(Received for publication, February 18, 1997, and in revised form, March 24, 1997)
From the Lymphocyte Activation Laboratory, Imperial Cancer Research Fund, 44 Lincoln's Inn Fields, London WC2A 3PX, United Kingdom and the § Laboratory for Physiological Chemistry, Utrecht University, Universiteitsweg 100, Utrecht 3584 CG , The Netherlands
Phosphatidylinositol 3-kinase (PI 3-kinase) is activated by the cytokine interleukin-2 (IL-2). We have used a constitutively active PI 3-kinase to identify IL-2-mediated signal transduction pathways directly regulated by PI 3-kinase in lymphoid cells. The serine/threonine protein kinase B (PKB)/Akt can act as a powerful oncogene in T cells, but its positioning in normal T cell responses has not been explored. Herein, we demonstrate that PKB is activated by IL-2 in a PI 3-kinase-dependent fashion. Importantly, PI 3-kinase signals are sufficient for PKB activation in IL-2-dependent T cells, and PKB is a target for PI 3-kinase signals in IL-2 activation pathways. The present study establishes also that PI 3-kinase signals or PKB signals are sufficient for activation of p70 S6 kinase in T cells. PI 3-kinase can contribute to, but is not sufficient for, activation of extracellular signal-regulated kinases (Erks) and Erk effector pathways. Therefore, PI 3-kinase is a selective regulator of serine/threonine kinase signal transduction pathways in T lymphocytes, and this enzyme provides a crucial link between the interleukin-2 receptor, the protooncogene PKB, and p70 S6 kinase.
The high affinity interleukin-2 receptor
(IL-2R),1 which comprises -,
-, and
-subunits controls G1 to S progression, T cell clonal
expansion, and functional differentiation (1-3). The IL-2R orchestrates downstream effector pathways by protein tyrosine kinase-dependent activation mechanisms engaging the Src
family tyrosine kinases Lck and Fyn (4) and the Janus kinases 1 and 3 (5-7). Signaling cascades integrated by the action of these tyrosine
kinases include activation of the Ras/Raf/extracellular-signal regulated kinase (Erk) pathway (8-10), activation of the transcription factors STAT3 and STAT5 (11), and the regulation of
phosphatidylinositol 3-kinase (PI 3-kinase) (12).
PI 3-kinase is a ubiquitously expressed enzyme that catalyzes the
phosphorylation of phosphoinositides at the D-3 hydroxyl of the
myo-inositol ring generating PI 3-phosphate, PI
3,4-bisphosphate, and PI 3,4,5-trisphosphate (13, 14). The form of PI
3-kinase involved in protein-tyrosine kinase-dependent
receptor signal transduction comprises a regulatory 85-kDa subunit that
contains two Src homology 2 domains and at its N terminus one Src
homology 3 domain and a catalytic 110-kDa subunit. Following IL-2R
stimulation, several mechanisms have been proposed to recruit PI
3-kinase to the plasma membrane, where its cellular substrate PI
4,5-bisphosphate is located: engagement of the IL-2R leads to binding
of the p85 regulatory subunit of PI 3-kinase to tyrosine 392 in the
IL-2R -chain (15); in addition, interleukin-2 (IL-2) stimulation results in the interaction of PI 3-kinase with the Src family kinases
Fyn (16) and Lck (17).
The activation of PI 3-kinase is a response that IL-2 shares with other cytokines that control lymphoid cell growth and development such as IL-4 and IL-7 (18, 19). It is also clear that PI 3-kinase activation is necessary for the growth- and differentiation-inducing properties of these cytokines (20-23). However, despite the pivotal role of PI 3-kinase in lymphoid cells, there is only a preliminary and incomplete understanding of the targets for this enzyme in the mitogenic signaling pathways regulated by the hematopoietin family of cytokines. To date, the identification of biochemical targets for PI 3-kinase in T cells stems mainly from studies employing the PI 3-kinase inhibitor wortmannin or the LY294002 compound (10, 20). Hence, IL-2 activation of the mitogen-activated protein (MAP) kinase Erk is sensitive to wortmannin (10). Similarly, IL-2 activation of the serine/threonine kinase p70 S6 kinase (p70S6k) is prevented by these PI 3-kinase inhibitors (20). In addition, IL-2 activation of p70S6k is impeded by the immunosupressant rapamycin, which targets another member of the PI 3-kinase family of enzymes, Frap (for FKBP12-rapamycin-associated protein) also termed "mammalian target of rapamycin" (mTor) (24, 25). Observations that wortmannin and rapamycin have identical inhibitory effects on IL-2 activation of p70S6k generated a model for the p70S6k signaling pathway in which PI 3-kinase acts as an upstream regulator of Frap (24, 25). However, this model has been challenged by a recent study showing that the action of Frap is directly inhibited by wortmannin and LY294002 (26). These results raise the issue of whether PI 3-kinase itself has any upstream regulatory role in p70S6k activation in T lymphocytes. Similar caution must be applied to interpretations of data involving PI 3-kinase in Erk activation in T cells. In this context, expression of an active PI 3-kinase is sufficient for Erk activation in Xenopus oocytes (27), but it would be fallacious to extrapolate data obtained in Xenopus cells to T cells, since the role of PI 3-kinase as an upstream regulator of kinase pathways can vary depending on the cell system; to this end, PI 3-kinase signals did not stimulate Erk activity in a variety of fibroblasts and in a monoblast cell line (28-31). Whether PI 3-kinase signals are sufficient to stimulate p70S6k or Erk activation in T cells awaits analysis.
