Roles of Tyrosine Phosphorylation and Cleavage of Protein Kinase Cδ in Its Protective Effect Against Tumor Necrosis Factor-related Apoptosis Inducing Ligand-induced Apoptosis*

Protein kinase Cδ (PKCδ) regulates cell apoptosis in a cell- and stimulus-specific manner. Here, we studied the role of PKCδ in the apoptotic effect of TRAIL in glioma cells. We found that transfection of the cells with a PKCδ kinase-dead mutant (K376R) or with a small interfering RNA targeting the PKCδ mRNA increased the apoptotic effect of tumor necrosis factor-related apoptosis inducing ligand (TRAIL), whereas overexpression of PKCδ decreased it. PKCδ acted downstream of caspase 8 and upstream of cytochrome c release from the mitochondria. TRAIL induced cleavage of PKCδ within 2–3 h of treatment, which was abolished by caspase 3, 8, and 9 inhibitors. The cleavage of PKCδ was essential for its protective effect because overexpression of a caspase-resistant mutant (PKCδD327A) did not protect glioma cells from TRAIL-induced apoptosis but rather increased it. TRAIL induced translocation of PKCδ to the perinuclear region and the endoplasmic reticulum and phosphorylation of PKCδ on tyrosine 155. Using a PKCδY155F mutant, we found that the phosphorylation of PKCδ on tyrosine 155 was essential for the cleavage of PKCδ in response to TRAIL and for its translocation to the endoplasmic reticulum. In addition, phosphorylation of PKCδ on tyrosine 155 was necessary for the activation of AKT in response to TRAIL. Our results indicate that PKCδ protects glioma cells from the apoptosis induced by TRAIL and implicate the phosphorylation of PKCδ on tyrosine 155 and its cleavage as essential factors in the anti-apoptotic effect of PKCδ.

Apoptosis is a genetically controlled process of selective cell deletion involved in normal cell development and turnover (1). Apoptosis is triggered by genotoxic reagents, stress responses, or ligation of membrane death receptors (2). The apoptotic response is characterized by extrinsic and intrinsic pathways (3) that can converge at the mitochondria leading to the release of cytochrome c and activation of caspase 9 and downstream effector caspases (4). Proteins that play a role in this process include the Bcl2 family (5) and the caspases, which are cysteine-dependent aspartate-specific proteases (6,7). In addition, the apoptotic response is also regulated by various kinases (8), among the best characterized being phosphoinositide 3-kinase/ Akt (9), members of the mitogen-activated protein kinase family (10), and various PKC isoforms (11).
PKC␦, 1 a ubiquitously expressed isoform of the novel PKC subfamily (12), has been shown to regulate cell apoptosis in various cellular systems (13). Its function and regulation are clearly complex, however, because PKC␦ can be either proapoptotic or anti-apoptotic, depending on the specific cellular system and the specific ligand. PKC␦ acts in a pro-apoptotic fashion in response to etoposide (14,15), cytosine arabinoside (16), and in response to UV (17) and ␥-radiation (18). The pro-apoptotic function of PKC␦ in these systems is typically associated with its cleavage by caspase 3 in the hinge region, which results in a phospholipid-independent activation of the enzyme (19 -22). It is noteworthy, that PKC␦ can induce apoptosis in a cleavage-independent (23) or a kinase-independent manner (24) in some instances, emphasizing the complex role of PKC␦ in the regulation of cell apoptosis.
In other systems, in contrast to the above examples, PKC␦ has been clearly shown to function in an anti-apoptotic fashion. Thus, PKC␦ played a role in the anti-apoptotic effect of fibroblast-like growth factor in granulosa cells (25) and in serumdeprived PC12 cells (26) and protected macrophages from nitric oxide-induced apoptosis (27). Factors that are associated with the effects of PKC␦ on cell apoptosis are its phosphorylation on tyrosine residues and its subcellular localization. Tyrosine phosphorylation of PKC␦ occurs in response to many apoptotic stimuli and it has been associated with the apoptotic function of PKC␦ (15, 28 -31). Similarly, PKC␦ translocates to the mitochondria (32), Golgi (31), and nucleus (15,33) in response to various stimuli and specific apoptosis-related PKC␦ substrates have been identified in the mitochondria (34) and nucleus (21,35).
Tumor necrosis factor-related apoptosis inducing ligand (TRAIL; Apo2 ligand) belongs to the tumor necrosis factor superfamily and induces apoptosis in many transformed cells including some glioma cell lines and primary cultures (36,37).
TRAIL acts by formation of the death-inducing signaling complex that is also common to other members of the death receptors (38). Caspase 8 is then activated at the death-inducing signaling complex and leads to two different apoptotic pathways: a mitochondrial-independent pathway via the effector caspases 3 and 7 and a mitochondrial-dependent pathway via activation of caspase 9 (39). The signal transduction pathways that are involved in the TRAIL apoptotic effect are not fully understood. Recent studies reported that TRAIL activates the transcriptional factor NF-B and c-Jun N-terminal kinase in various cellular systems (40) and that NF-B (41) and phosphatidylinositol 3-kinase/AKT (42) are involved in the resistance of some transformed cells to the apoptotic effect of TRAIL. In addition, PKC signaling, mainly in response to PMA activation, has been implicated in the attenuation of the apoptotic signal induced by TRAIL (43,44); however, the involvement of the PKC␦ isoform in response to TRAIL has not been described.
