4E-BP1, a repressor of mRNA translation,
is phosphorylated and inactivated
by the Akt(PKB) signaling pathway
Anne-Claude Gingras,1,4Scott G. Kennedy,2,3,4Maura A. O’Leary,3Nahum Sonenberg,1and
1Department of Biochemistry, McGill University Montreal, Quebec, Canada H3G 1Y6; and2Department of Pharmacological
and Physiological Sciences and3T he Ben May Institute for Cancer Research, T he University of Chicago,
Chicago, Illinois 60637 USA
Growth factors and hormones activate protein translation by phosphorylation and inactivation of the
translational repressors, the eIF4E-binding proteins (4E-BPs), through a wortmannin- and rapamycin-sensitive
signaling pathway. T he mechanism by which signals emanating from extracellular signals lead to
phosphorylation of 4E-BPs is not well understood. Here we demonstrate that the activity of the
serine/threonine kinase Akt/PKB is required in a signaling cascade that leads to phosphorylation and
inactivation of 4E-BP1. PI 3-kinase elicits the phosphorylation of 4E-BP1 in a wortmannin- and
rapamycin-sensitive manner, whereas activated Akt-mediated phosphorylation of 4E-BP1 is wortmannin
resistant but rapamycin sensitive. A dominant negative mutant of Akt blocks insulin-mediated
phosphorylation of 4E-BP1, indicating that Akt is required for the in vivo phosphorylation of 4E-BP1.
Importantly, an activated Akt induces phosphorylation of 4E-BP1 on the same sites that are phosphorylated
upon serum stimulation. Similar to what has been observed with serum and growth factors, phosphorylation
of 4E-BP1 by Akt inhibits the interaction between 4E-BP1 and eIF-4E. Furthermore, phosphorylation of 4E-BP1
by Akt requires the activity of FRAP/mT OR. FRAP/mT OR may lie downstream of Akt in this signaling
cascade. T hese results demonstrate that the PI 3-kinase-Akt signaling pathway, in concert with FRAP/mT OR,
induces the phosphorylation of 4E-BP1.
[Key Words: Protein synthesis; phosphorylation; PI 3-kinase; protein kinase B; eIF4E; FRAP/mT OR]
Received November 13, 1997; revised version accepted December 19, 1997.
Numerous cellular processes are controlled by extracel-
lular stimuli that activate signaling cascades. Many
stimuli activate common pathways, such as the well-
described Ras and phosphoinositide 3-kinase pathways.
Phosphoinositide 3-kinase (PI 3-kinase) is activated by
growth factor receptors after growth factor stimulation
and induces cell proliferation and cell survival (for re-
view, see Franke et al. 1997a; Vanhaesebroeck et al.
1997). Several downstream targets of PI 3-kinase have
been identified, including p70 ribosomal protein S6 ki-
nase (p70S6k) (Chou and Blenis 1995; Proud 1996), the
GT Pases Rac (Hawkins et al. 1995), certain protein ki-
nase C isoforms (Nakanishi et al. 1993; Akimoto et al.
1996), and the serine/threonine kinase Akt (also known
as protein kinase B-PKB) (Burgering and Coffer 1995;
Frankeet al. 1995). Upon activation by growth factors, PI
3-kinase phosphorylates the D3 position of phosphati-
dylinositols. T hese phospholipids act as second messen-
gers that mediate the diverse cellular functions of PI 3-
kinase, including activation of Akt (for review, see
Franke et al. 1997a; Hemmings 1997). Wortmannin, a PI
3-kinase inhibitor, blocks activation of Akt after stimu-
lation with growth factors, indicating that the activity of
PI 3-kinase is an obligatory step in Akt activation by
growth factors (Burgering and Coffer 1995; Franke et al.
1997b; Kennedy et al. 1997). T ranslocation of Akt to the
plasma membrane through its pleckstrin homology (PH)
domain is likely required for its activity (Hemmings
1997). Indeed, constitutivetargeting of Akt to theplasma
membrane is sufficient to promote its activation (Bur-
gering and Coffer 1995; Franke et al. 1997b; Kennedy et
al. 1997). Full activation of Akt by growth factors re-
quires the phosphorylation of threonine and serine resi-
dues by upstream kinases (Alessi et al. 1997; Stokoeet al.
1997). Both the upstream activating kinases and the re-
cruitment of Akt to theplasmamembranearethought to
be dependent on the products of PI 3-kinase (Cohen et al.
1997). Akt mediates some of the PI 3-kinase cellular re-
sponses, including protection from apoptosis induced by
4T hese authors made an equal contribution to the work.
5Corresponding author. Present address: Department of Molecular Ge-
netics, University of Illinois at Chicago, Chicago, Illinois 60607.
E-MAIL firstname.lastname@example.org; FAX (773) 702-6260.
502GENES & DEV ELOPMENT 12:502–513 © 1998 by Cold Spring Harbor Laboratory Press ISSN 0890-9369/98 $5.00; www.genesdev.org
serum and growth factor deprivation (Kennedy et al.
T he immunosuppressive drug rapamycin, an inhibitor
of G1cell cycle progression, blocks the activation of
p70S6k by growth factors and Akt (Chung et al. 1992;
Price et al. 1992; Burgering and Coffer 1995). Rapamycin
forms a complex with the immunophilin FKBP12 to gen-
erate a potent inhibitor of FRAP/mT OR (also termed
RAFT 1 or RAPT 1) (Brown and Schreiber 1996). FRAP/
mT OR activates p70S6k in vivo (Brown et al. 1995); it is
unclear, however, whether FRAP/mT OR is a down-
stream effector in thePI 3-kinaseandAkt signaling path-
way. FRAP, which is the mammalian homolog of the
Saccharomyces cerevisiae targets of rapamycin T OR1
and T OR2 (Hall 1996), is a member, together with AT M
and DNA–PK, of a recently characterized family of phos-
phatidylinositol kinases-related (PIK-related) kinases.
PIK-related kinases activation and mechanisms of action
remain unclear (Hoekstra 1997). FKBP rapamycin-
associated protein /mammalian target of rapamycin
(FRAP/mT OR) could provide a link between cell cycle
progression and the control of mRNA translation, as
rapamycin, which blocks thecell cyclein G1, also causes
a decrease in mRNA translation (Beretta et al. 1996;
Brown and Schreiber 1996). Consistent with this finding,
the yeast T OR has been demonstrated to regulate G1
progression through a translational mechanism (Barbet
et al. 1996).
Regulation of protein translation is an important as-
pect of the control of cell growth. A rate-limiting step in
this process is the binding of the mRNA to the small
ribosomal subunit, a step mediated by the eIF4 group of
initiation factors (for review, see Sonenberg 1996). eIF4F,
through its smaller subunit eIF4E, recognizes the cap
structure (m7GpppX, where X is any nucleotide) that is
present at the 5? end of all cellular, except organellar,
mRNAs. eIF4F, in conjunction with eIF4B, is thought to
unwind the secondary structure in the mRNA 5?-UT R to
facilitate ribosome binding (Sonenberg 1996). Overex-
pression of eIF4E in rodent cells leads to cellular trans-
formation and eIF4E has been implicated in cell cycle
control (Lazaris-Karatzas et al. 1990; Sonenberg 1996). In
addition, a role for eIF4E in cell survival has been pro-
posed, as NIH 3T 3 cells that express eIF4E ectopically
are refractory to apoptosis induced by serum deprivation
(Polunovsky et al. 1996). eIF4E is the target of a family of
translational repressors termed the 4E-BPs (for eIF4E-
Binding Proteins; also known as PHAS). T heserepressors
bind to eIF4E and prevent its association with eIF4G and
incorporation into the eIF4F complex, which leads to
inhibition of cap-dependent, but not cap-independent,
translation (Sonenberg 1996). Overexpression of 4E-BP1
or 4E-BP2 in cells transformedby eIF4E, Ha-v-ras or v-src
partially reverts their transformed phenotypes (Rousseau
et al. 1996).
Inhibition of translation by 4E-BPs is reversible. After
treatment of cells with serum, growth factors, or hor-
mones, 4E-BP1 (the prototype member of the family), is
hyperphosphorylated in a wortmannin- and rapamycin-
sensitive manner, and dissociates from eIF4E (Beretta et
al. 1996; von Manteuffel et al. 1996, 1997). T he rapid
increase in 4E-BP1 phosphorylation after serum or
growth factor stimulation provides a very attractive
mechanism for explaining the increase in translation
rates of several mRNAs after stimulation.
Because phosphorylation of 4E-BP1 is wortmannin
sensitive, and mutants in the PDGF receptor that fail to
activate PI 3-kinase also fail to phosphorylate 4E-BP1
after PDGF treatment (Beretta et al. 1996; von Man-
teuffel et al. 1996), it was suggested that phosphorylation
of 4E-BP1 by serum and growth factors is mediated by PI
3-kinase. However, it is not clear whether PI 3-kinase
lies directly upstream of 4E-BP1 in a phosphorylation
cascade. T his is an important question, particularly in
light of a recent report demonstrating that wortmannin
can inhibit FRAP/mT OR activity directly (Brunn et al.