We and others have recently reported that targeting the catalytic p110 subunit of PI 3-kinase to the plasma membrane generates a constitutively active enzyme that induces cellular accumulation of D-3 phosphoinositides (28-31). A constitutively active PI 3-kinase finally allows assessment of the relative contribution of PI 3-kinase-derived signals to a certain effector pathway, in particular whether PI 3-kinase activation is sufficient to promote a specific cellular response. In the present study, we have used a membrane-localized p110 construct, rCD2p110, that induces accumulation of cellular levels of PI 3,4-bisphosphate and PI 3,4,5-trisphosphate (29) as a tool to explore the regulation of serine/threonine kinase pathways by PI 3-kinase in T lymphocytes. We show that activation of PI 3-kinase is sufficient to stimulate p70S6k activity, although PI 3-kinase signals were not sufficient to induce activation of the MAP kinase Erk2 in T cells. The present study also characterizes a previously unrecognized IL-2-mediated signal transduction pathway in T cells that involves the serine/threonine protein kinase B (PKB) also known as c-Akt or Rac protein kinase (32-34). PKB was originally identified as the cellular homologue of the directly transforming oncogene of the murine retrovirus AKT8, which causes thymic lymphomas (35). Herein, we demonstrate that PKB is rapidly activated by IL-2 via a wortmannin- and LY294002-sensitive but rapamycin-insensitive pathway. PI 3-kinase signals alone were sufficient to activate PKB in T cells, and expression of a constitutively active PKB could stimulate the activity of p70S6k. Therefore, PI 3-kinase is a selective regulator of serine/threonine kinase signal transduction pathways in T lymphocytes, and this enzyme is an upstream regulator of the IL-2-activated kinases PKB and p70S6k.
Phorbol 12,13-dibutyrate (PdBu) and wortmannin
were from Calbiochem. LY294002 was a gift from Zeneca. PD098059 was
from New England Biolabs. Rapamycin was a gift from G. Thomas (FMI,
Basel). [14C]Acetyl coenzyme A (at 50 mCi/mmol),
[-32P]ATP (5000 Ci/mmol), and
125I-conjugated protein A were from Amersham Corp.
Ox34 monoclonal antibody (mAb) is raised against rat CD2 (rCD2) (29); 12CA5 mAb is reactive with hemagglutinin (HA), and 9E10 mAb is reactive with the Myc epitope (36); anti-human S6 kinase M5 antiserum (37) was from Santa Cruz Biotechnology; M1 antiserum reactive with p70S6k (37) was a gift from G. Thomas; Rac-PK-CT Ab (Upstate Biotechnology, Inc.) is reactive with PKB.
Plasmids and Reporter ConstructsHA-p70S6k (pBJ5) (38); HA-PKB (pSG5) and gagPKB (pSG5) (32); HA-Erk2 (pCEP4) (39); Myc-V12Rac (pEF), Myc-V12Cdc42 (pEF), and Myc-V14Rho (pEF) (40); and Ha-v-ras (pEF) (29) vector constructs have been described. The described rCD2p110, rCD2p110-R/P, and rCD2p85 chimeras (29) were subcloned into the pEF-BOS expression vector. The reporter plasmids Nlex.Elk-1 (pEF) and 2lexoptk.CAT (41) as well as Nlex.C2 (pMLV) (42) have been described.
Cell Culture and Transient TransfectionsThe Kit225 T leukemic cell line (43) was maintained in RPMI 1640 medium containing 10% heat-inactivated fetal calf serum supplemented with 20 ng/ml of recombinant IL-2 (rIL-2) (Eurocetus) under normal growth conditions. For IL-2 activation assays of endogenous proteins, Kit225 cells were washed three times with phosphate-buffered saline A to remove the IL-2 and cultured further in RPMI supplemented with 5% fetal calf serum in the absence of rIL-2 for 48-72 h prior to IL-2 activation assays. When Kit225 cells were transfected, cells were treated as above but only deprived of rIL-2 for 24 h prior to transfection.
