Evidence That Protein Kinase Cε Mediates Phorbol Ester Inhibition of Calphostin C- and Tumor Necrosis Factor-α-induced Apoptosis in U937 Histiocytic Lymphoma Cells*

Protein kinase C (PKC) activators, such as the tumor-promoting phorbol esters, have been reported to protect several cell lines from apoptosis induced by a variety of agents. Recent evidence suggests that PKCε is involved in protection of cardiac myocytes from hypoxia-induced cell death (Gray, M. O., Karliner, J. S., and Mochly-Rosen, D. (1997) J. Biol. Chem.272, 30945–30951). We investigated the protective effects of the phorbol ester 12-O-tetradecanoylphorbol-13-acetate (TPA) on U937 histiocytic lymphoma cells induced to undergo apoptosis by tumor necrosis factor-α (TNF-α) or by the specific PKC inhibitor calphostin C. U937 cells were transiently permeabilized with a peptide (εV1-2) derived from the V1 region of PKCε that has been reported to specifically block translocation of PKCε. The εV1-2 peptide blocked the inhibitory effect of TPA on both TNF-α- and calphostin C-induced apoptosis. A scrambled version of εV1-2 and a peptide reported to inhibit PKCβ translocation (βC2-4) had no effect on the ability of TPA to inhibit apoptosis. These results suggest that PKCε is required for the protective effect of TPA in TNF-α- and calphostin C-induced apoptosis. Furthermore, calphostin C reduced membrane-associated PKCε activity and immunoreactivity, suggesting that PKCε may play an important role in leukemic cell survival.

The phospholipid-dependent protein kinase C (PKC) 1 family of isozymes has a central role in the transduction of extracellular signals and has been implicated in tumor promotion (1). PKCs also appear to be important in the regulation of apoptosis, and several reports have indicated that PKC activation can inhibit apoptosis. For example, the tumor-promoting phorbol ester and potent PKC activator 12-O-tetradecanoylphorbol-13acetate (TPA) prevents apoptosis induced by tumor necrosis factor-␣ (TNF-␣), Fas, and microtubule-disrupting drugs in human lymphoma cells (2,3); by growth factor deprivation in leukemic and normal myeloid cells (4); and by anticancer drug treatment of leukemic cells and a human breast cancer cell line (5,6). TPA also inhibits apoptotic oligonucleosomal DNA degradation in isolated nuclei from thymocytes (7). Other PKC activators have also been reported to inhibit apoptosis. These include phorbol dibutyrate, which protects mouse spleen blast cells against interleukin-2 withdrawal (8), and mezerein and bryostatin-1, which block apoptosis induced by the sphingolipid ceramide (9). A synthetic analogue of the natural PKC activator diacylglycerol has also been reported to prevent apoptosis in human leukemic cells induced by ceramide and the isoprenoids farnesol and geranylgeraniol (9 -11).
It should be noted, however, that there are some reports that demonstrate that PKC activation can induce or increase apoptosis in some cell types (12)(13)(14). The effects of PKC activation on apoptosis may therefore be cell type-specific and could be determined by factors such as the rate of PKC down-regulation.
Evidence has recently emerged suggesting that specific PKC isozymes may be involved in the prevention of apoptosis. Mochly-Rosen and co-workers (15) have used peptides that specifically block the interaction of PKC subforms with membrane-associated binding proteins and have demonstrated that a peptide that inhibited membrane translocation of PKC⑀ blocked the protection of cardiac myocytes from hypoxia-induced cell death. In another study, apoptosis in human leukemic cells induced by TNF-␣, anti-Fas antibody, sphingomyelinase, and ceramide was associated with the redistribution of PKC␦ and PKC⑀ from the membrane to the cytosol. These redistributions were prevented by concentrations of TPA that blocked apoptosis (16).
