Protein kinase C (PKC) inhibits fas receptor-induced apoptosis through modulation of the loss of K+ and cell shrinkage. A role for PKC upstream of caspases.

Cell shrinkage and loss of intracellular K(+) are early requisite features for the activation of effector caspases and apoptotic nucleases in Fas receptor-mediated apoptosis of Jurkat cells, although the mechanisms responsible for both process remain unclear (Bortner, C. D., Hughes, F. M., Jr., and Cidlowski, J. A. (1997) J. Biol. Chem. 272, 32436-32442). We have now investigated the role of protein kinase C (PKC)-dependent signaling in the regulation of Fas-induced cell shrinkage and loss of K(+) during apoptosis. Anti-Fas induced cell shrinkage was blocked during PKC stimulation by the phorbol ester 12-O-tetradecanoylphorbol-3-acetate (PMA) and by bryostatin-1. Conversely, inhibition of PKC with Gö6976, enhanced the anti-Fas-mediated loss of cell volume. Analyses of mitochondrial membrane potential and DNA fragmentation revealed that the PKC-mediated effect observed in cell volume is propagated to these late features of apoptosis. Flow cytometric analyses and (86)Rb efflux experiments revealed that a primary effect of PKC appears to be on the modulation of Fas-induced K(+) efflux, since both PMA and bryostatin-1 inhibited extrusion of K(+) that occurs during Fas-mediated cell death, and Gö6976 exacerbated the effect of anti-Fas. Interestingly, high extracellular K(+) significantly blocked the effect of anti-Fas alone or anti-Fas combined with Gö6976, suggesting an underlying effect of PKC on K(+) loss. Western blot analyses showed the caspase-dependent proteolysis of PKC isotypes delta, epsilon, and theta in whole cell extracts from anti-Fas treated Jurkat T cells. However, stimulation of PKC by PMA or bryostatin-1 prevented this isotypic-specific PKC cleavage during apoptosis, providing further evidence that PKC itself exerts an upstream signal in apoptosis and controls the caspase-dependent proteolytic degradation of PKC isotypes. Finally, we show that PMA or bryostatin-1 prevents the activation of caspase-3 and caspase-8. Thus, this study shows that the protective effect that PKC stimulation exerts in the Fas-mediated apoptotic pathway occurs at a site upstream of caspases-3 and -8.

During apoptosis, cells activate an intrinsic death program that eventually eliminates them from the surrounding cells in a process that in vivo reduces the likelihood of an inflammatory response. Apoptosis induced by ligation of the cell surface Fas/ Apo1/CD95 receptor (1-4) plays a pivotal role in the physiological turnover of lymphoid cells, which dysfunction can lead to an overgrowth of lymphoid organs and to a variety of disease states (5). Studies on cell death induced by activation of the well defined Fas receptor system have indeed provided extensive information on the cellular and molecular bases of apoptosis. Cell death mediated by Fas receptor (for review, see Refs. 6 -8) is the result of the initiation and transmission of cellular signals that activate the inherent cell death machinery, which ultimately leads to the disassembly of the cell in a highly regulated process. In this context, activation of the Fas receptor by the Fas ligand molecule or anti-Fas antibody sequentially induces clustering and trimerization of the receptor, followed by its interaction with Fas-associated death domain (FADD) adaptor protein and subsequent recruitment of caspase-8 to the Fas receptor complex, which leads to the formation of deathinducing signaling complex (DISC) 1 (9), and initiates the activation of the caspase pathway (10 -12).
From lymphocytes to neurons, an incipient mark of apoptotic cells is their shrunken morphology (13)(14)(15)(16)(17)(18)(19). However, the cellular mechanisms involved in apoptotic cell shrinkage are not well understood. The maintenance of cell volume is an energydemanding process, the failure of which dramatically challenges cell function. During apoptosis, cells follow a program which success is dependent on many cellular functions that are required to fully complete the entire apoptotic cascade. Hence, this early apoptotic loss of cell volume has been proposed to be an active regulated process required for the apoptotic progression. Our laboratory and others have demonstrated that cell shrinkage and the loss of intracellular K ϩ are early and necessary features of apoptosis (13,14,19,20). In fact, in a total cell population forced to die by an apoptotic stimulus, only the shrunken/low intracellular [K ϩ ] cells show characteristics of apoptosis, i.e. DNA degradation and effector caspase activation (19). Activation of effector caspases plays a key role in the hierarchy of the cell death cascade (21). Caspase-8 is the most proximal caspase to the Fas receptor known to date, and its enzymatic action initiates the activation of the caspase cascade (10,11) in which a subsequent step is activation of caspase-3 * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
The PKC family comprises at least 11 serine/threonine kinase isozymes, with different constitutive expression among cell types, and which are believed to have non-redundant cellular functions (29 -31). PKCs have been classified in three groups, based on their structure and response mechanisms to regulatory factors. Conventional PKC (␣, ␤, ␤II, ␥) are Ca 2ϩdependent and in vivo activated by diacylglycerol (DAG) or by the synthetic phorbol ester 12-O-tetradecanoylphorbol-3-acetate (PMA). The novel isotypes (␦, ⑀, , , ) do not respond to Ca 2ϩ but are activated by DAG and PMA. Atypical PKCs (, , ) are insensitive to both Ca 2ϩ and DAG. PKC is responsible for transducing many cellular signals during a variety of cellular processes such as mitogenesis, cellular metabolism, differentiation, tumor promotion, and apoptosis. In addition, PKC is known to regulate the activity of different plasma membrane proteins, including transporters, channels, and plasma-membrane related cytoskeletal proteins (see, e.g., . In Fas-induced apoptosis, phorbol esters cause a marked attenuation of the cell death pathway (39,40), and the enzymatic kinase activity of PKC has been shown to be partially blocked in apoptotic Jurkat cells (41). Moreover, many other cellular models of apoptosis have been used to demonstrate that, during the transduction of cell death signals, there is selective inhibition/activation of PKC isotypes, depending on cell type and apoptotic stimuli considered (2,(41)(42)(43)(44). In fact, the specific cleavage of different PKC isotypes through caspasedependent mechanisms during apoptosis (2,42,(45)(46)(47)(48) provides evidence that PKC functions in the cell death program to positively or negatively regulate the process. Indeed, a variety of apoptotic stimuli have been shown to differentially induce rapid changes in the subcellular location of the PKC isoenzymes (41,44,49,50), suggesting also that the relative sublocation of PKC in the cell may be important in their specific cell death signaling. However, based on proteolytic inactivation/ activation of PKC isotypes mediated by catalytic activation of effector caspases, most studies have suggested that PKC modulates apoptosis in a pathway downstream of caspases (2,45,46,51,52).
