Regulation of cyclin D1 expression by mTORC1 signaling ... - Nature

Oncogene (2008) 27, 1106–1113

& 2008 Nature Publishing Group All rights reserved 0950-9232/08 $30.00 www.nature.com/onc

ORIGINAL ARTICLE

Regulation of cyclin D1 expression by mTORC1 signaling requires eukaryotic initiation factor 4E-binding protein 1 J Averous1, BD Fonseca and CG Proud Department of Biochemistry and Molecular Biology, University of British Columbia, Vancouver, BC, Canada

There is currently substantial interest in the regulation of cell function by mammalian target of rapamycin (mTOR), especially effects linked to the rapamycin-sensitive mTOR complex 1 (mTORC1). Rapamycin induces G1 arrest and blocks proliferation of many tumor cells, suggesting that the inhibition of mTORC1 signaling may be useful in cancer therapy. In MCF7 breast adenocarcinoma cells, rapamycin decreases levels of cyclin D1, without affecting cytoplasmic levels of its mRNA. In some cell–types, rapamycin does not affect cyclin D1 levels, whereas the starvation for leucine (which impairs mTORC1 signaling more profoundly than rapamycin) does. This pattern correlates with the behavior of eukaryotic initiation factor 4E-binding protein 1 (4E-BP1, an mTORC1 target that regulates translation initiation). siRNA-mediated knockdown of 4E-BP1 abrogates the effect of rapamycin on cyclin D1 expression and increases the polysomal association of the cyclin D1 mRNA. Our data identify 4E-BP1 as a key regulator of cyclin D1 expression, indicate that this effect is not mediated through the changes in cytoplasmic levels of cyclin D1 mRNA and suggest that, in some cell types, interfering with the amino acid input to mTORC1, rather than using rapamycin, may inhibit proliferation. Oncogene (2008) 27, 1106–1113; doi:10.1038/sj.onc.1210715; published online 27 August 2007 Keywords: rapamycin; translation initiation; 4E-BP1; cell cycle; cancer

Introduction The heightened interest in signaling through mammalian target of rapamycin (mTOR) reflects its role in many cellular functions and in human diseases including cancer (Easton and Houghton, 2006; Mamane et al., 2006; Sabatini 2006). This protein kinase occurs as two types of complex, that is, mTORC1 and mTORC2 (Wullschleger et al., 2006). So far, mTOR’s Correspondence: CG Proud, Department of Biochemistry and Molecular Biology, University of British Columbia, Life Sciences Centre, 2350 Health Sciences Mall, Vancouver, BC V6T 1Z3, Canada. E-mail: [email protected] 1 Current address: UMR 1019, Unite´ de Nutrition Humaine, INRA de Theix, 63122 Saint Genes Champanelle, France. Received 7 May 2007; revised 26 June 2007; accepted 1 July 2007; published online 27 August 2007

best-understood roles are those mediated by mTORC1, which is sensitive to the immunosuppressant drug rapamycin. mTORC1 signaling is activated by signaling via phosphatidylinositide 3-kinase and protein kinase B (PKB, also termed Akt) and by the Ras/Raf/ERK pathway (Corradetti and Guan, 2006). Because of its links to oncogenes and tumor suppressors, such as the lipid phosphatase PTEN, mTORC1 signaling is constitutively active in many tumor cells and their proliferation is inhibited by rapamycin (Easton and Houghton, 2006; Sabatini 2006). Rapamycin and its analogs appear to be good candidates for the use in cancer therapy, judging from their abilities to inhibit cell proliferation and/or induce apoptosis in certain tumor cells, and are currently in clinical trials (Easton and Houghton, 2006). However, there is as yet little information on how inhibition of mTORC1 signaling exerts its ‘anticancer effects’. This is important both for understanding mTORC1’s role in transformation and for developing new anticancer therapies. The best-described targets for mTOR are the ribosomal protein S6 kinases (S6Ks) and eukaryotic initiation factor 4E-binding protein 1 (4E-BP1, Raught et al., 2001). Hypophosphorylated 4E-BP1 binds eIF4E and so inhibits the formation of translation initiation complexes (eIF4F). eIF4F contains eIF4E (the cap-binding protein), eIF4G (the scaffold protein) and eIF4A (an RNA helicase). This complex is probably especially critical for the translation of mRNAs with extensive secondary structure in their 50 -untranslated regions (50 -UTRs). Artificial overexpression of eIF4E leads to transformation in cell lines (Lazaris-Karatzas and Sonenberg, 1992) and in animal models (Ruggero et al., 2004). Interestingly, eIF4E is expressed at high levels in many tumors (De Benedetti and Graff, 2004). Since enhanced expression of eIF4E increases eIF4F levels, eIF4E may transform cells by promoting the translation of specific mRNAs, for example, VEGF, cyclin D1 and cmyc (Mamane et al., 2006). This mechanism may explain the effect of mTOR inhibition on cell proliferation/ transformation, since blocking mTOR signaling should decrease eIF4F levels. However, there is so far little evidence that the inhibition of eIF4F formation by hypophosphorylated 4E-BP1 decreases expression of these proteins. mTOR promotes the phosphorylation of 4E-BP1 at several sites leading to its release from eIF4E, allowing