Kit225 cells were transfected by electroporation with 20-40 µg of plasmid DNA. The amounts of plasmid DNA were kept constant per cuvette by adding vector plasmid. Kit225 cells (1.5 × 107 cells/0.625 ml) were pulsed at 320 V and 960 microfarads using a Gene Pulser (Bio-Rad). The amounts of plasmid used were as follows (unless indicated otherwise): 7.5 µg of HA-p70S6k; 12.5 µg of HA-PKB; 10 µg of HA-Erk2; 20 µg of the plasmid pEF empty, rCD2p110, rCD2p110-R/P, Ha-v-ras, V12Rac, V12Cdc42, gagPKB, or rCD2p85; 7.5 µg of 2lexoptk.CAT; and 15 µg of pEFNlex.Elk-1 or pMLVNlex.C2. For gene reporter assays, cells were stimulated as indicated 2-4 h after transfection. Cells were collected 14-18 h after transfection.
Immunoprecipitation, p70S6k Assays, and Western Blot AnalysisAfter stimulations as indicated, Kit225 cells were lysed in lysis buffer 1 (120 mM NaCl, 50 mM Tris pH 8.0, 20 mM NaF, 1 mM benzamidine, 1 mM EDTA, 6 mM EGTA, 7.5 mM PPi, 15 mM p-nitrophenyl phosphate, 1% Nonidet P-40, 0.1 mM phenylmethylsulfonyl fluoride, and 0.1 mM Na3VO4). Cell extracts corresponding to 3 × 106 cell equivalents (nontransfected cells) or 1.5 × 106 live cell equivalents (transfected cells) were used for each immunoprecipitation. Postnuclear lysates were precleared with protein A cell suspension (Sigma) prior to incubation with 2 µg of 12CA5 mAbs or, for endogenous proteins, 1 µg of M5 Abs. Immune complexes were precipitated with protein G-Sepharose beads (Sigma) or, when M5 Abs were used, with protein A-Sepharose beads (Pharmacia Biotech Inc.). The immunoprecipitates were washed three times in lysis buffer 1 and once in p70S6k assay buffer (50 mM Mops, pH 7.2, 5 mM MgCl2, 0.1% Triton X-100) and assayed as described (37) using S6 as a substrate (a gift from G. Thomas). Proteins were resolved by SDS-PAGE. The lower part of the gel was dried, and 32P-labeled S6 proteins were detected by autoradiography. The levels of p70S6k protein in each immunoprecipitate were assessed by transferring the proteins in the upper part of the gel onto polyvinylidene difluoride membranes and performing Western blot analysis with 12CA5 mAbs or M1 Abs using the ECL detection system (Amersham). If the p70S6k protein levels in the immunoprecipitate were not equal, activities were normalized for p70S6k expression levels by quantitation of Western blots probed with M1 Abs followed by 125I-conjugated protein A (Amersham). Quantitation of incorporated 32P into S6 or of bound 125I-conjugated protein A was performed using a PhosphorImager (Molecular Dynamics). To test for effector protein expression in transfected cells, postnuclear cell extracts corresponding to 3 × 106 cell equivalents of the same extract as above were analyzed by Western blotting as described (44) using 9E10, Ox34, or specific Abs.
PKB AssaysCells were treated as for p70S6k assays except
that lysis buffer 2 (120 mM NaCl, 50 mM Hepes,
pH 7.4, 10 mM NaF, 1 mM EDTA, 40 mM
-glycerophosphate, pH 7.5, 1% Nonidet P-40, 0.1 mM
phenylmethylsulfonyl fluoride, 0.1 mM
Na3VO4) was used to lyse cells. To
immunoprecipitate endogenous PKB, 2 µg of Rac-PK-CT Abs were used.
The immunoprecipitates were washed twice in lysis buffer 2, twice in
high salt wash buffer (500 mM LiCl, 100 mM
Tris, pH 7.5, 1 mM EDTA, pH 7.5), and once in PKB assay
buffer (50 mM Tris, pH 7.5, 10 mM
MgCl2, 1 mM dithiothreitol). The reaction was
initiated by the addition of 15 µl of PKB reaction buffer containing
3 µCi of [
-32P]ATP, 50 µM ATP, 7.3 mM MgCl2, 730 µM dithiothreitol,
500 nM protein kinase inhibitor (Sigma), 40 mM
Tris, pH 7.5, and 2.5 µg of histone 2B (H2B) (Boehringer Mannheim).
After 30 min at 25 °C, the reaction was terminated by adding
reducing SDS-PAGE sample buffer and boiling. Proteins were resolved by
SDS-PAGE, and the gel was treated as for p70S6k assays. To detect PKB
proteins, Western blot analysis was performed with Rac-PK-CT Abs.
Cells and cell extracts were processed as for
p70S6k assays. HA-tagged Erk2 was immunoprecipitated with 12CA5 mAbs.
Precipitated immune complexes were washed three times with lysis buffer
1 and once with Erk wash buffer (30 mM Tris, pH 8.0, 20 mM MgCl2, 2 mM MnCl2).