In the present study, we investigated the role of PKC signaling in the modulation of the early cell shrinkage and loss of intracellular K ϩ that occurs during apoptosis and delineated a signaling pathway that controls these necessary features of programmed cell death. Our results show that PKC can modulate apoptosis in an apical phase of the Fas pathway by modulating the levels of intracellular K ϩ and cell shrinkage. In addition, PKC-dependent suppression of apoptosis involves inhibition of the most proximal caspase, caspase-8, as well as inhibition of a late caspase, caspase-3.

Reagents and Protocol of Cell Culture and Treatments
Jurkat T cells were cultured at 37°C and 7% CO 2 atmosphere in RPMI 1640 medium supplemented with 10% heat-inactivated fetal calf serum, 4 mM glutamine, 75 units/ml streptomycin, and 100 units/ml penicillin. In each experiment, Jurkat cells were plated to 0.5-0.8 ϫ 10 6 cells/ml in the presence or absence of 50 ng/ml anti-human Fas IgM, clone CH-11 (Kamiya Biomedical, Seattle, WA). High potassium/isotonic medium (high potassium medium) contained the same composition as RPMI 1640 medium except for [KCl], which was 102.7 mM (instead of 5.4 mM), and for [NaCl], which was 5.4 mM instead of 102.7 mM. The phorbol ester 12-O-tetradecanoylphorbol 13-acetate (PMA; synthetic analog of diacylglycerol), Gö6976 (Calbiochem, La Jolla, CA), and bryostatin-1 (Biomol, Plymouth Meeting, PA) were dissolved in Me 2 SO (final [Me 2 SO] never exceeded 0.1% in the culture medium). Concanavalin A (ConA; Sigma) was dissolved in RPMI 1640 medium and used at a final concentration of 20 g/ml in the cell culture. In all cases, cells were preincubated in the presence of the correspondent PKC modulator for 30 min. Subsequently, apoptosis was induced by addition of the anti-Fas antibody (anti-Fas), and cells were incubated as described above until sample processing. For caspase inhibition, caspase-3 inhibitor benzyloxycarbonyl-Asp-Glu-Val-Asp-fluoromethyl ketone (z-DEVD-fmk) and caspase-8 inhibitor benzyloxycarbonyl-Ile-Glu-Thr-Asp-fluoromethyl ketone (z-IETD-fmk) were purchased from Kamiya Biomedical, dissolved in Me 2 SO, and used at a final concentration of 25 M.

Flow Cytometry
Analysis of Cell Size-Changes in cell volume were assessed by flow cytometry analysis using a Becton Dickinson FACSort. Two ml of cell suspension were used per sample prior to addition of propidium iodide (PI, 10 g/ml final concentration; Sigma). Ten thousand cells were examined for each sample by excitation with 488 nm argon laser and by determining their position on a forward scatter versus side scatter dot plot. Light scatter in the forward direction is proportional to cell size, and light scattered at 90 o angle (side scatter) is proportional to cell density (53). Changes in the cell size were then examined on a forward scatter versus side scatter dot plot using CellQuest software (14). Cells with high PI fluorescence have lost plasma membrane integrity, and therefore were excluded from volume analysis to ensure the analysis of only viable cell population. In the control sample, a gate was drawn around the normal population and cell size distribution was then represented in histograms showing forward scatter versus cell number. The percentage of shrunken apoptotic cells was determined by analyses of forward scatter histograms from cells analyzed under each condition.
Analysis of DNA Content-For DNA content analysis, cells were fixed in cold 70% ethanol, pelleted, washed once in phosphate-buffered saline (PBS), and resuspended in PBS containing 20 g/ml PI and 1 mg/ml ribonuclease A (Sigma). Seven thousand and five hundred fixed cells were examined per each experimental condition by flow cytometry, and percentage of degraded DNA was determined by the number of cells displaying subdiploid (sub-G 1 ) DNA divided by the total number of cells examined.
Analysis of Mitochondrial Membrane Potential (MMP)-Changes in MMP were assessed by flow cytometry using the intramitochondrial dye JC-1 (Molecular Probes) using previously documented methodology (54). Briefly, 30 min prior to flow cytometric analyses, 10 M JC-1/ml of culture were added. Ten thousand cells were then examined per sample using 530 nm (FL-1) versus 585 nm (FL-2) fluorescence. JC-1 forms aggregates in cells, which leads to high values in FL-2 fluorescence, indicating a normal mitochondrial potential. Loss of MMP leads to reduction in FL-2 fluorescence (aggregate state of JC-1) and a concomitant increase in FL-1 fluorescence (monomeric state of JC-1). Data were converted to density plots using CellQuest software.