In the present study, we examined the role of PKC␦ in the apoptotic effect of TRAIL in glioma cells and analyzed the molecular mechanisms underlying its effects. We found that PKC␦ protected glioma cells from the apoptosis induced by TRAIL and that this effect was dependent on the phosphorylation of PKC␦ on tyrosine 155 and on its caspase-dependent cleavage.
Site-directed Mutagenesis of PKC␦-PKC␦ cloned into the pEGFP plasmid served as a template vector for the site-directed mutagenesis using the QuikChange Site-directed Mutagenesis Kit (Stratagene, La Jolla, CA). Conversion of tyrosine residues at sites 52, 155, and 187 into phenylalanine was performed as previously described (45). A PKC␦ K376R dominant mutant was generated as previously described (15). The caspase cleavage site of PKC␦, aspartate 327, was mutated to alanine using the following primers: sense, 5Ј-GTGACATCCTAGC-CAACAACGGGACC-3Ј and antisense, 5Ј-GGTCCCGTTGTTGGCTAG-GATGTCAC-3Ј. The mutation was confirmed by DNA sequencing. PKC␦ and the PKC␦ mutants were subcloned into the pCMV-Tag 2b expression vector (Stratagene) generating PKC␦ or PKC␦ mutants with a N-terminal FLAG tag. The activities and translocations of the PKC␦ tyrosine mutants and the PKC␦K376R were already characterized (15,28,45). The activity of the PKC␦D327A mutant was similar to that of PKC␦ WT using the immune complex kinase assay. Similarly, PKC␦D327A translocated to the plasma and perinuclear membranes in response to PMA, similar to the translocation of PKC␦ (data not shown).
Construction of PKC␦-GFP and PKC␦155 Fusion Proteins-cDNAs encoding the murine PKC␦ and the PKC␦Y155F mutant were fused into the N-terminal enhanced GFP vector pEGFP-N1 (Clontech Laboratories, Palo Alto, CA) as previously described (45). Briefly, the original pEGFP-N1 vector was modified by the insertion of a MluI site in the plasmid polylinker and the clones containing the GFP-PKC␦ or GFP-PKC␦Y155F were constructed by the excision of PKC␦ or PKC␦Y155F from pCMV tag 2b PKC plasmids by digestion with XhoI and MluI. The inserts were then ligated into the modified GFP vector using the same restriction sites. DNA sequencing of the GFP-PKC constructs confirmed the intended reading frame. The expression and activities of these plasmids were already described (15,28,45).
Construction of the PKC␦ Catalytic and Regulatory Domains-The construction of the PKC␦ regulatory and catalytic domains was performed using the following primers: PKC␦-reg sense, GATGCTCG-AGGCCACCATGGCACCCTTCCTTCGCATTT; PKC␦-reg, ASGATGA-CGCGTGTCTAGGATGTCACTCCCAGAGAC; PKC␦-cat sense, GATG-CTCGAGGCCACCATGAACAACGGGACCTATGGCAAG, and PKC␦cat, ASGGACGCGTAATGTCCAGGAATTGCT. The PKC␦ regulatory and catalytic fragments were subcloned into the EGFP and the pCMV tag 2b expression vectors.
Generation of PKC Chimeras-The PKC chimeras were generated by exchanging the regulatory and catalytic domains of PKC␣ and -␦ as previously described (15). PKC␣/␦ refers to the chimera with the PKC␣ regulatory domain and the PKC␦ catalytic domain, and PKC␦/␣ refers to the reciprocal chimera. The PKC cDNAs were subcloned into the pCMV tag 2b (Stratagene) generating PKC chimeras with an N-terminal FLAG tag. The expression of these chimeras and their activities were recently described (15,45).
A172 Glioma Cultures and Cell Transfection-A172 cells were grown in medium consisting of Dulbecco's modified Eagle's medium containing 10% heat-inactivated fetal calf serum, 2 mM glutamine, penicillin (50 units/ml), and streptomycin (0.05 mg/ml). The cells were transfected by electroporation using the Nucleofector device (Amaxa Inc., Gaithersburg, MD) either with the empty vector or with expression vectors for the PKC␦ and PKC␦ mutants. Transfection efficiency using nucleofection was about 80 -90%.
siRNA Transfections-siRNA duplexes (siRNAs) were obtained from Dharmacon (Lafayette, CO) and consisted of a pool of 4 PKC␦ siRNA duplexes. Transfection of siRNAs was performed using 50 nM PKC␦ or scrambled siRNAs and Oligofectamine (Invitrogen, Carlsbad, CA) according to the manufacturer's instructions. PKC␦ protein levels were determined using Western blot analysis.