Further reports suggest that PKC⑀ is anti-apoptotic in some cells. For example, in HT58 human lymphoblastic cells, TPA causes growth inhibition and down-regulation of PKC (50). These cells could be induced to undergo apoptosis by the general kinase inhibitor staurosporine, but only when the cells had been pretreated with low concentrations of TPA that downregulated PKC⑀, but not PKC␣ and PKC␤ (17). In oncogenetransformed rat embryo fibroblasts, susceptibility to the anticancer drug cisplatin is increased, whereas PKC⑀ expression is decreased compared with nontransformed cells. Stable transfection of these cells with PKC⑀ prevents cisplatin-induced apoptosis and also protects these cells against cisplatin cytotoxicity (17).
Taken together, these reports suggest that PKC⑀ may be of particular significance in the negative regulation of apoptosis, and this possibility has been explored further in this study. We have used U937 histiocytic lymphoma cells, in which apoptosis induced by ceramide, TNF-␣, and the PKC inhibitor calphostin C is strongly inhibited by TPA. In this study, we used peptides based on unique sequences within PKC isozymes that specifically block the binding of individual isozymes to anchoring proteins, termed RACK proteins (receptors for activated C-kinase). Recent work has demonstrated that a peptide corresponding to amino acids 14 -21 in the V1 region of PKC⑀ prevents phorbol ester-induced translocation of PKC⑀ and inhibits contraction in cultured cardiac myocytes (18). In a further study, the ⑀V1-2 peptide inhibited hypoxic preconditioning and phorbol ester-mediated protection of cardiac myocytes from hypoxia-induced cell death. In this study, we found that this peptide inhibited the ability of TPA to prevent apoptosis induced by TNF-␣ and calphostin C (calC). calC also reduced putative PKC⑀ activity and displaced PKC⑀ from the membrane to the cytosol.
Cell Culture-U937 human myeloid leukemic cells (American Type Culture Collection, Rockville, MD) were maintained in RPMI 1640 medium supplemented with 10% heat-inactivated fetal calf serum, 0.1 mM nonessential amino acids, 100 units/ml penicillin, and 100 mg/ml streptomycin. Cells were transferred to RPMI 1640 medium supplemented with 2.5% heat-inactivated fetal calf serum for treatment with TNF-␣ or calC. After addition of calC, cells were illuminated for 10 min with a 40-watt incandescent lamp at a distance of 30 cm.
Preparation of Cell Extracts-Cytosolic and solubilized particulate fractions were prepared by sonicating (3 ϫ 10 s) cells in buffer A (20 mM Tris-HCl (pH 7.4), 5 mM EGTA, 2 mM EDTA, 10 mM benzamidine, 0.01% leupeptin, 10 mM 2-mercaptoethanol, 2 mM phenylmethylsulfonyl fluoride, 0.1 unit/ml aprotinin, and 0.25 M sucrose). The sonicate was cleared of unlysed cells by centrifugation at 500 ϫ g for 5 min, and the cleared supernatant was centrifuged at 100,000 ϫ g for 20 min at 4°C. The pellet was sonicated in buffer A containing Triton X-100 (as indicated), incubated on ice for 30 min, and centrifuged as described before to produce a supernatant particulate fraction.
Western Blot Analysis-Aliquots of cytosolic and particulate fraction extracts (described above) were added to well loading buffer (20 mM Tris-HCl (pH 6.8), 40% sucrose, 6% SDS, and 10 mM 2-mercaptoethanol) (2 vol extract: 1 vol loading buffer) and heated for 5 min at 100°C. Aliquots (5 g of protein; determined by a modified method (19) of Lowry et al. (49)) were size-separated by SDS-polyacrylamide electrophoresis using 12% gels in a Bio-Rad minigel system (200 V, 45 min) and then electrophoretically transferred to Schleicher & Schuell nitrocellulose paper (0.2 m; 100 V, 90 min) as described (20). Membranes were blocked with 5% milk powder and incubated with the primary antibodies for 1 h at 37°C. To determine specificity, primary antibodies were preincubated for 2 h with 10-fold excess peptide against which the antibody was raised. After incubation with primary antibodies, the membranes were washed, incubated with the horseradish peroxidaseconjugated secondary antibody for 45 min, and washed again. The bound secondary antibody was detected with ECL reagent. Chemiluminescence was photographed with HYPERfilm-ECL.
