Complex Formation and Cooperation of Protein Kinase Cθ and Akt1/Protein Kinase Bα in the NF-κB Transactivation Cascade in Jurkat T Cells

Protein kinase Cθ (PKCθ) is known to induce NF-κB, an essential transcriptional element in T cell receptor/CD28-mediated interleukin-2 production but also T cell survival. Here we provide evidence that PKCθ is physically and functionally coupled to Akt1 in this signaling pathway. First, T cell receptor/CD3 ligation was sufficient to induce activation as well as plasma membrane recruitment of PKCθ. Second, PKCθ selectively cooperated with Akt1, known to act downstream of CD28 co-receptor signaling, in activating a NF-κB reporter in T cells. Third, Akt1 function was shown to be required for PKCθ-mediated NF-κB transactivation. Fourth, PKCθ co-immunoprecipitated with Akt1; however, neither Akt1 nor PKCθ served as a prominent substrate for each other in vitro as well as in intact T cells. Finally, plasma membrane targeting of PKCθ and Akt1 exerted synergistic transactivation of the I-κB kinase β/inhibitor of NF-κB/NF-κB signaling cascade independent of T cell activation. Taken together, these findings suggest a direct cross-talk between PKCθ and Akt1 in Jurkat T cells.

PKC 1 isoenzymes are thought to reside in the cytosol in an inactive conformation and translocate to the plasma membrane upon cell activation (1). PKC was shown to co-localize selectively with the TCR to the T cell synapse when antigen-specific T cells are engaged by their physiological ligand (2,3). The isoenzyme-selective recruitment of PKC to the plasma membrane was shown to be dependent on the Vav-1/Rac-mediated pathway (4,5). Functional studies of PKC revealed an early and essential role in the TCR/CD28-induced stimulation of mitogen-activated protein kinase c-Jun NH 2 -terminal kinase/ AP-1 and nuclear factor of activated T cells but also the IKK␤/ I-B/NF-B signaling cascade (for review see Ref. 6). Consistently, PKC was found to associate with an activated IKK complex in GM 1 -enriched lipid rafts during TCR/CD28-mediated T cell activation (7). Mouse genetic evidence (employing PKC-KO mice (8)) identified the TCR/CD3-induced activation of NF-B in the IL-2 gene promoter as the major physiological function of PKC; however, its biochemical relevance as well as the manner in which PKC becomes coupled to the TCR during antigen stimulation have been largely undefined.
Similarly to PKC, the serine/threonine kinase Akt (also known as protein kinase B) has been shown to contribute to NF-B (9). In peripheral mouse T cells, Akt could be activated in response to TCR stimulation and led to enhanced NF-B activation via accelerated degradation of the NF-B inhibitory protein I-B␣ (10). In Jurkat T cells, Akt was shown to be activated enzymatically in TCR/CD3 signal transduction involving Rac and PI3K function (11). Additionally, CD28 ligation has been reported to result in the strong association with, and activation of, PI3K, subsequently recruiting and enzymatically activating Akt in both the leukemic T cell line Jurkat and freshly isolated human peripheral blood-derived normal T lymphocytes (12). Consistently, Akt function was recently shown to provide the CD28 co-stimulatory signal in T cells (13).
However, Akt was not sufficient by itself to induce cytokine promoter and NF-B reporters in Jurkat T cells, since signals from other pathways, in particular phorbol ester, the pleiotropic PKC activator, were shown to be required (9), implicating a functional Akt-PKC connection. PKC has been implicated as the prime candidate for PKC function (6), including Akt cooperation (13) in T cells; however, the biochemical basis has not been resolved. Here we addressed the detailed relationship between PKC and PI3K/Akt in TCR/CD28-mediated signal transduction. Our observations indicate the Akt/PKC cooperation as a critical process in TCR/CD28-induced signaling. Consistently, plasma membrane targeting of both PKC and Akt (e.g. by recombinant NH 2 -terminal myristoylation motifs) is sufficient to bypass TCR/CD28 ligation in order to start the IKK␤/I-B/NF-B-signaling cascade. The results provide evidence for a model of a direct cross-talk between PKC and Akt, downstream of TCR/CD3 and CD28 co-receptor, respectively, in the critical NF-B signaling pathway.