Adenovirus Preparation and Infection-The AdEasy system was kindly provided by Dr. Vogelstein (The Johns Hopkins University School of Medicine, Baltimore, MD) (46). PKC␦ and the different PKC␦ mutants were first cloned into the pShuttle-CMV vector as previously described (47). The plasmids were then linearized by digestion with PmeI and were transformed into Escherichia coli BJ5183-AD-1 competent cells (Stratagene) carrying the pAdEasy-1 plasmid that encodes the adenovirus-5 backbone. Recombination was confirmed by restriction and PCR analyses. The linearized recombinant plasmids were transfected into HEK293 cells. Viruses were collected after 7-10 days and were further amplified. An adenovirus expressing the LacZ gene was used as a control.
Cells were incubated with 5 multiplicity of infection of the appropriate recombinant adenovirus vectors for 1 h. The medium was then replaced with fresh medium and the cells were used 24 -48 h post-infection.
Measurements of Cell Apoptosis-Cell apoptosis was measured using propidium iodide staining and analysis by flow cytometry (15) and ELISA (Cell Death Detection ELISA Kit) using anti-histone antibodies. Cells (1 ϫ 10 6 /ml) were plated in 6-well plates and subjected to the indicated treatments. Detached cells and trypsinized, adherent cells were pooled, fixed in 70% ethanol for 1 h on ice, washed with PBS, and treated for 15 min with RNase (50 M) at room temperature. Cells were then stained with propidium iodide (5 g/ml) and analyzed on a BD Biosciences cell sorter.
For anti-histone ELISA (Cell Death Detection ELISA kit), cellular extracts containing histone-associated DNA fragments were incubated in 96-well plates coated with anti-histone antibodies for 2 h. Plates were then washed and incubated with anti-DNA antibodies conjugated to peroxidase for an additional 2 h. Substrate solution was added and absorbance was measured at 405 nm.
Measurement of Caspase 8 Activity-Caspase 8 activity was measured using the caspase fluorescent substrate/inhibitor QuantiPak assay kit obtained from BIOMOL (Plymouth Meeting, PA) using the fluorescent substrate Ac-IETD-AMC according to the manufacturer's recommendations.
Cytochrome c Release-Cytochrome c release from the mitochondria was determined in the cytosolic fraction. Mitochondrial and cytosolic fractions were isolated using the ApoAlert Cell Fractionation Kit (Clontech, BD Biosciences) according to the manufacturer's instructions. Briefly, control and TRAIL-treated cells were centrifuged at 600 ϫ g for 5 min at 4°C. Cell pellets were resuspended with 0.8 ml of ice-cold fractionation buffer and incubated on ice for 10 min. Cells were then homogenized with an ice-cold Dounce homogenizer and centrifuged at 700 ϫ g for 10 min. The supernatants were then centrifuged at 10,000 ϫ g for 25 min at 4°C and the supernatants (cytosolic fraction) and pellets (mitochondrial fraction) were collected. Cytochrome c was identified in the cytosolic fraction by using a rabbit anti-cytochrome c antibody.
Immunoblot Analysis-Lysates were resolved by SDS-PAGE and transferred to nitrocellulose membranes. The membranes were blocked with 5% dry milk in phosphate-buffered saline and subsequently stained with primary antibody. Specific reactive bands were detected using a goat anti-rabbit or goat anti-mouse IgG conjugated to horseradish peroxidase (Bio-Rad) and the immunoreactive bands were visualized by the ECL Western blotting detection kit (Amersham Biosciences).
Immunofluorescence Staining-Cells were grown on glass coverslips. Following TRAIL treatment, the cells were washed with PBS and fixed in 4% paraformaldehyde for 15 min. Subsequently, cells were washed in PBS and, after blocking with staining buffer (2% bovine serum albumin and 0.1% Triton X-100 in PBS) for 30 min at room temperature, they were incubated with an anti-PKC␦ antibody. Following washes in PBS, cells were incubated with an anti-rabbit Alexa Fluor 488 antibody for an additional 60 min and mounted in FluoroGuard antifade reagent. Cells were viewed and photographed using confocal microscopy with ϫ63 magnification at an excitation wavelength of 488 nm. For the visualization of the ER, cells were incubated with mouse anti-calnexin antibody followed by incubation with anti-mouse Alexa Fluor 546 antibody. For analysis of the translocation of GFP-PKC␦ and the GFP-PKC␦Y155F mutant, cells were treated for various time points with TRAIL, fixed in 4% paraformaldehyde (PFA) for 10 min, and mounted in FluoroGuard antifade reagent.
Statistical Analysis-The results are presented as the mean Ϯ S.E. Data were analyzed using analysis of variance and a paired Student's t test to determine the level of significance between the different groups.

PKC␦ Protects Glioma Cells from the Apoptotic Effect of
TRAIL-PKC␦ regulates cell apoptosis in a cell and stimulusdependent manner (13). To examine the role of PKC␦ in the apoptotic effect of TRAIL we employed the A172 glioma cell line that exhibits high sensitivity to this ligand. Overexpression of a PKC␦ kinase-dead (KD) mutant (K376R) using an adenovirus vector ( Fig. 1A) did not affect the basal level of A172 cell apoptosis (Fig. 1B). However, it increased the apoptosis induced by TRAIL ( Fig. 1, B and C). Similar results were obtained with a siRNA targeting the PKC␦ mRNA. Transfection of the A172 cells with PKC␦ siRNA reduced the expression of PKC␦ by 80% after 3 days of transfection as compared with cells transfected with a scrambled siRNA (Fig. 1D), whereas it did not alter the expression of the other PKC isoforms (data not shown). The PKC␦ siRNA-transfected cells exhibited a higher degree of cell apoptosis in response to TRAIL as compared with cells transfected with a control scrambled siRNA (Fig. 1E). Similarly, the PKC␦ inhibitor rottlerin (5 M) increased the apoptotic effect of TRAIL by about 35% as compared with control TRAIL-treated cells (data not shown).