1996). Here we provide direct evidence that PI 3-kinase
and its downstream effector Akt lie in a pathway leading
to the in vivo phosphorylation of 4E-BPs. T his phos-
phorylation is sensitive to rapamycin. T he rapamycin
sensitivity can be overridden by coexpression of a rapa-
mycin-resistant mutant of FRAP/mT OR. T hus, FRAP/
mT OR may lie downstream of Akt in the 4E-BP1 phos-
P110?, the catalytic subunit of PI 3-kinase,
and its downstream effector Akt/PKB mediate
the phosphorylation of 4E-BP1
T o study the role of Akt in the phosphorylation of 4E-
BP1, a hemagglutinin-tagged 4E-BP1 (HA–4E-BP1) was
generated. Wefirst examined whether thetransiently ex-
pressed HA–4E-BP1 exhibits a change in electrophoretic
mobility after phosphorylation, as was observed for the
endogenous 4E-BP1. Human embryonic kidney (HEK)
293 cells were transfected transiently with a HA–4E-BP1
expression vector. After transfection, the cells were de-
prived of serum for 36 hr and then stimulated with in-
sulin for 30 min. Immunoblot analysis demonstrated a
clear shift in mobility of HA–4E-BP1 with insulin stimu-
lation (Fig. 1A, lanes 1,2). T he mobility shift was not
observed when cells were preincubated with either wort-
mannin or rapamycin (lanes 3,4), consistent with what
has been observed previously for endogenous 4E-BP1
(von Manteuffel et al. 1996).
Previous studies have indicated a role for PI-3-kinase
in the phosphorylation of 4E-BP1 by serum and growth
factors (Beretta et al. 1996; von Manteuffel et al. 1996).
However, it was also suggested that the effects of extra-
cellular stimuli on 4E-BP1 phosphorylation could be ex-
plained by direct activation of FRAP, as the in vitro au-
tokinase activity of FRAP is also inhibited by wortman-
nin (Brunn et al. 1996). T o examine whether PI 3-kinase
can affect the phosphorylation state of 4E-BP1, we tran-
siently cotransfected HA-4E-BP1 and PI 3-kinase expres-
sion vectors into serum-deprived 293 cells. Cotransfec-
tion of HA-4E-BP1 with the catalytic subunit of PI 3-
kinase p110? induced phosphorylation of 4E-BP1, as
manifestedby ashift in its mobility (Fig. 1B, lane2). T his
Akt regulates 4E-BP1 phosphorylation
GENES & DEV ELOPMENT503
shift in mobility is similar to that observed with insulin
stimulation (cf. Fig. 1A, lane 2, with Fig. 1B, lane 2). An
activated form of p110?, p110?caax (p110?*), which is
targeted to the plasma membrane by farnesylation
(Khwaja et al. 1997) (Fig. 1B, lane 3), also caused this
mobility shift (interestingly, overexpression of the wild-
type p110? in 293 cells is sufficient to induce 4E-BP1
mobility shift). T hus, PI 3-kinase by itself affects the
phosphorylation of 4E-BP1.
Next, we examined whether 4E-BP1 phosphorylation
is mediated by Akt. We used two forms of Akt: the wild
type c-Akt, and an activated form of Akt, MyrAkt.
MyrAkt is comprised of the entire coding sequence of
c-Akt fused in-frame to the Src myristoylation signal.
T his fusion protein is constitutively active, is indepen-
dent of growth factors, and is wortmannin resistant
(Ahmed et al. 1997; Kennedy et al. 1997; see below). Co-
transfection of either WT Akt or MyrAkt and HA–4E-BP1
expression vectors caused a mobility shift of 4E-BP1 (Fig.
1C). T hese results suggest that both PI 3-kinase and its
downstream effector Akt are intermediates in the signal-
ing pathway leading to 4E-BP1 phosphorylation. T ran-
sient transfections of 293 cells produces a high level of
c-Akt expression that was sufficient to elicit the 4E-BP1
mobility shift to the same extent as MyrAkt (Fig. 1C, cf.
lanes 3 and 4).
T o determine whether the endogenous PI 3-kinase/
Akt signaling pathway is involved in mediating the
phosphorylation of 4E-BP1 by growth factors, we used a
kinase-deficient mutant of Akt containing a point mu-
tation in the AT P-binding domain K179M [Akt(kin−)]
(Franke et al. 1995). When expressed at high levels, the
kinasedeadmutant actedin adominant negativefashion
to abolish the mobility shift of 4E-BP1 normally elicited
by insulin treatment (Fig. 1D). Insulin treatment induces
the mobility shift of 4E-BP1 (lanes 1,2). Coexpression of
Akt(kin−) significantly reduced the insulin-mediated
mobility shift (lanes 3,4). T hese results demonstrate that
endogenous Akt is required to transmit the signal lead-
ing to 4E-BP1 phosphorylation.
Recent results suggested the possibility that the wort-
mannin sensitivity of 4E-BP1 phosphorylation is attrib-
utable to direct inhibition of FRAP/mT OR activity by
wortmannin (Brunn et al. 1996). T herefore, we examined
the sensitivity of PI 3-kinase and Akt-mediated phos-
phorylation of 4B-BP1 to wortmannin and rapamycin.
Both wortmannin and rapamycin inhibited the ability of
an activated form of p110? to cause the mobility shift of
4E-BP1 (Fig. 2A). However, the mobility shift elicited by
theactivated Akt was wortmannin resistant but rapamy-
cin sensitive (Fig. 2B). T hese results indicate that the
wortmannin sensitivity of 4E-BP1 phosphorylation by
of 4E-BP1. (A) Insulin-mediated phosphorylation of 4E-
BP1 is both rapamycin and wortmannin sensitive. Hu-
man embryonic kidney (HEK) 293 cells were trans-
fected transiently with a hemaglutinin (HA) epitope-
tagged 4E-BP1 expression vector. After transfection,
cells weredeprivedof serum for 36 hr, andeither mock
treated(lane1) or stimulatedwith insulin (1 µg/ml) for
30 min (lanes 2–4) in the presence of either wortman-
nin, [200 nM (Wort.)] (lane 3) or rapamycin [20 ng/ml
(Rap.)] (lane 4). Cell extracts were prepared as de-
scribed in Materials and Methods and HA–4E-BP1 was
detected by immunoblot analysis with an anti-HA an-
tibody (12CA5). Molecular size markers (in kD) are
indicated. Arrows indicate the different phosphory-
latedisoforms of HA–4E-BP1. (B) T hecatalytic subunit
of PI 3-kinase p110? elicits phosphorylation of HA–
4E-BP1. HEK 293 cells were cotransfected with HA–
4E-BP1 expression vector along with one of the follow-
ing: control vector (lane 1), p110? expression vector
(lane 2), or p110?caax (p110?*) expression vector (lane
3). After transfection, cells were deprived of serum for
36 hr. HA–4E-BP1 was detected as described in A.
Small arrows indicate the different phosphorylation
forms of 4E-BP1. p110? and p110?caax were detected
as described in Materials and Methods. (C) Akt elicits
phosphorylation of HA–4E-BP1. HEK 293 cells were mock transfected (lane 1) or cotransfected with HA–4E-BP1 expression vector and
one of the following: control vector (lane 2), HA–c-Akt expression vector (lane 3), or HA–MyrAkt expression vector (lane 4). Cells were
deprived of serum for 36 hr. HA–4E-BP1 was detected as described above. Small arrows indicate the different phosphorylation forms
of 4E-BP1. (D) A kinase-deficient mutant of Akt inhibits phosphorylation of 4E-BP1 by insulin. HEK 293 cells were cotransfected with
a HA–4E-BPI expression vector (100 ng) and the following: control vector (lanes 1,2) or HA–AktK179M expression vector [Akt(kin−)]
(lanes 3,4). After transfection, cells were serum-deprived for 36 hr and then stimulated with 100 ng/ml of insulin for 45 min (lanes 2,4).
HA–4E-BP1 was detected as described above. HA–AktK179M was detected on the same immunoblot. Small arrows indicate the
different phosphorylated forms of 4E-BP1. T he results shown are representative of three independent experiments.
PI 3-kinase and Akt elicit phosphorylation
Gingras et al.
504GENES & DEV ELOPMENT
growth factors is attributable to inhibition of PI 3-kinase
and not inhibition of FRAP activity.
T aken together these results suggest a linear pathway
from growth factor receptors to the activation of PI 3-
kinase, which in turn activates Akt and leads to phos-
phorylation of 4E-BP1.