EXPERIMENTAL PROCEDURES
Reagents and Plasmids-MG132, LY294002, and Gö6850 were purchased from Alexis, Lausen, Switzerland. [␥-32 P]ATP was purchased from PerkinElmer Life Sciences, and PDBu and ionomycin were from Sigma. The monoclonal antibody anti-HA was obtained from BMB, Vienna, Austria; the antibodies that recognize PKC isoenzymes were from Signal Transduction Labs, Upstate Biotechnology, Inc., and Santa Cruz Biotechnology, respectively; the anti-Akt antibody, used for immunoprecipitated as well as the anti-Ser(P) 473 -Akt antibody were from Cell Signaling Technologies (New England Biolabs); and the anti-Akt mAb was from Signal Transduction Laboratories. The antibody used for immunoprecipitated PKC was obtained from Santa Cruz Biotechnology (sc-1875). The polyclonal phosphospecific PKC antibody was raised against phosphothreonine containing 8 amino acid peptide sequences coupled to keyhole limpet hemocyanin. The antibody was sequentially purified by protein A and affinity chromatography employing the threonine-phosphorylated immunogen. The antibodies used for stimulation of T cells, anti-human CD28 mAb, clone 28.2 were from PharMingen (San Diego, CA); the OKT3 mAb, the anti-TCR IgM (number 305), and the anti-CD28 IgM were a gift from A. Altman and A. Weiss, respectively. The NF-B-and IL-2-promoter luciferase reporter plasmids were obtained from M. Karin and G. R. Crabtree, respectively. Human PKC, rat PKC⑀, mouse PKC, and bovine PKC␣ wt cDNAs were subcloned into the pEF-neo, respectively, and constitutively active mutants thereof have been described (14). Recombinant baculovirus-expressed proteins of inactive and PDK1-phosphorylated, active Akt were obtained from Upstate Biotechnology, Inc. Recombinant baculovirus-expressed protein of PKC wt was expressed in Sf21 cells and purified with nickel-nitrilotriacetic acid-agarose (Qiagen).
Construction of Membrane-targeted PKC Mutant cDNAs-By using recombinant overlap-extension polymerase chain reaction, the nucleotide sequences encoding the Myr peptide (derived from v-Src, NH 2 -MGSSKSKPKDPSQR-COOH) has been introduced into the PKC cDNA (see Table I), generating pEF-PKC NH 2 -terminal fusion mutant Myr-PKC. Correct constructs have been identified by restriction analysis and partial DNA sequencing, using PKC cDNA and vector-specific sequencing primers.
Cells and Transfections-Jurkat T cells were maintained in RPMI medium supplemented with 10% fetal calf serum (Life Technologies, Inc.). Transient transfection of cells was performed by electroporation in a BTX T820 ElectroSquarePorator (ITC, Biotech, Heidelberg, Germany) apparatus using predetermined optimal conditions: 2 ϫ 10 7 cells at 450 V/cm and 5 pulses of 99 ms. Optimix medium (Equibio, Kent, UK) was used for studies of promoter reporter gene expression.
Cell Fractionation-1 ϫ 10 7 Jurkat T cells per assay point were transfected with 5-20 g of the various cDNA expression plasmids, encoding CD28 receptor, wild-type or mutant Myr PKC, or empty vector controls, as indicated, for subcellular fractionation experiments of PKC. After incubation for 21 h the cells were stimulated with solid-phase antibodies against CD3/TCR and/or CD28 for 15 min at 37°C or left unstimulated. Cell fractionation was performed by subsequent lysis in equivalent amounts of different buffers (without detergent-soluble, s fraction; containing 1% Nonidet P-40-particulate, pt fraction; containing 2% SDS-nonsoluble, ns fraction), as described (15).
Promoter Reporter Gene Analysis-Reporter gene expression was measured in co-transfection assays using 5 g of pSR␣-CD28, 15 g of the expression vector, and 15 g of the relevant promoter firefly luciferase reporter (RLU1) as described (15,16). For normalization, 0.3 g of the Renilla luciferase reporter vector pTK-Renilla-Luc (Promega, Medison, WI) (RLU2) has been used. After 24 h cells were stimulated with 50 ng/ml PDBu and 1 g/ml ionomycin or soluble anti-CD28 and anti-TCR IgM (10 g/ml each) for 16 h or left unstimulated, as indicated. Treatment of synthetic inhibitor compounds, as indicated, has been done at concentrations of 2.5 (MG132) and 25 M (LY294002) and 50 nM (Gö6850), respectively.

