Mitochondrial Translocation of Protein Kinase C δ in Phorbol Ester-induced Cytochrome c Release and Apoptosis*

Apoptosis is induced by the release of cytochromec from mitochondria to the cytoplasm. The present studies demonstrate that the phorbol ester 12-O-tetradecanoylphorbol-13-acetate (TPA) induces translocation of protein kinase C (PKC) δ from the cytoplasm to mitochondria. The results also show that translocation of PKCδ results in release of cytochrome c. The functional significance of this event is further supported by the demonstration that PKCδ translocation is required for TPA-induced apoptosis. These findings demonstrate that translocation of PKCδ to mitochondria is responsible, at least in part, for inducing cytochrome crelease and apoptosis.

overexpressed in HeLa and other cells, is sufficient to induce apoptosis (3). Proteolytic activation of PKC␦ has also been implicated in ultraviolet radiation-induced apoptosis of keratinocytes (4). These findings have supported a role for the PKC␦ isoform in the apoptotic response of cells to diverse stimuli.
The treatment of human myeloid leukemia cells with TPA is associated with the induction of apoptosis (5). Other studies have demonstrated that TPA induces apoptosis of human MCF-7 breast cancer cells and Jurkat T cells (6,7). These findings have suggested that TPA-induced activation of certain PKC isoforms confers signals that induce apoptosis. However, the identity of the PKC isoform(s) that are activated in TPAinduced apoptosis and the downstream signals that confer the apoptotic response are unknown. The available evidence indicates that at least two pathways induce apoptosis in the cellular response to other stimuli. One pathway is initiated by engagement of the tumor necrosis factor or Fas receptors and thereby activation of caspase 8 (8,9). In turn, caspase 8 cleaves Bid and induces cytochrome c release (10,11). Caspase 8 can also directly activate caspase 3 (12). In the second pathway, other signals, which importantly remain undefined, converge to induce the release of cytochrome c (13,14). Cytosolic cytochrome c binds to Apaf-1, induces the autoprocessing of caspase 9 and thereby the activation of caspase 3 (11,15). Neither of these pathways has been linked to TPA-induced apoptosis.
The present studies demonstrate that TPA treatment is associated with the translocation of cytoplasmic PKC␦ to mitochondria. The results show that translocation of PKC␦ induces the release of cytochrome c and the activation of caspase 3. These findings support a novel mechanism for TPA-induced cytochrome c release and apoptosis.
Isolation of Mitochondria-Cells were washed twice with phosphatebuffered saline (PBS), homogenized in buffer A (210 mM manitol, 70 mM sucrose, 5 mM HEPES, 1 mM EGTA) and 110 g/l digitonin in a glass homogenizer (Pyrex no. 7727-07) and centrifuged at 5000 ϫ g for 20 min. Pellets were resuspended in buffer A, homogenized in a small glass homogenizer (Pyrex no. 7726), and centrifuged at 2000 ϫ g for 5 min. Supernatant (S1) was collected and the pellet again homogenized in buffer A. Supernatant (S2) was collected after centrifugation at 2000 ϫ g for 5 min. Supernatants S1 and S2 were mixed and centrifuged at 11,000 ϫ g for 10 min. Mitochondrial pellets were disrupted in lysis buffer at 4°C for 30 min and then centrifuged at 15,000 ϫ g for 20 min. The concentration of mitochondrial proteins in the supernatant was determined using Bio-Rad protein estimation kit.
Isolation of the Cytosolic Fraction-Cells were washed twice with PBS, and the pellet was suspended in 5 ml of ice-cold buffer B (20 mM HEPES, pH 7.5, 1.5 mM MgCl 2 , 10 mM KCl, 1 mM EDTA, 1 mM EGTA, 1 mM dithiothreitol, 0.1 mM phenylmethylsulfonyl fluoride, and 10 g/ml leupeptin and aprotinin) containing 250 mM sucrose. The cells were homogenized by disrupting three times in a Dounce homogenizer in buffer B. After centrifugation for 5 min at 4°C, the supernatants then were centrifuged at 105,000 ϫ g for 30 min at 4°C. The resulting supernatant was used as the soluble cytosolic fraction.
Immunofluorescence Microscopy-Cells immobilized on slides were fixed with 3.7% formaldehyde, permeabilized with 0.2% Triton X-100, incubated with 20 ng of anti-PKC␦/slide and then Texas Red-conjugated goat anti-rabbit IgG (Southern Biotechnology Associates, Inc). Mitochondria were stained with 0.006 ng/slide of Mitotracker Green FM (Molecular Probes). The slides were analyzed using a Zeiss Auxiphot fluorescence microscope coupled to a CCD camera and a power Macintosh 8100. Image analysis was performed using the IPLab Spectrum 3.1 software (Signal Analytics).