The role of PKC␦ in the apoptotic effect of TRAIL was also demonstrated using cells overexpressing PKC␦. Overexpression of PKC␦ by infection with an adenovirus vector ( Fig. 1A) or by transfection (data not shown) decreased cell apoptosis in response to TRAIL by about 55% (Fig. 1C). Similar results were obtained with anti-histone ELISA (data not shown). Overexpression of PKC␦ did not significantly inhibit the activation of caspase 8 by TRAIL (Fig. 1F). However, it inhibited cytochrome c release to the cytosol (Fig. 1G), suggesting that PKC␦ acted downstream of caspase 8 and upstream of the mitochondria pathway.
Cleavage of PKC␦ by TRAIL-PKC␦ undergoes caspase 3-dependent cleavage in response to diverse apoptotic stimuli (15, Twenty-four hours thereafter, the expression of PKC␦ was determined using Western blot analysis (A), the cells were treated with TRAIL (100 ng/ml) for 5 h, and cell apoptosis was determined using PI staining and FACS analysis (B and C). In another set of experiments, cells were transfected with control (scrambled) or PKC␦ siRNAs using Oligofectamine. PKC␦ expression was determined following 72 h (D) and the apoptotic effect of TRAIL (100 ng/nl) was determined after 5 h of treatment (E). Caspase 8 activation was determined in cells treated with TRAIL for 45 min using the fluorescent substrate Ac-IETD-AMC as described under "Experimental Procedures" (F) and cytochrome c release from the mitochondria to the cytosol was determined using Western blot analysis (G). The results represent the mean Ϯ S.E. of triplicate measurements in each of four experiments (C, E, and F) or are from one representative experiment out of four similar experiments. (A, B, D, and G). 19,20). To examine the effect of TRAIL on the cleavage of PKC␦ we treated the A172 cells with TRAIL (100 ng/ml) for various periods of time and examined the cleavage of PKC␦ using an anti-PKC␦ antibody that recognized the catalytic domain. We found that TRAIL induced cleavage of PKC␦ and accumulation of the PKC␦ catalytic fragment after 2-3 h of treatment ( Fig.  2A). The time course of the cleavage of PKC␦ correlated with induction of cell apoptosis by TRAIL as indicated by poly(ADPribose) polymerase cleavage (Fig. 2B) and by PI staining and FACS analysis (data not shown).
Using caspase inhibitors, we found that the cleavage of PKC␦ was inhibited by caspase 3, 8, and 9 inhibitors (Fig. 2C). These results suggest that in the A172 cells TRAIL activates both the caspase 8 and mitochondria pathways.
The Cleavage of PKC␦ Is Essential for Its Protective Effect-The cleavage of PKC␦ generates a constitutively active catalytic fragment that has been associated with the pro-apoptotic function of PKC␦ in various cellular systems (19 -21). However, the role of PKC␦ cleavage in its protective effect has not been studied. To examine the role of PKC␦ cleavage in its protective effect against TRAIL-induced apoptosis, we constructed a caspase-resistant PKC␦ mutant in which the aspartic acid at position 327 was mutated to alanine. We overexpressed the PKC␦ WT and the PKC␦D327A mutant in the A172 cells and examined their cleavage and effects on the apoptosis induced by TRAIL. As presented in Fig. 3A, overexpressed PKC␦ underwent cleavage in response to TRAIL similar to the endogenous PKC␦, whereas the PKC␦D327A did not undergo cleavage and no accumulation of a PKC␦ catalytic fragment was observed in TRAIL-treated cells. In addition, we found that in contrast to the protective effect of the overexpressed PKC␦, the PKC␦ caspase mutant did not protect the A172 cells from TRAIL-induced apoptosis but rather increased it (Fig. 3B). These results suggest that the cleavage of PKC␦ is essential for the protective effect of PKC␦ and that the PKC␦D327A mutant acted as a dominant negative mutant of PKC␦ by inhibiting its cleavage and protective effect.
Our recent studies (15) and studies from other groups (14,33) indicate that PKC␦ plays an essential role in the apoptosis induced by etoposide mainly via its cleaved catalytic fragment. Thus, PKC␦ plays an opposite role in the apoptosis induced by TRAIL and etoposide. Because PKC␦ undergoes cleavage in response to both stimuli we also examined the effect of the PKC␦D327A mutant in the apoptotic effect of etoposide. The PKC␦D327A mutant failed to undergo cleavage in response to etoposide (Fig. 3C), similar to the results obtained with TRAIL. However, in this case the PKC␦ mutant protected the A172 cells from etoposide-induced apoptosis in contrast to its apoptotic effect in TRAIL-treated cells (Fig. 3D).
