Protein Kinase C Regulates Fas (CD95/APO-1) Expression¹

Activation of the TCR leads to the expression of a series of genes related to either apoptosis or cell proliferation (1, 2). Whether cell proliferation, differentiation, or apoptosis is favored depends upon the cellular stage and the extracellular milieu (2, 3, 4). The activation of immature T cells during thymic development induces apoptosis (5, 6, 7). Although TCR stimulation of resting peripheral T cells usually leads to proliferation and clonal expansion, primary activated mature T cells (8, 9, 10) and peripheral mature T cells from HIV-infected patients (11) are also sensitive to activation-induced cell death (AICD).⁴ Thus, AICD in T cells is a fundamental mechanism for maintaining immune tolerance and cellular homeostasis during T cell development, immune responses, and disease (12, 13).

Similar to immature thymic lymphocytes and activated mature T cells, T cell hybridomas also undergo AICD following TCR cross-linking (14, 15, 16, 17). Recent studies have demonstrated that AICD in T cell hybridomas is dependent upon the induction of Fas (CD95) and Fas ligand (FasL/CD95 ligand) (18, 19, 20). Importantly, competitive inhibition of the interaction between Fas and FasL with soluble Fas proteins effectively inhibited AICD (19). Therefore, Fas and FasL play a fundamental role in AICD in T cells and T cell hybridomas.

The activation of Fas through its interaction with FasL or with specific anti-Fas Abs triggers the activation of a cascade of proteases; these proteases, in turn, execute the apoptotic process (21). Interestingly, the apoptosis signals after Fas ligation are independent of macromolecular synthesis (22). Therefore, the expression of Fas is the key checkpoint for cells to commit to AICD. In many cellular systems, the expression of Fas is strictly controlled (23). An overexpression of Fas rendered cells highly sensitive to FasL-mediated killing (24, 25). However, the regulation of Fas expression is poorly understood. Thus, elucidation of the regulatory mechanisms for Fas expression could provide important information for a better understanding of the molecular mechanisms of apoptosis.

Ligation of the TCR by specific Ag peptide presented by the MHC on APCs initiates a biochemical cascade that involves the activation of protein tyrosine kinases and protein tyrosine phosphatases (1, 2). These enzymes change the phosphorylation status of a number of intracellular substrates participating in the signal transduction cascade, which leads to the activation of protein kinase C (PKC) and to Ca²⁺ mobilization (13). Since PKC activation and Ca²⁺ redistribution have been shown to be critical in TCR-mediated activation signals, we examined the role of these signals in Fas and FasL expression during AICD. An increase in cytosolic Ca²⁺ has been shown to be fundamental in the regulation of FasL expression, since the activation-induced expression of FasL could be completely inhibited by cyclosporin A (CsA) and FK506 (26). Nevertheless, the role of PKC in the regulation of Fas and FasL expression is not clear. We report that the activation of PKC without intracellular Ca²⁺ mobilization is sufficient to induce Fas expression. Interestingly, FasL expression requires both PKC activation and intracellular Ca²⁺ redistribution.

Murine T cell hybridoma (A1.1) cells (27), B cell lymphoma CH31 (a gift of Dr. David Scott, Holland Laboratory of the American Red Cross), T cell hybridoma KCIT and its mutants (kindly provided by Dr. Yongwon Choi, Rockefeller University, New York, NY), and human Jurkat cells were maintained in RPMI 1640 medium (Life Technologies, Gaithersburg, MD) that was supplemented with 2 mM l-glutamine, 50 mM 2-ME, 10% heat-inactivated FBS (Sigma, St. Louis, MO), and 10 mM gentamicin. In addition, rIL-2 was added to the CTLL2 culture at a concentration of 100 U/ml. Cultures were incubated at 37°C in humidified atmosphere with 5% CO₂. Ab to CD3 was produced by a hamster B cell hybridoma, 145-2C11ε (obtained from Dr. Jeffery Bluestone, University of Chicago, Chicago, IL). PMA, 1-oleoyl-2-acetyl-sn-glycerol (OAG), N-(2-guanidinoethyl)-5-isoquinolinesulfonamide (HA 1004), 1-(5-isoquinolinesulfonyl)-2-methylpiperazine (H7), Gö6976, bisindolylmaleimide I, calphostin C, and ionomycin were purchased from Calbiochem (San Diego, CA). An inhibitor of intracellular Ca²⁺ mobilization, 8-(diethylamino)-octyl-3,4,5-trimethoxybenzoate hydrochloride (TMB-8), was obtained from Sigma. A PKC isoform-typing kit was obtained from Transduction Laboratory (Lexington, KY). The other chemicals used were of the purest grade available from Sigma.