4E-BP1 couples mTORC1 to cyclin D1 expression J Averous et al

1107

eIF4F formation. Phosphorylation at Ser65 and Thr70 modulates the binding of 4E-BP1 to eIF4E directly. Their phosphorylation depends upon 4E-BP1’s C-terminal TOS motif that binds raptor, a component of mTORC1. Phosphorylation of Thr37/46 is required for the modification of Thr70 and Ser65, reflecting the hierarchical phosphorylation of 4E-BP1 (see Gingras et al., 2001), and depends upon 4E-BP1’s N-terminal RAIP motif (Tee and Proud, 2002). Phosphorylation of Thr37/46 is profoundly inhibited by starving cells of amino acids, which inactivates mTOR signaling (Wang et al., 2005). Although eIF4F is thought to play a key role in translation initiation, rapamycin generally only weakly decreases global protein synthesis. Nonetheless, rapamycin strongly affects cell proliferation, cell growth and angiogenesis (Averous and Proud, 2006; Easton and Houghton, 2006; Mamane et al., 2006; Sabatini, 2006). One possible explanation is that mTORC1 only strongly regulates the translation of certain mRNAs. Here, we focused our attention on cyclin D1, a key regulator of the G1/S transition of the cell cycle. Overexpression of eIF4E increases cyclin D1 protein levels (Rosenwald et al., 1993). Conversely, rapamycin, which induces a G1/S arrest, inhibits the serum-induced increase in cyclin D1 protein (Albers et al., 1993; Hashemolhosseini et al., 1998). The control of cyclin D1 mRNA translation is complex, involving both cap-dependent and cap-independent mechanisms (Gera et al., 2004). Previous data suggested roles for S6Ks (Koziczak and Hynes, 2004) and eIF4E/4E-BP1

0

1

2

4

Rapa (h)

(Rosenwald et al., 1993; Guan et al., 2007) in the regulation of cyclin D1 expression. Here, we set out to clarify the roles of these mTORC1 targets in controlling cyclin D1 expression. We show that 4E-BP1 is absolutely required for rapamycin-induced suppression of cyclin D1 levels in MCF7 cells and that 4E-BP1 controls the association of its mRNA with polysomes. Results Rapamycin decreases cyclin D1 levels in MCF7 cells Rapamycin treatment of MCF7 cells for up to 4 h markedly decreased the cyclin D1 protein levels (Figure 1a) and caused the dephosphorylation of 4E-BP1 (manifested by a shift to faster-migrating, hypophosphorylated a/b species and loss of the g form, Figure 1b). However, there was little change in the overall phosphorylation of Thr37/46, which is rather insensitive to rapamycin (Wang et al., 2005). The dephosphorylation of 4E-BP1 led to its increased binding to eIF4E and decreased the association of eIF4E with eIF4G (Figure 1c). The fact that rapamycin did not affect cytoplasmic cyclin D1 mRNA levels suggests that the decrease in cyclin D1 protein may reflect impaired translation of its message. In HEK293 cells, cyclin D1 levels are decreased by leucine starvation, but not by rapamycin HEK293 cells have extensively been used to study mTORC1 signaling. In the presence of serum, rapamycin 0

1

2

4

Rapa (h) γ β α

Cyclin D1

4E-BP1 (P)37-46

Tubulin

120

Rapa (h) eIF4E 4E-BP1 eIF4G

0

2h rapa

4E-BP1

100

320

eIF4G

100

6

Relative amount of cyclin D1 mRNA/ beta actin mRNA

2

m7GTP pull down

0

100 80 60 40 20 0 0

1 2 Rapa (h)