Direct Measurement of Membrane-associated PKC Activity-PKC activity in the particulate fraction was measured using the method of Durkin et al. (21) with modifications. This method does not require a preliminary extraction with subsequent reconstitution and artificial activation of the enzyme with Ca 2ϩ and phosphatidylserine. PKC activity in the particulate fraction was measured using a PKC peptide substrate based on the epidermal growth factor receptor or a highly selective peptide substrate based on neurogranin-(28 -43) that allows measurement of PKC in crude extracts, and a PKC inhibitor peptide based on the pseudosubstrate-(19 -31) region of PKC.
The assay reaction mixture contained 4 -7 g of protein from the particulate fraction in assay buffer (50 mM Tris-HCl (pH 7.5), 5 mM MgCl 2 , 1 M CaCl 2 , 100 M sodium vanadate, 100 M sodium pyrophosphate, 1 mM sodium fluoride, and 100 M phenylmethylsulfonyl fluoride) and 50 M peptide substrate in a final volume of 90 l. The reaction was started by the addition of 10 l of 500 M [␥-32 P]ATP (220 cpm/pmol in Tris buffer, 0.5 Ci/assay) and stopped after 10 min by the addition of 10 l of 75 mM orthophosphoric acid. The samples were clarified by centrifugation at 16,000 ϫ g for 5 min, and a 90-l sample of each supernatant was applied to Whatman P-81 paper (2 cm 2 ). The papers were washed twice in 75 mM orthophosphoric acid with gentle agitation for 10 min. The radioactivity bound to the washed papers was determined by liquid scintillation counting.
Assessment of DNA Fragmentation-Qualitative assessment of DNA degradation was performed by agarose gel electrophoresis (22). Pelleted cells (5 ϫ 10 6 ) were fixed by suspension in 70% ethanol at Ϫ20°C overnight. Cells were then centrifuged at 200 ϫ g for 5 min, and the ethanol was thoroughly removed. Cell pellets were resuspended in 40 l of phosphate/citrate buffer (192 parts 0.2 M Na 2 HPO 4 and 8 parts 0.1 M citric acid (pH 7.8)) at room temperature for 30 min. After centrifugation at 1000 ϫ g for 5 min, the supernatant was transferred to new tubes, dried by vacuum in a SpeedVac (Savant Instruments, Inc., Farmingdale, NY), and reconstituted with 15 l of sterile distilled water. The DNA extract was then incubated with the addition of 3 l of 0.25% Nonidet P-40 and 3 l of RNase (1 mg/ml) for 30 min at 37°C. The extract was incubated for a further 30 min with proteinase K (3 l, 1 mg/ml). After this digestion, 5 l of loading buffer (0.25% bromphenol blue, 0.25% xylene cyanol, and 30% glycerol) was added, and the extract was fractionated by electrophoresis on 1.8% agarose at 5 V/cm for 3 h. DNA was visualized under UV light after staining with 0.5 g/ml ethidium bromide.
Quantitation of Apoptosis-U937 cells (5 ϫ 10 5 /ml) were treated with test compounds in RPMI 1640 medium containing 2.5% fetal calf serum. Chromatin was visualized by fluorescent microscopy after a 30min incubation with the cell-permeable fluorochrome Hoechst 33258 (10 g/ml). The proportion of cells undergoing apoptosis at a given time was determined by counting cells exhibiting two or more membrane blebs and brightly stained condensed and fragmented chromatin. At least 200 cells were counted for each sample. The proportion of cells undergoing apoptosis was also determined using flow cytometry (23). Flow cytometry results are the percentage of cells (10,000) containing subdiploid DNA.
