these sites by mTOR signalling. The dashed arrows denote phosphorylation events that depend on mTOR but may not be catalysed directly by mTOR.
consistent with the idea that electrostatic repulsion between the phosphoserine and negative charges on the capped RNA contribute to weakening cap-eIF4E binding (Zuberek et al., 2003, 2004). One may speculate that this serves to allow additional ribosomal initiation complexes to bind the mRNA while the previous one is still engaged in scanning, or that it may permit eIF4E to be released from one mRNA in order to bind others, whose translation has been stimulated by other mechan-isms (Scheper and Proud, 2002). Both could contribute to the activation of mRNA translation. However, there are so far no data on the function of eIF4E phospho-rylation in mRNA translation in mammalian cells.
The kinases that phosphorylate eIF4E are the Mnks (MAP kinase-interacting kinases or MAP kinase signal-integrating kinases) (Scheper and Proud, 2002; Ueda et al., 2004; Figure 4a). There are two Mnk genes in rodents and man. In man, each gives rise to two proteins differing at their C-termini. The longer form of Mnk1 (corresponding to the only Mnk1 species identified in mice) is tightly regulated by MAP kinase signalling, through the ERK and p38 MAP kinase a/b pathways (Fukunaga and Hunter, 1997; Waskiewicz et al., 1997) (these enzymes bind its C-terminus and switch it on). In contrast, murine Mnk2 (or the longer form of the human enzyme) has high basal activity that is affected little by activation or inhibition of the ERK pathway (Scheper et al., 2001), probably because it can bind to activated ERK, allowing Mnk2 to remain active in cells that have not been stimulated or have even been treated with compounds that block ERK activation (Parra et al., 2005). It binds relatively less well to p38 MAP kinase a/b than Mnk1 does. Given that the ERK pathway is activated by oncoproteins such as Ras and Raf, it is possible that the Mnks and eIF4E phospho-rylation play roles in cell transformation. Indeed, it has been shown that overexpression of a mutant of eIF4E in which Ser209has been altered to alanine is much less efficient than wildtype eIF4E in transforming NIH 3T3 cells, as judged by its ability to induce the formation of foci (Topisirovic et al., 2004). The overexpression of wild type, but not mutant, eIF4E is also accompanied by an increase in cyclin D1 levels. The fact the Mnk inhibi-tor CGP57380 was able to decrease cyclin D1 expression reinforced the idea that the phosphorylation of eIF4E may participate in the process of transformation.
The short forms of human Mnk1 and 2 (termed Mnk1b/2b) are partially nuclear (Scheper et al., 2003; O’Loghlen et al., 2004): in the case of Mnk1, this is due to the absence of the nuclear export signal found in Mnk1a. The Mnks may thus have substrates that are nuclear: these would include eIF4E itself, which is partially nuclear (Strudwick and Borden, 2002) (see below). The N-termini of all Mnk isoforms contains a polybasic sequence that serves both to recruit the Mnks to eIF4G and to interact with importin a and allow their nuclear import (Waskiewicz et al., 1999; Parra-Palau et al., 2003). The association of Mnks with eIF4G is important for the efficient phosphorylation of eIF4E in vivo, presumably because, by binding both kinase and substrate, eIF4G brings these proteins together in the
required orientation (as well as increasing their effective local concentrations) (Pyronnet et al., 1999).
A second mode of regulation of eIF4E is through its regulated binding to partner proteins, as described next. This involves mTOR.
Binding partners for eIF4E: roles for eIF4E in mRNA translation, transport and degradation
A surface on the so-called ‘dorsal’ face of eIF4E plays a key role in its binding to several partner proteins, one residue, Trp73, being especially important – mutation of this residue to alanine abolishes these interactions. These partners include the translation initiation scaffold protein eIF4G, the eIF4E-binding proteins (4E-BPs) and the protein 4E-T, which was first reported as a nucleocytoplasmic shuttling protein for eIF4E (Dostie et al., 2000). All these proteins contain similar motifs involved in their binding to eIF4E (Figure 4a). There are two eIF4G genes in mammals, eIF4GI and eIF4GII
(Gingras et al., 1999). The eIF4GI protein is much better characterized than eIF4GII. It exists as multiple
isoforms that differ at their N-termini (Bradley et al., 2002; Byrd et al., 2005).
eIF4G binds a number of other proteins, which include the Mnks (through a region in the extreme C-terminus of eIF4G), the RNA helicase eIF4A, the multisubunit factor eIF3 (which brings to together the 40S ribosomal subunit and eIF4G and its partners) and the poly(A)-binding protein, PABP. The eIF4E/eIF4G/ eIF4A subcomplex is often referred to as eIF4F. The helicase function of eIF4A is thought to be of particular importance for the translation of mRNAs that have long and structured 50-untranslated regions (50-UTRs), which require ‘unwinding’ in order for ribosomes to be able to traverse the 50-UTR and reach the start codon (Gingras et al., 1999). Increased engagement of eIF4E with eIF4G (formation of ‘eIF4F’) will, accord-ing to this idea, favour the translation of such structured mRNAs. There is (limited) experimental evidence in favour of this (see, e.g., Koromilas et al., 1992), as reviewed more fully in Mamane et al. (2004). A number of transformation-related proteins are encoded by ‘less competitive’ mRNAs, including cyclin D1, c-myc and vascular endothelial-growth factor (VEGF; reviewed by Graff and Zimmer (2003) and Mamane et al. (2004)). Their translation is expected to be increased by the ‘activation’ of eIF4E that results from activated mTOR signalling and the release of eIF4E from inhibition by 4E-BP1.