Both the Regulatory and Catalytic Domains of PKC␦ Are Required for Its Protective Effect-The regulatory and catalytic domains of PKC have been associated with the regulation of cell apoptosis (19,20,48), however, their role in the protective effect of PKC␦ has not been explored. To examine the relative contributions of the different domains of PKC␦ to its protective effect we employed PKC chimeras between the regulatory and catalytic domains of PKC␣ and -␦ combined at the highly conserved hinge region. We found that overexpression of PKC␣ did not affect the apoptotic effect of TRAIL, whereas PKC␦ decreased it as already described (Fig. 1C). Cells overexpressing chimeras with the regulatory domain of PKC␦ (PKC␦/␣) or the catalytic domain of PKC␦ (PKC␣/␦) exhibited a similar degree of apoptosis to that of CV cells and none of the chimeras resembled cells overexpressing PKC␦ (Fig. 4A), suggesting that both domains of PKC␦ are required for its protective effect.
To further examine the roles of the regulatory and catalytic domains of PKC␦ in its protective effect we overexpressed these two domains in the A172 cells and examined their effects on cell apoptosis in the absence and presence of TRAIL. We found that overexpression of the PKC␦ catalytic (PKC␦-cat) domain induced some cell death in the A172 cells, whereas no apoptosis was observed with the PKC␦ regulatory domain (PKC␦-reg, Fig. 4B). The PKC␦-reg did not alter the apoptotic response of the cells to TRAIL (Fig. 4B), whereas the PKC␦-cat slightly enhanced it. These results further support the results obtained with the PKC␦ chimeras that both domains are required for the protective effect of PKC␦.
Using the PKC␦-cat and PKC␦-reg fragments tagged to GFP, we found that the catalytic domain of PKC␦ accumulated in the cytosol and to a large degree in the nucleus, whereas the regulatory domain of PKC␦ was expressed throughout the cell (Fig.  4C). The localization of the PKC␦-cat and -reg domains was not significantly altered by TRAIL treatment (data not shown).
TRAIL Induces Translocation of PKC␦ to the Perinuclear Region-One of the important factors in the diverse apoptotic functions of PKC␦ is its distinct patterns of translocation in response to different stimuli (13). To examine the effect of TRAIL on the translocation of PKC␦, we treated A172 cells with TRAIL for various periods of time and followed the localization of the endogenous PKC␦ using immunofluorescence staining and confocal microscopy. In control cells, PKC␦ was largely localized to the cytosol with some expression in the nucleus (Fig. 5A). Treatment with TRAIL induced translocation of PKC␦ to the perinuclear region and in most of the cells this translocation was followed by the exit of PKC␦ from the nucleus. Translocation was first observed after 15 min, was increased after 30 min, and persisted up to 1 h following TRAIL treatment. After that time PKC␦ was mainly localized in the cytosol (Fig. 5A)  following TRAIL treatment we used the ER marker calnexin (49). As shown in Fig. 5B, the ER marker calnexin stained the perinuclear region in a similar pattern to that of PKC␦ in the TRAIL-treated cells. Merged images clearly showed co-localization of the green fluorescence of PKC␦ and of the red fluorescence of calnexin, suggesting that TRAIL induced translocation of PKC␦ to the ER. Similar results were observed using anti-KDEL antibody (Stressgen) as an ER marker (data not shown).
TRAIL Induces Phosphorylation of PKC␦ on Tyrosine 155-Another important factor for the pro-and anti-apoptotic effects of PKC␦ is its phosphorylation on specific tyrosine residues (13,15,28,31). PKC␦ undergoes tyrosine phosphorylation in response to various apoptotic stimuli such as H 2 O 2 , ceramide , ␥-irradiation, etoposide, and infection with Sindbis virus (15, 28 -31). To examine the effect of TRAIL on tyrosine phosphorylation of PKC␦, we treated A172 cells with TRAIL for 1-120 min and the phosphorylation of PKC␦ was determined using phospho-specific antibodies against tyrosines 52, 155, and 187. As illustrated in Fig. 6, TRAIL induced phosphorylation of tyrosine 155 on PKC␦. Initial phosphorylation was observed after 5 min, maximal phosphorylation was obtained after 60 min and was decreased thereafter. In contrast, TRAIL did not increase the phosphorylation of tyrosines 187 (Fig. 6) or 52 (data not shown) in A172 cells.
Phosphorylation of PKC␦ on Tyrosine 155 Is Essential for Its Protective Effect-To examine the role of tyrosine phosphorylation of PKC␦ in its protective effect against TRAIL-induced apoptosis, we employed a PKC␦ mutant in which tyrosine 155 was mutated to phenylalanine (PKC␦Y155F). We overexpressed PKC␦ WT and the PKC␦Y155F mutant in A172 cells and examined the apoptosis of the transfected cells in response to TRAIL. As already described (Fig. 1C), overexpression of PKC␦ reduced cell apoptosis. In contrast, overexpression of the PKC␦Y155F increased apoptosis of the A172 cells by 10 -15% as compared with CV cells (Fig. 7A), suggesting that the phos-phorylation of PKC␦ on tyrosine 155 is essential for the protective effect of PKC␦ and that PKC␦Y155F acted as a dominant negative of PKC␦. Similarly, we found that overexpression of PKC␦ decreased the expression of active caspase 3, whereas overexpression of PKC␦Y155F slightly increased it as compared with CV cells (Fig. 7B). This apoptotic effect was specific for the PKC␦Y155F mutant because overexpression of other PKC␦ tyrosine mutants, PKC␦Y52F and PKC␦Y187F, protected the A172 cells against TRAIL-induced apoptosis, similar to the effect of PKC␦ (Fig. 7C).