Permeabilization-In initial experiments, U937 cells were transiently permeabilized using the TransPort TM kit according to the manufacturer's instructions. In subsequent experiments, cells were permeabilized by electroporation. U937 cells were suspended at 10 ϫ 10 6 cells/ml in serum-free RPMI 1640 medium, and 800-l aliquots were electroporated with or without the indicated peptides at 330 microfarads and 200 V at room temperature. The electroporated cells were kept at room temperature for 10 min and then transferred to RPMI 1640 medium containing 2.5% fetal calf serum to a final concentration of 5 ϫ 10 5 cells/ml for treatment.

Characterization of Calphostin C-induced Apoptosis in U937
Cells-In the presence of light, calC is a highly specific PKC inhibitor that interacts with the regulatory domain of the enzyme and inhibits phorbol ester binding at high concentrations (24). calC was chosen for this study because, unlike TNF-␣ or ceramide, calC induced apoptosis in a large proportion of cells (Fig. 1), thus making biochemical studies more interpretable. Apoptosis induced by low concentrations of calC was acutely Left panels, proteins were extracted with 2% Triton X-100 from whole cell lysates. Western blot analysis was performed as described under "Experimental Procedures." The bands were detected (Ab only) at positions corresponding to partially purified PKC isozymes from rat brain and were not detected by primary antibodies (Ab) that had been preincubated with the peptide against which the antibodies had been raised (Ab ϩ peptide). Right panels, shown is the distribution of the PKC isozymes between the cytosol and a particulate extract solubilized with Triton X-100. The results are representative of two different experiments.
blocked by TPA (Figs. 1 and 2). The protective effect of TPA was less after prolonged incubations (Fig. 1), possibly because of down-regulation of PKC. calC induced apoptosis in U937 cells in a dose-dependent manner and generated characteristic oligonucleosomal DNA fragmentation (Fig. 3), membrane blebbing (data not shown), and chromatin condensation (Fig. 2). Using time-lapse microcinematography, it was observed that calC-treated U937 cells underwent fragmentation into clusters of apoptotic bodies in a similar way to cells treated with TNF-␣ (data not shown).
PKC Isozymes and Distribution in U937 Cells-The main intracellular receptors for TPA are the PKC family of isozymes. Because U937 cells have been reported to contain different profiles of PKC isozymes, we screened our U937 cells using immunoblotting. U937 cells have been reported to express PKC␣, PKC␤I, PKC␦, and PKC, (25); PKC␣, PKC␤, PKC⑀, and PKC (26,27); or PKC␤I, PKC␤II, PKC⑀, and PKC (28). The U937 cells used in this study expressed PKC␤I, PKC␤II, PKC⑀, and PKC and very low levels of PKC␦, but did not contain detectable levels of PKC␣ or PKC (Fig. 4). PKC␤I and PKC␤II were predominantly located in the cytosol, whereas PKC⑀ and PKC were distributed between the cytosolic and particulate fractions (Fig. 4).
TPA Protection against calC-and TNF-␣-induced Apoptosis Is Inhibited by ⑀V1-2-To determine whether PKC⑀ transloca-tion is required for the protective effect of TPA, we used a peptide derived from the V1 region of PKC⑀ (⑀V1-2) that specifically inhibits the TPA-induced membrane translocation of this isoform of PKC. The ⑀V1-2 peptide blocks the interaction between PKC⑀ and anchoring proteins that have been termed RACK proteins. Transient permeabilization with this peptide has been demonstrated to block TPA-induced contraction and protection against hypoxia-induced cell death in cardiac myocytes (15,18). Furthermore, expression of the V1 fragment of PKC⑀ from which ⑀V1-2 was derived blocked TPA enhancement of neurite outgrowth in PC-12 rat neuroblastoma cells (29). Similar peptides have been identified with specificity for PKC␦ (29) and PKC␤ (␤C2-4) (30). Clearly, such peptides provide a powerful tool to evaluate the biological roles of individual PKC subforms.