Quantitation of Apoptosis by Flow Cytometric Analysis-Cells were harvested, washed twice with PBS, and fixed with 80% ethanol. Cells (10 6 cells/ml) were washed and incubated with propidium iodide (2.5 g/ml) and RNase (50 g/ml). FACScan (Becton Dickinson) was used to assess cells with sub-G 1 DNA content.

RESULTS AND DISCUSSION
PKC isoforms regulate diverse cellular processes but are not known as effectors of mitochondria (1). To determine whether PKC regulates mitochondrial function, human U-937 cells were treated with TPA to activate PKC. PKC translocation was assessed by subjecting cytoplasmic and mitochondrial fractions to immunoblotting with anti-PKC antibodies. The results demonstrate that TPA treatment is associated with decreases in cytoplasmic PKC␦ and concomitant increases in mitochondrial PKC␦ (Fig. 1A). As controls, the cytoplasmic and mitochondrial fractions were also subjected to immunoblotting with anti-actin and anti-Hsp60 to ensure purity of the preparations (Fig. 1A). By contrast to translocation of PKC␦, TPA had no detectable effect on cytoplasmic or mitochondrial levels of PKC␥ and PKC ( Fig. 1B and data not shown). The immunoblots were scanned to calculate percent PKC␦ translocation to mitochondria. The results demonstrate that approximately 40% of PKC␦ translocates to mitochondria in response to TPA.
The demonstration that PKC␦ also translocates to mitochondria in TPA-treated MCF-7 cells indicates that the finding is not restricted to certain cell types ( Fig. 2A). In addition, to confirm the subcellular redistribution of PKC␦ in TPA-treated cells, we visualized intracellular fluorescence with a CCD camera and image analyzer. Examination of fluorescence markers in control cells showed distinct patterns for anti-PKC␦ (red signal) and a mitochondrion-selective dye (Mitotracker; green signal) (Fig. 2B). The demonstration that TPA induces a marked change in fluorescence signals (red and green 3 yellow/ orange) supported translocation of PKC␦ to mitochondria (Fig.  2B). These findings obtained by immunofluorescence microscopy thus confirm the results of PKC␦ redistribution found by subcellular fractionation.
To determine whether the natural product bryostatin, which activates PKC (19), also induces the translocation of PKC␦, mitochondrial lysates from U-937 cells treated with 100 nM bryostatin were subjected to immunoblot analysis with anti-PKC␦. As a control, mitochondrial lysates were also subjected to immunoblot analysis with anti-PKC. The results demonstrate that, in contrast to PKC, treatment with bryostatin was associated with translocation of PKC␦ to mitochondria (Fig.  3A). PLC is activated by cell membrane-initiated signaling pathways and, by conferring the hydrolysis of phosphatidylinositol or phosphatidylcholine, results in the formation of DAG (20). To determine whether PLC induces the translocation of PKC␦, mitochondrial lysates from U-937 cells treated with 0.5 unit/ml PLC were subjected to immunoblot analysis with anti-PKC␦. The results demonstrate that treatment with PLC is associated with translocation of PKC␦ to mitochondria (Fig.  3B). To confirm the involvement of DAG in mitochondrial translocation of PKC␦, cells were treated with a cell permeable-DAG (DOG) (21). Immunoblot analysis of DOG-treated cell lysates demonstrated that DOG induced the translocation of PKC␦ to mitochondria (Fig. 3C). These findings indicate that, like TPA, treatment with bryostatin, PLC, and DOG is associated with redistribution of cytosolic PKC␦ to mitochondria.