Phosphorylation on Tyrosine 155 Is Essential for the Translocation and Cleavage of PKC␦-One of the possible mechanisms for the different effects of PKC␦ and the PKC␦Y155F mutant on TRAIL-induced apoptosis is their differential subcellular localization following TRAIL treatment. To examine this point, we followed the translocation of PKC␦ and the PKC␦Y155F mutant tagged with GFP in TRAIL-treated cells. A172 cells were transfected with GFP constructs and the translocation of PKC␦ and PKC␦Y155F was monitored as a function of time following TRAIL treatment. Similar to the endogenous PKC␦, PKC␦-GFP exited the nucleus and translocated to the ER after 15 min of TRAIL treatment and this translocation persisted up to 60 min thereafter (Fig. 8A). In contrast, no significant translocation was observed in the A172 cells transfected with the PKC␦Y155F-GFP mutant, and in TRAILtreated cells the PKC␦Y155F-GFP mutant was localized to the cytoplasm and nucleus similar to the control untreated cells (Fig. 8A). The translocation of the PKC␦Y155F mutant in response to PMA and etoposide was also examined. We found that PMA induced translocation of the PKC␦Y155F mutant to the plasma and perinuclear membranes, similar to translocation of the PKC␦-GFP WT (Fig. 8B). Etoposide induces nuclear translocation of PKC␦. We found that the PKC␦Y155F-GFP translocated to the nucleus in response to etoposide, similar to PKC␦-GFP (Fig. 8B). The cleavage of PKC␦ was essential for its protective effect. We therefore examined whether phosphorylation of PKC␦ on tyrosine 155 was associated with the cleavage of PKC␦ by TRAIL. Using cells overexpressing PKC␦ WT and various PKC␦ tyrosine mutants, we found that TRAIL induced the cleavage of PKC␦ and that of the PKC␦Y187F and PKC␦Y52F mutants but did not induce cleavage of the PKC␦Y155F (Fig.  8C). Thus, phosphorylation of PKC␦ on tyrosine 155 was essential for the cleavage of PKC␦ in response to TRAIL. In contrast, phosphorylation of PKC␦ on tyrosine 155 was not essential for the cleavage of PKC␦ by etoposide (data not shown), suggesting a specific role of this phosphorylation site in the apoptotic pathway activated by TRAIL.
PKC␦ Activates pAKT in TRAIL-treated Cells-To further delineate the mechanisms underlying the protective effect of PKC␦ and the role of tyrosine 155 phosphorylation in its function, we examined the effect of PKC␦ on the activation of AKT because this signaling pathway has been implicated in the apoptotic response of various cells to TRAIL (50). We found that, in the A172 cells, TRAIL induced a small and transient increase in the phosphorylation of AKT similar to recent reports (51). Overexpression of PKC␦ increased the phosphorylation of AKT in untreated cells; this phosphorylation was further increased in response to TRAIL and persisted up to 60 min thereafter (Fig. 9). This effect of PKC␦ was abolished when tyrosine 155 was mutated to phenylalanine, suggesting that phosphorylation of tyrosine 155 was involved in the AKT phosphorylation induced by PKC␦ in TRAIL-treated cells. DISCUSSION In this study we explored the role of PKC␦ in the apoptotic effect of TRAIL. Various studies have reported that PKC regulates the apoptotic function of TRAIL using either the PKC activator PMA or various PKC inhibitors (43,44); however, the role of PKC␦ in the effect of TRAIL has not been explored. In an attempt to delineate the molecular mechanisms underlying the anti-apoptotic effect of PKC␦, we focused on three factors that are central for the diverse apoptotic function of PKC␦, its cleavage, localization, and tyrosine phosphorylation (13).
We found that PKC␦ protected the A172 cells from the apoptosis induced by TRAIL. Thus, the inhibition of PKC␦ activity or expression increased, whereas overexpression of PKC␦ attenuated the apoptotic effect of TRAIL. PKC␦ has been extensively studied for its effects on the regulation of apoptosis in various cellular systems (13,52). In many of these studies PKC␦ acted as a pro-apoptotic kinase and it contributed to the apoptotic effects of a large and diverse apoptotic stimuli, including etoposide (14,15), ␥ radiation (18), UV radiation (17), and ceramide (31). However, in other systems PKC␦ clearly acted in an anti-apoptotic fashion (25,26). Thus, PKC␦ protected macrophages from NO-induced apoptosis (27) and glioma cells from the apoptosis induced by Sindbis virus infection (28). The basis for the anti-and pro-apoptotic effects of PKC␦ is not understood and, although various downstream targets have been associated with the pro-apoptotic effect of PKC␦, less is known about the mechanisms involved in its anti-apoptotic effect. We found that PKC␦ inhibited the release of cytochrome c from the mitochondria and the cleavage of caspase 3, whereas it did not alter the activation of caspase 8 induced by TRAIL. Thus, PKC␦ acted downstream of caspase 8 activation and upstream of the mitochondrial pathway.