Figs. 5 and 6 show the effects of the ⑀V1-2 peptide on the TPA inhibition of calC-induced apoptosis. Apoptosis was quantified both by Hoechst chromatin staining and by fluorescence-activated cell sorter analysis. The ⑀V1-2 peptide attenuated the protective effect of TPA against calC-or TNF-␣-induced apoptosis, and the effect was dose-dependent (Fig. 5). However, the peptide had no effect on the levels of apoptosis induced by calC or TNF-␣ and did not induce apoptosis in untreated cells (data not shown). No significant effect was observed with a scrambled peptide containing the same amino acids as ⑀V1-2 or with a peptide that blocks PKC␤I and PKC␤II binding to RACK proteins (␤C2-4 ( Figs. 5 and 6). These data strongly support a role for PKC⑀ in mediating the protective effect of TPA against calC-and TNF-␣-induced apoptosis. It should be noted that a low concentration of TPA (10 nM) was used in these experiments, a concentration that gave only partial protection against apoptosis (Fig. 5). The ⑀V1-2 peptide was less effective at a higher concentration of TPA (100 nM) (data not shown), as reported previously in cardiac myocytes (18).
A prediction from these data is that the ⑀V1-2 peptide should specifically inhibit the TPA-induced translocation of PKC⑀ in U937 cells. Such inhibition has been reported in cardiac myocytes (15,18). However, in a number of experiments including a time course (0 -60 min) and several concentrations of Triton X-100, the ⑀V1-2 peptide did not significantly inhibit 10 nM TPA-induced translocation of PKC⑀ by Western blotting (data not shown). A similar lack of effect of the ⑀V1-2 peptide was found in preliminary experiments with immunofluorescence localization (data not shown). We suggest two reasons for the inability to demonstrate an inhibitory effect on PKC⑀ translocation. First, U937 cells have a high basal level of particulate PKC⑀ (Figs. 4 and 7) (26), which generates a high background level of immunoreactivity. Second, the peptide blocks only the binding of PKC⑀ to RACK proteins. Interaction of PKC⑀ with other binding proteins will not be inhibited. For example, PKC␣, PKC⑀, and PKC have been reported to bind to caveolin (31). PKC also binds to AKAP79 (32) and to PICK1 (33). Consequently, any effects of ⑀V1-2 on PKC⑀ translocation will depend on the relative proportions of RACK and other binding proteins in a particular cell type. We are currently investigating possible effects of the peptide on PKC⑀ intracellular localization in the presence of TPA using confocal microscopy.
Membrane-associated PKC⑀ Is Active in Untreated U937 Cells and Is Greatly Reduced during Calphostin C-induced Apoptosis-The above results suggest that the binding of PKC⑀ to RACK proteins mediates the protective effect of TPA against both calC-and TNF-␣-induced apoptosis in U937 cells. These results further suggest that the anti-apoptotic effect of TPA requires the phosphorylation of target proteins by PKC⑀, as PKC only binds to RACK proteins in the presence of activating cofactors (34). Recent reports also suggest that PKC⑀ plays a role in the survival of cells that have not been stimulated with agents that directly activate PKC. Expression of PKC⑀, but not of PKC␦, extended the survival of interleukin-3 (IL-3)-dependent cells in the absence of the cytokine (35), and stable expression of PKC⑀ protected transformed rat embryo fibroblasts from cisplatin-induced apoptosis (17). However, we found that the ⑀V1-2 peptide had no effect on the levels of apoptosis induced by calC and TNF-␣ or in untreated cells, suggesting that association of PKC⑀ with RACK proteins is not involved in U937 cell survival. PKC⑀ does, however, appear to be inhibited in TNF-␣-, Fas-and ceramide-induced apoptosis, where PKC⑀ is displaced from the particulate fraction to the cytosol (16). We therefore determined if calC also displaced PKC⑀ from the membrane and whether membrane-associated PKC⑀ was active in U937 cells.