A DNA fragmentation assay was conducted essentially as described previously (15). Briefly, 5 × 10⁵ cells were suspended in 30 μl of Ca²⁺-free PBS and lysed by the addition of 30 μl of lysis buffer (80 mM EDTA, 200 mM Tris (pH 8.0), 1.6% sodium lauryl sarcosinate, and 1 mg/ml proteinase K). Cell lysates were incubated at 50°C for 4 h followed by the addition of RNase A to a final concentration of 0.2 mg/ml. After incubation at 37°C for another 30 min, the resulting DNA samples were fractionated by agarose gel electrophoresis and visualized under UV light after staining with ethidium bromide.

Genomic DNA integrity was also analyzed by determining the DNA content in the nuclei by flow cytometry as described previously (28). Cells were fixed with 70% ethanol for 30 min at 4°C, followed by two washes with PBS. The fixed cells were then incubated in PBS containing propidium iodide (Sigma) at 50 μg/ml and RNase (Boehringer Mannheim, Indianapolis, IN) at 0.1 mg/ml at room temperature for 30 min. DNA content was determined by flow cytometry on a FACScan (Becton Dickinson, San Jose, CA). The FL2 intensity was plotted as histograms on a linear scale.

The TCR activation-induced production of IL-2 in culture supernatants was measured using a Cytoscreen Immunoassay Kit obtained from BioSource (Camarillo, CA). The murine rIL-2 included in the kit was diluted to generate a standard curve. An ELISA was performed according to the manufacturer’s instructions. Briefly, culture supernatants were diluted in the standard diluent buffer included in the kit. Standard dilutions and samples were incubated on the first anti-IL-2 Ab-coated microtiter plates for 1.5 h at 37°C. After washing with the washing buffer (PBS plus 0.2% Tween 20), the bound IL-2 was detected with a biotinylated second anti-IL-2 Ab and streptavidin-peroxidase. The amount of IL-2 was determined by the addition of tetramethyl benzidine.

Activation-induced Fas expression on murine T cell hybridomas was assessed by determining the sensitivity to killing by anti-Fas Ab (JO2) or by L cells expressing sense FasL but not antisense FasL (kindly provided by Dr. T. A. Ferguson, Washington University School of Medicine, St. Louis, MO). Briefly, cells were incubated with JO2 or L cells after appropriate treatments. Apoptosis was determined by flow cytometric DNA content analysis as described above.

Cell viability was measured by MTT conversion (29). Briefly, cells were incubated in complete RPMI 1640 medium in 96-well plates. After treatments as indicated, 10 μl of MTT (5 mg/ml in H₂O) was introduced, and cells were incubated at 37°C for 4 h followed by addition of 100 μl of acid-isopropanol (0.04 N HCl). After the reduced MTT crystals were dissolved, the plate was immediately scanned by an ELISA reader with a 595-nM filter.

Total RNA was isolated with affinity columns (Qiagen, Chatsworth, CA) according to the protocol recommended by the manufacturer. RNA samples were fractionated on 1% agarose/2.2 M formaldehyde denaturing gel and transferred onto a Nytran membrane (Schleicher and Schuell, Keene, NH). The DNA probes (mouse Fas and FasL were provided by Dr. Shigekazu Nagata, Osaka Bioscience Institute, Osaka, Japan; TDAG51 was obtained from Dr. Yongwon Choi) were labeled by random priming (Boehringer Mannheim) according to the manufacturer’s instructions. Prehybridization and hybridization were conducted at 42°C in a solution containing 5× SSC (10× SSC is 1.5 M NaCl, and 0.15 M sodium citrate), 2.5 mM EDTA, 0.1% SDS, 5× Denhardt’s solution, 2 mM sodium pyrophosphate, 50 mM sodium phosphate, and 50% formamide. After washing with 0.2× SSC and 0.1% SDS at 56°C for 1 h, hybridization signals were detected by autoradiography.