Akt mediates the phosphorylation of 4E-BP1
and 4E-BP2 in vivo
T o confirm that the gel mobility shift observed by co-
transfection of Akt with 4E-BP1 is attributable to an in-
crease in 4E-BP1 phosphorylation, we first established a
stable293-cell lineoverexpressing MyrAkt. 293–MyrAkt
or wild-type 293 cells were then labeled metabolically
with [32P]orthophosphate, and 4E-BP1 and 4E-BP2 were
immunoprecipitatedandsubjectedto SDS-PAGE andau-
toradiography. In control cells, two phosphorylated iso-
forms termed ? and ? (the fastest migrating isoform, ?, is
unphosphorylated and hence not detected by32P label-
ing) were detected (Fig. 3A, lane 1). After serum or insu-
lin stimulation (lanes 2,3) a 2.5-fold increase in32P in-
corporation was observed, and two isoforms of slower
mobility (? and ?) appeared. Rapamycin or wortmannin
treatment abrogated the effect of serum and insulin on
4E-BP1 phosphorylation (lanes 4,5). In 293–MyrAkt cells,
the four phosphorylated species (?, ?, ?, ?) were present
in the absence of stimuli (lane 6, 2.5-fold more total32P
incorporation than in starved 293 cells). T he pattern of
phosphorylation is very similar to that of 293 cells
stimulated with serum or insulin (cf. lane 6 with lanes 2
and 3). In 293–MyrAkt cells, the phosphorylation of 4E-
BP1 was sensitive to rapamycin treatment, but com-
pletely resistant to wortmannin treatment (lanes 7,8). A
similar effect was observed for 4E-BP2. 4E-BP2 was im-
munoprecipitated from the lysates as 4E-BP1. 4E-BP2 is
phosphorylated on fewer residues than 4E-BP1 (A.-C.
Gingras and N. Sonenberg, unpubl.), and only one iso-
form incorporating32P is detected by SDS-PAGE (Fig.
3B). T otal32P incorporation in 4E-BP2 was increased ∼2-
to 2.5-fold in serum- and insulin-stimulated cells (Fig.
3B, cf. lanes 2 and 3 with lane 1). T his increase was
diminished (1.5-fold) by rapamycin treatment and abol-
ished by wortmannin treatment (lanes 4,5). In MyrAkt-
expressing 293 cells, 4E-BP2 phosphorylation was in-
creased approximately twofold, as compared to serum
deprived 293 cells (cf. lane 6 with lane 1). As with 4E-
BP1, the increase in 4E-BP2 phosphorylation in MyrAkt
cells was rapamycin sensitive, but wortmannin insensi-
tive (lanes 7,8). Although the effect of rapamycin on 4E-
BP2 phosphorylation was modest (1.5-fold) in the experi-
ment presented here, this inhibitory effect was repro-
duced several times in 293 cells, with an inhibition
varying from 1.5- to 3-fold. T his is similar to the inhibi-
tion of 4E-BP1 phosphorylation by rapamycin.
T hese results further confirm that the differences in
the mobility shift observed for 4E-BP1 in Figures 1 and 2
are attributable to changes in the phosphorylation state
of 4E-BP1. T o determine whether the effects observed on
4E-BP1 phosphorylation are specific to 293 cells, the
same experiment was repeated in Rat1a cells stably ex-
pressing MyrAkt (Kennedy et al. 1997). Similar results to
that observed with 293 cells (two- to threefold increase
in phosphorylation) were obtained for both 4E-BP1 and
4E-BP2, and MyrAkt-induced phosphorylation was also
sensitive to rapamycin, but resistant to wortmannin
(data not shown).
T o determine whether Akt-mediated phosphorylation
of 4E-BP1 occurs on the in vivo phosphorylation sites,
phosphopeptide maps were performed on32P orthophos-
phate-labeled 4E-BP1. T en phosphopeptides were de-
tected in serum-starved 293 cells (Fig. 4A, labeled 1–10
in theorder of decreasing intensity). When 293 cells were
stimulated with serum, the intensity of some of the
spots greatly increased (Fig. 4B, spots 8 and 9), whereas
some new spots (11–14) appeared. Peptide 14 was not
reproducibly detected in other experiments and will not
be discussed further. Rapamycin treatment caused the
decrease or disappearance of two serum-dependent spots
(11 and 12) and of spots 8 and 9 (Fig. 4C). T he effects of
mycin on 4E-BP1 phosphorylation by Akt
and PI 3-kinase. (A) Phosphorylation of 4E-
BP1 by p110?caax is both wortmannin and
rapamycin sensitive. HEK 293 cells were
cotransfected with HA–4E-BP1 expression
vector and one of the following: control
vector (lane 1) or p110?caax (p110?*) ex-
pression vector (lanes 2–4). After transfec-
tion, cells were deprived of serum for 36 hr
and treated with either wortmannin [200
nM (Wort.); lane 3] or rapamycin [20 ng/ml
(Rap.); lane 4]. Cell extracts were prepared
and HA–4E–BP1 was detected as described
Effect of wortmannin and rapa-
in Fig. 1. Arrows indicate different phosphorylation isoforms of 4E-BP1. (B) Phosphorylation of 4E-BP1 induced by an activated Akt is
wortmannin resistant but rapamycin sensitive. HEK 293 cells were cotransfected with HA–4E-BP1 expression vector and with control
vector (lane 1), or with HA–MyrAkt expression vector (lanes 2–4). After transfection, cells were deprived of serum for 36 hr. Cells were
treatedand4E-BP1 was detectedas describedin A. HA–MyrAkt was detectedon thesameimmunoblot. Small arrows indicatedifferent
phosphorylation forms of 4E-BP1. T he figure is representative of two independent experiments.
Akt regulates 4E-BP1 phosphorylation
GENES & DEV ELOPMENT505
wortmannin were even more striking, resulting, for ex-
ample, in the diminution of spot 7 (Fig. 4D). Serum-
starved 293–MyrAkt cells exhibited a phosphopeptide
pattern almost identical to that of 293 cells stimulated
with serum. Phosphopeptides 11–13 were present in the
absenceof serum, although theincreasein spots 9 and13
was not as marked as in serum-treated cells (cf. E with A
and B). As expected, rapamycin treatment caused a
marked decrease in spots 11 and 12 (Fig. 4F). Wortman-
nin treatment had practically no effect on phosphopep-
tides 8, 11, 12 (Fig. 4G). T aken together, these data indi-
cate that the increase in phosphorylation of 4E-BP1 in
293–MyrAkt cells occurs on physiologically relevant
sites and confirm our data that MyrAkt expression con-
fers resistance to wortmannin, but not to rapamycin on
T o determine whether the changes in 4E-BP1 phos-
phorylation correlate with 4E-BP1 activity, the interac-
tion of 4E-BP1 with eIF4E was examined, using cap-af-
finity chromatography. 293 and other human cells con-
tain multiple 4E-BP1 isoforms (Gingras et al. 1996; von
Manteuffel et al. 1996). T he isoform pattern observed in
mouseandrat cells is much simpler, becauseof asmaller
number and better separation of the isoforms. T hus, we
performed the analysis in stably transfected Rat1a/
MyrAkt and wild-typeRat1a cells. In rodent cells, a total
of three 4E-BP1 forms can be detected: (1) a hyperphos-
phorylated slow migrating isoform (?), which does not
interact with eIF4E; (2) a middle form (?), which is phos-
phorylated and binds eIF4E with a low affinity; and (3) an
unphosphorylated fast migrating species (?), which in-
teracts very strongly with eIF4E (Lin et al. 1995; Beretta
et al. 1996; Gingras et al. 1996). Rat1a and Rat1a MyrAkt
cells were serum starved and then serum stimulated
(Rat1a). A Western blot performed on total extracts in-
dicated that the three isoforms (?, ?, and ?) were present
in serum-starved cells (Fig. 5A, lane 1, the ? form is
predominant), only the two slower migrating forms (?
and ?) were detected in serum-stimulated Rat1a cells
(lane 2, the ? form is predominant). Only the hyperphos-
phorylated form (?) was detected in Rat1a MyrAkt cells
(lane 3; significantly less 4E-BP1 was present consis-
tently in the MyrAkt cells, for reasons that are not im-
mediately clear), indicating that 4E-BP1 is hyperphos-
phorylated. A cap-affinity isolation of eIF4E was con-
ducted. In serum-starved Rat1a cells, there was a
significant amount of 4E-BP1 (isoforms ? and ?) that
bound eIF4E (Fig. 5B, lane 1). Binding was abolished in
serum-stimulated cells (lane 2), and in Rat1a (MyrAkt)
cells (lane 3). T aken together, these results suggest that
phosphorylation of 4E-BP1 in MyrAkt cells prevents its
association with eIF4E and that Akt plays a pivotal role
in regulating 4E-BP1 activity in cells.
Akt cannot directly phosphorylate 4E-BP1
T o address the question of whether Akt can phosphory-
late 4E-BP1 directly, in vitro kinase assay was used. T he
kinase reaction was performed using histone H2B (as a
control substrate) (Kennedy et al. 1997). Histone H2B
phosphorylation was increased gradually with time of
incubation with immunoprecipitated HA–MyrAkt (Fig.
6, lanes 1–3). After a 45-min incubation, background
phosphorylation of H2B was detected even with an im-
munoprecipitate from mock transfected cells (lane 7). In
contrast, only background phosphorylation of GST –4E-
BP1 was observed upon incubation with immunoprecipi-
tated HA–MyrAkt (lanes 4–6, and lane 8). Phosphoryla-
tion of 4E-BP1 was not observed even after prolonged
exposure (data not shown). T herefore, we conclude that
Akt cannot phosphorylate 4E-BP1 in vitro and is un-
likely to serve as the kinase that directly phosphorylates
4E-BP1 in vivo (see also below).
creased in 293 MyrAkt cells and is resistant to wort-
mannin treatment. Cells were labeled with [32P]ortho-
phosphate as described in Materials and Methods and
4E-BP1 (A) and 4E-BP2 (B) were immunoprecipitated
successively with polyclonal antibodies, separated by
SDS-PAGE, transferred to Immobilon-PSQand sub-
jected to autoradiography. Different phosphorylated
isoforms are indicated for 4E-BP1.