4

Figure 1 Rapamycin decreases cyclin D1 expression and eIF4F formation in MCF7 cells. (a) MCF7 cells were treated with 100 nM rapamycin for the indicated times. Cell lysates were analysed by sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS– PAGE)/western blotting using antisera for cyclin D1 and tubulin (as loading control). (b) As shown in (a), western blotting using antisera for phospho-4E-BP1 (Thr37/46). 4E-BP1 runs as several distinct bands (a–g; indicated by arrowheads), the fastest moving one (a) being the least phosphorylated. (c) MCF7 cells were treated with 100 nM rapamycin for 2 h. Proteins isolated by affinity purification on m7GTP-Sepharose were analysed by SDS–PAGE/western blotting using the indicated antisera. Quantification of the data is presented below: data are normalized to the signal for eIF4E (arbitrary units, time zero ¼ 100). (d) As shown in (a), cytoplasmic RNA was used to perform real-time RT–PCR experiments using primers for cyclin D1 orb-actin. Expression of cyclin D1 mRNA is normalized to b-actin mRNA levels and expressed relative to the value at zero time (arbitrary units, n ¼ 3). Oncogene


1108

only slightly decreased the phosphorylation of several sites in 4E-BP1, but did cause the complete dephosphorylation of S6, another target of mTORC1 (Figure 2a), and the inactivation of the S6Ks (data not shown). Rapamycin had no apparent effect on the association of eIF4E with 4E-BP1 and little effect on eIF4E/eIF4G binding (Figure 2b). The single most effective amino acid at regulating mTORC1 is leucine (Kimball and Jefferson, 2006). Leucine starvation led to a marked dephosphorylation of Thr70/Ser65 and a modest dephosphorylation of Thr37/46 (Figure 2a). Leucine starvation increased eIF4E/ 4E-BP1 binding and decreased eIF4E/eIF4G complexes

Ctl

-Leu

(Figure 2b). Rapamycin treatment did not decrease cyclin D1 protein levels (Figure 2c), but leucine starvation did (Figure 2d). Conversely, re-supplying leucine led to a rapid re-accumulation of cyclin D1 (Figure 2e). Leucine starvation regulates the transcription or stability of several mRNAs (Bruhat et al., 1997; Peng et al., 2002). However, cytoplasmic cyclin D1 mRNA levels were unchanged even after 4 h leucine starvation (Figure 2f), although cyclin D1 protein levels had decreased substantially (Figure 2d). Clearly, leucine is a potent regulator of both cyclin D1 expression and 4E-BP1 phosphorylation in HEK293

m7GTP pull down

Rapa γ 4E -BP1 β (P)37-46 α * 4E-BP1 (P)65

Ctl

-Leu

Rapa eIF4E

4E-BP1

4E-BP1 (P)70

eIF 4G

S6 (P)240/244

Ctl

S6

0

1

2

4

+Rapa

4E-BP1 100

169

94

100

ND

54

eIF4G Rapa (h)

-Leu

0

1

2

4

Cyclin D1

Cyclin D1

Tubulin

1

2

4

8 +

Leu (h) Cyclin D1

Tubulin T0

Relative amount of cyclin D1 mRNA/ β-actin mRNA

Tubulin

0

-Leu (h)

160 140 120 100 80 60 40 20 0

0

1

2

4

Time (h) Figure 2 Leucine starvation decreases cyclin D1 expression and eIF4F formation in HEK293 cells. (a) HEK293 cells were starved of leucine or treated with 100 nM rapamycin for 2 h. Samples of total protein extract (30 mg) were analysed by sodium dodecyl sulfate– polyacrylamide gel electrophoresis (SDS–PAGE)/western blotting using the indicated antisera. Arrows labeled a–g denote differentially phosphorylated forms of 4E-BP1, while the starred arrow indicates 4E-BP2, which cross-reacts with the (P)Thr37/46 antibody. (b) As shown in (a). Proteins isolated on m7GTP-Sepharose were analysed by SDS–PAGE/western blotting using the indicated antisera. Quantification of the data is presented below, normalized to the signal for eIF4E (arbitrary units, t0 ¼ 100). (c, d) HEK293 cells were treated with 100 nM rapamycin (c) or starved of leucine (d) for the times indicated. Total protein (30 mg) was analysed by SDS–PAGE/western blotting for cyclin D1 and tubulin (loading control). (e) HEK293 cells were starved of leucine for 16 h. Leucine was re-added and cells were harvested at the indicated times. Samples of total protein (30 mg) were analysed by SDS–PAGE/western blotting as indicated. (f) HEK293 cells were starved of leucine for the times indicated. Cytoplasmic RNA was used to perform real-time RT–PCR experiments using primers for cyclin D1 orb-actin. Expression of cyclin D1 mRNA was normalized to b-actin mRNA levels (arbitrary units, t0 ¼ 100; n ¼ 3). Oncogene