PKC Is Recruited and Catalytically Activated by TCR/CD3
Cross-linking-Initially, the distinct role(s) of TCR/CD3 and/or CD28 in engagement of endogenous PKC in Jurkat as well as peripheral T cells were investigated. As a first result, ligation of TCR/CD3 was shown to be sufficient to result in maximal plasma membrane recruitment of PKC in Jurkat T cells (using both IgG as well as IgM-agonistic antibody clones coupled to beads). CD28 co-cross-linking did not significantly enhance TCR/CD3-induced PKC membrane translocation ( Fig. 1), even after recombinant CD28 overexpression (not shown). However, and in part consistent with a previous report (17), soluble anti-CD3 antibodies proved less effective in inducing PKC plasma membrane recruitment (not shown).
Next, and more importantly, PKC in situ activity was monitored, employing anti-phospho-(p)PKC immunoblotting of endogenous PKC immunoprecipitates derived from whole cell extracts. The PKC autophosphorylation-specific antibody used and characterized in this study selectively recognized PKC in its active state. Phorbol ester stimulation induced a marked immunoreactivity of endogenous PKC in isolated human PBMCs ( Fig. 2A). Consistently, phorbol ester induced rapid immunoreactivity of highly purified recombinant enzyme preparations in in vitro autophosphorylation assays (Fig. 2B). This activation-induced phospho-status of PKC was shown to be dependent on the catalytic activity of PKC; only immunoprecipitates of transiently overexpressed and phorbol ester activated wild-type (wt) but not kinase-dead mutant PKCK409R enzyme were recognized by this phosphospecific PKC antibody (Fig. 2C). Finally, the established CA mutant PKCA148E (14) exhibited a constitutive high level of phosphorylation, as detected by this anti-(p)PKC antibody, independent of any In order to amplify and clone in a single step fusion mutant PKC-cDNA encoding the desired Myr sequences, PKC hybrid primers plus the appropriate anchor primers (in order to ensure efficient amplification of the fusion mutant PKC-PCR fragment) have been used.

5Ј-TTTGGAGCGGCCGCTACTCAGCGGCCAACACACA-3Ј
a ANP, anchor primer, used in combination with the appropriate hybrid primer.

PKC/Akt1 Cross-talk
activation stimulus (Fig. 2C). Employing this anti-active-(p)PKC antibody, the distinct role(s) of TCR/CD3 and/or CD28 engagement in the phosphorylation status of endogenous PKC in T cells were investigated. Although, Jurkat T cells exhibited some basal levels of endogenous PKC phosphorylation, it was significantly induced upon TCR/CD3 or PDBu control stimulation (Fig. 2D). However, and consistent to data in Fig. 1, PKC phosphorylation did not show any significant dependence on CD28 co-stimulation in these cells, even after recombinant CD28 overexpression (not shown). Similarly, ligation of CD3 (but not CD28) significantly induced phosphorylation of endogenous PKC in human PBMCs (not shown). Taken together, stimulation via the TCR/CD3 receptor, similar to PDBu stimulation, results in the marked cytosol-to-membrane translocation and in situ catalytic activation, as determined by the autophosphorylation status of PKC in T cells. CA-PKC A148E Mutant Function Requires PI3K-As reported previously, only PKCA148E by itself (but none of other CA mutants of T cell-expressed PKC isoforms) was sufficient to stimulate the NF-B reporter in Jurkat T cells (15,17,18) but not NIH3T3, COS-1 (15), and EL4 cells (14). Since among those cells, only Jurkat T cells demonstrate complete loss-of-function in phosphatase and tensin homolog (PTEN) (19), an established negative regulator of PI3K signaling (20), we next investigated potential functional interaction of PKCA148E and PI3K signal transduction. Additionally, PI3K is an established major signaling molecule downstream of the CD28 (21), the co-receptor shown to be functionally required for maximal NF-B transactivation (for review see Ref. 6).