4E-BP1 and 4E-BP232P incorporation is in-
Gingras et al.
506GENES & DEV ELOPMENT
FRAP/mTOR activity is required for 4E-BP1
phosphorylation by Akt
Previous data and our results suggested that FRAP/
mT OR is required for 4E-BP1 phosphorylation. T he
transfection of wild-type FRAP into serum-deprived
wild-type 293 cells did not result in a mobility shift of
4E-BP1 (S.G. Kennedy and N. Hay, unpubl.). T o evaluate
the role of FRAP/mT OR in our system, we transfected
wild-type FRAP and a rapamycin-resistant mutant
(S2035T ) form of FRAP (Brown et al. 1995) together with
HA–4E-BP1 into 293 cells stably expressing MyrAkt.
T he transiently transfected HA–4E-BP1 exhibited a mo-
bility shift even after 36 hr of serum deprivation (Fig. 7,
lane 1). T hese results are consistent with the increased
phosphorylation of 4E-BP1 observed in serum-deprived
293–MyrAkt cells (see Fig. 3). Phosphorylation of 4E-BP1
in 293–MyrAkt cells was unchanged by transfection
with wild-type FRAP and FRAP(S2035T ) (Fig. 7, lanes
2,5) and was resistant to wortmannin treatment (lanes
inhibits interaction with eIF4E. Rat1a and
Rat1a/MyrAkt cells were incubated in 0.5%
FCS overnight. Rat1a cells were then treated
with 20% FCS for 40 min. Cells were lysed by
freeze–thaw cycles and extracts were either
heat treated (total extract; 100 µg) or incu-
bated (750 µg) with m7GDP–agarose resin, as
described in Materials and Methods. Samples
were separated by SDS-PAGE and 4E-BP1 pro-
tein was analyzed by Western blotting. (A) T o-
tal extract (100 µg). (B) Material bound to the
m7GDP–agaroseresin. Positions of the4E-BP1
isoforms are indicated.
Phosphorylation of 4E-BP1 by Akt
BP1 in 293 MyrAkt cells is identical to that
of serum-stimulated 293 cells.
4E-BP1 (Fig. 3) was excised from an Immobi-
lon membrane, digested with trypsin–chy-
motrypsin, andanalyzedby two-dimensional
phosphopeptide mapping, as described in
Materials and Methods. HEK 293 cells (A–D)
and HEK 293/ MyrAkt cells (E–G) were de-
prived of serum for 36 hr. Cells were labeled
with32P as described in Materials and Meth-
ods. (A,E) Untreated cells (B–D,F,G) were
treated as follows: with 15% FCS for 30 min
(B); pretreated with rapamycin (20 ng/ml) for
20 min before addition of FCS (C); pretreated
with wortmannin (100 nM) for 20 min before
addition of FCS (D); with rapamycin (20 ng/
ml) for 20 min (F); with wortmannin (100 nM)
for 20 min (G).
T he phosphopeptide map of 4E-
Akt regulates 4E-BP1 phosphorylation
GENES & DEV ELOPMENT507
3,6). However, phosphorylation of 4E-BP1 in 293–
MyrAkt cells transfected with wild-type FRAP was still
sensitive to rapamycin treatment (lane 4), whereas trans-
fection of FRAP(S2035T ), conferred rapamycin resis-
tance to 4E-BP1 phosphorylation (lane 7). T herefore, we
conclude that Akt-induced phosphorylation of 4E-BP1
requires FRAP/mT OR.
T ranslation rates are modulated in response to growth
factors, hormones, and mitogens. A major target for ex-
tracellular stimuli is the translation initiation factor
eIF4E. Its activation correlates with the phosphorylation
and inactivation of 4E-BPs, as well as with its own phos-
phorylation on serine 209 (Whalen et al. 1996). Initial
studies reported that 4E-BP1 is a direct downstream tar-
get of mitogen-activated protein kinase (MAPK) (Lin et
al. 1994). Subsequent studies using pharmacological
agents such as wortmannin and LY294002, which in-
hibit PI 3-kinase activity, and mutants of the PDGF re-
ceptor that cannot bind PI 3-kinase, have indicated that
PI 3-kinase is required for the phosphorylation and inac-
tivation of 4E-BP1, and precluded MAPK as an upstream
regulator of 4E-BP1 (Beretta et al. 1996; von Manteuffel
et al. 1996). However, a recent report showing that wort-
mannin and LY294002 inhibit the kinase activity of
FRAP/mT OR, which is required for 4E-BP1 phosphory-
lation, challengedtheideathat PI 3-kinaseis requiredfor
4E-BP1 phosphorylation (Brunn et al. 1996). In the pre-
sent study we have provided direct evidence that growth
factors mediate phosphorylation of 4E-BP1 with PI 3-
kinase, which is the wortmannin-sensitive component
in this signaling pathway (Figs. 1B and2A). Moreover, we
have demonstrated that the downstream effector of PI
3-kinase, Akt, is a critical intermediate in the signal
transduction pathway leading from growth factors to the
phosphorylation of 4E-BP1. Both the wild-type catalytic
subunit of PI 3-kinase and the wild-type c-Akt promote
the phosphorylation of 4E-BP1 when overexpressed tran-
siently in 293 cells. Wefoundthat aconstitutively active
form of Akt promotes phosphorylation of 4E-BP1 in the
absence of growth factors, in a wortmannin-resistant
manner, whereas a dominant-interfering mutant of Akt
blocks theability of insulin to inducephosphorylation of
4E-BP1 (Fig. 1D). An activated Akt can also mediate the
in vivo phosphorylation of 4E-BP2 (Fig. 3B). T he phos-
phorylation of 4E-BP1 by an activated Akt and in re-
sponse to growth factors occurred at apparently identical
sites, as was demonstrated by phosphopeptide mapping
(Fig. 4A,E). An activated Akt inhibits the binding of 4E-
BP1 to eIF4E, even in the absence of growth factors, and
therefore is presumed to increase eIF4E-dependent trans-
lation (Fig. 5).
T he fact that identical phosphorylation sites on 4E-
BP1 are diminished by rapamycin in both serum-induced
and Akt-induced 4E-BP1 phosphorylation (Fig. 4) implies
that the rapamycin-sensitive component lies down-
stream of Akt in this signaling cascade. Indeed, a rapa-
mycin-resistant mutant of FRAP/mT OR confers rapa-
mycin resistance to 4E-BP1 phosphorylation induced by
an activated Akt (Fig. 7). Although these results strongly
suggest a linear pathway from PI 3-kinase through Akt
and to FRAP/mT OR, we cannot exclude a parallel sig-
naling pathway that includes FRAP/mT OR and subse-
quently, that converges into a common downstream ef-
fector leading to phosphorylation of 4E-BP1 (see also Fig.
8). A recent study, which showed that FRAP/mT OR im-
4E-BP1. HEK 293 MyrAkt cells were cotransfected with HA–
4E-BP1 expression vector and with control vector (lane 1) or
with wild-type epitope-tagged FLAG–FRAP expression vector
(lanes 2–4), or a rapamycin-resistant mutant FLAG–FRAP
S2035T expression vector (lane 5–7). After transfection, cells
were serum deprived of for 36 hr and were either left untreated
(lanes 1,2,5) or treated with wortmannin, [200 nM (Wort.); lanes
3,6] or rapamycin [20 ng/ml (Rap.); lanes 4,7]. Cell extracts were
prepared, and HA–4E-BP1 was detected as described in Fig. 1A.
Equal amounts of extract from the same experiment were used
for detection of FLAG–FRAP with anti-Flag monoclonal anti-
bodies. FLAG–FRAP is indicated by the large arrow. Small ar-
rows indicate different phosphorylation states of 4E-BP1. T he
results shown in this figure are representative of three indepen-
Akt requires FRAP activity for phosphorylation of
cells were transfected transiently with HA–MyrAkt expression
vector or with vector alone. Forty-eight hours after transfection,
MyrAkt was immunoprecipitated with an anti-HA antibody
(HA.11). Immunoprecipitates from mock transfected (lanes 7,8)
or from HA–MyrAkt transfected cells (lanes 1–6) were used for
kinase reactions, as described in Materials and Methods. GST –
4E-BP1 (2 µg) and histoneH2B (2 µg) wereused as substrates and
incubated with the immunoprecipitates for the indicated times.
Samples were analyzed by SDS-PAGE. An equal amount of im-
munoprecipitate was used for each reaction. Equal protein load-
ing was visualized by Coomassie Blue staining. T he results
shown are representative of two independent experiments.
Akt does not phosphorylate4E-BP1 in vitro. HEK 293
Gingras et al.