The activity of eIF4A is enhanced by eIF4B (an RNA-binding protein) and eIF4B is a substrate for S6K1 (Raught et al., 2004). Its phosphorylation by S6K1 favours its binding to eIF3 (Holz et al., 2005). Indeed, S6K1 is actually recruited to the translational machinery through an interaction with eIF3 (Holz et al., 2005). eIF4G is subject to rapamycin-sensitive phos-phorylation at several sites although the functional significance of these phosphorylations and the kinase(s) responsible for them are so far unknown (Raught et al., 6428
2000). Nevertheless, it has recently been reported that insulin stimulates the association of eIF3 with eIF4G and that this is inhibited by rapamycin (Harris et al., 2006). It has not so far been reported whether cells from S6K1/ S6K2/ mice show any defect in protein
synthesis.
Interestingly, overexpression of other components of eIF4F is also associated with transformation. The mRNA for eIF4A is highly expressed in some tumours or tumour cells (Eberle et al., 1997; Shuda et al., 2000). Artificial overexpression of eIF4GI in
NIH3T3 cells leads to anchorage-independent cell proliferation and tumour formation in nude mice (Fukuchi-Shimogori et al., 1997), which is consistent with a key role for eIF4F in events that lead to transformation. Thus, inhibiting mTOR signalling and thereby promoting the loss of such complexes may restrain transformation, in an analogous way to the effect of overexpressing 4E-BPs, which also prevent eIF4F complex formation (Rousseau et al., 1996a). Consistent with this, expression of phosphorylation-defective mutants of 4E-BP1 suppressed the tumori-genicity of breast carcinoma cells and the spontaneous loss of expression of these 4E-BP1 mutants correlated with reversion of the cells to a malignant phenotype (Avdulov et al., 2004).
4E-T also contains the type of eIF4E-binding motif found in eIF4GI/II and the 4E-BPs (Dostie
et al., 2000). It was initially described as a shuttling protein for eIF4E, mediating its import into the nucleus. Recent data indicate that it also recruits eIF4E to P (processing)-bodies, which are sites for mRNA degradation (Ferraiuolo et al., 2005). 4E-T is a phosphoprotein, although the role (if any) of phospho-rylation in controlling its function or localization is unclear.
Control of the 4E-BPs by mTOR
The regulation of the 4E-BP1 is much better understood than that of 4E-BP2/3. This protein undergoes phos-phorylation at seven sites, at least four of which are linked to mTOR signalling (Mothe-Satney et al., 2000a, b; Gingras et al., 2001; Wang et al., 2005). These are threonines 37, 46 and 70, and serine 65 (numbering based on human 4E-BP1; numbers for the rodent proteins differ by1) (Figure 4b). Ser65 and Thr70 lie close to the eIF4E-binding site and, in many cells, their phosphorylation is stimulated by insulin (etc) in a rapamycin-sensitive manner. In some cells, such as CHO cells, they are constitutively phosphorylated (Wang et al., 2005). In HEK 293 cells, for example, their phosphorylation is dependent upon the C-terminal TOS motif (Beugnet et al., 2003). Although there are data that suggest that phosphorylation of this site is required for release of eIF4E from 4E-BP1, the role of phosphorylation of Ser65 is still controversial (Ferguson et al., 2003). Molecular dynamics findings (Tomoo et al., 2005) and earlier biophysical data suggest that phosphorylation of Ser65 and Thr70 is
insufficient to bring about release from eIF4E. Phosphorylation of both Ser65 and Thr70 depends upon the prior phosphorylation of the N-terminal threonines, and modification of Thr46 may precede and be required for phosphorylation of Thr37 (Mothe-Satney et al., 2000b; Gingras et al., 2001; Herbert et al., 2002). The phosphorylation of Thr70 and Ser65 in human 4E-BP1 depends upon a further site, Ser101 (Wang et al., 2003). 4E-BP1 is thus subject to a complex series of hierarchical phosphorylation events (summa-rized in Figure 4b).
The phosphorylation of the N-terminal threonines in 4E-BP1 and 4E-BP2 depends upon a motif in the N-terminus of 4E-BP1, which includes the sequence Arg-Ala-Ile-Pro (hence ‘RAIP’ motif) (Tee and Proud, 2002; Beugnet et al., 2003) (Figure 4b). Their phosphorylation is not greatly affected by inactivation of the TOS motif and is rather insensitive to rapamycin (Wang et al., 2005). This could suggest that it is mediated indepen-dently of mTOR. However, several lines of evidence suggest that their phosphorylation is mediated by mTOR: (i) it is inhibited by starvation of cells for amino acids, in common with effects mediated by mTORC1; (ii) it is activated by Rheb, a stimulator of mTOR (or least, mTORC1); (iii) it is suppressed by TSC1/2, the GTPase-activating complex for Rheb; (iv) its is sensitive to inhibitors of the kinase activity of mTOR, such as wortmannin, and (v) it is decreased in cells in which mTOR levels have been knocked down (Wang et al., 2005). The insensitivity to rapamycin could suggest that mTORC2 mediates the phosphorylation of Thr37/46, but the other data argue against this: for example, amino acids are not known to regulate mTORC2. mTORC2 cannot phosphorylate 4E-BP1 on Thr 37/46 (Fonseca and Proud, unpublished data). Thus, it seems likely that the phosphorylation of Thr37/ 46 in 4E-BP1 is mediated by mTOR through a mechanism that is independent of both raptor (mTORC1) and rictor (mTORC2). Identifying how this works would generate important new data on mTOR signalling as well as perhaps providing a novel and specific approach to inhibiting 4E-BP1 phosphorylation and thus eIF4E function, without inhibiting with S6Ks and promoting PI 3-kinase signalling.