Measurement of Intracellular Potassium-Changes of intracellular K ϩ levels were assessed in Jurkat cells using a Becton Dickinson FACSort and the fluorescent K ϩ -binding dye, benzofuran isophalate (PBFI-AM; Molecular Probes, Eugene, OR) as described previously (14). In brief, cells were incubated in the presence of 5 M PBFI-AM from a fresh stock solution made dissolving PBFI-AM in equal volumes of 25% (w/v) Pluronic F-127 (Molecular Probes) and Me 2 SO. One hour prior to cytometric analysis, 1 ml of cells were harvested, resuspended in RPMI 1640 medium, and loaded with the fluorescent dye to allow the compound to reach equilibrium between the extra and intracellular compartments. Immediately prior to flow cytometric analysis, 10 g/ml PI were added to each sample. Fluorescence of PBFI-AM and PI were measured by excitation at 340 -350 and 488 nm, respectively. Ten thousand cells were analyzed per sample on a PBFI (K ϩ ) versus PI dot plot and represented as PBFI (K ϩ ) fluorescence versus forward scatter histograms. Cells displaying high PI fluorescence were gated and excluded from ion analysis. In the control Jurkat population, a gate was drawn around PBFI (K ϩ )-positive cells and the percentage of cells with decrease in PBFI (K ϩ ) fluorescence was determined under each condition by comparison to the control cell population.

Measurement of 86 Rb ϩ Efflux
For K ϩ efflux measurements, 86 Rb ( 86 RbCl, Amersham Pharmacia Biotech) was used as isotopic tracer of K ϩ . Jurkat cells (0.5-0.8 ϫ 10 6 cells/ml) were loaded for 18 h with RPMI 1640 medium containing 2 Ci/ml 86 RbCl. After two washes with normal RPMI medium, cells were preincubated in the presence of PMA, bryostatin-1, or Gö6976 and subsequently with 50 ng/ml anti-Fas for 3 h. At 1-h intervals, triplicates of 1-ml aliquots of cells from each treatment were harvested and pelleted. The pellet was washed with RPMI medium, resuspended in RPMI containing 0.5% Triton X-100, and radioactivity counted in a LS 6500 scintillation counter (Beckman Coulter). The average of cpm of 86 Rb in the pellet fraction in three independent experiments run in triplicate was determined under each condition.

Western Blot Analysis of PKC Isotypes
Jurkat T cells were collected by gentle centrifugation and washed with cold PBS. Pelleted cells were immediately placed in ice, resuspended in lysis buffer (20 mM Tris-HCl, pH 7.5, 2 mM EDTA, 150 mM NaCl, and 0.5% Triton X-100) containing a mixture of protease inhibitors (1 M pepstatin, 1 M leupeptin, 1 g/ml aprotinin, 1 M pepstatin, and 1 mM phenylmethylsulfonyl fluoride), and processed with a Dounce homogenizer. After 15 min of centrifugation at 16,000 ϫ g, the supernatant was collected and assayed for protein concentration by the method of Bradford (55) using the Bio-Rad system (Richmond, CA). Twenty to 50 g of protein/sample were equally diluted in Laemmli loading buffer and denatured at 99°C for 5 min, followed by 2 h of electrophoresis in 12% SDS-polyacrylamide gel (NOVEX, San Diego, CA) at 120 V. Gels were then electrophoretically transferred to nitrocellulose membranes (Schleicher & Schuell) at 42 V for 1.5 h, and stained with Ponceau S (Sigma) to verify the equal amount and quality of protein between lanes prior to Western blot analyses. Membranes were blocked overnight at 4°C in Tris-buffered saline (TBS) containing 0.05% Tween (Sigma) and 5% of nonfat dry milk. Next, PKC-specific antibodies (Transduction Laboratories, Lexington, KY) recognizing epitopes for PKC-␣, -, -␦, and -⑀ were diluted 1:500 in TBS, 0.05% Tween, 0.5% milk, and membranes were blotted with the corresponding antibody for 1 h at room temperature. Blots were washed three times with TBS, 0.05% Tween and incubated for 1 h with peroxidase-linked anti-mouse IgG (Amersham Pharmacia Biotech ECL) diluted 1:5000 in TBS, 0.05% Tween, 0.5% milk. After three washes with TBS, membranes were treated with ECL chemiluminescence detection system, exposed to hyperfilm (ECL, NEN Life Science Products), and developed.

In Vitro Protease Assay
Caspase-3-like and caspase-8-like activities from Jurkat lysates were measured using a colorimetric assay (Lab Biovision, Palo Alto, CA) according to the manufacturer's instructions. In brief, cells were pelleted and resuspended in chilled lysis buffer for 10 min. Supernatant was collected by 15-min centrifugation at 16,000 ϫ g, kept on ice, and protein extract concentration determined by the Bradford method. For caspase-3-like activity, 40 g of Jurkat cell extract were incubated for 1 h at 37°C in the caspase buffer containing acetyl-Asp-Glu-Val-Aspp-nitroanilide. After 1 h of incubation (linearity of the assay ranged to 2 h of incubation; data not shown), release of the chromogenic compound p-nitroanilide cleaved from the substrate was measured at 400 nm using a Beckman DU 650 spectrophotometer. Determination of caspase-8-like activity was carried out using the same approach, but acetyl-Ile-Glu-Thr-Asp-p-nitro-anilide was used as the caspase-8 substrate.

PKC Modulates Fas-induced Cell
Shrinkage-Previous studies in several model systems, including Fas-mediated Jurkat T cell apoptosis, have revealed that PKC stimulation inhibits programmed cell death, but the underlying mechanisms of this effect are not well understood. We have now examined the potential role of PKC in volume regulation and ionic changes that occur during apoptosis, by monitoring changes in cell volume by flow cytometry, which permits quantification of the  Table I) and increased to 24.7 Ϯ 1.8% after 3 h of anti-Fas exposure (Fig.  2B). Activation of PKC in Jurkat cells with 20 nM PMA (56) or 10 nM bryostatin-1 (57) (tumor and non-tumor promoter PKC activators, respectively) did not produce any significant effect on the cell size in the control cells (Fig. 2, C and E, respectively). However, in anti-Fas-treated cells, cell shrinkage was significantly reduced by PKC stimulation with PMA (Fig. 2D) and bryostatin-1 (Fig. 2F).