1109 siCTL

cells, whereas rapamycin is not. Since rapamycin does strongly impair S6K activity, S6Ks do not contribute significantly to regulating cyclin D1 expression in these cells. The above data suggest that 4E-BP1 contributes to mTOR-dependent control of cyclin D1 expression. To determine whether binding of 4E-BP1 to eIF4E does indeed regulate cyclin D1 expression, we overexpressed two forms of 4E-BP1 in HEK293 cells: a wild-type 4E-BP1 and a mutant (AAAA/F113A), which is mutated in both the RAIP and the TOS motifs. This mutant cannot be phosphorylated and binds ‘constitutively’ to eIF4E (Tee and Proud, 2002; Wang et al., 2005). Similar levels of eIF4E/4E-BP1 complexes were seen in serum-grown cells expressing the two different 4E-BP1 species (Figure 3a). Wild-type and mutant 4E-BP1 each decreased cyclin D1 protein levels (Figure 3b).

0

1

siBP1

2

4

0

1

2

4

Rapa (h) γ α

m7GTP pull-down

siCTL 0

siBP1

2

0

2

Rapa (h)

m7GTP-bound material

eIF4G

4E-BP1

eIF4G

100

10

210

96

4E-BP1 100

225

18

88

siCTL 0

1

2

siBP1 4

0

1

2

4

Rapa (h) Cyclin D1

Tubulin 100 101 58

49 130 119 115 132 Cyclin D1 (t0= 100)

Myc-4E-BP1 AAAA/ F113A

Empty vector

Myc-4E-BP1 WT

Figure 4 Knocking-down 4E-BP1 abrogates the inhibitory effect of rapamycin on cyclin D1 expression. MCF7 cells were transfected with control siRNA or siRNA directed against 4E-BP1. Cells were treated with 100 nM rapamycin for the indicated times. (a) Total protein (30 mg) was analysed by sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS–PAGE)/western blotting for 4E-BP1. (b) Proteins isolated by affinity purification on m7GTPSepharose were analysed by SDS–PAGE/western blotting using the indicated antisera. Quantification of the data is presented below: data were normalized to eIF4E (arbitrary units, t0 ¼ 100). (c) Total protein (30 mg) was analysed by SDS–PAGE/western blotting for cyclin D1 and tubulin, as loading control (n ¼ 4).

Myc-4E-BP1 AAAA/ F113A

Myc-4E-BP1 WT

Empty vector

4E-BP1 controls the association of the cyclin D1 mRNA with polysomes To assess whether 4E-BP1 actually regulates the initiation of translation of the cyclin D1 mRNA, we examined

4E-BP1

Tubulin

eIF4E

Knocking down 4E-BP1 abolishes the effect of rapamycin on cyclin D1 levels To address directly the contribution of 4E-BP1 to controlling cyclin D1 expression, we knocked down its expression in MCF7 cells by siRNA. This very efficiently decreased 4E-BP1 levels (Figure 4a). Cells showed no morphological changes and proliferated normally. In 4E-BP1 knock-down cells, eIF4G remained bound to eIF4E even after rapamycin treatment (Figure 4b). A trace of 4E-BP1 bound to eIF4E under these conditions, similar to the amount seen in control cells without rapamycin (Figure 4b). This likely explains why rapamycin caused a small decrease in eIF4G/eIF4E binding in the knock-down cells (Figure 4b). Knocking down 4E-BP1 eliminated the rapamycininduced decrease in cyclin D1 protein levels, even after 4 h rapamycin treatment (Figure 4c), suggesting that in MCF7 cells 4E-BP1 is the main regulator of cyclin D1 expression in response to inhibition of mTORC1.