Whereas 4E-BP1 is mainly cytoplasmic, 4E-BP3 is also found in the nucleus (Kleijn et al., 2002). It shares this property with its partner, eIF4E, although the biological significance of this is unclear. 4E-BP2 and 4E-BP3 both possess C-terminal TOS motifs that are identical in sequence to that in 4E-BP1. Only 4E-BP2 contains a RAIP motif, which apparently contributes to controlling 4E-BP2 phosphorylation by amino acids in a similar way to its role in controlling 4E-BP1 (Wang et al., 2005).
eIF4E: roles in cell transformation and survival
eIF4E may be described as a proto-oncogene. There is compelling evidence that it plays a role in the processes that lead to cell transformation. Its expression is
increased in several kinds of tumors (reviewed in Bjornsti and Houghton, 2004; De Benedetti and Graff, 2004). Moreover, the overexpression of eIF4E can induce cellular transformation in vitro and in vivo (Lazaris-Karatzas et al., 1990; Ruggero et al., 2004; Wendel et al., 2004). Using a murine model, eIF4E was found to recapitulate the effect of PKB overexpression in terms of tumorigenesis and drug resistance (Wendel et al., 2004). PKB overexpression did not increase the level of eIF4E but led to activation of mTOR and hyperphosphorylation of 4E-BP1. This results in release (activation) of eIF4E and to increased formation of eIF4F complexes. While rapamycin was effective in restoring the sensitivity of PKB-expressing cells to proapoptotic stimuli, it did not do so in eIF4E-overexpressing cells. The data suggest that eIF4E may play a key role in the mechanisms by which PKB transforms cells, and can be interpreted as suggesting that the ‘target’ through which rapamycin acts to sensitize cells to proapoptotic stimuli is likely to be 4E-BP1/eIF4E. Thus, control of the translational machinery may be a key element of the mechanism by which PKB transforms cells.
The overexpression of eIF4E does not lead to a global increase in mRNA translation but rather to the increased expression of specific proteins such as ODC, c-myc or cyclin D1. In several cases, their mRNAs possess a long and structured 50-UTRs and, as mentioned earlier, the translation of such mRNAs is thought to be highly dependent on the eIF4F complex (for reviews see Clemens, 2004; Mamane et al., 2004). Several of these mRNAs encode proteins that are involved or implicated in tumorigenicity (see review Mamane et al., Oncogene 2004). The c-myc mRNA, in particular, has been studied extensively. In a recent study, Ruggero et al. (2004) have shown that over-expression of eIF4E under the control of the b-actin promoter leads to tumour formation. When these mice were crossed with others expressing c-myc under the control of the immunoglobulin heavy-chain enhancer element, lymphoma formation was greatly accelerated. It appears that enhanced expression of eIF4E not only upregulates expression of c-myc but also that eIF4E cooperates with it in the process of transformation. Although c-myc is an oncogene, it can also promote apoptosis, and this effect is antagonized by eIF4E overexpression which exerts antiapoptotic effects (Li et al., 2003). On the other hand, c-myc antagonizes the effect of eIF4E overexpression on cellular senescence. In addition to the reported ability of c-myc to increase the level of expression of other proteins involved in translation (reviewed in Schmidt, 2004), it likely also cooperates with eIF4E in transformation by virtue of its ability to counter the effect of eIF4E on cellular senescence. It is also important to consider that eIF4E overexpression may play a different kind of role in cellular transformation, independently of its effect on translation.
A fraction of the total cellular eIF4E is found within the nucleus (reviewed by Strudwick and Borden, 2002). In this compartment, it associates with the promyelocytic
leukaemia protein, PML (Cohen et al., 2001; Kentsis et al., 2001), which is also associated with Mnk2b (Scheper et al., 2003). Nuclear eIF4E may be involved in the export of certain mRNAs to the cytoplasm (Rousseau et al., 1996b; Culjkovic et al., 2005): over-expression of eIF4E enhances the over-expression of cyclin D1 not by increasing the translation of its mRNA directly, but rather by increasing the level of the cyclin D1 mRNA in the cytoplasm (Rousseau et al., 1996b). The mechanism by which eIF4E does this is unclear, but it involves a feature in the 30-UTR of the cyclin D1 mRNA (Culjkovic et al., 2005), as well as the cap-binding activity of eIF4E. Recent kinetic studies suggest that eIF4E may interact with additional regions of the mRNA, via interactions with positively charged side chains in the groove in the 3-D structure of eIF4E where the mRNA may be located (Slepenkov et al., 2006). However, the increase in cyclin D1 expression caused by enhanced levels of eIF4E does not require the ability of eIF4E to engage in translation, as illustrated by the ability of the Trp73Ala mutant of eIF4E, which cannot bind eIF4G, to increase cyclin D1 levels (Topisirovic et al., 2002, 2003). 4E-T binds eIF4E through a similar motif to that found in eIF4G, so this residue should be needed for binding of eIF4E to 4E-T and for nuclear import of eIF4E. This prompts the question: how does eIF4E(W73A) get into the nucleus to promote export of the cyclin D1 message? It would be important also to define whether the binding of 4E-BP1 to eIF4E could interferes with eIF4E localization and so the transport of cyclin D1 mRNA, it will define a new role for mTOR signalling. Nevertheless, the existing data provoke the idea that one way in which PML exerts its tumour-suppressor function is by decreasing the binding of eIF4E to the cyclin D1 mRNA (PML diminishes the cap-binding ability of eIF4E), thereby impairing export of this mRNA, and inhibiting the G1/S transition of the
cell cycle (Cohen et al., 2001).