TRAIL induced cleavage of PKC␦ within 2-3 h post-treatment and this cleavage was mediated by caspase 3 downstream of caspase 8 and 9 activation. Using a caspase-resistant PKC␦ mutant (PKC␦D327A) we found that, in contrast to PKC␦ WT, which was cleaved by TRAIL and attenuated its apoptotic effect, the PKC␦D327A mutant did not undergo cleavage and increased the apoptotic effect of TRAIL. The enhanced apoptotic effect of this mutant was specific to TRAIL because the same mutant attenuated the apoptotic effect of etoposide. These results suggest that the PKC␦D327A acted as a domi-nant negative of PKC␦ and that the cleavage of PKC␦ was essential for the protective effect of PKC␦ in TRAIL-treated cells. Our results demonstrating a protective effect of the cleaved PKC␦ against TRAIL-induced apoptosis are the first to show that cleavage of PKC␦ can provide anti-apoptotic signals. Similar results were recently reported for the cleaved PKC⑀ in tumor necrosis factor-␣-treated MCF-7 cells (53); thus also in the case of PKC⑀ the cleaved product can exert both pro-and anti-apoptotic effects (53,54).
The cleavage of PKC␦ is well documented and it occurs in response to many apoptotic stimuli such as etoposide (15), cisplatin (55), and UV radiation (56). The caspase-dependent cleavage of PKC␦ leads to the generation of a constitutively active catalytic fragment that is associated with the apoptotic function of PKC␦ in response to various apoptotic stimuli (15,19,20,55,56). Moreover, overexpression of the PKC␦ catalytic fragment has been reported to induce cell apoptosis in various systems by phosphorylating apoptosis related proteins such as p73␤ and DNA-PK (21,22). The fate and function of the regulatory domain released following the caspase-dependent cleavage are less understood. However, in a recent study Schultz et al. (48) reported that the regulatory domain of PKC exerted an apoptotic function when overexpressed, suggesting that both the regulatory and catalytic domains of PKC can regulate cell apoptosis. The cleavage of PKC␦ and the roles of its catalytic and regulatory domains have not been studied previously in cellular systems where PKC␦ acted as an anti-apoptotic kinase.
In an attempt to determine the roles of the regulatory and catalytic domains of PKC␦ in its protective effect against TRAIL-induced apoptosis, we employed PKC chimeras between the regulatory and catalytic domains of PKC␦ and PKC␣. We found that both domains were required for the protective effect of PKC␦, because neither of the chimeras acted like PKC␦ to protect the A172 cells from TRAIL-induced apoptosis. The involvement of both domains was further confirmed by the experiment of overexpressing the PKC␦-reg and PKC␦cat because neither of them protected the A172 cells from the apoptotic effect of TRAIL. These results are similar to our recent results in which the two domains were required for the apoptotic effect of PKC␦ in response to etoposide (15). Unlike these two systems, the isoform specificity of PKC␦ for inhibition of glioma cell proliferation depended only on the regulatory domain (45). The mechanisms that are involved in the protective effect of the catalytic domain in this case as compared with its apoptotic effects in other systems are not understood. However, it is noteworthy that the localization of the catalytic domain in TRAIL-treated cells was outside the nucleus as compared with its nuclear localization in response to apoptotic stimuli or when overexpressed (14,15,33). Thus, the phosphorylation of different substrates, anti-apoptotic as compared with pro-apoptotic, as a result of this differential localization could account for the different effects of the catalytic domain.
TRAIL induced translocation of PKC␦ to the ER within 15 min of treatment. PKC␦ has been reported to undergo differential translocation to distinct subcellular sites in response to diverse apoptotic stimuli. Thus, PKC␦ was translocated to the nucleus in response to etoposide (14,15), to the mitochondria in response to PMA (32) and H 2 O 2 (57), and to the Golgi in response to ceramide (31). Interestingly, we recently found that PKC␦ translocated to the ER in response to infection of glioma cells with Sindbis virus (28). Similar to its effect in TRAILtreated cells, PKC␦ protected glioma cells from the apoptosis induced by SV infection (28). It is currently not clear what is the function of PKC␦ in the ER and which proteins can be phosphorylated by PKC␦ at this site. One candidate is Bcl2, which resides in the ER in addition to its mitochondrial local- ization (58) and has been shown to be a PKC substrate (59) and to protect cells from TRAIL-induced apoptosis (60).