508GENES & DEV ELOPMENT
munoprecipitated from cells is able to phosphorylate 4E-
BP1 in vitro (Brunn et al. 1997), is consistent with a
linear signaling pathway leading from growth factor re-
ceptor to PI 3-kinase, Akt, and FRAP/mT OR. However,
the observation that immunoprecipitated FRAP/mT OR
can phosphorylate 4E-BP1 does not preclude the possi-
bility that a downstream effector of both FRAP/mT OR
and Akt can be coimmunoprecipitated with FRAP/
mT OR and consequently, responsible for 4E-BP1 phos-
phorylation in vitro.
Our studies delineate the signaling pathway leading
from growth factor receptor to the phosphorylation of
4E-BP1 and activation of eIF4E and provide direct evi-
dence that PI 3-kinase, Akt, and FRAP/mT OR are key
regulators in this pathway (Fig. 8). Akt has been reported
to be activated by the lipid products of PI 3-kinase
through its PH domain (Frankeet al. 1997b). Morerecent
studies showed that at least two upstream kinases are
required for the full activation of Akt. One kinase,
termed PDK1, is activated directly by the lipid products
of PI 3-kinase and is responsible for the phosphorylation
of T hr-308 in Akt. T he other kinase is required to phos-
phorylate Ser-473 in Akt (Alessi et al. 1997; Stokoe et al.
1997). Weshowedthat Akt cannot phosphorylate4E-BP1
directly. It is not clear, however, whether Akt activates
FRAP/mT OR directly or modulates the activity of an-
other intermediate such as phosphatase or a kinase.
PI 3-kinase, Akt, and FRAP/mT OR were also impli-
cated in the activation of p70S6k (Brown et al. 1995;
Burgering and Coffer 1995; Reif et al. 1997; Fig. 8). Ex-
periments using three different variants of p70S6k
showed that overexpression of p70S6k inhibits the phos-
phorylation of 4E-BP1 by insulin probably by sequester-
ing a common upstream activator. T hese results also
suggest that the 4E-BP1 phosphorylation pathway bifur-
cates immediately upstream of p70S6k (von Manteuffel
et al. 1997). Because the phosphorylation of both p70S6k
and 4E-BP1 is inhibited by rapamycin, the common up-
stream activator could be the target of rapamycin FRAP/
mT OR. FRAP/mT OR was shown to mediate the in vivo
phosphorylation of both p70S6k and 4E-BP1 (this work;
Brown et al. 1995; Brunn et al. 1997). It is unlikely, how-
ever, that p70S6k and 4E-BP1 are phosphorylated by the
same kinase because the motifs flanking possible 4E-BP1
phosphorylation sites differ from the rapamycin-sensi-
tive p70S6k sites (Moser et al. 1997). T he large size of
FRAP/mT OR (289 kD) raises the possibility (as depicted
in Fig. 8) that it may act as a scaffold protein, which can
interact with multiple kinases having different inputs.
T hus, one of these kinases could be the p70S6k kinase
and another kinase could be the 4E-BP1 kinase.
Akt, apoptosis, cell proliferation, and protein synthesis
Akt appears to have multiple downstream effectors
(Marte and Downward 1997). In this study we demon-
strated that 4E-BP1 is a new downstream target of Akt.
Akt can promote cell survival; it blocks apoptosis accel-
erated by c-Myc (Kauffmann-Zeh et al. 1997; Kennedy et
al. 1997), by UV irradiation (Kulik et al. 1997), by growth
factor withdrawal in fibroblasts (Kennedy et al. 1997), in
cerebellum neurons (Dudek et al. 1997), and in hemapoi-
etic cells (Ahmed et al. 1997), and by a decrease in cell
matrix adhesion (Khwaja et al. 1997). T he ability of Akt
to promote survival is not dependent on changes in the
steady-state levels of Bcl-2 and BclxL (Kennedy et al.
1997). However, Akt inhibits the activity of caspase-3-
like proteases that execute the cell death pathway
(Kennedy et al. 1997). T he ability of Akt to promote cell
survival might be dependent on the synergistic effects of
its multiple downstream effectors. Interestingly, at least
two of its downstream targets, p70S6k and 4E-BP1, are
involved in the control of translation. In many cases of
apoptosis where cells have already been programmed to
die, inhibition of protein synthesis either does not affect
the rate of cell death or augments cell death as in tumor
necrosis factor (T NF), Fas, or Myc-induced apoptosis
(Evan et al. 1992; Polunovsky et al. 1994; Wagner et al.
1994; Natoli et al. 1995; Foote et al. 1996; Karsan et al.
1996; Reinartz et al. 1996). In this regard, it is notewor-
thy that rapamycin has a pronounced effect on the trans-
lation of insulin-like growth factor II (IGF–II) (Nielsen et
al. 1995), which acts as a survival factor (Christofori et
an increase in protein synthesis. PI 3-kinase, Akt, and FRAP/
mT OR are downstream effectors of growth factor receptors that
lead to phosphorylation of the 4E-BPs, and subsequent activa-
tion of eIF4E. For details see Discussion.
A model illustrating thesignaling cascades leading to
Akt regulates 4E-BP1 phosphorylation
GENES & DEV ELOPMENT509
al. 1994; Ueda and Ganem 1996). However, treatment
with rapamycin alone is not sufficient to induce apopto-
sis (Yao and Cooper 1996; Kauffmann-Zeh et al. 1997),
although it was reported that rapamycin can accelerate
apoptosis under certain conditions (Shi et al. 1995). It is
also intriguing that ectopic expression of eIF4E can res-
cue cells from c-Myc and growth factor withdrawal-in-
duced apoptosis (Polunovsky et al. 1996), suggesting that
at least in part, the pathway uncovered in these studies
could be relevant to the ability of Akt to promote sur-
Because both eIF4E and Akt were also shown to play a
role in cell proliferation (Lazaris-Karatzas et al. 1990;
Cheng et al. 1997), and 4E-BP1 is a negative regulator of
cell growth (Rousseau et al. 1996), it is possible that at
least in part Akt activity in cell proliferation is depen-
dent on the activation of eIF4E. As was shown for eIF4E,
Akt is overexpressed and amplified in certain tumors
(Bellacosa et al. 1995; Kerekatte et al. 1995; Cheng et al.
1996; Nathan et al. 1997). Recent studies in Caenorhab-
ditis elegans identified the mammalian equivalents of
insulin receptor and PI 3-kinase as modulators of longev-
ity, probably through regulation of metabolism and pro-
tein synthesis (Morris et al. 1996; Kimura et al. 1997).
Because Akt is conserved in C. elegans (Waterston et al.
1992), it is possible that Akt and its downstream effec-
tors, leading to protein synthesis, are also downstream
components of insulin receptor/PI 3-kinase-mediated
longevity. Because of its pivotal role in cell survival and
proliferation, modulation of Akt activity in vivo might
have an impact on therapies of cancer and degenerative
diseases. 4E-BPs, the new downstream targets of Akt
identified in these studies, can facilitate research on the
activity of this multipotent kinase because the mobility-
shift assay of 4E-BP1 can provide a simple and attractive
read-out for Akt activity in vivo.
Materials and methods
Plasmids and antibodies
T he human 4E-BP1 coding sequence was amplified by PCR and
introduced in-frame into the cytomegalovirus (CMV)-based vec-
tor pACT AG-2 (a kind gift from A. Charest and M. T remblay,
McGill University, Montreal, Canada) to express a fusion pro-
tein with three amino-terminal HA tags. HA–c-Akt, HA–c-Akt
K179M (Franke et al. 1995), and HA–MyrAkt (Ahmed et al.
1997) expression vectors were generously provided by Philip
T sichlis andAlfonsoBellacosa(Fox ChaseCancer Center, Phila-
delphia, PA). T he BamHI/BglII fragment of HA–c-Akt K179M
was introduced into pCDNA3 (Invitrogen) to generate the c-Akt
(kin−) expression vector used in this study. FRAP and
FRAP(S2035T ) (Brown et al. 1995) expression vectors were gen-
erously provided by Stuart Schreiber and Eric Brown (Harvard
University, Cambridge, MA). Myc–epitope-tagged p110? and
p110?*(Khwaja et al. 1997) expression vectors were generously
provided by Julian Downward [Imperial Cancer Research Fund
(ICRF), London, UK]. Antibody 11208 against human 4E-BP1
was described previously (Gingras et al. 1996). Antibody 11209
is a rabbit polyclonal antibody (Pocono Rabbit Farm, Canaden-
sis, PA) raised against human 4E-BP1 (expressed as a GST fusion
protein). Neither antibody cross-reacts with 4E-BP2 in immu-
noprecipitation studies. Anti-4E-BP2 rabbit polyclonal antibody
was raised against a GST –HMK–4E-BP2 fusion protein (de-
scribed in Pause et al. 1994). T he 4E-BP2 crude antisera cross-
reacts (as indicated by Western blotting and by immunoprecipi-
tation studies) with 4E-BP1 and, to a lesser extent, with a novel
4E-BP family member, 4E-BP3. T he anti-HA antibody 12CA5
(mouse monoclonal) was concentrated from tissue culture su-
pernatant using protein G–Sepharose beads (Pharmacia). T he
anti-HA mouse monoclonal antibody HA.11 was purchased
from BabCO and was used at a dilution of 1:1000. T he anti-Flag
mousemonoclonal antibody M5 was purchasedfrom Kodak and
was used at a dilution of 1:400. T he anti-Myc–epitope mouse
monoclonal antibody 9E10 was used at a dilution of 1:500 for
immunoprecipitation and 1:1000 for Western blotting.