The reaction was initiated by the addition of 10 µl of Erk reaction
buffer containing 4 µCi of [-32P]ATP, 20 µM ATP, 20 mM MgCl2, 2 mM MnCl2, 5 mM
p-nitrophenyl phosphate, 500 nM protein kinase
inhibitor, 30 mM Tris, pH 8.0, and 15 µg of myelin basic
protein (Sigma). After 30 min at 37 °C, the reaction was terminated
by adding reducing SDS-PAGE sample buffer and boiling. Proteins were
resolved by SDS-PAGE, and the gel was treated as for p70S6k assays. To
detect Erk proteins, Western blot analysis was performed with 12CA5
mAbs as primary Ab, rabbit anti-mouse IgG as secondary Ab, and
125I-conjugated protein A.
Fourteen to 16 h after inductions, as indicated, Kit225 T cells were harvested and cells were lysed in 200 µl of lysis buffer (0.65% Nonidet P-40, 10 mM Tris, pH 8, 1 mM EDTA, 150 mM NaCl). Gene expression assays were carried out as described (45). The data are presented as percentage of conversion.
For our studies, we used the well characterized human
IL-2-dependent T cell line Kit225. The data in Fig.
1A show that p70S6k activity is low in
quiescent rIL-2-deprived Kit225 cells but can be rapidly stimulated by
rIL-2. The activity of p70S6k is increased in response to phorbol
esters that stimulate protein kinase C (PKC) (Fig. 1A). We
asked whether PI 3-kinase signals could substitute for IL-2 in inducing
p70S6k activity. We have shown recently that plasma membrane targeting
of p110, the catalytic subunit of PI 3-kinase, generates an enzyme that
is constitutively active in vivo (29). Our membrane-targeted
active PI 3-kinase construct comprises a chimera of the extracellular
and transmembrane domains of the rCD2 antigen fused to the p110
catalytic domain of PI 3-kinase, rCD2p110. As a control, we used a
rCD2p110 molecule that contained in the catalytic subunit of PI
3-kinase an inactivating point mutation (R1130P) in the ATP binding
site (46) that abolishes its in vivo and in vitro
lipid kinase activity, rCD2p110-R/P. To assess the effects of PI
3-kinase signals on p70S6k activity, rIL-2-starved Kit225 cells were
co-transfected with HA epitope-tagged p70S6k and rCD2p110. Cell surface
expression of the rCD2p110 chimera was confirmed by flow cytometric
immunofluorescence analysis with rCD2 mAbs (data not shown). The
HA-tagged p70S6k was immunoprecipitated from transiently transfected
cells and assayed for its ability to phosphorylate S6 ribosomal
subunits (Fig. 1B). Expression of the active PI 3-kinase,
rCD2p110, resulted in constitutive IL-2-independent p70S6k activation
(Fig. 1B). p70S6k was not constitutively activated in cells
expressing "kinase-dead" rCD2p110-R/P, confirming that the p70S6k
activation requires the kinase activity of the p110 subunit (Fig.
1C). The expression of rCD2p110-R/P was noted in some
experiments to suppress rIL-2 inducibility of p70S6k, indicating that
this chimera may be an inhibitory mutant of PI 3-kinase pathways.
IL-2 and Membrane-localized PI 3-Kinase Activate PKB
PKB can
be activated by receptor tyrosine kinases such as the platelet-derived
growth factor receptor and has been identified as a target of PI
3-kinase in fibroblasts (28, 32, 33). However, whether this pathway is
conserved in the hematopoietic system has not been explored. In
particular, although PKB can become oncogenic and initiate thymic
tumors, its regulation and significance for normal T cell growth
processes is not known. Since cytokine receptors have essential
functions in the development and maintenance of the hematopoietic
system, we were interested to assess whether members of the
hematopoietin receptor family, such as the prototypical IL-2R, regulate
PKB. To examine whether IL-2 activates PKB, immunoprecipitates of this
kinase were prepared from rIL-2-deprived and rIL-2-activated Kit225
cells and subjected to in vitro kinase assays using H2B as a
substrate. The data in Fig. 2A show that IL-2
induced a rapid activation of PKB. A 2-3-fold increase in enzyme
activity over basal levels was sustained for more than 60 min in
response to rIL-2. PKB activity is regulated by phosphorylation as
indicated by the reduced electrophoretic mobility of PKB isolated from
rIL-2-activated cells (Fig. 2A). PKB activity was not
induced by exposure of Kit225 cells to phorbol esters that activate PKC
(Fig. 2A). The data in Fig. 2B show the failure
of rIL-2 to stimulate PKB in cells pretreated with LY294002 or
wortmannin, two well characterized PI 3-kinase inhibitors that bind to
the ATP or lipid binding sites on the p110 catalytic subunit, respectively. These inhibitors also prevent the autokinase activity of
Frap/mTor (26), a member of the PI 3-kinase family (47), which is the
cellular target for the drug rapamycin and which prevents
IL-2-coordinated cell cycle progression and proliferation of T
lymphocytes (24, 25). Frap activity is absolutely required for p70S6k
action in T cells (24, 25). We therefore assessed whether Frap function
was necessary for IL-2-induced stimulation of PKB. Rapamycin had no
effect on IL-2-triggered activation of PKB (Fig. 2B),
although rapamycin completely abolished IL-2- or PI 3-kinase-controlled
induction of p70S6k (data not shown). Thus, the inhibition of PKB by
wortmannin and LY294002 cannot be caused by prevention of Frap activity
and indicate that IL-2 regulation of PKB employs PI 3-kinase.