4E-BP1 should counter the ability of eIF4E to promote cell transformation. It has been shown in cells in culture that overexpression of 4E-BP1 can reduce the tumorigenic effect of eIF4E overexpression (Avdulov et al., 2004). Moreover, the same authors have implanted mammary cancer cell MDA-MB-468 trans-fected with 4E-BP1 or a mutant of 4E-BP1, which constitutively binds to eIF4E, into mice. Expression of wildtype 4E-BP1 decreased tumour size and expression of the mutant had an even larger effect. Lynch et al. (2004) have also shown that the overexpression of a mutant of 4E-BP1 inhibits transformation induces by eIF4E and overexpression. However, this mutant is also able to block the transformation induces by c-myc overexpression in Rat1 cells. They observe a decrease of colony formation in soft agar assay and also of tumour size in nude mice implanted with the Rat1 cells. The authors have shown that this effect of the 4E-BP1 mutant involved a slower G1 progression and did not decrease global protein synthesis. This last result reinforced the idea that ‘activation’ of eIF4E does not increase the level of general protein synthesis but rather enhances the translation of specific mRNAs that encode 6430
proteins which play an important role, for example, in the cell cycle. Conversely, overexpression of 4E-BP1 enhanced the ability of rapamycin to inhibit the proliferation of PC3 (prostate cancer) cells. This effect was not apparent in cells overexpressing a mutant of 4E-BP1 that cannot bind to or inhibit eIF4E (Beugnet and Proud, unpublished data).
4E-BPs and apoptosis
Several lines of data point to links between 4E-BP1 (and perhaps other 4E-BPs) and programmed cell death (reviewed in Proud, 2005). For example, expres-sion of 4E-BP1 promoted apoptosis of Ras-transformed fibroblasts (Polunovsky et al., 2000). Treatment of cells with cell-permeant peptides that inhibit eIF4E in a similar way to the 4E-BPs leads to rapid cell death (Herbert et al., 2000). Conversely, as noted above, expression of eIF4E exerts antiapoptotic effects (see, e.g., Polunovsky et al., 1996; Li et al., 2003, 2004; Wendel et al., 2004).
DNA damage, which can cause apoptosis, has two types of effect on 4E-BP1. Firstly, prior to the onset of apoptosis (as gauged, e.g., by caspase activation), mTOR signalling is inhibited, causing increased binding of 4E-BP1 to eIF4E and inhibition of eIF4F formation (Tee and Proud, 2000, 2001), which may impair the translation of antiapoptotic mRNAs and enhance the translation of mRNAs for proapoptotic proteins such as APAF-1, a component of the apoptosome which is encoded by an mRNA that contains an internal ribosome entry segment (IRES) that allows eIF4E-independent translation (Coldwell et al., 2000). p53 may be involved in mediating this effect, as activation of p53 leads to inhibition of mTOR signalling (Horton et al., 2002; Feng et al., 2005) and dephosphorylation and degradation of eIF4GI and 4E-BP1 (Constantinou and
Clemens, 2005). However, the data in the last paper suggest that p53 activation inhibits protein synthesis through mechanisms that are independent of rapamycin-sensitive mTOR function.
Secondly, following caspase activation, 4E-BP1 is cleaved, removing the section containing the RAIP motif (Tee and Proud, 2002). 4E-BP1 therefore becomes dephosphorylated, causing it to bind to eIF4E in a manner that is ‘irreversible’, in the sense that it is no longer possible to bring about the phosphorylation-mediated release of 4E-BP1. This will further serve to shut down protein synthesis in an ordered manner and, perhaps, also promote the upregulation of pro-apoptotic proteins encoded by IRES-containing mRNAs.
Other targets of mTOR signalling
There are no other direct targets for mTOR signalling that can really be described as ‘well-characterized’ in terms of their control by mTOR. The transcription factor STAT3 has been reported to be regulated by mTOR in terms both of its phosphorylation and its
activity. In one report (Yokogami et al., 2000), rapamycin was shown to block the phosphoryla-tion of Tyr727 in STAT3 that is induced by ciliary neurotrophic factor, and STAT3 activity. In response to interferon (IFN)-g, STAT3 undergoes dephosphoryla-tion at another site, and this effect is also inhibited by rapamycin (Fang et al., 2006). In this case, rapamycin actually inhibited IFNg-induced apoptosis. The links between mTOR signalling and the control of the (de)phosphorylation of STAT3 are unknown. However, since the phosphorylation and function of STAT3 are dysregulated (e.g., through constitutive tyrosine phos-phorylation) in a range of human cancers (Klampfer, 2006), inhibition of mTOR signalling may be of value in the context of such tumours.
As mentioned above, it is well-established that mTOR controls the translation of the subset of mRNAs that include those encoding ribosomal proteins and some elongation factors (Meyuhas and Hornstein, 2000). In response to rapamycin treatment, these mRNAs shift out of polyribosomes indicating impairment of their translation (reviewed by Meyuhas and Hornstein, 2000). This control is conferred by a sequence in their extreme 5-ends comprising a series of pyrimidines (hence 50-tract
of oligopyrimidines; 50-TOP). However, it is not known how this motif confers control by mTOR or how mTOR modulates their translation.
mTOR is also reported to regulate the transcription of rDNA genes (genes encoding ribosomal RNA (Hannan et al., 2003; Mayer et al., 2004)), as described elsewhere in this volume. A role for mTOR in controlling the synthesis or ribosomal proteins and rRNA is clearly logical to link, for example, amino-acid availability and anabolic/mitogenic signalling to ribosome biogenesis.