β

Myc-4E-BP1

Cyclin D1

eIF4E

Tubulin

Figure 3 4E-BP1 overexpression decreases cyclin D1 levels. HEK293 cells were transfected with the indicated vectors (AAAA/F113A mutant of 4E-BP1 lacks regulatory motifs essential for its phosphorylation). Cells were harvested, 24 h later. (a) Proteins isolated on m7GTPSepharose were analysed by sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS–PAGE)/western blotting; top section: antisera for 4E-BP1 (the band shown corresponds to myc-4E-BP1), bottom section: antisera for eIF4E (endogenous and exogenous forms indicated). (b) Samples of total protein (30 mg) were analysed by SDS–PAGE/western blotting for cyclin D1 and tubulin (n ¼ 2). Oncogene


1110

siRNA control - rapamycin

siRNA control + rapamycin

siRNA 4E-BP1 - rapamycin

siRNA 4E-BP1 + rapamycin

40/60/80S

40S 60S 80S

A254

sedimentation

Polysomes

polysomes

1800

mRNA amount, arbitrary unit

1600

SI CTL CTL SI CTL RAP SI BP CTL SI BP RAP

1400 1200 1000 800 600 400 200 0 2/3

4/5

6/7

8/9

10/11

12/13

14/15

16/17

Figure 5 Knocking-down 4E-BP1 increases the polysomal association of cyclin D1 mRNA. MCF7 cells, or MCF7 cells in which 4EBP1 had been knocked down by siRNA, were treated with rapamycin where indicated (100 nM, 2 h). (a) Cell lysates were subjected to sucrose density centrifugation. Absorbance (254 nm) of the displaced gradient was monitored and fractions were collected. Positions of 40, 60 and 80S ribosomal particles and polysomes are indicated. (b) Fractions (paired as indicated) were analysed by real-time RT–PCR for cyclin D1 mRNA.

polysomal levels of this message in MCF7 cells in which 4E-BP1 expression had been knocked down. Ribosomes/ polysomes were resolved on sucrose density gradients (Figure 5a), fractions were collected and cyclin D1 mRNA levels were assessed by real-time PCR. Knocking down 4E-BP1 had no effect on the overall proportion of ribosomes in polysomes (Figure 5a) but markedly increased the levels of polysome-associated cyclin D1 mRNA. This is, therefore, a selective effect on the cyclin D1 mRNA. Total amounts of cytoplasmic cyclin D1 mRNA were similar under all conditions (not shown). The increase in polysomal cyclin D1 mRNA was accompanied by a modest increase (30%) in cyclin D1 protein (Figure 4c). Thus, 4E-BP1 normally restricts the association of the cyclin D1 mRNA with polysomes in the serum-fed cells. In control cells, rapamycin decreased the amount of polysomes and caused a concomitant increase in inactive 80S ribosomes. This effect was slightly decreased in 4EBP1 knock-down cells (Figure 5a). Surprisingly, rapamycin treatment induced a rightward shift of the cyclin D1 mRNA in the polysomal region in control and 4EBP1 knock-down cells (Figure 5b). Strikingly, following Oncogene

rapamycin treatment, more cyclin D1 mRNA remained in the polysomal region in 4E-BP1 knock-down cells than in controls (Figure 5b). These data illustrate the complexity of the regulation of cyclin D1 mRNA translation, which likely includes cap-dependent and cap-independent mechanisms. Importantly, they reinforce the idea that 4E-BP1 plays a crucial role in controlling cyclin D1 expression. Discussion The inhibition of cyclin D1 expression by rapamycin was previously attributed to enhanced mRNA and protein degradation (Hashemolhosseini et al., 1998). The effect on mRNA levels was modest, but that on cyclin D1 protein turnover was larger. Those experiments were mainly performed in the serum-starved cells, while we used the serum-fed cells. Stable overexpression of a mutant 4E-BP1, which binds constitutively to eIF4E, is associated with the decreases in cyclin D1 protein and mRNA (Jiang et al., 2003). However, that approach provides no information on the role of