Cell culture and viral infection
Cell culture was performed as described previously (Wagner et
al. 1994; Kennedy et al. 1997). Ecotropic MyrAkt retrovirus was
made by transient transfection of Bosc23 cells as described pre-
viously (Wagner et al. 1994). Virus was used to infect PA317
cells andastableproducer cell linewas generated. T heresultant
amphotropic virus was used to infect HEK 293 cells, and stable
clones wereselected with G418 (500 µg/ml, Wagner et al. 1994).
Clones were pooled and maintained as the stable cell line 293–
T ransient transfections were conducted using either Lipofect-
AMINE (GIBCO BRL) or calcium phosphate, as indicated in the
figure legends. For LipofectAMINE transfection, HEK 293 cells
were plated at 1 × 106/6-cm plate and grown overnight in
DMEM/10% FCS. Cells were rinsed once with PBS and placed
in 600 µl of DMEM. Four hundred nanograms of HA 4E-BP1 and
2.5 µg of thep110? and Akt expression constructs wereadded to
600 µl of DMEM containing 25 µl of LipofectAMINE and incu-
bated at room temperature for 30 min. T his reaction mixture
was added to cells and incubated at 37°C for 4 hr. DMEM with
20% FCS (3 ml) was added for 2 hr, then cells were rinsed once
with PBS and incubated in DMEM overnight. T he medium was
changed and cell extracts were prepared after 48 hr. Calcium
phosphate transfection was conducted according to (Sambrook
et al. 1989). T he calcium phosphate/DNA precipitate was in-
cubated with cells overnight in DMEM with 10% FCS. Cells
were rinsed once with PBS and placed in DMEM. T he following
day cells were rinsed again with DMEM.
Extract preparation and Western blotting
Cells were rinsed twice with cold buffer a [20 mM T ris-HCl (pH
7.5), 100 mM KCl, 20 mM ?-glycerolphosphate, 1 mM DT T , 0.25
mM Na3VO4, 10 mM NaF, 1 mM EDT A, 1 mM EGT A, 10 nM
okadaic acid, 1 mM phenylmethylsulfonyl fluoride] and scraped
into a minimal volume of the same buffer. Lysis was performed
by three freeze–thaw cycles. Cell debris was pelleted by cen-
trifugation, and the protein concentration in the supernatant
was measured using the Bio-Rad assay. For analysis of endog-
enous 4E-BP1, 75 µg of total cell extract was incubated at 100°C
for 7 min to enrich for 4E-BP1, which is heat-stable. Samples
were incubated on ice for 5 min, and precipitated material was
removed by centrifugation (13,000 rpm, 5 min). When neces-
sary, the extract was T CA precipitated (Pause et al. 1994).
Laemmli sample buffer was added to the supernatant, which
was then subjected to SDS-15% PAGE. For analysis of trans-
fected HA–4E-BP1, cells were lysed by three freeze–thaw cycles,
and 50 µg of protein was analyzed by SDS–10% PAGE. Western
Gingras et al.
510 GENES & DEV ELOPMENT
blotting, using chemiluminescence detection, was performed as
described (Gingras et al. 1996) using either the anti-4E-BP1
11208 antibody (1:1500) or the anti-HA 12CA5 monoclonal an-
tibody (0.5 µg/ml). For analysis of p110?, p110? and p110?*
expression vectors weretransfected into HEK 293 cells. After 36
hr cells were lysed into 500 µl of RIPA buffer [50 mM T ris-HCl
(pH 7.5), 150 mM NaCl, 1% Nonidet P-40, 0.5% sodium deoxy-
cholate, 0.1% SDS] containing 1 mM PMSF and subjected to
immunoprecipitation for 3 hr with 1 µg of the mouse monoclo-
nal anti-Myc antibody 9E10. Immunoprecipitates were rinsed
three times with RIPA buffer and resuspended in Laemmli
buffer. Samples were boiled and subjected to SDS–8% PAGE.
p110? was detected by incubation with the 9E10 antibody.
Chromatography on m7GDP-agarose
Cell extracts prepared by three freeze–thaw cycles, in buffer
containing 20 mM HEPES–KOH (pH 7.5), 75 mM KCl and 1 mM
EDT A, were incubated for 1 hr with m7GDP coupled to agarose
adipic resin [30 µl of packed beads per reaction; beads were
prepared according to Edery et al. (1988)]. Beads were spun down
in amicrofuge(3000 rpm, 30 sec), washed T HREE times with 20
volumes of the same buffer, and resuspended in Laemmli
sample buffer. Samples were then analyzed by SDS-PAGE and
Western blotting as described above.
Metabolic labeling and immunoprecipitation
HEK 293 cells starved for 30–36 hr or Rat1a cells starved for
16–24 hr were incubated at 37°C for 3 hr in serum-free DMEM
containing 0.5 mCi/ml [32P]orthophosphate (DuPont NEN;
3000 mCi/mmole). Rapamycin (20 ng/ml) or wortmannin (100
nM) were added for 20 min, followed by the addition of dialyzed
FBS (15%; GIBCO) for 30 min. T he medium was removed and
the cells were rinsed twice in cold PBS. Cells were lysed in lysis
buffer [10% glycerol, 50 mM T ris (pH 7.5), 60 mM KCl, 2 mM
acid), 1% T riton X-100, 2 mM DT T , 20 mM ?-glycerolphosphate,
10 nM okadaic acid] for 30 min at 4°C. Lysate was harvested by
scraping and cell debris was removed by centrifugation. T otal
radioactivity in thelysatewas monitored by spotting 1, 2, 5, and
10 µl of the extract onto a phosphocellulose (P81) paper, which
was washed extensively with 75 mM phosphoric acid and dried.
Bound radioactivity was measured by scintillation counting
(Whalen et al. 1996). T he extract (equivalent quantities of ra-
dioactivity) was precleared by incubation with protein A beads
(50 µl per ml of extract) with end-over-end rotation at 4°C for 1
hr. T he supernatant was transferred to a fresh tube, together
with 30 µl of 11209 crude antisera per milliliter of extract, and
incubated for 3 hr at 4°C. Protein A beads (30 µl packed beads)
were added and incubation end-over-end was carried out for 2 hr
at 4°C. Beads were spun down (microfuge, 6000 rpm, 2 min) and
washed two times in lysis buffer, 2 times in RIPA buffer, and 2
times in LiCl solution (200 mM LiCl, 1 mM DT T ). In some
experiments, the supernatant (unbound fraction) was further
incubated with 11211 anti-4E-BP2 antibody as for 4E-BP1. Im-
munoprecipitated material was subjected to SDS–15% PAGE
and transferred to PVDF membranes (Immobilon-P or Immobi-
lon-PSQ, Millipore), which were dried and autoradiographed. Ra-
dioactive bands corresponding to 4E-BPs were excised and
T ryptic–chymotryptic digestion of 4E-BP1 immobilized on the
PVDF membranes was performed essentially as described (van
der Geer and Hunter 1994), with the following modifications.
T he digest was performed using a 200:1 mixture of T PCK-
treated trypsin and chymotrypsin (5 µg, Worthington) for 10 hr,
followed by the addition of 5 µg enzyme for 3 hr. T he sample
was then lyophilized (speed-vac, Savant), resuspended in 200 µl
of water, lyophilized again, resuspended in 100 µl of water, ly-
ophilized a third time, resuspended in 100 µl (pH 1.9) buffer
(2.5% vol/vol formic acid 88% and 7.8% vol/vol glacial acetic
acid) and lyophilized a fourth time. For chromatography, first
dimension (electrophoresis) was performed in pH 1.9 buffer us-
ing the HT LE 7000 apparatus (CBS Scientific); second dimen-
sion was performed in phosphochromatography buffer (37.5%
vol/vol n-butanol, 25% vol/vol pyridine, 7.5% vol/vol glacial
acetic acid). Plastic-coated cellulose thin-layer chromatography
plates (Kodak; 20 cm × 20 cm) were used.
Akt kinase assay
HEK 293 1 × 106cells were transfected with 3 µg of HA–
MyrAkt DNA using LipofectAMINE. Cells were grown
in 10% DMEM for 48 hr, then lysed into Akt lysis buffer
(Kennedy et al. 1997). Extracts were incubated with 3 µg
of monoclonal mouse anti-HA (BabCO, HA.11) for 3 hr.
Immunoprecipitations andkinasereactions with histone
H2B (2 µg) and GST –4E-BP1 (2 µg) were conducted ac-
cording to (Franke et al. 1995).
We are grateful to S. Schreiber, A. Bellacosa, P. T sichlis, and J.
Downward for plasmids and antibodies. We thank C. Palfrey, B.
Raught, and S. Conzen for critical comments and valuable dis-
cussions. T his work was supported by American Cancer Society
(ACS) grant CB-133 and by National Institutes of Health (NIH)
grant CA71874 (N.H.), and by the National Cancer Institute of
Canada (N.S.). N.S. is a Medical Research Council of Canada
Distinguished Scientist and a Howard Hughes Medical Institute
International Scholar. A.-C.G. was supported by a Natural Sci-
ence and Engineering of Canada 1967 Studentship. S.G.K. was
supported by training grant GM07151.