To determine whether inhibition of PKC could also exert modulation of anti-Fas-mediated cell shrinkage, cells were treated with a specific inhibitor of PKC named Gö6976, a bisindolylmaleimide compound that acts as competitive inhibitor for the ATP-binding site of PKC (58). Treatment of control cells with 5 M Gö6976 (Fig. 2G) did not produce changes in the cell volume. However, anti-Fas-induced cell shrinkage was significantly triggered by PKC inhibition with Gö6976 (Fig. 2H), reaching 36.8 Ϯ 3.3% in the total cell population. Comparable results were found using GF109203X to inhibit PKC, but we observed a slight toxic effect of this compound in the cell population (data not shown) and proceeded with the study using Gö6976.

Fas-induced Changes of MMP and DNA Degradation Are
Modulated by PKC -We next examined the effect of PKC modulation on later events in the Fas pathway, i.e. loss of mitochondrial membrane potential and DNA degradation. Fasinduced changes in the morphology and transmembrane potential of mitochondria have been shown to occur prior to nuclear events and after caspase-8 activation (59 -62). More recently, it has been shown that the loss of MMP occurs in the shrunken population of Fas apoptotic cells (54). In order to determine whether Fas receptor-induced changes in the MMP are modulated by a PKC-dependent process, we assessed the MMP during apoptosis in the presence of PMA, bryostatin-1, and Gö6976 using JC-1, a dye that has been proven to be specific for measuring changes in the MMP (59). Intracellular fluorescence in Jurkat cells was monitored after loading the cells with JC-1 (see "Materials and Methods"). High FL-2 fluorescence (585 nm) corresponds to the aggregated form of JC-1, and is proportional to an intact MMP. Loss of MMP leads to a reduction in FL-2 and a concomitant gain of cells that exhibit high FL-1 fluorescence (530 nm). Treatment of Jurkat cells with anti-Fas (Fig. 3) increases the amount of cells that have lost their MMP, which is proportional to the number of cells that have lost FL-2 fluorescence (R2 region shown in Fig. 3): 37.3% during anti-Fas treatment versus 7.2% in the normal population. PKC stimulation by either PMA or bryostatin-1 led to a significant inhibition of mitochondrial depolarization induced by anti-Fas in Jurkat cells (Fig. 3, D and F). Additionally, Gö6976 caused an enhanced loss of MMP during anti-Fas-treatment (Fig. 3H). Once again, PKC stimulation or inhibition alone did not exert  2) were gated as shrunken cells. Averaged number of shrunken cells Ϯ S.E. of 9 -12 independent experiments are summarized in the first column. For MMP analyses, the percentage of cells displaying loss of MMP was determined by gating cell population displaying increase in fluorescence at 535 nm (R2 region shown in plots of Fig. 3). Data of MMP summarizes the mean Ϯ S.E. of at least three separate experiments. For analyses of DNA histograms (Fig. 4), the amount of cells displaying subdiploid peak of DNA under each experimental condition was determined and mean Ϯ S.E. calculated from at least four separate analyses. Results were analyzed by Student's t test. *, significantly different from control at p Ͻ 0.05; **, significantly different from control at p Ͻ 0.001.  significant changes in the MMP (Fig. 3, C, E, and G). These results suggest that PKC plays a role upstream of mitochondrial membrane depolarization during Fas-induced events.
Fragmentation of genomic DNA is a late event in apoptosis known to occur downstream of mitochondrial events. Fragmentation of DNA in fixed cells can be detected by analyzing DNA content histograms, which display subdiploid DNA (Fig. 4). The number of cells containing subdiploid DNA increases during anti-Fas-treatment (Fig. 4B) compared with control (untreated) Jurkat cells (Fig. 4A), which is consistent with the reported induction of apoptosis and DNA fragmentation that anti-Fas treatment produces (63). PKC stimulation did not affect DNA degradation in control cells (Fig. 4, C and E). However, anti-Fas-induced DNA degradation was significantly attenuated by PKC stimulation with 20 nM PMA or 10 nM bryostatin-1 (Fig. 4,  D and F, respectively). In contrast, 5 M Gö6976 enhanced DNA degradation during anti-Fas-mediated apoptosis, with no effect in the control population (Fig. 4, H and G, respectively). These results provide evidence that PKC activity influences apoptosis from an early event (loss of cell volume) through a signal that is transduced to a later event (loss of MMP) and a final point in apoptosis (nucleic acid degradation). Table I summarizes data of cell volume, loss of MMP, and DNA content analyses of similar experiments performed over a period of several weeks. Although different passages of Jurkat cells were used in the study, the consistency of these observations is verified by the low variability of the data.
Receptor-dependent Activation of PKC Interferes with Fasinduced Apoptosis-Activation of PKC is an essential pathway for T cell activation, which occurs upon triggering the TCR-CD3 receptor complex expressed in the plasma membrane of lymphocytes. Stimulation of PKC via TCR-CD3 complex induces mitogenesis and plays a pivotal role in a variety of biological responses mediated by T cells, including antigen recognition. Thus, we tested whether or not stimulation of PKC via induction of membrane receptor signaling could lead to an inhibition of Fas-induced apoptosis similar to that observed by direct PKC stimulation with PMA or bryostatin-1. Jurkat cells were stimulated with ConA for 30 min prior to addition of the Fas antibody. ConA causes engagement and stimulation of the T cell receptor (TCR-CD3) complex (64,65) and activates endogenous PKC through activation of phospholipase C within minutes, with a subsequent elevation of intracellular levels of diacylglycerol (66), the physiological activator of PKC. After 3 h of treatment, forward scatter analyses of cells treated with anti-Fas antibody in the presence or absence of ConA were undertaken by flow cytometry. In these experiments, ConA (20 g/ml) alone did not induce loss of cell volume (Fig. 5), and mimicked the inhibitory effect of PMA and bryostatin-1 (Fig. 2, Table I) on Fas-induced cell shrinkage. To address whether the protective effect of ConA was expanded to other features of apoptosis, MMP and DNA fragmentation were also analyzed during treatment of cells with ConA. As shown in Fig. 5, both loss of MMP and DNA degradation were also attenuated during Fas-induced apoptosis in the presence of the lectin.