Conclusions and perspectives
mTOR is one of the main regulators of cell growth and proliferation. Its functions in controlling the translational machinery are summarized in Figure 5. mTOR is located at the cross road of several major signalling pathways (including PI 3-kinase, Erk and AMPK) and is able to integrate a large panel of stress signal such as nutrient deprivation, energy depletion and oxidative or hypoxic stress. It can thereby modulate numerous cellular functions. Inhibition of mTOR by rapamycin and its analogs appears to be an efficient way, in certain cases, to stop the growth of tumour cells and/or to promote apoptosis. The use of rapamycin as an anticancer drug is now being evaluated on patients with solid or haematological tumours.
However, it is important to consider that the effectiveness of inhibition of mTOR very likely differs substantially between different types of tumours. It appears logical that one of the main determinants of this sensitivity is the level of activity of mTOR itself. By virtue of its location in the signalling network, its activity is regulated by several tumour suppressors such
as PTEN or LKB1. It has been also described that p53 can inhibit mTOR activity (Feng et al., 2005). As a consequence of this, mTOR activity is enhanced in different kinds of cancers. One may hypothesize that the inhibition of different targets of mTOR by rapamycin will have differing levels of importance according to the type of cancer. These points raise a question: is it appropriate to inhibit all the (rapamycin-sensitive) targets of mTOR or just the one(s) which seems to play an active role in the transformation or in the main-tenance of the malignant phenotype. The first option is the one currently being tested. Concerning the second possibility, it will first be necessary to identify the relevant other targets of mTOR and then to generate specific inhibitors that interfere with the target or its control by mTOR.
This notion that the inhibition of a specific target and not all the targets could be beneficial is illustrated with the example of S6K1 inhibition, which may enhance the antiapoptotic effect of the PI3 K pathway due the absence of the negative feed-back control of IRS1 by S6K1 (see review Manning). In the case of 4E-BP1, it may be envisaged that its activation may also lead to the increased translation of antiapoptotic proteins. This notion that in some case mTOR inhibition could lead to resistance to apoptosis despite a growth arrest should be studied with attention. For example, although most studies describe rapamycin as a pro-apoptotic agent, it has been already observed that rapamycin can increase levels of Bcl-2 and thus exert an antiapoptotic effect (Calastretti et al., 2001a, b). It may be that the so-called mTOR inhibition elicited by a stress leads to the activation of apoptotic effectors and this activation is not reproduced by rapamycin treatment.
Among the best-known targets of mTOR, 4E-BP1 presents the most direct link with cancer by regulating the level of key proteins such as c-myc, VEGF and
cyclin D1 through its control over eIF4E availability. It will be interesting to determine which other proteins exhibit similar regulation. Notably, one may suggest that HIF is one of them. HIF is a transcription factor involved in the adaptation to hypoxia. HIF is known to be downregulated by rapamycin treatment. HIF ex-pression increases when the tumour suppressor VHL (von Hippel-Lindau), a ubiquitin ligase, is mutated (Thomas et al., 2006). In that case, mTOR inhibition is particularly efficient in inhibiting cell proliferation. Mutating the 50-UTR of the HIF mRNA abolishes the effect of inhibiting mTOR (see comments in Choo and Blenis (2006)). Another interesting point is that, despite the fact that mTOR is described as being inhibited by hypoxia, HIF expression is actually enhanced in hypoxia. In fact, it appears that residual mTOR activity is maintained under hypoxic conditions and allows the translation of specific mRNA such as HIF (Hui et al., 2006). In their work, Hui et al. observed a less profound effect of hypoxia on 4E-BP1 phosphorylation than on S6K1 phosphorylation. One can thus envisage the use of rapamycin treatment for tumours presenting a low level of vascularization even if the level of mTOR activity is low. It would be also interesting to study the regulation of the translation of proteins involved in the control of autophagy. It is known than mTOR regulates this process (Meijer and Codogno, 2004) and there is a growing interest in the role of autophagy in cancer (Ng and Huang, 2005). It has been shown that the oncogene Bcl-2 maintains cell survival by inhibiting autophagy (Pattingre and Levine, 2006).
It is important to understand more clearly the signalling events that operate downstream of mTOR. One example is the phosphorylation of 4E-BP1, notably the rapamycin-insensitive regulation which is not dependent on mTORC2. This pathway appears to be sensitive to amino acids and so could perhaps
5’-cap 40S 60S 40S 60S 40S 60S 40S 5’-UTR 3’-UTR AA AA A AA A eIF4E eIF4E 4E-BP1 P P P P P 4E-BP1 P P P eIF4G P P 60S 40S eIF4B eIF3 P PP mTOR S6K eEF2 eEF2 eEF2 kinase RIBOSOME BIOGENESIS 60S 40S ?? ?? ?? 5’-cap Overall translational capacity ?? ??
Figure 5 Overview of the links between mTOR and the translational machinery. The thick black line represents an mRNA, with its 50-cap, start codon (gold star) and other features shown. Questions marks denote signaling connections or functional consequences that
are not understood. Solid lines indicate signaling connections including phosphorylation events, while dashed straight lines signify association or dissociation events. The figure is entirely schematic, with no attempt to depict relative sizes of components, or actual numbers or locations of phosphorylation sites.
involve the class III PI3 K, but this remains to be established. The importance of this rapamycin-insensitive regulation is enforced by the work of Constantinou and Clemens (2005). Using a tempera-ture-sensitive mutant of p53 in erythroleukaemia cells, they show that p53 activation can cause the
dephosphorylation of 4E-BP1 at the rapamycin-resis-tant site(s). The other rapamycin-insensitive pathway that involves mTORC2 is also a new field of investiga-tion, which will perhaps allow the development of new approaches to inhibit mTOR functions in a more specific manner.
References
Alessi DR, Kozlowski MT, Weng Q-P, Morrice N, Avruch J. (1998). Curr Biol 8: 69–81.