T he publication costs of this article were defrayed in part by
payment of page charges. T his article must therefore be hereby
marked ‘‘advertisement’’ in accordance with 18 USC section
1734 solely to indicate this fact.
Ahmed, N.N., H.L. Grimes, A. Bellacosa, T .O. Chan, and P.N.
T sichlis. 1997. T ransduction of interleukin-2 antiapoptotic
and proliferative signal via Akt protein kinase. Proc. Natl.
Acad. Sci. 94: 3627–3632.
Akimoto, K., R. T akahashi, S. Moriya, N. Nishioka, J. T akay-
anagi, K. Kimura, Y. Fukui, S. Osada, K. Mizuno, S. Hirai, A.
Kazlauskas, and S. Ohno. 1996. EGF and PDGF receptors
activate atypical PKC lambda through phosphatidylinositol
3-kinase. EMBO J. 15: 788–798.
Alessi, D., S. James, C.P. Downes, A. Holmes, P. Gaffney, C.
Reese, andP. Cohen. 1997. Characterization of a3-phosphoi-
nositide-dependent protein kinase which phosphorylates
and activates protein kinase B?. Curr. Biol. 7: 261–269.
Barbet, N.C., U. Schneider, S.B. Helliwell, I. Stansfield, M.F.
T uite, and M.N. Hall. 1996. T or controls translation initia-
tion and early G1 progression in yeast. Mol. Biol. Cell 7: 25–
Bellacosa, A., D.D. Feo, A.K. Godwin, D.W. Bell, J.Q. Cheng,
Akt regulates 4E-BP1 phosphorylation
GENES & DEV ELOPMENT 511
D.A. Altomare, M. Wan, L. Dubeau, G. Scambia, V. Masci-
ullo, G. Ferrandina, P.B. Panici, S. Mancuso, G. Neri, andJ.R.
T esta. 1995. Molecular alterations of the AKT 2 oncogene in
ovarian and breast carcinomas. Intern. J. Cancer 64: 280–
Beretta, L., A.C. Gingras, Y.V. Svitkin, M.N. Hall, and N.
Sonenberg. 1996. Rapamycin blocks the phosphorylation of
4E-BP1 and inhibits cap-dependent initiation of translation.
EMBO J. 15: 658–664.
Brown, E.J. and S.L. Schreiber. 1996. A signaling pathway to
translational control. Cell 86: 517–520.
Brown, E.J., P.A. Beal, C.T . Keith, J. Chen, T .B. Shin, and S.L.
Schreiber. 1995. Control of p70 S6 kinase by kinase activity
of FRAP in vivo. Nature 377: 441–446.
Brunn, G.J., J. Williams, C. Sabers, G. Wiederrecht, J.C.J.
Lawrence, and R.T . Abraham. 1996. Direct inhibition of the
signaling functions of the mammalian target of rapamycin
by the phosphoinositide 3-kinase inhibitors, wortmannin
and LY294002. EMBO J. 15: 5256–5267.
Brunn, G.J., C.C. Hudson, A. Sekulic, J.M. Williams, H. Hosoi,
P.J. Houghton, J. Lawrence, Jr., and R.T . Abraham. 1997.
Phosphorylation of thetranslational repressor PHAS-I by the
mammalian target of rapamycin. Science 277: 99–101.
Burgering, B.M. and P.J. Coffer. 1995. Protein kinase B (c-Akt) in
phosphatidylinositol-3-OH kinase signal transduction. Na-
ture 376: 599–602.
Cheng, J.Q., B. Ruggeri, W.M. Klein, G. Sonoda, D.A. Altomare,
D.K. Watson, and J.R. T esta. 1996. Amplification of AKT 2 in
human pancreatic cells and inhibition of AKT 2 expression
and tumorigenicity by antisense RNA. Proc. Natl. Acad. Sci.
Cheng, J.Q., D.A. Altomare, M.A. Klein, W.C. Lee, G.D. Kruh,
N.A. Lissy, and J.R. T esta. 1997. T ransforming activity and
mitosis-related expression of the AKT 2 oncogene: Evidence
suggesting a link between cell cycle regulation and oncogen-
esis. Oncogene 14: 2793–2801.
Chou, M.M. and J. Blenis. 1995. T he 70 kDa S6 kinase: Regu-
lation of a kinase with multiple roles in mitogenic signaling.
Curr. Opin. Cell Biol. 7: 806–814.
Christofori, G., P. Nalik, andD. Hanahan. 1994. A secondsignal
supplied by insulin-like growth factor II in oncogene-in-
duced tumorigenesis. Nature 369: 414–418.
Chung, J., C.J. Kuo, G.R. Crabtree, and J. Blenis. 1992. Rapamy-
cin-FKBP specifically blocks growth-dependent activation of
and signaling by the 70 kd S6 protein kinases. Cell 69: 1227–
Cohen, P., D.R. Alessi, and D.A.E. Cross. 1997. PDK1, one of
the missing links in insulin signal transduction? Fed. Eur.
Biochem. Soc. Lett. 410: 3–10.
Dudek, H., S.R. Datta, T .F. Franke, M.J. Birnbaum, R. Yao, G.M.
Cooper, R.A. Segal, D.R. Kaplan, and M.E. Greenberg. 1997.
Regulation of neuronal survival by the serine-threonine pro-
tein kinase Akt. Science 275: 661–665.
Edery, I., M. Altmann, and N. Sonenberg. 1988. High-level syn-
thesis in Escherichia coli of functional cap-binding eukary-
otic initiation factor eIE-4E and affinity purification using a
simplified cap-analog resin. Gene 74: 517–525.
Evan, G.I., A.H. Wyllie, C.S. Gilbert, T .D. Littlewood, H. Land,
M. Brooks, C.M. Waters, L.Z. Penn, and D.C. Hancock.
1992. Induction of apoptosis in fibroblasts by c-myc protein.
Cell 69: 119–128.
Foote, L.C., T .J. Schneider, G.M. Fischer, J.K. Wang, B. Rasmus-
sen, K.A. Campbell, D.H. Lynch, S.T . Ju, A. Marshak-Roth-
stein, and T .L. Rothstein. 1996. Intracellular signaling for
inducibleantigen receptor-mediatedFas resistancein B cells.
J. Immunol. 157: 1878–1885.
Franke, T .F., S. Yang, T .O. Chan, K. Datta, A. Kazlauskas, D.K.
Morrison, D.R. Kaplan, and P.N. T sichlis. 1995. T he protein
kinase encoded by the Akt proto-oncogene is a target of the
Franke, T .F., D.R. Kaplan, and L.C. Cantley. 1997a. PI3K:
Downstream AKT ion blocks apoptosis. Cell 88: 435–437.
Franke, T .F., D.R. Kaplan, L.C. Cantley, and A.T . T oker. 1997b.
Direct regulation of the Akt proto-oncogene product by
phosphatidylinositol-3,4-bisphosphate. Science 275: 665–667.
Gingras, A.C., Y. Svitkin, G.J. Belsham, A. Pause, and N. Sonen-
berg. 1996. Activation of thetranslational suppressor 4E-BP1
following infection with encephalomyocarditis virus and po-
liovirus. Proc. Natl. Acad. Sci. 93: 5578–5583.
Hall, M.N. 1996. T he T OR signalling pathway and growth con-
trol in yeast. Biochem. Soc. Transac. 24: 234–239.
Hawkins, P.T ., A. Eguinoa, R.G. Qui, D. Stokoe, F.T . Cooke, R.
Walter, S. Wennstrom, W.L. Claesson, T . Evans, M. Symons,
and L. Stephens. 1995. PDGF stimulates an increase in Rac–
GT P via the activation of phosphoinositide 3-kinase. Curr.
Biol. 5: 393–403.
Hemmings, B.A. 1997. Akt signaling: Linking membraneevents
to life and death decisions. Science 275: 628–630.
Hoekstra, M.F. 1997. Responses to DNA damage and regulation
of cell cycle checkpoints by the AT M protein kinase family.
Curr. Opin. Genet. Dev. 7: 170–175.
Karsan, A., E. Yee, and J.M. Harlan. 1996. Endothelial cell death
induced by tumor necrosis factor- alpha is inhibited by the
Bcl-2 family member, A1. J. Biol. Chem. 271: 27201–27204.
Kauffmann-Zeh, A., P. Rodriguez-Viciana, E. Ulrich, C. Gilbert,
P. Coffer, J. Downward, and G. I. Evan. 1997. Suppression of
c-Myc-induced apoptosis by Ras signalling through PI(3)K
and PKB. Nature 385: 544–548.
Kennedy, S.G., A.J. Wagner, S.D. Conzen, J. Jordan, A. Bellacosa,
P.N. T sichlis, and N. Hay. 1997. T he PI 3-kinase/Akt sig-
naling pathway delivers an anti-apoptotic signal. Genes &
Dev. 11: 701–713.
Kerekatte, V., K. Smiley, B. Hu, A. Smith, F. Gelder, and A. De
Benedetti. 1995. T he proto-oncogene/translation factor
eIF4E: A survey of its expression in breast carcinomas. In-
tern. J. Cancer 64: 27–31.