Regulation of Intracellular [K ϩ ] by PKC during Apoptosis-Our laboratory previously demonstrated that cell shrinkage during Fas receptor-induced apoptosis is accompanied by a significant loss of intracellular K ϩ , which provokes a change in the intracellular ionic environment that may be required for caspase activation and nuclease activity (19). To determine if the mechanism of PKC-mediated inhibition of Fas receptor- Jurkat T cells were treated as described in Fig. 2 legend, fixed in 70% ethanol, and stained with 20 g/ml PI containing ribonuclease. Cell cycle histograms were generated by flow cytometric analyses of DNA content represented versus cell number. The vertical bar was drawn to gate cells containing subdiploid (fragmented) DNA. Histograms are representative of at least four independent experiments. mediated apoptosis occurs through a modulatory effect of the intracellular K ϩ levels, changes in this ion were analyzed using PFBI-AM (K ϩ ) (a fluorescent potassium-binding dye), and PI to eliminate cells with a loss of membrane integrity (see "Materials and Methods") ( Fig. 6). To investigate if cell volume modulation is coupled to changes in intracellular K ϩ levels, flow cytometric analyses are shown by representing forward scatter versus PBFI fluorescence. A significant decrease of PFBI fluorescence and a concomitant loss of cell volume occurs only in anti-Fas-treated Jurkat cells compared with the distribution in the normal cell population (Fig. 6, B and A, respectively). However, the effect of anti-Fas on the K ϩ movement was attenuated by PKC stimulation with PMA or bryostatin-1 (Fig. 6,  D and F). In contrast to these results, inhibition of PKC with Gö6976 enhanced the anti-Fas-induced loss of K ϩ (Fig. 6H). Table II summarizes analyses of flow cytometric data, in which cells displaying PFBI fluorescence values below the control cell population shown in Fig. 6 were gated and counted as cells with decreased K ϩ concentration, as described under "Materials and Methods." Importantly, our results indicate that neither of the PKC modulators triggered any significant change in the intracellular [K ϩ ] in Jurkat cells in the absence of anti-Fas (Fig. 6,  D, F, and H). This supports our premise that PKC-induced resistance to anti-Fas-induced apoptosis may involve the modulation of ionic transport activity that is specifically required during Fas receptor-induced apoptosis.
Anti-Fas Promotes Efflux of 86 Rb (K ϩ ), Which Can Be Blocked during PKC Activation-The results shown in the previous section suggested a potential PKC-dependent modulation of the cell death pathway by blocking K ϩ efflux induced by anti-Fas antibody. To investigate this question, we determined 86 Rb efflux from Jurkat cells prelabeled with 86 Rb (see "Materials and Methods"), which is used as a marker of K ϩ . Extrusion of 86 Rb from control preloaded cells was assessed for a period of 3 h (Fig. 7), showing that Fas antibody induces efflux of K ϩ . The amount of 86 Rb remaining in the cells after 3 h of incubation in 86 Rb free-RPMI medium represented approximately 31% of the initial intracellular content of the isotope. Treatment of cells with anti-Fas significantly enhanced 86 Rb efflux (12% of the initial content), confirming that a primary effect of the anti-Fas-induced loss of K ϩ occurs via extrusion of the ion (14). Interestingly, both PMA and bryostatin-1 interfered with the Fas-induced efflux of 86 Rb, as represented in Fig.   7 by a retention of intracellular levels of rubidium to 22% and 21% (respectively) of the initial radioactive load during anti-Fas treatment. In contrast, the presence of Gö6976 dramati-FIG. 6. Fas-induced movement of K ؉ can be modulated by PKC. Jurkat T cells were treated with or without 20 nM PMA, 10 nM bryostatin-1, or 5 M Gö6976 for 30 min and, next, with or without 50 ng/ml anti-Fas for 3 h. For intracellular PBFI-AM (K ϩ ) fluorescence measurements, 5 M PBFI-AM and 10 g/ml PI were added in 1 ml of cells 1 h and 5 min prior to flow analyses, respectively. Samples were then analyzed in a FACSVantage flow cytometer on PBFI (K ϩ ) versus PI fluorescence dot plot to eliminate cells PI positive (nonviable). In order to examine whether apoptotic cell shrinkage was coupled to changes in K ϩ content, viable cells were then examined and represented (A-H) on a forward scatter light (cell size) versus PBFI (K ϩ ) three-dimensional plot by flow cytometry. Data are representative of at least seven separate experiments.

FIG. 5. Treatment of Jurkat cells with concanavalin A has a protective effect on anti-Fas-induced cell shrinkage, mitochondrial depolarization, and DNA degradation.
Jurkat T cells were pretreated for 30 min with or without a dose of 20 g/ml ConA, which activates PKC by binding to the TCR-CD3 complex of T cells (see "Results"). After subsequent treatment of cells with anti-Fas antibody for 3 h, cells were then analyzed for cell size (upper panels), mitochondrial membrane potential (middle panels), and DNA fragmentation (lower panels), as previously described. Numbers show percentages of shrunken cells, cells with loss of MMP, and cells with degraded DNA. Histograms are representative of three separate experiments run in triplicate, from which mean Ϯ S.E. were calculated.
cally exacerbated Fas-induced 86 Rb efflux from the cells since the amount of radioactivity present in the cells after 3 h of treatment represented the 4% of the initial content. 86 Rb efflux data at 1-and 2-h time points (data not shown) were consistent with the results shown at the 3-h time point. These results suggest that PKC has, at least, a primary point of modulation of Fas-induced cell death at a mechanism of extrusion of intracellular K ϩ that is selectively activated during apoptosis.