Ali SM, Sabatini DM. (2005). J Biol Chem 280: 19445–19448. Arora S, Yang JM, Kinzy TG, Utsumi R, Okamoto T,
Kitayama T et al. (2003). Cancer Res 63: 6894–6899. Avdulov S, Li S, Michalek V, Burrichter D, Peterson M,
Perlman DM et al. (2004). Cancer Cell 5: 553–563. Avruch J, Belham C, Weng Q, Hara K, Yonezawa K. (2001).
Prog Mol Subcell Biol 26: 115–154.
Bagaglio DM, Cheng EH, Gorelick FS, Mitsui K, Nairn AC, Hait WN. (1993). Cancer Res 53: 2260–2264.
Bagaglio DM, Hait WN. (1994). Cell Growth Differ 5: 1403–1408.
Barlund M, Monni O, Kononen J, Cornelison R, Torhorst J, Sauter G et al. (2000). Cancer Res 60: 5340–5344.
Beugnet A, Wang X, Proud CG. (2003). J Biol Chem 278: 40722.
Bjornsti MA, Houghton PJ. (2004). Cancer Cell 5: 519–523. Bradley CA, Padovan JC, Thompson TL, Benoit CA,
Chait BT, Rhoads RE. (2002). J Biol Chem 277: 12559–12571. Browne GJ, Proud CG. (2002). Eur J Biochem 269: 5360–5368. Browne GJ, Proud CG. (2004). Mol Cell Biol 24: 2986–2997. Byfield MP, Murray JT, Backer JM. (2005). J Biol Chem 280:
33076–33082.
Byrd MP, Zamora M, Lloyd RE. (2005). J Biol Chem 280: 18610–18622.
Calastretti A, Bevilacqua A, Ceriani C, Vigano S, Zancai P, Capaccioli S et al. (2001a). Oncogene 20: 6172–6180. Calastretti A, Rancati F, Ceriani MC, Asnaghi L, Canti G,
Nicolin A. (2001b). Eur J Cancer 37: 2121–2128.
Carlberg U, Nilsson A, Nygard O. (1990). Eur J Biochem 191: 639–645.
Cheng EH, Gorelick FS, Czernik AJ, Bagaglio DM, Hait WN. (1995). Cell Growth Differ 6: 615–621.
Choo AY, Blenis J. (2006). Cancer Cell 9: 77–79. Clemens MJ. (2004). Oncogene 23: 3180–3188.
Cohen N, Sharma M, Kentsis A, Perez JM, Strudwick S, Borden KLB. (2001). EMBO J 20: 4547–4559.
Coldwell MJ, Mitchell SA, Stoneley M, MacFarlane M, Willis AE. (2000). Oncogene 19: 899–905.
Constantinou C, Clemens MJ. (2005). Oncogene 24: 4839–4850.
Culjkovic B, Topisirovic I, Skrabanek L, Ruiz-Gutierrez M, Borden KL. (2005). J Cell Biol 169: 245–256.
Davies SP, Reddy H, Caivano M, Cohen P. (2000). Biochem J 351: 95–105.
De Benedetti A, Graff JR. (2004). Oncogene 23: 3189–3199. Dostie J, Ferraiuolo M, Pause A, Adam SA, Sonenberg N.
(2000). EMBO J 19: 3142–3156.
Eberle J, Krasagakis K, Orfanos CE. (1997). Int J Cancer 71: 396–401.
Fang P, Hwa V, Rosenfeld RG. (2006). Exp Cell Res 312: 1229–1239.
Feng Z, Zhang H, Levine AJ, Jin S. (2005). Proc Natl Acad Sci USA 102: 8204–8209.
Ferguson G, Mothe-Satney I, Lawrence Jr JC. (2003). J Biol Chem 278: 47459–47465.
Ferraiuolo MA, Basak S, Dostie J, Murray EL, Schoenberg DR, Sonenberg N. (2005). J Cell Biol 170: 913–924.
Fingar DC, Richardson CJ, Tee AR, Cheatham L, Tsou C, Blenis J. (2004). Mol Cell Biol 24: 200–216.
Fingar DC, Salama S, Tsou C, Harlow E, Blenis J. (2002). Genes Dev 16: 1472–1487.
Flynn A, Proud CG. (1995). J Biol Chem 270: 21684–21688. Fukuchi-Shimogori T, Ishii I, Kashiwagi K, Mashiba H,
Ekimoto J, Igarashi K. (1997). Cancer Res 57: 5041–5044. Fukunaga R, Hunter T. (1997). EMBO J 16: 1921–1933. Gingras A-C, Raught B, Gygi SP, Niedzwieka A, Miron M,
Burley SK et al. (2001). Genes Dev 15: 2852–2864.
Gingras A-C, Raught B, Sonenberg N. (1999). Annu Rev Biochem 68: 913–963.
Graff JR, Zimmer SG. (2003). Clin Exp Metastasis 20: 265–273.
Gressner AM, Wool IG. (1974). Biochem Biophys Res Commun 60: 1482–1490.
Guertin DA, Sabatini DM. (2005). Trends Mol Med 11: 353–361.
Hait WN, Jin S, Yang JM. (2006). Clin Cancer Res 12: 1961–1965.
Hannan KM, Brandenburger Y, Jenkins A, Sharkey K, Cavanaugh A, Rothblum L et al. (2003). Mol Cell Biol 23: 8862–8877.
Harris TE, Chi A, Shabanowitz J, Hunt DF, Rhoads RE, Lawrence Jr JC. (2006). EMBO J 25: 1659–1668.
Herbert TP, Fahraeus R, Prescott A, Lane DP, Proud CG. (2000). Curr Biol 10: 793–796.
Herbert TP, Tee AR, Proud CG. (2002). J Biol Chem 277: 11591–11596.