Interleukin-2 activates PKB: PI 3-kinase activity is necessary for IL-2-mediated activation of PKB, and PI 3-kinase signals are sufficient to stimulate PKB activity in Kit225 T cells. A, interleukin-2 activates PKB. Kit 225 cells were deprived of rIL-2 for 68 h and treated with 20 ng/ml rIL-2 or 50 ng/ml PdBu for the indicated times, and PKB activation/phosphorylation was assessed by electromobility shift assays (top). PKB was precipitated from lysates with Rac-PK-CT Abs and PKB activity was analyzed in immune complex kinase assays using H2B as a substrate. [32P]Phosphate incorporation into H2B (middle) was quantified (graph) using a PhosphorImager and is expressed in arbitrary units. Protein levels of PKB present in immune complexes were assessed in parallel by Western blotting with Rac-PK-CT Abs (bottom). B, the PI 3-kinase inhibitors wortmannin and LY294002, but not rapamycin, inhibit IL-2-dependent activation of PKB. Kit225 cells starved of rIL-2 for 72 h were pretreated for 30 min with the vehicle dimethyl sulfoxide (DMSO), 20 ng/ml rapamycin, 5 µM LY294002, or 100 nM wortmannin and then stimulated with rIL-2 for the times indicated before lysis. PKB activity was measured in immune complex kinase assays using H2B as a substrate. C, PI 3-kinase signals trigger a potent stimulation of PKB activity. Kit225 cells were co-transfected with HA-PKB plasmids and vector plasmid (empty), or plasmids encoding for rCD2p110, rCD2p110-R/P, Ha-v-ras, or V12Rac as indicated. HA-PKB activity was analyzed in anti-HA tag immune complex kinase assays using H2B as a substrate. B and C, [32P]phosphate incorporation into H2B (top) was quantified (graph) using a PhosphorImager and is expressed in arbitrary units. Protein levels of PKB (B) or HA-PKB (C) present in immune complexes were assessed in parallel by Western blotting with Rac-PK-CT Abs (bottom).
To investigate directly whether PI 3-kinase signals are sufficient to activate PKB, rIL-20-deprived Kit225 cells were co-transfected with either rCD2p110 or rCD2p110-R/P expression vectors together with an expression vector encoding HA epitope-tagged PKB. In addition, the ability of activated forms of the small GTPases Ha-v-ras and V12Rac to activate PKB was assessed. Immunoprecipitates of HA-tagged PKB were assayed for kinase activity using H2B as a substrate. The constitutively active PI 3-kinase rCD2p110 induced a robust activation of PKB (Fig. 2C). This stimulatory effect of rCD2p110 was dependent on the kinase activity of the chimera, since co-expression of kinase-inactive rCD2p110-R/P did not stimulate PKB activity. As observed previously in other cell systems (28, 33, 48), co-expression of activated Ha-v-ras but not of active V12Rac led to a moderate rise in PKB activity in Kit225 T cells.
Co-expression of an Activated Form of PKB Stimulates p70S6k in Kit225 T Cellsp70S6k is activated by multiple serine/threonine
phosphorylation in response to mitogenic stimuli. The retroviral
oncogene v-Akt is a chimeric molecule, consisting of the retroviral Gag protein fused to the N terminus of c-Akt, which is myristoylated, and
hence v-Akt is predominantly found at the plasma membrane, which may
give raise to its oncogenicity (49). The expression of constitutively
active PKB, gagPKB, has been described as activating p70S6k in Rat-1
cells (32) and COS1 cells (33). Nevertheless, the ability of phorbol
esters to stimulate p70S6k without any discernible activation of PKB
indicated that PKB-independent pathways for activation of p70S6k must
exist in T cells. To determine the role of PKB in p70S6k activation in
T cells, rIL-2-deprived Kit225 cells were co-transfected with a gagPKB
expression vector together with an expression vector encoding HA
epitope-tagged p70S6k. p70S6k activity was analyzed in anti-HA tag
immune complexes with S6 ribosomal subunits as a substrate.