By using two independent approaches, i.e. PTEN functional reconstitution of PTEN-deficient Jurkat T cells and pharmacological PI3K inhibition, we demonstrated that endogenous PI3K activity is required for PKC-mediated NF-B transactivation. First, PKCA148E function was markedly inhibited by PTEN expression as compared with an inert protein expression control (Fig. 3A). Second, treatment of LY294002, a selective PI3K inhibitor, abrogated NF-B transactivation (Fig. 3A) as well as the established PKCA148E/CaN-induced IL-2 promoter induction (Fig. 3B), (22,23).
Akt Function Modulates PKC-mediated NF-B Transactivation-In order to define the PH-containing signaling protein acting downstream of PI3K, Akt wt and Akt mutants have been employed to investigate the NF-B signaling cascade(s) induced by PKC. As shown in Fig. 3C, kinase-active Akt overexpression enhanced the effect of PKC. In this respect, the Akt PH domain appears indispensable for PKC cooperation, since a PH deletion mutant of Akt completely failed to enhance PKC-mediated NF-B induction (Fig. 3C). This may indicate the Akt PH domain is critical for the PKC interaction, similar to a reported Akt-PKC␦ interaction (24). Conversely, DN mutants of Akt inhibited PKCA148E-induced NF-B activity. The biological relevance of this inhibition is confirmed by the finding that a DN-Akt also significantly reduced the activationinduced NF-B signal (Fig. 4B).
Expression of combinations of CA mutant PKCA148E and defined wild-type and mutant IKK␤ proteins, the established downstream effector of PKC (7,17,18), demonstrated comparable results in parallel experiments (Fig. 3D). Overexpression of IKK␤ wt enhanced the effect of PKC. Consistently, the DN mutants IKK␤K/M and (as a positive control) PKCK409R inhibited PKCA148E action, respectively. An involvement of Rac1 GTPase function in the PKCA148E-mediated NF-B induction has also been found ( Fig. 3B and see Ref. 5). Combined, PKC action appeared to be functionally mediated via Akt and IKK␤ induction, the latter also shown by IKK␤ kinase complex assays (17,18).
Additionally and to define further PKC subfamily involved in Akt-mediated NF-B induction, Jurkat T cells were co-transfected with expression vectors encoding Akt as well as representative CA-PKC mutant isoforms, e.g. PKC␣, -⑀, -, and -. These experiments identify predominantly PKC but also at a reduced level PKC⑀ (but not PKC␣ & ) as PKC isoenzymes capable to cooperate with Akt in Jurkat T cells (not shown). In this regard, a role of PKC⑀ in NF-B signal induction has been described previously (25).
PKC and Akt1 Functions Are Both Required for Maximal NF-B Transactivation-Next, overexpression experiments of wt and DN mutants of PKC and Akt in phorbol ester/ionomy-cin-stimulated Jurkat T cells identified both PKC and Akt as predominant mediators of NF-B activation. Overexpression of wt enzymes of PKC or Akt led to 4-fold increase of activationinduced NF-B activity over that obtained in vector controls (Fig. 4A). Consistently, expression of kinase-dead PKC K409R or Akt triple mutant K179A,T308A,S473A, both shown to act in a dominant negative fashion in intact cells (9,14), significantly reduced the activation-induced NF-B activity (Fig. 4A). Similarly, reduced effects were obtained under more physiological conditions, since expression of DN mutants of Akt and PKC were able to decrease the CD3/CD28-induced NF-B signal. However, no complete inhibition could be achieved most likely due to multiple converging signals. Consistently, co-expression of both PKC wt and Akt wt did further enhance the CD3/ CD28-induced NF-B activity (Fig. 4B).
Complex Formation of PKC and Akt-As a next step, we investigated whether Akt is a potential downstream substrate of PKC. In vitro complex kinase assays showed that the phosphorylation status of neither PKC nor Akt was prominently affected by each other (Fig. 5A). PKC transphosphorylationspecific inhibitor Gö6850 (which does not affect PKC-auto- phosphorylation) 2 did not decrease (but rather increased) Akt phosphorylation status, excluding PKC as a prominent protein kinase of Akt. More importantly, no increase of Akt-Ser 473 phosphorylation could be observed in PKCA148E-transfected cells compared with the control (Fig. 5B). Vice versa, in vitro phosphorylation of PKC was not modulated by Akt (both preactivated or not by PDK1 at Thr 308 ) (Fig. 5A). Co-expression of neither Myr-Akt nor DN-Akt mutants in Jurkat T cells showed any modulation of phosphorylation of Myr-PKC (Fig. 5C). Consistently, phorbol ester-induced autophosphorylation of endogenous PKC protein was unaffected by DN-Akt expression (Fig. 5D).