PKC Inhibition Enhances Anti-Fas-induced Apoptosis through a [K ϩ ]-dependent Mechanism-As shown in Figs. 2-5, PKC inhibition can trigger apoptotic features that occur during anti-Fas treatment. Moreover, data from K ϩ (PFBI-AM) analysis and 86 Rb efflux experiments suggest that modulation of the movement of potassium across the membrane could play an important role in the mechanism of PKC-mediated regulation of Fas-induced apoptosis. To further investigate whether inhibition of PKC enhances apoptosis through modulation of K ϩ movement across the plasma membrane, we cultured Jurkat cells in an isosmotic medium containing high levels of K ϩ (see "Materials and Methods") and changes in cell volume and DNA content were determined by flow cytometry. The presence of high concentration of K ϩ in the culture medium blocks the net efflux of this ion by elimination of the normal electrochemical gradient of K ϩ in the cell membrane (67). Fas-induced cell shrinkage and DNA fragmentation were partially blocked in the high potassium medium (Fig. 8). Cell shrinkage was attenuated during anti-Fas treatment of cells exposed to high potassium medium (24% in normal medium versus 12.4% in high potassium medium) with a consistent decrease in DNA degradation (16.3% in normal medium versus 9.1% in high potassium medium). When cells were exposed to Gö6976 (PKC inhibition) and anti-Fas, cell shrinkage and DNA fragmentation were triggered to 49.9% and 27.7%, respectively, in the normal RPMI medium. However, this potent effect of Gö6976 was reduced in high potassium medium since only 14.1% of anti-Fas-treated cells were shrunken (compared with the shown 49.9%) and only 10.7% exhibited DNA fragmentation (compared with 27.7%). These results show in sum that, during inhibition of PKC, anti-Fas-induced apoptosis is not enhanced under conditions in which net loss of intracellular K ϩ is inhibited.
Fas Receptor-induced Isotypic PKC Cleavage Is Inhibited by Jurkat cells treated with PKC modulators To determine percentage of cells that had change in PBFI (K ϩ ) fluorescence, a gate on PBFI (K ϩ ) fluorescence versus forward scatter (cell size) was set in the control population and was compared in plots generated from each treatment. Percentage of cells Ϯ S.E. displaying decreased PBFI (K ϩ ) was averaged from seven independent experiments. Results were analyzed by Student's t test. *, significantly different from control at p Ͻ 0.05; **, significantly different from control at p Ͻ 0.001.

Treatment
Percentage of cells with decrease in PBFI (K ϩ ) fluorescence  86 RbCl to load cells with 86 Rb, a physiological tracer of K ϩ . After two washes with normal RPMI medium, equal amount of cells per treatment were preincubated in the presence of 20 nM PMA, 10 nM bryostatin-1, or 5 M Gö6976. At 1-h intervals after addition of 50 ng/ml anti-Fas, triplicates of 1-ml aliquots of cells were lysated and radioactivity measured. The average Ϯ S.E. of 86 Rb counts in the pelleted cells from triplicates was determined under each condition, and the percentage of intracellular 86 Rb was calculated respect to the initial content (radioactive counts) after loading. The plot shows percentage of radioactivity present at 3-h time point under each condition of averaged triplicates with S.E. bars from one of three representative experiments run separately. The results were reproduced in three independent experiments with equivalent data. *, significantly different from control at p Ͻ 0.05. PKC Stimulation-Limited proteolysis of PKC isotypes by caspase-3 dependent mechanisms during apoptosis has been extensively reported (2,42,45,46,52). The generation of PKC mutants resistant to this proteolytic fragmentation has provided experimental evidence that this cleavage may be necessary in the cell death program (46). Expression of PKC-isoforms was analyzed using antibodies directed to PKC-, -⑀, -␦, and -␣ isotypes, which are constitutively expressed in Jurkat cells (56,68). Cleavage of isotypes , ⑀, ␦, but not ␣, in anti-Fas-treated cells was detected by Western blot analysis (Fig. 9). PKC-␣ isotype remained intact among treatments and was used as control since it has been shown that this isotype is not cleaved during Fas receptor apoptosis in Jurkat cells (47). Proteolysis of , ⑀, and ␦ was abrogated in the presence of benzyloxycarbonyl-Val-Ala-Asp-fluoromethyl ketone (a general caspase inhibitor that blocks the Fas cascade) or z-DEVD-fmk (a caspase-3-like activity inhibitor) (data not shown). Importantly, we show for the first time that PKC stimulation with either PMA or bryostatin-1 significantly inhibited the generation of proteolytic fragments of ⑀, ␦, and isotypes in cells exposed to anti-Fas, providing evidence for potential PKC-dependent modulation of the caspase-3 enzyme, known to be responsible of the cleavage of these isotypes during apoptosis (45,47,52).

PKC Stimulation Leads to Inhibition of Caspase-3-like and Caspase-8-like
Activities-To study if the cellular activity of caspases could be modulated by PKC, activation of PKC in Jurkat cells was induced using PMA or bryostatin-1 during anti-Fas treatment. Caspase-3-like activity and its upstream activator, caspase-8-like activity (Fig. 10, A and B, respectively), were measured in fresh Jurkat-lysates from cells cultured in different conditions. Using Asp-Glu-Val-Asp-p-nitroanilide as substrate, treatment of cell with anti-Fas led to a 6-fold increase in caspase-3-like activity (Fig. 10A). However, the induction was inhibited in cells treated with PMA or bryostatin-1, to the same level as cells cultured in the presence of 25 M z-DEVD-fmk caspase-3-like inhibitor and anti-Fas (Fig.  10A). These results suggest that PKC can modulate apoptosis through a mechanism involving regulation of the activation of this protease. Since caspase-3 is cleaved and activated by caspase-8 in the Fas pathway (22), caspase-8-like activity was also assayed in Jurkat cells using Ile-Glu-Thr-Asp-p-nitroanilide as substrate for the enzyme. Treatment of cells with anti-Fas (Fig. 10B) showed increased proteolytic activity of caspase-8. However, PMA or bryostatin-1 partially blocked Fas receptor-induced activation of the enzyme. In contrast to these results, inhibition of PKC with Gö6976 resulted in a significant increase of both caspase-3-like and caspase-8-like activities detected in Fas-induced cell lysates. This observation is consistent with our previous results in that PKC inhibition triggers Fas-dependent apoptosis in the cell population. Using z-IETD-fmk as inhibitor of caspase-8, partial inhibition of the caspase-8-like activity was found. PKC stimulation inhibited caspase-3-like activity more effectively than caspase-8-like activity, indicating that PKC might be controlling apoptosis by regulating the caspase cascade from the most proximal caspase to the Fas receptor (caspase-8), which could amplify the signal to downstream caspases. DISCUSSION The loss of cell volume is a classical hallmark of apoptosis in many species and in a variety of cell types, which suggests that it is an important, conserved process in programmed cell death. It is unlikely that apoptotic cell shrinkage occurs by a general failure of the volume regulatory mechanisms in the cells, since maintenance and changes of cell volume are controlled and mediated by a number of cell volume regulatory mechanisms including ion transport activity, osmolyte accumulation, cytoskeletal reorganization, metabolism, and expression of appropriate genes (69 -72). A variety of second messengers including PKC, Ca 2ϩ , calmodulin-dependent protein kinase, cAMP, and cAMP-dependent protein kinase have been shown to mediate cell volume regulation (73).