Co-expression of constitutively active PKB induced a strong activation
of p70S6k that was comparable with increases in p70S6k activity seen by
co-expression of rCD2p110 (Fig. 3A). A
rCD2p85 construct that does not regulate cellular levels of D-3
phosphoinositides (29) did not stimulate p70S6k. In contrast to data
described in fibroblasts (48), co-expression of V12Rac and V12Cdc42 had
no effect on p70S6k activity in Kit225 cells (Fig. 3A). To
confirm that V12Rac and V12Cdc42 are active in Kit225 cells, we tested
their ability to activate the stress-activated protein kinases, also
known as c-Jun N-terminal kinases (50, 51). Stress-activated protein
kinase and hence Rac/Cdc42 activity can be measured by the ability of
stress-activated protein kinases to phosphorylate the transcription
factor ATF-2 (42). To monitor ATF-2 transcriptional activity, a fusion
protein comprising the N terminus of ATF-2 (termed C2) linked to the
LexA repressor (42) was co-transfected into Kit225 cells together with
a LexA operator-controlled chloramphenicol acetyltransferase (CAT)
reporter gene. V12Rac and V12Cdc42 potently stimulated ATF-2/LexA-C2
transcriptional activity in Kit225 T cells (Fig. 3B). This
response was specific, since the expression of V14Rho, which does not
regulate stress-activated protein kinases (50, 51), did not induce
ATF-2/LexA-C2-controlled gene expression. Therefore, the GTPases V12Rac
and V12Cdc42 are active and can stimulate Rac-/Cdc42-regulated
signaling pathways in Kit225 T cells.
IL-2 Regulates the Transcription Factor Elk-1 in Kit225 Cells in a PI 3-Kinase-dependent Fashion
In T cells, a PI
3-kinase-sensitive pathway for regulating the activity of Erk kinase
(Mek) and the Erks has been reported to co-exist alongside the PI
3-kinase/p70S6k pathway (10). While p70S6k is thought to exert its
mitogenic function by controlling translation initiation and protein
synthesis, the MAP kinase Erk is implicated in regulating the
phosphorylation and activity of certain transcription factors. One well
characterized cellular substrate for Erks in fibroblasts and T cells is
the transcription factor Elk-1 (40, 52, 53). We therefore tested the
ability of IL-2 to regulate Elk-1 transcriptional activity and hence
Erk in Kit225 cells. To monitor Elk-1 transcriptional activity, a fusion protein comprising the C terminus of Elk-1 linked to the LexA
repressor (41) was co-transfected into Kit225 cells with a LexA
operator-controlled CAT reporter gene. The data in Fig. 4A demonstrate that IL-2 can regulate Elk-1
transcriptional activity in Kit225 cells. To confirm that Elk-1
transactivation is induced by a Mek/Erk-sensitive pathway, we
investigated the ability of the well characterized inhibitor of Mek
activation, PD098059 (54), to prevent IL-2-mediated activation of
Elk-1. Treatment of Kit225 cells with the PD098059 component inhibited
stimulation of Elk-1 transcriptional activity triggered by rIL-2 (Fig.
4B). Moreover, rIL-2-induction of Elk-1 activity was
prevented by the PI 3-kinase inhibitor wortmannin in a
dose-dependent manner (Fig. 4C), which corroborates earlier studies indicating that Erk activation by IL-2
requires PI 3-kinase function (10). Treatment of Kit225 cells with
rapamycin did not affect Elk-1 transactivation in Kit225 cells (Fig.
4D). To assess whether constitutively active PI 3-kinase and
the in vivo production of D-3 phosphoinositides could induce MAP kinase signaling pathways in T cells, the ability of rCD2p110 to
induce transcriptional activation of Elk-1 was analyzed. LexA-Elk-1 transcriptional activity was low in quiescent Kit225 cells but could be
instigated by co-expression of active Ha-v-ras and by stimulation with
phorbol esters, whereas expression of rCD2p110 did not stimulate Elk-1
transactivation (Fig. 4E). However, rCD2p110 signals could
potentiate phorbol ester induction of the transcriptional activity of
Elk-1. This potentiating effect was not observed in cells expressing
the kinase-dead rCD2p110-R/P and was thus dependent on the integrity of
the lipid kinase and the cellular production of D-3
phosphoinositides. Moreover, gagPKB cannot mimic the effects of PI
3-kinase on the Erk/Elk-1 pathway (Fig. 4E).
PI 3-Kinase Signals Synergize with Phorbol Esters to Induce Erk Activity in Kit225 Cells
To assess the effect of
membrane-localized PI 3-kinase on Erk activity directly, rIL-2-deprived
Kit225 cells were co-transfected with expression vectors encoding
rCD2p110 and HA epitope-tagged p42 Erk2, and cells were stimulated with
phorbol esters or left untreated. Co-expression of rCD2p110 did not
stimulate Erk2 activity, although Erk2 could be activated by
co-expressing the activated Ras, Ha-v-ras (Fig. 5).