Activation of PI3K frequently involves the PIP 3 -mediated recruitment of Akt to the plasma membrane, similar to the TCR/CD3-induced translocation of PKC (see Fig. 1). Thus, it was tempting to speculate that a physical interaction may underlie the observed functional cooperation of these two protein kinases. To assess this hypothesis, we performed co-immunoprecipitation of PKC and Akt in vitro as well as in intact cells. As a result, purified recombinant Akt (both preactivated or not by PDK1 at Thr 308 ) was found to associate with purified recombinant PKC in vitro indicating a direct physical PKC-Akt interaction (Fig. 6A). Consistently, reverse co-immunopre-cipitation confirmed this association (Fig. 6B); however, a minor nonspecific PKC fraction has been repeatedly found to stick to the beads independently of immunoprecipitated Akt (Fig. 6B, compare the 3rd lane to the 1st and 2nd lanes). In Jurkat T cells the association of Akt with endogenous PKC was found to be constitutive (Fig. 6C). Identically, no modulation of this constitutive Akt/PKC interaction by TCR/CD28 or PDBu stimulation was observed in reverse immunoprecipitation results (not shown). Combined, these observations demonstrate that PKC and Akt associate in preformed complexes in vivo; however, enzymatic activities of neither Akt nor PKC appear to be significantly modulated by each other. We therefore speculated that the observed PKC-Akt interaction may specify cellular location and impose integration with other signaling systems, e.g. local substrate availability as well as exposure to allosteric activators. ristoylation signal (Fig. 7A). Similar myristoylation signal protein modifications have been made successfully for PKC␣ (26) and -isoenzymes (27). Overexpression in Jurkat T cells revealed substantial plasma membrane targeting but also accumulation in the detergent-ns fraction of Myr-PKC shown by cellular fractionation (Fig. 7B) as well as immunofluorescence analysis in ectopic expression studies (not shown). In contrast, CA-PKCA149E mutant was found to reside predominantly in the ns fraction (Fig. 7B). This subcellular location indicated its different nature of activation, as observed by its high transactivation capability (Fig. 7C). Additionally, accumulation of activated forms of PKC in the detergent ns fractions may implicate a particular signaling function at this location.
Several signaling molecules, among them p56 lck , known to bind to and phosphorylate PKC at tyrosine 90 are localized in this subdomains in intact T cells (28).
Functional studies of Myr-PKC resulted in significant stimulation the IL-2 signaling pathway (65-fold, Fig. 7C), as compared with PKC-wt, excluding a simple overexpression artifact. To control for non-PKC-specific myristoylation effect and/or decoy effect on the N-myristoyltransferase machinery, another myristoylation fusion mutant protein, Myr-ERK2 (fully characterized in Ref. 29), was used in our experimental setup. Myr-ERK-2 (Fig. 7C) as well as the expressed catalytic subfragment of PKC (found to reside exclusively in the cytosolic fraction) had no effect on IL-2 signaling (not shown). PKCCAAX, a COOH-terminal farnesylation and palmitoylation signal fusion mutant, revealed substantial plasma membrane targeting but no transactivation of its downstream effectors. 2 These results indicate a gene-specific Myr-PKC effect on its downstream targets and demonstrate Myr-PKC as a novel agonist-independent PKC mutant. Myr-PKC similar to the established CA mutant PKCA148E (but not PKCwt) exhibited high levels of basal phosphorylation, as detected by the anti-active(p)PKC antibody (Fig. 7D), confirming constitutive autophosphorylation.
However, and despite full in situ catalytic activity of PKC, as monitored by the anti-active(p)PKC result (Fig. 7D), Myr-PKC fusion mutant by itself was significantly less potent to transactivate (in combination with an ionophore) the IL-2 promoter than the established CA pseudosubstrate acidic exchange mutant PKCA148E (14) (Fig. 7C). This observation suggested to us that additional activation-dependent signal(s) are required to cooperate with Myr-PKC for maximal NF-B signaling in these cells.