Murine T cell hybridoma A1.1 cells commit to AICD in a Fas- and FasL-dependent manner (15, 19). AICD occurs in a period of 8 h, with a sequential induction of Fas at 2 h and FasL at 4 h following TCR cross-linking (19). Ligation of the TCR initiates a signal transduction cascade in which Ca²⁺ influx and the activation of PKC have been shown to be the critical points in activation-induced cytokine production and proliferation in T cells. Thus, the combination of PMA and Ca²⁺ ionophore mimics the signals from the TCR (13). Here, we examined the role of PKC in the regulation of Fas and FasL expression during AICD.

We initially tested the requirement of PKC activation and Ca²⁺ influx for AICD in T cell hybridomas. As shown in Figure 1, treating A1.1 cells with the Ca²⁺ ionophere, ionomycin, plus either PMA or OAG readily induced apoptosis as detected by genomic DNA fragmentation (Fig. 1,A) and MTT viability assay (Fig. 1 B). When applied individually, neither ionomycin, PMA, nor OAG was sufficient to induce apoptosis. Similar to the activation of primary T cells (30), both PKC activation and Ca²⁺ influx are required for AICD in the A1.1 hybridoma.

To further investigate the role of PKC in the pathway leading to AICD, we examined the effect of several relatively specific PKC inhibitors (31). We found that these inhibitors could effectively block TCR ligation-induced AICD in A1.1 hybridomas as demonstrated by DNA content analysis (Fig. 1,C) and genomic DNA oligonucleosomal fragmentation (Fig. 1,D). Furthermore, we have performed PKC isoform typing and found that α, γ, δ, ε, ι, and μ were present in A1.1 cells, while isoforms β, ζ, and θ, were not found (Fig. 1 E). Thus, several PKC isoforms are present in T cell hybridoma A1.1 cells and are required for activation-induced apoptosis.

The activation-induced expression of Fas and FasL is absolutely required for AICD in T cell hybridomas, since soluble Fas protein could completely block AICD (19). Unactivated A1.1 cells express a low level of Fas, which are insensitive to anti-Fas Ab- or FasL-expressing fibroblast-induced apoptosis (Fig. 2). Interestingly, although the application of PMA or OAG alone could not induce apoptosis in A1.1 cells (Fig. 1), these treatments primed A1.1 to undergo increased Fas ligation-induced apoptosis as detected by DNA content analysis (Fig. 2). Specifically, A1.1 cells were treated with PMA (Fig. 2,A) or OAG (Fig. 2,B) for 2 h and then incubated with L cells that had been transfected with either sense FasL or antisense FasL (Fig. 2,A) or with or without anti-Fas (J02, Fig. 2 B) for 12 h. PKC activation with PMA or OAG sensitized A1.1 cells to Fas ligation-induced apoptosis as assessed by DNA content analysis. There are two possible explanations to account for this result, i.e., PKC activation either increases Fas expression or sensitizes the T cell hybridoma cells to Fas-mediated death signals.

We have reported that AICD in T cell hybridomas requires de novo synthesis of macromolecules, since apoptosis could be suppressed by the inhibition of either protein or RNA synthesis (15). However, recent studies have shown that the apoptosis triggered by anti-Fas Ab is independent of macromolecular synthesis once the Fas receptor is expressed (22). Since treatment with phorbol esters sensitized cells to anti-Fas killing (Fig. 2), the effect of PKC activation might induce Fas expression.

Therefore, we characterized the phorbol ester induction of Fas expression. A1.1 cells were treated with PMA in the presence or absence of ionomycin for 3 h, and RNA was isolated. Electrophoretically fractionated RNA was examined for Fas and FasL expression by Northern blot hybridization. As shown in Figure 3,A, the expression of Fas mRNA could be induced by phorbol ester alone, whereas the expression of FasL required both the activation of PKC and Ca²⁺ mobilization. To further confirm the role of PKC activation in Fas expression, we treated A1.1 cells with different concentrations of PMA (0.1 ng-1000 ng/ml). The maximal induction of Fas expression was observed at 5 ng/ml. The expression level was comparable with the level induced by CD3 ligation (Fig. 3 B). On the other hand, when the same blots were examined for the expression of FasL, we found that PMA alone does not have any effect on the induction of FasL. PMA plus ionomycin or anti-CD3, however, induced the expression of FasL. Experiments with OAG, which is a membrane permeable analogue of 1,2-diacylglycerol (32), showed results that were similar to those obtained with PMA (data not shown).