PKC⑀ immunoreactivity was divided between the particulate and cytosolic fractions of proliferating U937 cells, as has been reported by others (26). Particulate PKC⑀ immunoreactivity was reduced at 10 min by calC and TNF-␣ (Fig. 7) and ceramide (data not shown) and was associated with increased immunoreactivity in the cytosol (Fig. 7). In calC-treated cells, the re- duction in particulate-associated PKC⑀ was sustained for at least 6 h (data not shown). This effect was reversed by TPA (Fig. 7). calC induced apoptosis and displaced particulate PKC⑀ only when U937 cells were exposed to the inhibitor in the presence of light (Fig. 8). This suggests that the effects of calC may be due to direct inhibition of PKC, as calC must be illuminated to inhibit the enzyme (36). Particulate levels of PKC were not changed by either TNF-␣ or calC (data not shown), a result that is consistent with a recent report (16).
To determine if membrane-associated PKC⑀ was active and whether PKC⑀ activity was reduced by calC, particulate fractions were assayed directly using a method that does not utilize reconstitution and artificial activation of the enzyme. This method was previously used to demonstrate that particulateassociated PKC was not catalytically active during excitatory amino acid-induced death of primary cortical neurons (21). The assay involved preincubating particulate fractions with or without a PKC inhibitor peptide and assaying for kinase activity using a peptide substrate. As shown in Table I, particulate fractions of U937 cells catalyzed the phosphorylation of a peptide derived from the epidermal growth factor receptor. Phosphorylation was linear for up to 10 min (data not shown) and was largely inhibited by inclusion of a PKC inhibitor based on PKC pseudosubstrate- (19 -31) (Table I). Putative PKC phosphorylation of the peptide substrate was decreased 70% by particulate fractions from calC-treated cells, and this reduction was mostly prevented by TPA (Table I). The residual PKC activity that was not inhibited by calC may have been due to PKC (16). Similar results were obtained in one experiment using neurogranin- (28 -43) as the substrate, which is a highly specific substrate for PKC (data not shown).

DISCUSSION
Much evidence has accumulated that suggests a role for PKC in the inhibition of apoptosis. The calcium-independent isoform PKC⑀ has recently been implicated in the protection of cardiac myocytes from hypoxia-induced cell death (15). The studies presented here suggest that PKC⑀ is also required for the inhibition of apoptosis by the tumor-promoting phorbol ester TPA. The major evidence for this is that the ⑀V1-2 peptide derived from the V1 region of PKC⑀ blocked TPA protection against both calC-and TNF-␣ induced apoptosis. A peptide made from scrambled amino acids present in the ⑀V1-2 peptide and a peptide specific for PKC␤ had no effect on TPA protection. We were unable, however, to consistently demonstrate an effect of ⑀V1-2 on TPA-induced translocation of PKC⑀. This result may be partly due to the high level of particulate PKC⑀ in U937 cells (26), which generates a high background level of immunoreactivity. Furthermore, the ⑀V1-2 peptide will not block the interaction of PKC⑀ with other binding proteins such as caveolin (31), AKAP79 (32), and PICK1 (33). RACK proteins, however, apparently bind a large proportion of translocated PKC (34), so it is also possible that the peptides that inhibit binding to RACK proteins allow PKC to associate with proteins in the particulate fraction that are not bound under normal conditions. This phenomenon may be cell type-specific, and we are currently investigating this using immunolocalization and confocal microscopy. Studies in other cells, however, have established the specificity of ⑀V1-2 for inhibiting PKC⑀ translocation. In neonatal rat cardiomyocytes, the ⑀V1-2 peptide inhibited TPA-induced translocation of PKC⑀ measured by confocal microscopy and hypoxia-induced PKC⑀ translocation measured by immunoblotting (15,18). It is therefore likely that TPA-induced binding of PKC⑀ to RACK proteins is required for the suppression of apoptosis.