1111

endogenous 4E-BP1 in regulating cyclin D1 in response to impaired mTORC1 signaling. Here, we show that knocking-down 4E-BP1 markedly increases the polysomal association of the cyclin D1 mRNA and eliminates the effect of rapamycin treatment on cyclin D1 levels. However, this differs from the conclusions of a study performed in MDA-MB-453 cells (Koziczak and Hynes, 2004). The elimination of S6K expression by siRNA modestly decreased (20–30%) cyclin D1 expression. S6K associates with translation initiation complexes and phosphorylates eIF4B (Holz et al., 2005; Shahbazian et al., 2006), which might explain these observations. However, our study suggests that S6K is not a primary regulator of cyclin D1 expression: (i) rapamycin treatment eliminates S6K activity in HEK293 cells but does not affect cyclin D1 levels and (ii) elimination of 4E-BP1 is enough to abolish the effect of rapamycin on cyclin D1 expression. eIF4E can mediate nuclear export of the cyclin D1 mRNA (Rousseau et al., 1996). Entry of eIF4E into the nucleus appears to be mediated by 4E-T (Dostie et al., 2000), which contains a similar eIF4E-binding motif to that in 4E-BP1. Therefore, binding of 4E-BP1 to eIF4E could prevent eIF4E entering the nucleus, potentially affecting the transport of the cyclin D1 mRNA into the cytoplasm. However, we find that interfering with mTORC1 signaling does not alter cytoplasmic levels of cyclin D1 mRNA. This strongly suggests that the mechanism by which 4E-BP1 or mTORC1 inhibition impairs cyclin D1 expression does not involve decreased mRNA transport. An alternative model is that the inhibition of translation of the cyclin D1 mRNA is involved, for example, due to the loss of eIF4F complexes. Our data support this idea, since knocking down 4E-BP1 enhances the polysomal association of the cyclin D1 mRNA without increasing overall polysome levels. Furthermore, knocking-down 4E-BP1 essentially eliminated the rapamycin-induced decrease in cyclin D1 protein levels. Surprisingly, rapamycin caused a shift of the cyclin D1 mRNA into a faster-sedimenting material (in control and 4E-BP1 knock-down cells). The reason for this is unclear, but control of cyclin D1 mRNA translation is clearly complex. For example, the 50 -UTR of the cyclin D1 mRNA apparently possesses an internal ribosome entry site, which is particularly active in cells showing low Akt activity (Gera et al., 2004). Further work, beyond the scope of this study, is required to clarify the mechanisms controlling the cyclin D1 mRNA translation. It is important to note that in serum-fed HEK293 cells, rapamycin does not efficiently elicit the formation of inactive eIF4E/4E-BP1 complexes, even though it blocks other mTORC1 functions (for example, S6 phosphorylation). This could explain why the proliferation of some tumor cells is resistant to rapamycin. In such cases, specifically targeting the mechanism by which 4E-BP1 phosphorylation is regulated (for example, by amino acids) might be more effective than rapamycin. Further work is required to achieve a better understanding of the regulation of 4E-BP1.

Alternatively, inhibiting eIF4E/4G binding may be a valuable approach: a small molecule inhibitor of the eIF4E/eIF4G interaction was recently shown to decrease the expression of ‘weak’ mRNAs such as c-myc and inhibit the proliferation of cancer cell lines (Moerke et al., 2007). These points are of particular importance given the substantial interest in the roles of eIF4E and mTOR signaling in cancer. In summary, this study shows that 4E-BP1 plays a key role in coupling cyclin D1 expression to mTORC1 signaling. This is consistent with the findings that the overexpression of a ‘constitutively-active’ form of 4EBP1 in MCF7 cells led to cell cycle arrest and decreased the expression of cyclin D1 (Jiang et al., 2003) and that protein kinase C-mediated inhibition of cyclin D1 expression involves the activation (dephosphorylation) of 4E-BP1 (Hizli et al., 2006). Other mRNAs may be regulated in a similar way to the cyclin D1 mRNA, that is, their identification is an important goal.

Materials and methods The procedures for cell culture, transfection and lysis, and for sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS–PAGE) and immunoblotting are described in the Supplementary data. Isolation of proteins in m7GTP-Sepharose m7GTP-Sepharose 4B (Amersham Biosciences Inc., Buckinghamshire, UK) was mixed with two volumes of Sepharose CL-4B. Twenty microliters of the mixture was added to 0.2– 0.5 mg of cell lysate. The lysates were then mixed with the beads for 1 h at 41C and the beads were pelleted by centrifugation at 1000 g for 2 min. The beads were then washed three times in the extraction buffer. Proteins were removed from the m7GTP matrix by boiling in SDS loading buffer and the supernatant analysed by SDS–PAGE. Sucrose density gradient Cells were lysed with the following buffer: 50 mM HEPES (pH 7), 7 mM MgCl2, 100 mM KCl, 0.1% Triton X-100. Lysates were then centrifuged for 10 min at 16 000 g. A fraction of the supernatant (3 mg of protein) was applied to a 20–50% (w/v) sucrose gradient (w/w) prepared in 50 mM HEPES (pH 7), 2 mM MgCl2, 100 mM KCl, which was then centrifuged at 110 000 g for 3 h in a Beckman SW-41Ti rotor. Each gradient was fractionated using a Brandel model 184 fractionator using upward displacement by 60% (w/v) sucrose. The A254 of the displaced gradient was analysed in an ISCO Type 6 Optical Unit. Fractions of 0.5 ml were collected. RNA extraction RNA was purified from cytoplasmic extract or sucrose density gradient fractions. SDS and proteinase K were added at the respective final concentrations of 1% and 50 mg/ml. Samples were then incubated for 30 min at 371C. One volume of phenol:chloroform:isoamylalcohol (25:24:1) was added, the samples were centrifuged for 10 min at 16 000 g. The aqueous phase was mixed with 0.2 volume of 7.5 M ammonium acetate and 2 volume of absolute ethanol before precipitation for 1 h at 201C. Samples were centrifuged for 15 min at 16 000 g at 41C. Pellets were washed with 70% ethanol and resuspended in water. Oncogene