Membrane Targeting of PKC Exaggerates the TCR/CD28induced NF-B Activation-Consistently with this hypothesis, stimulation by the TCR and/or CD28 in Myr-PKC (and much less PKC-wt)-transfected cells further enhanced the induction of the NF-B reporter severalfold over the vector-transfected cells (Fig. 8). Given a near-maximal observed transcriptional NF-B activity in just CD28 ligation-induced and Myr-PKCexpressing T cells, subcellular location of PKC to the plasma membrane appears to bypass the TCR/CD3 stimulation requirement for NF-B transactivation. Although it is evident FIG. 7. Characterization of membrane-targeted PKC fusion mutant, Myr-PKC. A, Myr-PKC was constructed by fusing a p60 srcderived myristoylation signal to the NH 2 terminus of PKC-wt. The CA-PKCA148E mutant is shown as well for comparison. B, Jurkat T cells were transfected with the indicated pEF-neo empty vector control, PKC-wt, Myr-PKC, or expression plasmids, respectively. Cytosol (soluble (s)), membrane (particulate (pt)), and cytoskeletal fractions (ns) were prepared and resolved by SDS-PAGE, and the relative distribution of PKC in each fraction was determined by immunoblotting. As positive evaluation for cell fractionation, immunostaining of p59 fyn predominantly recognized the particulate fraction. C, Jurkat cells were electroporated with the IL-2 promoter/luciferase reporter plus the indicated PKC or control expression constructs. 21 h posttransfection, the cells were harvested after 16 h of stimulation with ionomycin (1 g/ml), and reporter activity was determined, using luciferase reporter gene expression, as described (15). Statistical analysis of three independent experiments was evaluated as the means of fold induction Ϯ S.E. Expression of all transgenes has been confirmed by immunoblotting (not shown). The Myr-ERK2 (29) revealed a similar subcellular distribution as the Myr PKC fusion mutant (not shown). The constitutively active PKCA148E has been described (14). D, PKC immunoprecipitations of Jurkat T cell lysates, transfected 24 h earlier with pEF-neo control vector, PKC-wt, Myr-PKC, or PKCA148E expression vectors (as indicated) were subjected to immunoblotting with the anti-(p)PKC or anti-PKC antibodies, respectively. that TCR/CD3 stimulation is sufficient to recruit and activate PKC (see Fig. 1), the role of CD28 co-stimulation in the NF-B induction remains undefined. One likely explanation is that expression of Myr-PKC mutants mimic in part the TCR/CD3 signal leading to NF-B activation and that CD28 ligation induces a distinct signaling pathway able to functionally cooperate with PKC. In this regard, CD28 ligation has been reported to induce Akt (via PI3K action) in T lymphocytes (12,13). Consistently, induction of NF-B by expression of the established CA mutant PKCA148E was enhanced by CD28 but not TCR/CD3 ligation (see Refs. 13 and 18 and data not shown).

Membrane Targeting of Both PKC and Akt Leads to a Synergistic Transactivation of the IKK␤/I-B/NF-B Signaling
Cascade-Along this hypothesis, we employed our constitutively membrane-targeted Myr-PKC mutant in combinations with Akt, Vav1, PDK1, or emt/itk. Consistent with a crucial role of subcellular location, Myr-PKC dramatically synergized with Myr-Akt (but also at a reduced level with Akt-wt) to activate the NF-B reporter (Fig. 9, A and B). In contrast, the Myr mutants of Akt and PKC had only very little activities by themselves, and no potentiation of Myr-PKC action was observed by co-expression of any other PH domain-containing signaling molecule used (Fig. 9C). Particularly, no change in TCR/CD3-induced translocation as well as activation-induced autophosphorylation of endogenous PKC has been observed by PTEN expression (not shown), making a simple, direct effect of PI3K via PDK1, the established PKC kinase (27,30), on PKC unlikely. Consistently, no major changes in activation loop phosphorylation of PKC by PDK-1 could be detected (see similar findings for PKC␦ in Ref. 31). 2 Along this line, phosphorylation status of Myr-PKC was found not to be affected by PTEN expression (not shown).