The binding of PKC to RACK proteins requires the presence of PKC activating cofactors (34). This suggests that suppression of apoptosis by TPA requires that PKC⑀ not only bind to RACK proteins, but that PKC⑀ also phosphorylate target proteins. One potential target for PKC⑀ in the TPA-induced suppression of apoptosis is the Bcl-2 protein, which requires phosphorylation on putative PKC phosphorylation sites for its apoptosis-suppressive function. Serine-to-alanine mutations at putative PKC phosphorylation sites in Bcl-2 prevented the PKC activator bryostatin-1 from blocking apoptosis induced by IL-3 withdrawal (37). Expression of Bcl-2 may also be regulated by PKC. In IL-3-dependent B cells, phorbol ester treatment increased Bcl-2 expression via a cyclic adenosine monophosphate response element in the bcl-2 promoter. This element is bound by cAMP response element-binding protein, which is phosphorylated by PKC after phorbol ester stimulation on the same serine residue that is phosphorylated when IL-3 is pres-   (38,39). A role for PKC in regulating the levels of Bcl-2 protein is further supported by a report showing that PKC inhibition resulted in apoptosis and a reduction in Bcl-2 in glioma cells (40). Furthermore, overexpression of PKC⑀, but not of PKC␦, protected IL-3-dependent TF-1 cells from apoptosis and increased Bcl-2 levels, but did not abrogate IL-3 dependence (35). Overexpression of Bcl-2 in IL-3-dependent hematopoietic cells also did not cause IL-3 independence, suggesting that the anti-apoptotic signal transduction cascade from IL-3 may operate via PKC⑀ and Bcl-2 (41). Overexpression of PKC⑀ has also been shown to protect transformed rat embryo fibroblasts against cisplatin-induced apoptosis (17).
The overexpression data suggest that PKC⑀ may have a role in cell survival where PKC has not been activated by phorbol ester, and it is therefore intriguing that diverse apoptotic stimuli displace PKC⑀ from the membrane (16). We stress, however, that there is as yet no direct evidence that this displacement is causally linked to apoptosis. Although we provided evidence that particulate PKC⑀ had catalytic activity, the ⑀V1-2 peptide did not induce apoptosis or increase the sensitivity of the cells to TNF-␣ or calC. These results suggest that if PKC⑀ does protect cells from apoptosis in the absence of TPA, then this does not involve a PKC⑀-RACK interaction. The effect of calC on displacement of particulate PKC⑀ could be a direct result of calC binding to the enzyme. This may explain the lack of an effect on PKC, which is not inhibited by calC (16). Alternatively, the calC effect could be indirect. For example, calC has been reported to elevate ceramide levels in WEHI-231 cells (42), and exogenous ceramide has been reported to displace particulate PKC⑀ (16).
Although cell-permeable ceramides induce apoptosis in U937 cells (2,43) and other cells (44 -46), it has not been demonstrated that ceramide production is required for TNF-␣-and Fas-induced apoptosis in these cells. However, TPA prevents the generation of ceramide and inhibits apoptosis in TNF-␣treated U937 cells (47), and TPA also blocks apoptosis, sphingomyelin hydrolysis, and ceramide generation in cells induced to undergo apoptosis with ionizing radiation (48). Therefore, in these models, inhibition of apoptosis by PKC activation correlates with reduced ceramide accumulation. Consequently, one possible target for TPA-activated PKC is sphingomyelinase, resulting in an inhibition of sphingomyelin hydrolysis. This suggestion is supported by the observation that apoptosis following exposure to PKC inhibitors correlates with the activation of a neutral sphingomyelinase and an accumulation of ceramide (21). However, TPA is also able to block apoptosis induced by exogenous ceramide (2). Although this is nonphysiological, it does suggest the possibility of a regulatory target that is downstream of ceramide production. One possible target is sphingosine kinase, which, when activated, results in a reduction in ceramide and an accumulation of sphingosine 1-phosphate (SPP). Although ceramide levels are reduced by this process, the generation of SPP also appears to be important, as exogenous SPP blocks ceramide-and TNF-␣-induced apoptosis (43). SPP also blocks apoptosis induced in U937 cells by calC, suggesting that SPP is a possible mediator of antiapoptotic signals originating from PKC (43).