The current understanding of the participation of PKC in apoptosis typically places this kinase downstream of caspases, which is consistent with the fact that different PKC isotypes are cleaved in a caspase-3-dependent manner during apoptosis (2,42,(45)(46)(47). However, our data indicate that PKC regulates apoptosis from an early phase in the Fas cascade, by a mechanism that involves modulation of loss of K ϩ , cell shrinkage and caspase activity. The process of cell shrinkage during Fas receptor-induced apoptosis appears to be closely regulated by PKC (Fig. 2); activation of PKC by PMA, bryostatin-1, or ConA inhibited cell shrinkage, and conversely, PKC inhibition enhanced the loss of cell volume. PKC stimulation did not inhibit Fas-induced cell shrinkage of all the cells. Although not addressed in the present study, we speculate that intrinsic differences within the cell population in their sensitivity or susceptibility to respond to PKC stimulation may exist, possibly due to different levels of PKC activity/expression among the FIG. 9. Cleavage of PKC isotypes in Jurkat cells during apoptosis and effect of PKC stimulation. Jurkat cells were preincubated with or without 20 nM PMA or 10 nM bryostatin-1 for 30 min. Cells were then treated with or without 50 ng/ml anti-Fas to induce apoptosis. After 3 h of exposure to anti-Fas antibody, cells were harvested and protein extracted to perform Western blot analyses of PKC isotypes ␦, ⑀, , and ␣. Twenty to 50 g of protein were equally loaded per lane. Cells treated with anti-Fas display PKC fragment of the correspondent isotype, except for ␣, which was used as a control since this isotype is known to remain uncleaved during Fas-induced apoptosis of Jurkat cells. Each Western analysis shown is representative of at least three separate experiments.
cells. The mechanism of inhibition of Fas-induced cell shrinkage by activation of TCR-CD3 receptor with ConA (shown in Fig. 5) might be mediated by PKC. Resistance of Fas-induced cell death via activation of TCR-CD3 receptor has been previously shown to occur (74) by a mechanism that involves at least MAPK (mitogen-activated protein kinase) signaling (75). Although the mechanism by which ConA attenuates Fas-induced cell death has not been further investigated in this study, the observed inhibitory effect of ConA on the Fas-induced apoptotic features (Fig. 5) raises the possibility that activation of PKC upon stimulation of the TCR-CD3 receptor complex might be counteracting apoptotic signals induced by the Fas receptor. This notion is supported by the observation that the protective effect of ConA on Fas-induced cell shrinkage was blocked by the PKC inhibitor Gö6976 (data not shown). The inactive analog of PMA, 4␣-phorbol 12,13-didecanoate, failed to induce any effect in the cells (data not shown). Regulation of the loss of cell volume plays a pivotal role in the Fas pathway since blocking cell shrinkage attenuates the cell death program (14). Therefore, our results suggest that stimulated PKC can control the process of cell death by blocking the incipient and required loss of cell volume.
Analyses of MMP (Fig. 3) and DNA fragmentation (Fig. 4) support the notion that PKC activity regulates later events in the Fas pathway as well. Although there is an extensive amount of information about the importance of mitochondrial changes during apoptosis (76 -78), the precise control of this process remains unclear. Our laboratory has recently demonstrated that loss of MMP occurs in shrunken apoptotic cells and that the shrinkage occurs prior to the loss of MMP (54). Our findings are consistent with this, since PKC can modulate both apoptotic cell shrinkage and the loss of MMP. In addition, since caspase-8 activation has been shown to mediate mitochondrial changes in the Fas pathway (79,80), it is plausible that PKC modulation of the MMP occurs through modulation of caspase-8. In this context, PMA seemed to cause slight mitochondrial depolarization (Fig. 3C and Table I), an effect that was not found using bryostatin-1 to stimulate PKC. Although we have not experimentally addressed this question, PMA has recently been shown to selectively target PKC ␦ isotype to mitochondria and induce depolarization of the mitochondrial membrane (81). Despite that both PMA and bryostatin-1 are specific activators of PKC, they are structurally different, and divergences in their mechanism of promoting PKC activation and intracellular translocation have been previously reported in leukemia cells (82) and in other cell types (i.e. Ref. 83).