To determine whether activation of PKC␦ is required for translocation to mitochondria, we transfected cells with a vector expressing GFP-tagged PKC␦. Immunoblot analysis with anti-GFP demonstrated no detectable PKC␦ in the mitochondrial fraction from cells transfected with an empty GFP vector (Fig. 4A). By contrast, transfection of kinase-active GFP-PKC␦ was associated with PKC␦ expression in mitochondria (Fig.  4A). Moreover, treatment of the GFP-PKC␦-transfected cells with TPA resulted in further increases in levels of mitochondrial PKC␦ (Fig. 4A). Significantly, transfection of kinase-inactive GFP-PKC␦(K-R) had no effect on expression of mitochon-drial PKC␦ (Fig. 4A). In addition, overexpression of GFP-PKC␦(K-R) blocked the TPA-induced translocation of PKC␦ to mitochondria (Fig. 4A). To demonstrate that PKC␦(K-R) specifically blocks endogenous PKC␦ activity, and not that of other isoforms of PKC, 293T cells were transiently transfected with GFP-PKC␦ or GFP-PKC␦(K-R). Following transfection, cell lysates were subjected to immunoprecipitation with anti-PKC␦, anti-PKC, anti-PKC, anti-PKC, anti-PKC, or anti-PKC⑀. The precipitates were assayed in in vitro kinase assays using H1 histone as substrate. The results demonstrate that, in contrast to PKC, PKC, PKC, or PKC, overexpression of PKC␦(K-R) specifically inhibits the activity of endogenous PKC␦ (Fig. 4B). Of note, the results also indicate that overexpression of PKC␦(K-R) is associated with slight inhibition of the phosphorylated and active PKC⑀ (Fig. 4B). PKC␦ consists of an N-terminal regulatory domain (RD) and a C-terminal catalytically active fragment (2). MCF-7 cells stably transfected to express the 35-kDa RD exhibit attenuation of TPA-induced PKC␦ activity. 2 Translocation of PKC␦ to mitochondria was also attenuated in TPA-treated MCF-7/PKC␦RD cells as compared with that in MCF-7 cells expressing the empty neo vector (Fig. 5A). Other studies were performed with rottlerin, a selective inhibitor of PKC␦ activation (22). Treatment of U-937 cells with rottlerin abrogated TPA-induced localization of PKC␦ to mitochondria (Fig. 5B). These findings collectively demonstrate that PKC␦ activation is necessary for its translocation to mitochondria.
The potential role of PKC␦ translocation was explored by assessing mitochondrial release of cytochrome c. Whereas diverse apoptotic signals induce cytochrome c release, phorbol ester treatment of cells has not been associated with this event. Immunoblot analysis of cytoplasmic fractions with anti-cytochrome c demonstrated that TPA treatment of U-937 cells is  associated with cytochrome c release (Fig. 6A). Similar results were obtained when U-937 cells were treated with PLC or DOG (Fig. 6, B and C). To determine whether PKC␦ functions in inducing cytochrome c release, we pretreated U-937 cells with rottlerin before adding TPA. Of note, treatment of cells with rottlerin alone is associated with cytotoxic effects that contribute to a detectable release of cytochrome c (Fig. 7A). By contrast, analysis of cytoplasmic lysates demonstrated that rottlerin significantly blocks TPA-induced cytochrome c release (Fig. 7A). As these findings indicate that the PKC␦ kinase function is required for TPA-induced release of cytochrome c, 293T cells were transfected to express GFP, GFP-PKC␦, or GFP-PKC␦(K-R) and then treated with TPA. Immunoblotting of the cytoplasmic fraction from GFP positive cells demonstrated abrogation of TPA-induced cytochrome c release in cells expressing PKC␦(K-R) compared with that in cells transfected with the GFP-PKC␦ vector (Fig. 7B). Taken together, these results and those obtained for PKC␦ translocation support a role for PKC␦ in the mitochondrial release of cytochrome c.
The release of cytochrome c from mitochondria triggers ac-tivation of caspases and induction of apoptosis (23). To determine whether TPA-induced PKC␦ translocation and thereby cytochrome c release contributes to apoptosis, U-937 cells treated with rottlerin and TPA were assayed for sub-G 1 DNA content. The results demonstrate that treatment with rottlerin alone induces a low level of apoptosis (Fig. 8A). By contrast, the apoptotic response of U-937 cells to TPA was significantly attenuated by inhibition of PKC␦ with rottlerin (Fig. 8A). Moreover, treatment of MCF-7/neo cells with TPA was also associated with the induction of apoptosis (Fig. 8B). By contrast, the apoptotic response to TPA was significantly attenuated in MCF-7/PKC␦RD cells (Fig. 8B). Taken together with our other findings, these results support a role for TPA-induced localization of PKC␦ to mitochondria and in the induction of apoptosis. Previous work has demonstrated that TPA treatment is associated with translocation of PKC␦ to the cell membrane (24). The present studies demonstrate that TPA treatment of diverse cell types is associated with translocation of PKC␦ to mitochondria. These findings have been confirmed by cell fractionation and immunofluorescence studies. The results further demonstrate that the PKC␦ kinase function is necessary for TPA-induced mitochondrial localization. The functional significance of PKC␦ translocation to mitochondria is supported by the finding that this event is linked to mitochondrial release of cytochrome c. Moreover, the results demonstrate that abrogation of PKC␦ translocation to mitochondria significantly inhibits TPA-induced apoptosis. These findings thus support a model in which TPA induces the release of cytochrome c and thereby apoptosis by a PKC␦-dependent mechanism.