1112 Reverse transcription real-time PCR Reverse transcription and real-time PCR were performed as previously described (Averous et al., 2005). Primers for human cyclin D1 were forward primer 50 -ACCTGGATGCTGGA GGTCT; reverse primer, 50 -GCTCCATTTGCAGCAGCTC. Primers for human b-actin were forward primer, 50 -TCC CTGGAGAAGAGCTACGA; reverse primer, 50 -AGCA CTGTGTTGGCGTACAG.

4E-BP1, eIF4E-binding protein 1; eIF, eukaryotic initiation factor; m7GTP, 7-methyl GTP; mTOR, mammalian target of rapamycin; mTORC, mTOR complex; S6K, S6 kinase.

siRNA Stealth RNAi (Invitrogen, Carlsbad, CA, USA) directed against 4E-BP1 (sequence 50 -UCUAUGACCGGAAAUUCCUGAUG GA) was transfected at 40 nM. A Stealth RNAi with the same nucleotide composition was used as negative control (sequence 50 -UCUCCAGAAGGCUUAAGUCUUAGGA).

We thank Dr Maria Buxade´ and Dr Josep Parra for helpful discussions and advice. We are grateful to Dr Herman Ziltener and Dr Klaus Gossens for access to the Light-Cycler. This work was supported by funding from AstraZeneca. BDF gratefully acknowledges support from the Morton Trust/Dundee School of Life Sciences Alumnus Fund.

Abbreviations

Acknowledgements

References Albers MW, Williams RT, Brown EJ, Tanaka A, Hall FL, Schreiber SL. (1993). FKBP-rapamycin inhibits a cyclindependent kinase activity and a cyclin D1-Cdk association in early G1 of an osteosarcoma cell line. J Biol Chem 268: 22825–22829. Averous J, Maurin AC, Bruhat A, Jousse C, Arliguie C, Fafournoux P. (2005). Induction of IGFBP-1 expression by amino acid deprivation of HepG2 human hepatoma cells involves both a transcriptional activation and an mRNA stabilization due to its 30 UTR. FEBS Lett 579: 2609–2614. Averous J, Proud CG. (2006). When translation meets transformation: the mTOR story. Oncogene 25: 6423–6435. Bruhat A, Jousse C, Wang XZ, Ron D, Ferrara M, Fafournoux P. (1997). Amino acid limitation induces expression of CHOP, a CCAAT/enhancer binding proteinrelated gene, at both transcriptional and post-transcriptional levels. J Biol Chem 272: 17588–17593. Corradetti MN, Guan KL. (2006). Upstream of the mammalian target of rapamycin: do all roads pass through mTOR? Oncogene 25: 6347–6360. De Benedetti A, Graff JR. (2004). eIF-4E expression and its role in malignancies and metastases. Oncogene 23: 3189–3199. Dostie J, Ferraiuolo M, Pause A, Adam SA, Sonenberg N. (2000). A novel shuttling protein, 4E-T, mediates the nuclear import of the mRNA 50 cap-binding protein, eIF4E. EMBO J 19: 3142–3156. Easton JB, Houghton PJ. (2006). mTOR and cancer therapy. Oncogene 25: 6436–6446. Gera JF, Mellinghoff IK, Shi Y, Rettig MB, Tran C, Hsu JH et al. (2004). AKT activity determines sensitivity to mammalian target of rapamycin (mTOR) inhibitors by regulating cyclin D1 and c-myc expression. J Biol Chem 279: 2737–2746. Gingras A-C, Raught B, Gygi SP, Niedzwieka A, Miron M, Burley SK et al. (2001). Hierarchical phosphorylation of the translation inhibitor 4E-BP1. Genes Dev 15: 2852–2864. Guan L, Song K, Pysz MA, Curry KJ, Hizli AA, Danielpour D et al. (2007). Protein kinase C-mediated down-regulation of cyclin D1 involves activation of the translational repressor 4E-BP1 via a phosphoinositide 3-kinase/ Akt-independent, protein phosphatase 2A-dependent mechanism in intestinal epithelial cells. J Biol Chem 282: 14213–14225. Hashemolhosseini S, Nagamine Y, Morley SJ, Desrivieres S, Mercep L, Ferrari S. (1998). Rapamycin inhibition of Oncogene