These results thus confirm the data in Fig. 4E indicating
that PI 3-kinase signals are not sufficient to activate the Erk/Elk-1
pathway. The data in Fig. 5 demonstrate that active PI 3-kinase
markedly potentiated the level of Erk2 activation triggered by phorbol
esters, an effect that was not observed in cells expressing the
kinase-dead rCD2p110-R/P. PI 3-kinase signals did not enhance IL-2
activation responses on Erk (data not shown). Taken together, the
results in Figs. 4E and 5 clearly demonstrate that although
PI 3-kinase signals are not sufficient for Erk/Elk-1 activation, they
can synergize with phorbol esters to induce a maximal response. These
results are concordant with a model where PI 3-kinase signals bifurcate
to activate the PKB/rapamycin-sensitive/p70S6k pathway and
independently contribute to the Mek/Erk/Elk-1 pathway via an as yet
undefined mechanism (see Fig. 6).
The present study has used a membrane-targeted, constitutively active, catalytic subunit of PI 3-kinase as a tool to identify direct targets of PI 3-kinase action in IL-2 signal transduction pathways. We demonstrate that the serine/threonine kinase PKB/Akt can be activated by the cytokine IL-2 via a PI 3-kinase-dependent pathway. Importantly, PI 3-kinase signals alone are sufficient to activate PKB in T cells, demonstrating that PI 3-kinase acts as an upstream regulator of this serine/threonine kinase in lymphoid cells. PKB contains an N-terminal pleckstrin homology domain that can directly bind D-3 phosphoinositides (33, 55, 56), which may contribute to the regulation of the enzyme. Since PI 3-kinase signals are sufficient to substitute for IL-2 in PKB activation, PKB could be a direct target for PI 3-kinase signals during IL-2 signal transduction. PKB/c-Akt is highly expressed in the thymus (57), and the oncogenic form of this kinase causes thymic malignancies. Therefore, PKB has a pivotal role in controlling T cell proliferation/differentiation. The present data identify one function for PKB in T cells; PKB action is sufficient to stimulate p70S6k. Moreover, PI 3-kinase signals are sufficient for activation of p70S6k, which stresses the close link between PI 3-kinase and PKB in regulating p70S6k activity in T cells. Questions regarding the selectivity of the inhibitors that were first used to define a role for PI 3-kinase in T cell biology have challenged the involvement of this enzyme in the regulation of p70S6k in T cells (26). The present data resolve this controversy and provide unequivocal evidence that PI 3-kinase can function as an upstream regulator of p70S6k in T cells.
Results obtained recently with p110 constructs that were membrane-targeted by myristoylation or farnesylation signals showed that PI 3-kinase signals are sufficient to activate PKB and p70S6k in COS cells (28). It has also been shown in fibroblasts that the GTPases Rac and Cdc42 induce p70S6k activation. We find no evidence for Rac/Cdc42 activation of p70S6k in T cells, indicating that cells of different lineages can differ markedly in their cellular mechanisms for kinase activation. Nevertheless, the present data show a striking conservation of the PI 3-kinase/PKB/p70S6k link in human T cells and simian fibroblasts. The conservation of the PI 3-kinase/PKB/p70S6k signaling cascade in T cells implies a physiological importance of this pathway, which has guaranteed its evolutionary conservation.
The role of PI 3-kinase as an upstream regulator of the Erk kinase
pathways can also vary depending on the cell system; expression of an
active PI 3-kinase is sufficient for Erk activation in
Xenopus oocytes (27) but not in fibroblasts or monoblasts
(28-31). The present data show directly that PI 3-kinase can have a
positive regulatory role in Erk activation in T cells (see Fig. 6).
However, PI 3-kinase signals alone fail to stimulate Erk signaling
pathways but markedly potentiate Erk responses in combination with
phorbol esters. Erk regulation of downstream nuclear targets is hereby analyzed using the transactivation capacity of the ternary complex factor Elk-1, a well characterized substrate for Erks in fibroblasts and Jurkat T cells (40, 53). We establish that Elk-1 is regulated by
IL-2 via Mek- and PI 3-kinase-sensitive pathways. Furthermore, as
observed in direct Erk activation assays, PI 3-kinase signals potently
enhance phorbol ester induction of Elk-1 transcriptional activity. PI
3-kinase signals may thus be required for activation of MAP kinase
pathways in IL-2-dependent T cells, but they are not
sufficient and hence are one component of a more complex signaling network. We have not yet explored the PI 3-kinase effector pathways involved in Erk activation, although previous data have excluded the
involvement of the Frap/p70S6k pathway, since activation of Erk is not
sensitive to rapamycin inhibition. Moreover, PKB, which is a potent
activator of p70S6k, cannot mimic the effects of activated PI 3-kinase
on the Erk/Elk-1 pathway. These data best fit a model in which PI
3-kinase regulation of MAP kinases and PKB/p70S6k bifurcate prior to
activation of PKB (Fig. 6). Members of the Rho family of GTPases can
potentiate Erk activation pathways in fibroblasts (39). Activation of
PI 3-kinase is sufficient to induce cytoskeletal rearrangements
mediated by the GTPase Rac and Rho in Swiss 3T3 cells (29) and thus has
the potential to regulate Rac/Rho signaling pathways in T cells.