However, a strict dependence on high PIP 3 levels (via endogenous PI3K activity) has been observed for Myr-PKC/Akt-wtmediated induction of the NF-B reporter (Fig. 9D), as already observed for CA-PKCA148E function (Fig. 3, A and B). Presumably this is through PTEN-mediated reduction of PIP 3 levels and therefore inactivation of Akt wt in PTEN-expressing Jurkat T cells. Consistently, synergistic cooperation of membrane-targeted mutants Myr-PKC/Myr-Akt was affected to a much lower degree by PTEN expression (Fig. 9D). The Myr-Akt mutant has been shown to be activated mostly independent of PI3K function (32). The still observed partial (37%) inhibition effect is most likely due to a Myr-Akt fusion mutant pool, not being post-translationally myristoylated and therefore still dependent on high PIP 3 level for activation. However, Myr-PKC/ Myr-Akt mediated NF-B induction was still found to involve IKK␤ (as shown by DN-IKK␤-mediated inhibition, Fig. 10A) and was inhibited by the proteasome inhibitor MG132 (which blocks I-B degradation, Fig. 10B). Consistently, TCR/CD28 signal transduction (as well as CA PKCA149E function) was shown to activate the transcription factor NF-B via IKK␤mediated phosphorylation and subsequent degradation of I-B by the proteasome/ubiquitin pathway (7,15,17,18).
Additionally, the PKC/Akt cooperation in NF-B induction was dependent on the enzymatic activity of PKC, indicated by PKC inhibitor Gö6850-mediated functional abrogation (Fig.  10B). Consistently, no synergy of Myr-Akt with the "kinasedead" double mutant PKCA148E-K409R, DN-mutant PKCK409R, as well as catalytic or regulatory subfragments of PKC has been found (not shown). Together with the significant decrease of phorbol ester/ionomycin-and TCR/CD28-mediated induction of NF-B reporter by kinase-dead DN mutants of PKC and Akt (Fig. 4, A and B), these findings indicate that the catalytic activity but not a simple scaffold-function of PKC is required for the observed Akt cooperation.
Finally and in contrast to a strict dependence of PKCA148E-induced NF-B activity on Rac1 function (see Fig. 3D), membrane-targeted Myr mutants of PKC and Akt1 are capable of inducing NF-B signaling independent of DN-Rac1 N17 GTPase expression (Fig. 10A), indicating that Vav/Rac is involved mainly in recruiting PKC and/or Akt to the plasma membrane (5) but not in downstream signaling events. Consistently, phorbol ester, a direct PKC stimulation agent, can bypass the functional defects in T cells deficient for Vav1, WASP, or Rac1 function (33,34).
Our results identify an association between PKC and Akt and reveal functional cross-talk between PKC and PI3K/Akt in the TCR/CD28 signal transduction, which involves NF-B activation. This is consistent with the hypothesis that recruitment of both PKC and Akt to the plasma membrane appears to be the primary mechanism leading to activation of the NF-B pathway. Indeed, the recovery of Akt in co-immunoprecipitation experiments with PKC and the fact that plasma membrane location of both PKC and Akt can bypass TCR/ CD28 signaling as well as Rac1 GTPase function to start the IKK␤ signaling cascade (Fig. 10A) strongly argue in favor of this hypothesis.
T cells from transgenic mice expressing a CA-Akt mutant displayed resistance to a variety of apoptotic stimuli (10). The Akt-mediated protective effect is shown to be mediated by either NF-B activation (10) or phosphorylation of BAD at Ser 136 (35,36). Interestingly, phosphorylated BAD is sequestered in the cytoplasm to 14-3-3 (the 14-3-3 isoform shown to interact with cytosolic PKC (37)). Similarly, activated PKC has recently been shown to provide survival signals in T cells, which can be accounted in part by its ability to induce, via RSK2, BAD phosphorylation (38).
A future main question, however, remains concerning the direct target(s)/substrate(s) for PKC and Akt which impose integration with the NF-B pathway and potentially other signaling systems. Subsequently, NF-B target genes are not only involved in the T cell immune response (IL-2 and IL-2R␣) but also the inflammatory response (tumor necrosis factor-␣ and -␤, IL-1, and IL-6), cell adhesion (I-CAM, V-CAM, and E-selectin), and cell growth (p53, Ras, and c-Myc) (39). Taken together, our present findings might provide an interesting mechanism whereby the status of PKC and Akt in the given T cell may have broad implications for T cell growth and survival and, consequently, affect T cell fate during clonotypic expansion.