Holz MK, Ballif BA, Gygi SP, Blenis J. (2005). Cell 123: 569–580.
Horton LE, Bushell M, Barth-Baus D, Tilleray VJ, Clemens MJ, Hensold JO. (2002). Oncogene 21: 5325–5334. Houghton PJ, Huang S. (2004). Curr Top Microbiol Immunol
279: 339–359.
Hui AS, Bauer AL, Striet JB, Schnell PO, Czyzyk-Krzeska MF. (2006). FASEB J 20: 466–475.
Jacinto E, Loewith R, Schmidt A, Lin S, Ruegg MA, Hall A et al.(2004). Nat Cell Biol 6: 1122–1128.
Joshi B, Cai AL, Keiper BD, Minich WB, Mendez R, Beach CM et al. (1995). J Biol Chem 270: 14597–14603. Kentsis A, Campbell Dwyer E, Perez JM, Sharma M, Chen A,
Pan ZQ et al. (2001). J Mol Biol 312: 609–623.
Kimball SR, Jefferson LS. (2006). J Nutr 136: 227S–231S. Klampfer L. (2006). Curr Cancer Drug Targets 6: 107–121. Kleijn M, Scheper GC, Wilson ML, Tee AR, Proud CG.
(2002). FEBS Lett 532: 319–323.
Klionsky DJ, Emr SD. (2000). Science 290: 1717–1721. Knebel A, Morrice N, Cohen P. (2001). EMBO J 20:
4360–4369.
Koh H, Jee K, Lee B, Kim J, Kim D, Yun YH et al. (1999). Oncogene 18: 5115–5119.
Koromilas AE, Lazaris-Karatzas A, Sonenberg N. (1992). EMBO J 11: 4153–4158.
Lazaris-Karatzas A, Montine KS, Sonenberg N. (1990). Nature 345: 544–547.
Li S, Perlman DM, Peterson MS, Burrichter D, Avdulov S, Polunovsky VA et al. (2004). J Biol Chem 279: 21312–21317. Li S, Takasu T, Perlman DM, Peterson MS, Burrichter D,
Avdulov S et al. (2003). J Biol Chem 278: 3015–3022. Lynch M, Fitzgerald C, Johnston KA, Wang S, Schmidt EV.
(2004). J Biol Chem 279: 3327–3339.
Ma L, Chen Z, Erdjument-Bromage H, Tempst P, Pandolfi PP. (2005). Cell 121: 179–193.
Mamane Y, Petroulakis E, Rong L, Yoshida K, Ler LW, Sonenberg N. (2004). Oncogene 23: 3172–3179.
Manning BD. (2004). J Cell Biol 167: 399–403.
Mayer C, Zhao J, Yuan X, Grummt I. (2004). Genes Dev 18: 423–434.
Meijer AJ, Codogno P. (2004). Int J Biochem Cell Biol 36: 2445–2462.
Meyuhas O, Hornstein E. (2000). In: Sonenberg N, Hershey JWB, Mathews MB (eds). Cold Spring Harbor Laboratory Press: Cold Spring Harbor, NY, pp. 671–693. Mothe-Satney I, Brunn GJ, McMahon LP, Capaldo CT,
Abraham RT, Lawrence JC. (2000a). J Biol Chem 275: 33836–33843.
Mothe-Satney I, Yang D, Fadden P, Haystead TAJ, Lawrence JC. (2000b). Mol Cell Biol 20: 3558–3567.
Ng G, Huang J. (2005). Mol Carcinog 43: 183–187.
Nobukuni T, Joaquin M, Roccio M, Dann SG, Kim SY, Gulati P et al. (2005). Proc Natl Acad Sci USA 102: 14238–14243.
O’Loghlen A, Gonzalez VM, Pineiro D, Perez-Morgado MI, Salinas M, Martin ME. (2004). Exp Cell Res 299: 343–355. Parmer TG, Ward MD, Hait WN. (1997). Cell Growth Differ
8: 327–334.
Parmer TG, Ward MD, Yurkow EJ, Vyas VH, Kearney TJ, Hait WN. (1999). Br J Cancer 79: 59–64.
Parra JL, Buxade M, Proud CG. (2005). J Biol Chem 280: 37623–37633.
Parra-Palau JL, Scheper GC, Wilson ML, Proud CG. (2003). J Biol Chem 278: 44197–44204.
Pattingre S, Levine B. (2006). Cancer Res 66: 2885–2888. Pende M, Kozma SC, Jaquet M, Oorschot V, Burcelin R,
Le Marchand-Brustel Y et al. (2000). Nature 408: 994–997. Pende M, Um SH, Mieulet V, Sticker M, Goss VL, Mestan J
et al.(2004). Mol Cell Biol 24: 3112–3124.
Polunovsky VA, Gingras AC, Sonenberg N, Peterson M, Tan A, Rubins JB et al. (2000). J Biol Chem 275: 24776–24780. Polunovsky VA, Rosenwald IB, Tan AT, White J, Chiang L,
Sonenburg N et al. (1996). Mol and Cell Biol 16: 6573–6581. Proud CG. (2005). Cell Death Differ 12: 541–546.
Pullen N, Dennis PB, Andjelkovic M, Dufner A, Kozma SC, Hemmings BA et al. (1998). Science 279: 707–710.
Pyronnet S, Imataka H, Gingras AC, Fukunaga R, Hunter T, Sonenberg N. (1999). EMBO J 18: 270–279.
Raught B, Gingras AC, Gygi SP, Imataka H, Morino S, Gradi A et al. (2000). EMBO J 19: 434–444.