the G1 to S transition is mediated by effects on cyclin d1 mRNA and protein stability. J Biol Chem 273: 14424–14429. Hizli AA, Black AR, Pysz MA, Black JD. (2006). Protein kinase C alpha signaling inhibits cyclin D1 translation in intestinal epithelial cells. J Biol Chem 281: 14596–14603. Holz MK, Ballif BA, Gygi SP, Blenis J. (2005). mTOR and S6K1 mediate assembly of the translation preinitiation complex through dynamic protein interchange and ordered phosphorylation events. Cell 123: 569–580. Jiang H, Coleman J, Miskimins R, Miskimins WK. (2003). Expression of constitutively active 4EBP-1 enhances p27Kip1 expression and inhibits proliferation of MCF7 breast cancer cells. Cancer Cell Int 3: 2. Kimball SR, Jefferson LS. (2006). Signaling pathways and molecular mechanisms through which branched-chain amino acids mediate translational control of protein synthesis. J Nutr 136: 227S–231S. Koziczak M, Hynes NE. (2004). Cooperation between fibroblast growth factor receptor-4 and ErbB2 in regulation of cyclin D1 translation. J Biol Chem 279: 50004–50011. Lazaris-Karatzas A, Sonenberg N. (1992). The mRNA 50 capbinding protein, eIF-4E, cooperates with v-myc or E1A in the transformation of primary rodent fibroblasts. Mol Cell Biol 12: 1234–1238. Mamane Y, Petroulakis E, Le Bacquer O, Sonenberg N. (2006). mTOR, translation initiation and cancer. Oncogene 25: 6416–6422. Moerke NJ, Aktas H, Chen H, Cantel S, Reibarkh MY, Fahmy A et al. (2007). Small-molecule inhibition of the interaction between the translation initiation factors eIF4E and eIF4G. Cell 128: 257–267. Peng T, Golub TR, Sabatini DM. (2002). The immunosuppressant rapamycin mimics a starvation-like signal distinct from amino acid and glucose deprivation. Mol Cell Biol 21: 5575–5584. Raught B, Gingras A-C, Sonenberg N. (2001). The target of rapamycin (TOR) proteins. Proc Natl Acad Sci USA 98: 7037–7044. Rosenwald IB, Lazaris-Karatzas A, Sonenberg N, Schmidt EV. (1993). Elevated levels of cyclin D1 protein in response to increased expression of eukaryotic initiation factor 4E. Mol Cell Biol 13: 7358–7363. Rousseau D, Kaspar R, Rosenwald I, Gehrke L, Sonenberg N. (1996). Translation initiation of ornithine decarboxylase and nucleocytoplasmic transport of cyclin D1 messenger-RNA


1113 are increased in cells overexpressing eukaryotic initiation factor 4E. Proc Natl Acad Sci USA 93: 1065–1070. Ruggero D, Montanaro L, Ma L, Xu W, Londei P et al. (2004). The translation factor eIF-4E promotes tumor formation and cooperates with c-Myc in lymphomagenesis. Nat Med 10: 484–486. Sabatini DM. (2006). mTOR and cancer: insights into a complex relationship. Nat Rev Cancer 6: 729–734. Shahbazian D, Roux PP, Mieulet V, Cohen MS, Raught B, Taunton J et al. (2006). The mTOR/PI3K and MAPK pathways converge on eIF4B to control its phosphorylation and activity. EMBO J 25: 2781–2791.

Tee AR, Proud CG. (2002). Caspase cleavage of initiation factor 4E-binding protein 1 yields a dominant inhibitor of cap-dependent translation and reveals a novel regulatory motif. Mol Cell Biol 22: 1674–1683. Wang X, Beugnet A, Murakami M, Yamanaka S, Proud CG. (2005). Distinct signaling events downstream of mTOR cooperate to mediate the effects of amino acids and insulin on initiation factor 4E-binding proteins. Mol Cell Biol 25: 2558–2572. Wullschleger S, Loewith R, Hall MN. (2006). TOR signaling in growth and metabolism. Cell 124: 471–484.

Supplementary Information accompanies the paper on the Oncogene website (http://www.nature.com/onc).

Oncogene