Next, we isolated RNA from A1.1 cells that had been treated with PMA for different times; we found that the PMA-induced Fas expression was detectable at 2 h posttreatment. This increase was sustained for at least 12 h (data not shown), while the Fas expression that was induced by anti-CD3 diminished at 5 to 6 h postactivation. This sustained expression of Fas induced by PMA is compatible with the extended effect of PMA on PKC activation (33, 34).

To examine whether the activation of PKC alone could induce Fas in cell types outside the T cell lineage, we treated various cell lines with PMA for 3 h and analyzed Fas expression by Northern blot hybridization. We found that PMA could induce Fas expression in a murine B cell lymphoma, CH31, in a human breast cancer cell line, MDA231, in a human ovarian cancer cell line, Hey, in a human cervical cancer, HeLa, and in a human prostate cancer, DU145 (data not shown). Since it has recently been shown that treating Jurkat cells with PMA increases their resistance to Fas ligation-induced apoptosis (34), we have tested whether PMA also induces the expression of Fas in these cells. As shown in Figure 3 C, we found indeed that Jurkat cells were also induced to increase the expression of Fas. Therefore, the effect of PMA on the sensitivity of Jurkat cells to Fas-mediated apoptosis is not exerted by regulating Fas expression; rather the effect of PMA is exerted by interfering Fas signals. Nevertheless, our results in A1.1 cells are consistent with the recent observation of Wong et al. (35), who have shown that PKC treatment increased the sensitivity of T cell hybridomas to Fas-mediated apoptosis (the double-band pattern of human Fas has been shown previously). Thus, PKC plays a general role in regulating Fas expression.

Although it has been well-established that the activation of T cells through the TCR activates both PKC and Ca²⁺ influx, our studies on the effect of activation of PKC with PMA do not necessarily establish a physiologic role for PKC during TCR signaling. To confirm the role of PKC in activation-induced Fas expression through the TCR, we treated A1.1 cells with anti-CD3 in the presence or absence of the PKC inhibitor, H7, or its analogue, HA1004, for 3 h; total RNA was isolated. H-7 has a greater ability to inhibit PKC than cyclic nucleotide-dependent kinase. Conversely, HA1004 has a greater ability to inhibit cyclic nucleotide-dependent kinase than PKC (34). As shown in Figure 5, Northern blot analysis revealed that H7 completely blocked activation-induced Fas expression, while HA1004 had no effect on TCR activation-induced Fas expression in A1.1 cells. Similar results were also obtained with the relatively specific PKC inhibitors calphostin C and Gö6983 (data not shown).

We have previously reported that blocking Ca²⁺-dependent calcineurin with CsA completely inhibits AICD in T cell hybridomas. As reported recently (26), cyclosporin could inhibit activation-induced FasL but not Fas expression (data not shown). However, cyclosporin has no effect on the Fas expression induced by TCR ligation or PMA treatment (data not shown). Thus, calcineurin and downstream events, while important for FasL expression, are not required for Fas expression. Since cyclosporin does not have any effect on cytosolic Ca²⁺ levels and thus does not interfere with the interaction between PKC and Ca²⁺, we tested the effect of an inhibitor of intracellular Ca²⁺ redistribution, TMB-8 (36), on TCR cross-linking-induced AICD. We found that TMB-8 completely blocked AICD (Fig. 4,A) and FasL expression. However, TMB-8 did not inhibit the activation-induced expression of Fas (Fig. 4 B). Thus, activation-induced Fas expression is independent of the redistribution of intracellular Ca²⁺.

PKC signaling in T cells is a complex process involving multiple enzyme cascades. Although our data clearly demonstrated a vital role for PKC in AICD, the PKC downstream effectors have not yet been elucidated. It has recently been reported that TDAG51, a newly identified potential transcription factor, is required for activation-induced Fas expression in T cell hybridomas (17). Thus, we determined the relationship between PKC and TDAG51 in our system. When the expression of TDAG51 was analyzed in A1.1 cells, we found that this gene is not expressed in unactivated A1.1 cells. However, its expression is induced upon activation through the TCR or by PMA alone. Similar to Fas, the expression of TDAG51 also has a requirement for PKC (Fig. 5,A). In addition, we examined the expression of Fas in another T cell hybridoma, KCIT, and its mutant bearing the mutation of TDAG51. We found that PMA induced Fas expression in KCIT cells, but not in the TDAG51 mutant. Interestingly, transfecting the mutant with wild-type TGAG51 restored its response to PMA-induced Fas expression (Fig. 5 B). Therefore, TDAG51 is a downstream effector in the PKC-mediated pathway leading to the induction of Fas expression.