Accordingly, it is possible that Rac/Rho family GTPases could mediate
PI 3-kinase regulation of Erk. However, several other candidate
in vivo targets for D-3 phosphoinositides have been proposed
including members of the novel PKC family (58) and the atypical PKC
family, PKC- (59) and PKC-
(60), which has recently been
implicated as a regulator of Mek and Erk activity in COS cells
(61).
p70S6k plays a key role in cellular growth control mechanisms by coordinating protein biosynthesis via phosphorylation of the S6 subunit of 40 S ribosomes or via regulation of the activity of the eukaryotic initiation factor 4E binding protein, 4E-BP1 (24, 25, 62). Expression of an activated PKB can stimulate p70S6k activity in T cells, indicating that PKB substrates are part of the p70S6k activation pathways. Moreover, given the ability of PI 3-kinase signals to stimulate PKB and p70S6k, it seems probable that PKB mediates the PI 3-kinase activation of p70S6k in T cells. The immunosuppressive drug rapamycin inhibits the cell cycle progression and proliferation of T lymphocytes and has been shown previously to block IL-2 activation of p70S6k. Rapamycin forms a complex with the intracellular protein FKBP12, which subsequently provides a high affinity inhibitor of Frap. The Frap kinase is a member of the PI 3-kinase family of enzymes (47) and plays an established, but poorly defined, role as an upstream regulator of p70S6k (24, 25). Rapamycin prevents the activation of p70S6k induced in T cells by the constitutively active PI 3-kinase, rCD2p110, (data not shown) or by active PKB (32, 33), thus indicating that PI 3-kinase or PKB activation signals cannot bypass the role of Frap in p70S6k activation pathways. A simple interpretation of these data is that PI 3-kinase and PKB activation of p70S6k is mediated by Frap, although the possibility cannot be excluded that Frap regulates p70S6k by an essential signaling pathway operating in parallel with PI 3-kinase/PKB signals (Fig. 6). Frap controls p70S6k activation by regulating the phosphorylation of key residues in the enzyme (63, 64). Nevertheless, p70S6k is not a direct substrate for Frap, and some intermediate p70S6k kinase(s), as yet uncharacterized, must be invoked to explain the role of Frap in p70S6k activation. Although the evidence that PKB mediates PI 3-kinase effects on p70S6k are compelling, these data do not exclude the possibility that there are PKB-independent mechanisms for p70S6k activation of T cells. In this context, the present data show that activation of PKC by phorbol esters stimulates p70S6k without any discernible stimulatory effect on PKB.
Recent studies showing that cytokine activation of serine kinases is important for the regulation of apoptosis (65, 66) have focused attention on cytokine-induced serine kinase cascades. PI 3-kinase and PKB have been implicated in the prevention of apoptosis in other cell systems (67, 68). The present study demonstrates that PI 3-kinase can couple the IL-2R to a selective subset of serine/threonine kinase pathways in T cells, and in this respect, the PI 3-kinase/PKB link is intriguing, since PKB mediates activation of the Frap/p70S6k pathway but may also regulate other kinase cascades that bifurcate from the PKB/p70S6k pathway including glycogen synthase kinase-3 (GSK3) signaling pathways (69). Therefore, PI 3-kinase and/or PKB have the potential for pleiotropic functions in T cells, and their downstream effectors may include additional serine/threonine kinases evoked by IL-2R engagement.
Finally, PI 3-kinase is activated by members of the cytokine receptor family such as the IL-2R, the IL-4 receptor, the IL-7 receptor, and the IL-13 receptor. Signaling pathways regulated by PI 3-kinase can hence have an impact on lymphocyte biology at multiple points. Accordingly, it is important to establish the function of this enzyme in lymphoid cells. The IL-2R is a prototypical member of this hematopoietin receptor family. The present results directly define PI 3-kinase function in T cells and position PKB for the first time in a physiologically relevant cytokine-induced signal transduction pathway in lymphoid cells. The model described herein may also be applicable to serine/threonine kinase pathways regulated by other receptors that activate PI 3-kinase in T cells.
We thank George Thomas and Richard Treisman for reagents.
Recently, Alessi and colleagues described the purification of two upstream kinases that are likely to mediate PKB activation. The activity of at least one of these upstream kinases, PKD1, is regulated by binding D-3 polyphosphate phosphoinositides (Alessi, D. R., James, S. R., Downes, C. P., Holmes, A. B., Gaffney, P. R. J., Reese, C. B., and Cohen, P. (1997) Curr. Biol. 7, 261-269).