The analyses of changes in the intracellular K ϩ levels shown in Fig. 5 provide evidence that the modulatory effect of PKC on cell shrinkage, and ultimately, on the cell death program, occurs through modulation of K ϩ movement across the plasma membrane. In particular, we show for the first time that extrusion of K ϩ (Fig. 7) is one of the components of K ϩ movement modulated by activation of the Fas receptor, and that PKC is enabled to regulate this efflux component. Supporting these results, high [K ϩ ] in the culture medium (Fig. 8) not only attenuated the Fas receptor-induced cell shrinkage and DNA fragmentation (14), but abrogated the potent enhancement of apoptosis that Gö6976 produces in the anti-Fas-treated cells. Since there is evidence that a decrease in intracellular [K ϩ ] during apoptosis leads to the necessary intracellular ionic environment for the activation of both caspases and nucleases (19), we propose that PKC may be capable to regulate the apoptotic pathway, including the proteolytic cascade, by controlling in a precise manner balance K ϩ movement during apoptosis. Therefore, our results on PKC-dependent modulation of K ϩ efflux are also consistent with our finding that caspase-8-and caspase-3-like activities are down-regulated in the presence of anti-Fas during PKC stimulation. The protective effect of PKC stimulation during Fas-induced apoptosis might be mediated by the phosphorylation of ion channels regulated via Ser/Thr kinases in lymphocytes (84). Although the voltage-dependent Kv1.3 channel has been shown to be inactivated during Fas-induced apoptosis by tyrosine phosphorylation of the protein (85), our results suggest that along with this mechanism, a specific channel(s)/ion-coupled transporter(s) extruding K ϩ is activated during Fas-induced death and can be regulated by PKC. Further research on ionic movement induced is being conducted in our laboratory to explore the specific pathways involved in the Fas-induced K ϩ efflux. An alternative mechanism for the protective role of PKC activation could be the existence of interference between PKC and Fas receptor signaling, which by extension would block Fas-dependent events. This proposed mechanism is supported by observations from our laboratory, which showed that Fas-induced cell shrinkage and loss of K ϩ were totally blocked by benzyloxycarbonyl-Val-Ala-Asp-fluoromethyl ketone and only partially by z-DEVD-fmk or z-IETD-fmk (54), suggesting that initial transducing activity of the Fas receptor is necessary for the initiation of the loss of K ϩ and cell shrinkage which probably occur upstream of caspase-8 activation. Thus, we propose that a mechanism of PKC-dependent modulation of the Fas pathway occurs between the process of activation of the Fas receptor and the generation of the active form of caspase-8. Although it remains to be investigated, we do not exclude a direct effect of PKC activation on the Fas receptor. The cytoplasmic domain of the Fas receptor is rich in serine and threonine residues, which can be phosphorylated (8,86), but the biological significance of this phosphorylation is unknown to date.
Recently, Scaffidi et al. (79) have investigated the role of PKC in the suppression of Fas-induced apoptosis in type II Jurkat cells and demonstrated that one inhibitory action of PKC resides in a process between DISC formation and mitochondrial depolarization (87). These data obtained using a different experimental approach than ours are consistent with our new findings, which indicate that PKC can block K ϩ efflux and cell shrinkage. Since both of these phenotypic responses (K ϩ efflux and cell shrinkage) are necessary for achievement of both mitochondrial depolarization and effector caspase activation (19,54), we propose that PKC can act to multiple steps to suppress apoptosis. To date, the protective effect of PKC on Fas-induced cell death has been suggested to occur through two mechanisms. In a first study, Scaffidi et al. (87) showed that PKC acts upstream of caspase-8-dependent cleavage of BID. In a second study, Ruiz-Ruiz and colleagues (88) demonstrated that PKC controls clustering of Fas receptor in the plasma membrane, leading to the speculation that PKC may induce alterations in the cell morphology and cytoskeletal network that presumably can block the activation of the Fas cascade. These observations are both consistent with our results that show that PKC-dependent modulation of the Fas pathway seems to occur prior to caspase-8 activation. In addition, our results suggest that a third mechanism may operate in this network where PKC can target activation or repression of ion channels and inhibit recruitment/activation of caspase-8 to the Fas receptor complex and/or block the initial loss of intracellular K ϩ and cell shrinkage. Both of these effects could interfere with cytoskeletal reorganization necessary for the Fas receptor to cluster and signal death. The mechanisms that couple the activation of the Fas receptor or DISC formation to the loss of K ϩ and cell shrinkage remain unknown, however.
Interestingly, Eriksson and co-workers (89) have proposed that MAPK-dependent signaling operates in the resistance to Fas receptor-mediated cell death through a mechanism occurring upstream of caspase-8. Since MAPK signaling can be directly modulated by PKC, we do not discard the existence of other signaling intermediates, including MAPK, between the Fas receptor activation process and PKC action in the inhibition of apoptosis. In addition, activation of phosphatidylinositol 3Ј-kinase has been proposed as a mechanism of resistance to Fas-induced apoptosis through inhibition of caspase-8 cleavage (90).
Although other studies demonstrate a post-caspase participation of PKC during apoptosis (2,42,45,46,52), our results suggest that there is also an earlier regulatory mechanism exerted by PKC upstream of caspases in the Fas receptormediated death machinery. This is the first report showing caspase-8 and -3 activities under PKC control in Fas receptormediated apoptosis, with a consequent inhibition of proteolytic cleavage of PKC isotypes. Our data are not, however, inconsistent with the possibility that PKC might be regulating only caspase-8 and exerting a consequent inhibitory effect in downstream caspases. A recent report describes a PKC-dependent inhibition of caspase-3 activity in apoptosis induced by cis-diamminedichloroplatinum(II) in HeLa cells (91), but no mechanism is proposed in the study. Moreover, it should be mentioned that there are divergences between the anti and proapoptotic effects of PKC shown in the literature that seem to be dependent on cell type, apoptotic stimulus, and the PKC isotype considered. This heterogeneity could be explained on the bases of different constitutive expression patterns of individual PKC isotypes depending on cell type, state of cellular differentiation, and subcellular location of isotypes (30). Although not explored in this study, the resistance to Fas receptor-mediated cell death mediated by PKC in Jurkat T cells might be the result of a differential involvement and modulation of PKC isotypes constitutively expressed in T cells. However, an important issue to be considered is that PKC modulation can turn cells under apoptotic insults to death or to life, which may have critical implications in the in vivo regulation of apoptosis.