Phosphorylation of PKC on tyrosine residues has been im-plicated as an important mode of regulation of the activity and function of PKC␦ and various stimuli phosphorylate PKC␦ on distinct tyrosine residues (45,61,62). Indeed, phosphorylation of PKC␦ on specific tyrosine residues occurs in response to PMA (63), platelet-derived growth factor (39,64), EGF (65) (28). We found that TRAIL induced phosphorylation of PKC␦ on tyrosine 155, whereas no phosphorylation was observed on tyrosines 52 or 187. The phosphorylation of tyrosine 155 was essential for the anti-apoptotic effect of PKC␦ in TRAIL-treated cells, because a mutation in this tyrosine residue attenuated the protective effect of PKC␦ against TRAILinduced apoptosis. Interestingly, tyrosine 155 has been recently associated with the protective effect of PKC␦ against Sindbis virus-induced apoptosis (28). The phosphorylation of PKC␦ on tyrosine 155 occurred prior to the translocation of PKC␦ to the ER and was essential for the translocation of PKC␦ in response to TRAIL. In contrast, the phosphorylation on tyrosine 155 was not essential for the  translocation of PKC␦ to the plasma and perinuclear membranes in response to PMA or to the nucleus in response to etoposide. Translocation of PKC is a hallmark of its activation and it occurs via binding of PKC to specific scaffold proteins (RACKS) (67). The role of tyrosine phosphorylation of PKC in its translocation has been explored in various cellular systems. Phosphorylation of atypical PKC at Tyr 256 induced exposure of the arginine-rich NLS, which resulted in nuclear translocation of the atypical PKC (68). In contrast, the phosphorylation of PKC␦ on tyrosines 311 and 322 did not play a role in the translocation of PKC␦ to the Golgi in response to ceramide (31). Similarly, we recently reported that the phosphorylation of PKC␦ on tyrosines 187 and 64 was not essential for the nuclear translocation of PKC␦ in response to etoposide although it was essential for its apoptotic effect (15). The mechanisms by which phosphorylation of PKC␦ is involved in the translocation of PKC␦ to the ER is currently not understood, but conformational changes or association of the phosphorylated PKC␦ with another protein via SH2 or PTB domains are possible mechanisms.
In addition to its translocation to the ER, phosphorylation on tyrosine 155 was also essential for the cleavage of PKC␦ by TRAIL. The translocation of PKC␦ to the ER was transient and preceded the cleavage of PKC␦, suggesting that PKC␦ was cleaved outside of the ER. Interestingly, the inability of the PKC␦Y155F mutant to undergo cleavage by TRAIL was specific to this apoptotic stimulus because this mutation did not interfere with the ability of PKC␦ to undergo cleavage in response to etoposide. Similarly, mutations in tyrosines 187 and 52 did not attenuate the cleavage of PKC␦ in response to TRAIL.
TRAIL has been reported to activate both pro-and antiapoptotic signaling pathways in various cellular systems including ERK (51), c-Jun N-terminal kinase (40), AKT (51), and NF-B (40). AKT has been implicated as an important inhibitor of TRAIL-induced apoptosis (50,69). AKT exerts its anti-apoptotic effect by phosphorylating pro-apoptotic proteins such as BAD, caspase 9, and forkhead transcription factors or by activating anti-apoptotic pathways such as NF-B or the expression of FLIP (70). We found that PKC␦ increased the phosphorylation of AKT in control and TRAIL-treated cells and that this effect was dependent on the phosphorylation of PKC␦ on tyrosine 155. The mechanism by which the tyrosine-phosphorylated form of PKC␦ enhances the activation of AKT is currently not understood. One possible mechanism is an inhibitory effect of PKC␦ on phosphatase activity, similar to the recent results reported for PKC␣ (71). Alternative mechanisms include the activation of PDK-1 or phosphatidylinositol 3-kinase by PKC␦.
In summary, we demonstrated that PKC␦ protected glioma cells from TRAIL-induced apoptosis and that the phosphorylation of PKC␦ on tyrosine 155, its translocation to the ER, and its cleavage were associated with the protective effect of this isoform. PKC␦ acted downstream of caspase 8 activation and inhibited the mitochondrial pathway, probably via activation of AKT. Cleavage of PKC␦ and its translocation to the ER contributed to its protective effects by mechanisms that are still not understood.
Although PKC␦ has been extensively studied as a key kinase in the regulation of cell apoptosis, the mechanisms underlying its anti-apoptotic effects are not well understood. Our results provide new information regarding the factors that play a role in the anti-apoptotic effects of PKC␦ and implicate tyrosine phosphorylation of PKC␦ on tyrosine 155 as a key factor in the regulation of AKT phosphorylation and cell survival in glioma cells.
In addition to characterizing the anti-apoptotic role of PKC␦ in TRAIL-induced apoptosis, our results also have important implications for understanding the factors that regulate the apoptotic function of this key signaling molecule. Table I summarizes the localization, phosphorylation, and cleavage of PKC␦ in response to TRAIL and etoposide where PKC␦ acts as an anti-or pro-apoptotic kinase, respectively. Both TRAIL and etoposide induce phosphorylation of PKC␦ albeit on different residues and in both cases tyrosine phosphorylation is essential for the apoptotic function of PKC␦. Tyrosine phosphorylation on distinct tyrosine residues can modulate the subcellular localization of PKC␦, its affinity toward different substrates, and its association with other proteins via SH2 domains. These results clearly position tyrosine phosphorylation of PKC␦ as a pivotal factor in the regulation of the diverse functions of PKC␦ in cell apoptosis.