Khwaja, A., P. Rodriguez-Viciana, S. Wenstrom, P.H. Warne,
and J. Downward. 1997. Matrix adhesion and Ras transfor-
mation both activate a phosphoinositide 3-OH kinase and
protein kinase B/Akt cellular survival pathway. EMBO J.
Kimura, K.D., H.A. T issenbaum, Y. Liu, and G. Ruvkun. 1997.
daf-2, an insulin receptor-like gene that regulates longevity
and diapause in Caenorhabditis elegans. Science 277: 942–
Kulik, G., A. Klippel, and M.J. Weber. 1997. Antiapoptotic sig-
nalling by the insulin-like growth factor I receptor, phospha-
tidylinositol 3-kinase, and Akt. Mol. Cell. Biol. 17: 1595–
Lazaris-Karatzas, A., K.S. Montine, and N. Sonenberg. 1990.
Malignant transformation by eukaryotic initiation factor
subunit that binds to mRNA 5? cap. Nature 345: 544–547.
Lin, T .A., X. Kong, T .A. Haystead, A. Pause, G. Belsham, N.
Sonenberg, and J.C.J. Lawrence. 1994. PHAS-I as a link be-
tween mitogen-activated protein kinase and translation ini-
tiation. Science 266: 653–656.
Lin, T .A., X. Kong, A.R. Saltiel, P.J. Blackshear, and J.C.J.
Lawrence. 1995. Control of PHAS-I by insulin in 3T 3-L1
adipocytes. Synthesis, degradation, andphosphorylation by a
rapamycin- sensitive and mitogen-activated protein kinase-
independent pathway. J. Biol. Chem. 270: 18531–18538.
Gingras et al.
512GENES & DEV ELOPMENT
Marte, B.M. and J. Downward. 1997. PKB/Akt: Connecting
phosphoinositide 3-kinase to cell survival and beyond.
Trends Biochem. Sci. 22: 355–358.
Morris, J.Z., H.A. T issenbaum, and G. Ruvkun. 1996. a phos-
phatidylinositol-3-OH kinasefamily member regulating lon-
gevity and diapause in Caenorhabditis elegans. Nature
Moser, B.A., P.B. Dennis, N. Pullen, R.B. Pearson, N.A. Willi-
amson, R.E. Wettenhall, S.C. Kozma, and G. T homas. 1997.
Dual requirement for a newly identified phosphorylation
site in p70s6k. Mol. Cell. Biol. 17: 5648–5655.
Nakanishi, H., K.A. Brewer, and J.H. Exton. 1993. Activation of
the ? isozyme of protein kinase C by phosphatidylinositol
3,4,5-trisphosphate. J. Biol. Chem. 268: 13–16.
Nathan, C.A., L. Liu, B.D. Li, F.W. Abreo, I. Nandy, and A. De
Benedetti. 1997. Detection of the proto-oncogene eIF4E in
surgical margins may predict recurrence in head and neck
cancer. Oncogene 15: 579–584.
Natoli, G., A. Ianni, A. Costanzo, G. De Petrillo, I. Ilari, P.
Chirillo, C. Balsano, and M. Levrero. 1995. Resistance to
Fas-mediated apoptosis in human hepatoma cells. Oncogene
Nielsen, F.C., L. Ostergaard, J. Nielsen, and J. Christiansen.
1995. Growth-dependent translation of IGF-II mRNA by a
rapamycin-sensitive pathway. Nature 377: 358–362.
Pause, A., G.J. Belsham, A.C. Gingras, O. Donze, T .A. Lin,
H.C.J. Lawrence, and N. Sonenberg. 1994. Insulin-dependent
stimulation of protein synthesis by phosphorylation of a
regulator of 5?cap function. Nature 371: 762–767.
Polunovsky, V.A., I.B. Rosenwald, A.T . T an, J. White, L. Chiang,
N. Sonenberg, and P.B. Bitterman. 1996. T ranslational con-
trol of programmed cell death: Eukaryotic translation initia-
tion factor 4E blocks apoptosis in growth-factor-restricted
fibroblasts with physiologically expressed or deregulated
Myc. Mol. Cell. Biol. 16: 6573–6581.
Polunovsky, V.A., C.H. Wendt, D.H. Ingbar, M.S. Peterson, and
P.B. Bitterman. 1994. Induction of endothelial cell apoptosis
by T NF alpha: Modulation by inhibitors of protein synthesis.
Exper. Cell Res. 214: 584–594.
Price, D.J., J.R. Grove, V. Calvo, J. Avruch, and B.E. Bierer. 1992.
Rapamycin-induced inhibition of the 70-kilodalton S6 pro-
tein kinase. Science 257: 973–977.
Proud, C.G. 1996. p70 S6 kinase: An enigma with variations.
Trends Biochem. Sci. 21: 181–185.
Reif, K., B. Burgering, and D. Cantrell. 1997. Phosphatidyl-
inositol 3-kinase links the interleukin-2 receptor to protein
kinase B and p70 S6 kinase. J. Biol. Chem. 272: 14426–
Reinartz, J., M.J. Bechtel, and M.D. Kramer. 1996. T umor ne-
crosis factor-alpha-induced apoptosis in a human keratino-
cyte cell line (HaCaT ) is counteracted by transforming
growth factor-alpha. Exper. Cell Res. 228: 334–340.
Rousseau, D., A.C. Gingras, A. Pause, and N. Sonenberg. 1996.
T he eIF4E-binding proteins 1 and 2 are negative regulators of
cell growth. Oncogene 13: 2415–2420.
Sambrook, J., E.F. Fritsch, and T . Maniatis. 1989. Molecular
cloning: A laboratory manual, 2nd ed., pp. 16.33–16.35.
Cold Spring Harbor Laboratory Press, Cold Spring Harbor,
Shi, Y., A. Frankel, L.G. Radvanyi, L.Z. Penn, R.G. Miller, and
G.B. Mills. 1995. Rapamycin enhances apoptosis and in-
creases sensitivity to cisplatin in vitro. Cancer Res.
Sonenberg, N. 1996. T ranslational control. (ed. J.W.B. Hershey,
M.B. Mathews, and N. Sonenberg), pp. 245–269. Cold Spring
Harbor Laboratory Press, Cold Spring Harbor, NY.
Stokoe, D., L.R. Stephens, T . Copeland, P.R.J. Gaffney, C.B. Re-
ese, G.F. Painter, A.B. Holmes, F. McCormick, and P.T .
Hawkins. 1997. Dual role of phosphatidylinositol-3,4,5-tri-
sphosphate in the activation of protein kinase B. Science
Ueda, K. and D. Ganem. 1996. Apoptosis is induced by N-myc
expression in hepatocytes, a frequent event in hepadnavirus
oncogenesis, and is blocked by insulin-like growth factor II.
J. Virol. 70: 1375–1383.
van der Geer, P. and T . Hunter. 1994. Phosphopeptide mapping
and phosphoamino acid analysis by electrophoresis and
chromatography on thin-layer cellulose plates. Electropho-
resis 15: 544–554.
Vanhaesebroeck, B., S.J. Leevers, G. Panayotou, and M.D. Wa-
terfield. 1997. Phosphoinositide 3-kinases: A conserved fam-
ily of signal transducers. Trends Biochem. Sci. 22: 267–272.
von Manteuffel, S.R., A.C. Gingras, X.F. Ming, N. Sonenberg,
and G. T homas. 1996. 4EBP-I phosphorylation is mediated
by the FRAP-p70S6K pathway and is independent of mito-
von Manteuffel, S., P.B. Dennis, N. Pullen, A.C. Gingras, N.
Sonenberg, and G. T homas. 1997. T he insulin-induced sig-
nalling pathway leading to S6 and initiation factor 4E bind-
ing protein 1 phosphorylation bifurcates at a rapamycin-sen-
sitive point immediately upstream of p70s6k. Mol. Cell. Biol.
Wagner, A.J., J.M. Kokontis, and N. Hay. 1994. Myc-mediated
apoptosis requires wild-type p53 in a manner independent of
cell cycle arrest and the ability of p53 to induce p21waf1/cip1.
Genes & Dev. 8: 2817–2830.
Waterston, R., C. Martin, M. Craxton, C. Huynh, A. Coulson, L.
Hillier, R. Durbin, P. Green, R. Shownkeen, N. Halloran, et
al. 1992. A survey of expressed genes in Caenorhabditis el-
egans. Nature Genet. 1: 114–123.
Whalen, S.G., A.C. Gingras, L. Amankwa, S. Mader, P.E. Bran-
ton, R. Aebersold, and N. Sonenberg. 1996. Phosphorylation
of eIF-4E on serine 209 by protein kinase C is inhibited by
the translational repressors, 4E-binding proteins. J. Biol.
Chem. 271: 11831–11837.
Yao, R. and G.M. Cooper. 1996. Growth factor-dependent sur-
vival of rodent fibroblasts requires phosphatidylinositol 3-
kinase but is independent of pp70S6Kactivity. Oncogene
Akt regulates 4E-BP1 phosphorylation
GENES & DEV ELOPMENT 513