Raught B, Peiretti F, Gingras AC, Livingstone M, Shahbazian D, Mayeur GL et al. (2004). EMBO J 23: 1761–1769. Redpath NT, Foulstone EJ, Proud CG. (1996). EMBO J 15:
2291–2297.
Rousseau D, Gingras A-C, Pause A, Sonenberg N. (1996a). Oncogene 13: 2415–2420.
Rousseau D, Kaspar R, Rosenwald I, Gehrke L, Sonenberg N. (1996b). Proc Natl Acad Sci USA 93: 1065–1070.
Roux PP, Blenis J. (2004). Microbiol Mol Biol Rev 68: 320–344. Ruggero D, Montanaro L, Ma L, Xu W, Londei P,
Cordon-Cardo C et al. (2004). Nat Med 10: 484–486.
Ruvinsky I, Meyuhas O. (2006). Trends Biochem Sci 31: 342–348.
Ruvinsky I, Sharon N, Lerer T, Cohen H, Stolovich-Rain M, Nir T et al. (2005). Genes Dev 19: 2199–2211.
Ryazanov AG, Ward MD, Mendola CE, Pavur KS, Dorovkov MV, Wiedmann M et al. (1997). Proc Natl Acad Sci USA 94: 4884–4889.
Sarbassov dD, Ali SM, Sabatini DM. (2005). Curr Opin Cell Biol 17: 596–603.
Sarbassov dD, Ali SM, Sengupta S, Sheen JH, Hsu PP, Bagley AF et al. (2006). Mol Cell 22: 159–168.
Schalm SS, Blenis J. (2002). Curr Biol 12: 632–639.
Schalm SS, Tee AR, Blenis J. (2005). J Biol Chem 280: 11101–11106.
Scheper GC, Morrice NA, Kleijn M, Proud CG. (2001). Mol Cell Biol 21: 743–754.
Scheper GC, Parra J-L, Wilson ML, van Kollenburg B, Vertegaal ACO, Han Z-G et al. (2003). Mol Cell Biol 23: 5692–5705.
Scheper GC, Proud CG. (2002). Eur J Biochem 269: 5350–5359.
Scheper GC, van Kollenburg B, Hu J, Luo Y, Goss DJ, Proud CG. (2002). J Biol Chem 277: 3303–3309.
Schmidt EV. (2004). Oncogene 23: 3217–3221.
Shuda M, Kondoh N, Tanaka K, Ryo A, Wakatsuki T, Hada A et al. (2000). Anticancer Res 20: 2489–2494. Slepenkov SV, Darzynkiewicz E, Rhoads RE. (2006). J Biol
Chem 281: 14927–14938.
Smith EM, Finn SG, Tee AR, Browne GJ, Proud CG. (2005). J Biol Chem 280: 18717–18727.
Strudwick S, Borden KL. (2002). Differentiation 70: 10–22. Surace EI, Lusis E, Haipek CA, Gutmann DH. (2004). Ann
Neurol 56: 295–298.
Tee AR, Proud CG. (2000). Oncogene 19: 3021–3031. Tee AR, Proud CG. (2001). Cell Death Differ 8: 841–849. Tee AR, Proud CG. (2002). Mol Cell Biol 22: 1674–1683. Thomas GV, Tran C, Mellinghoff IK, Welsbie DS, Chan E,
Fueger B et al. (2006). Nat Med 12: 122–127.
Tomoo K, Matsushita Y, Fujisaki H, Abiko F, Shen X, Taniguchi T et al. (2005). Biochim Biophys Acta 1753: 191–208.
Topisirovic I, Capili AD, Borden KL. (2002). Mol Cell Biol 22: 6183–6198.
Topisirovic I, Culjkovic B, Cohen N, Perez JM, Skrabanek L, Borden KL. (2003). EMBO J 22: 689–703.
Topisirovic I, Ruiz-Gutierrez M, Borden KL. (2004). Cancer Res 64: 8639–8642.
Tuhackova Z, Sovova V, Sloncova E, Proud CG. (1999). Int J Cancer 81: 963–969.
Ueda T, Watanabe-Fukunaga R, Fukuyama H, Nagata S, Fukunaga R. (2004). Mol Cell Biol 24: 6539–6549.
Wang X, Beugnet A, Murakami M, Yamanaka S, Proud CG. (2005). Mol Cell Biol 25: 2558–2572.
Wang X, Li W, Parra J-L, Beugnet A, Proud CG. (2003). Mol Cell Biol 23: 1546–1557.
Wang X, Li W, Williams M, Terada N, Alessi DR, Proud CG. (2001). EMBO J 20: 4370–4379.
Waskiewicz AJ, Flynn A, Proud CG, Cooper JA. (1997). EMBO J 16: 1909–1920.
Waskiewicz AJ, Johnson JC, Penn B, Mahalingam M, Kimball SR, Cooper JA. (1999). Mol Cell Biol 19: 1871–1880.
Wendel HG, De Stanchina E, Fridman JS, Malina A, Ray S, Kogan S et al. (2004). Nature 428: 332–337.
Wu H, Yang JM, Jin S, Zhang H, Hait WN. (2006). Cancer Res 66: 3015–3023.
Wullschleger S, Loewith R, Hall MN. (2006). Cell 124: 471–484.
Yokogami K, Wakisaka S, Avruch J, Reeves SA. (2000). Curr Biol 10: 47–50.
Zhou HY, Wong AS. (2006). Endocrinology 147: 2557–2566.
Zuberek J, Jemielity J, Jablonowska A, Stepinski J, Dadlez M, Stolarski R et al. (2004). Biochemistry 43: 5370–5379. Zuberek J, Wyslouch-Cieszynska A, Niedzwiecka A,
Dadlez M, Stepinski J, Augustyniak W et al. (2003). RNA 9: 52–61.