It is clear that the expression of Fas and FasL is responsible for apoptosis in some cell types, not only for the elimination of activated T cells after mounting a proper immune response (12, 37) but also for the maintenance of immune-privileged sites (38, 39). AICD in T cells requires de novo macromolecular synthesis, including the up-regulation of Fas and FasL, whose interaction then activates the apoptosis program (19, 20, 21). Since it has been shown that Fas ligation-induced apoptosis is independent of macromolecule synthesis (22), the activation of Fas and FasL genes is a critical step in initiating AICD. It has been well-established in T cell hybridomas that the expression of Fas and FasL increases after activation (19, 17). Furthermore, blocking the interaction of Fas and FasL with Fas fusion proteins could completely block AICD (19). Alternatively, T cells can be activated by calcium ionophere and phorbol ester. This combination implies that both calcium redistribution and the translocation of PKC are required for T cell activation (30). Since the signals leading to the expression of Fas still remain largely unknown, we investigated the role of Ca²⁺ and PKC in the regulation of Fas expression. Our data show that Fas expression is solely dependent upon PKC activation and does not require Ca²⁺ redistribution. This effect of PKC is exerted through TDAG51. PKC-mediated Fas expression was also observed in the murine B cell lymphoma CH31 and in several human cancer cell lines. Therefore, PKC activation may be a critical component in the regulation of Fas expression.

Recent studies have demonstrated that FasL expression could be inhibited by CsA (26), retinoic acid (40), antioxidants (41), and dexamethasone (42). In accordance with our previous results on the requirement of c-myc during AICD, we have recently shown that the inhibition of c-myc expression also inhibits FasL expression (43). Interestingly, none of the FasL modulatory reagents had any effect on the expression of Fas, indicating that the expression of Fas and FasL are regulated by distinct pathways.

We activated A1.1 cells with anti-CD3 Ab-coated plastic in calcium-free medium supplemented with different concentrations of calcium and assayed DNA fragmentation at 12 h poststimulation. We found that the extent of DNA fragmentation was dependent upon the calcium concentration in the medium (data not shown). Similar results were obtained by supplementing with different concentrations of EGTA. Thus, these experiments clearly demonstrated that extracellular Ca²⁺ is essential for AICD in the T cell hybridoma. It has been shown that the blocking of the Ca²⁺ signal by CsA could completely inhibit FasL, but not Fas, expression. However, it is not known whether Ca²⁺ alone is sufficient in FasL induction. We have found that blocking intracellular Ca²⁺ with TMB-8 completely inhibited FasL expression but did not affect Fas. We have performed experiments in which A1.1 cells were treated with 2000 nM of PMA for 12 h. We found that this treatment blocked activation-induced apoptosis in these cells (data not shown). Although this treatment blocked activation-induced FasL expression, it did not show effects on Fas expression. Since the Ca²⁺ PKC isoforms are more sensitive to PMA depletion, we conclude that activation-induced Fas expression does not require Ca²⁺ based on these findings and the above data combined with our observation that PMA alone could induce maximal Fas expression. Thus, our data strongly suggest that FasL expression requires both PKC and Ca²⁺.

We have shown that the diacylglycerol analogue OAG, which stimulates PKC, induces apoptosis in A1.1 cells when added together with ionomycin. Further, we have shown that the inhibition of PKC could prevent activation and inhibit anti-CD3-induced Fas expression and apoptosis in the T cell hybridoma. This observation indicates that PKC is involved in activation-induced Fas expression. When we down-regulated PKC function by preincubating cells with high concentrations of PMA, we were able to show again that activation-induced Fas expression required PKC (data not shown).

TDAG51 is a newly identified potential transcription factor that is involved in the regulation of activation-induced Fas expression in the T hybridoma KMIs-8.3.5. This hybridoma undergoes Fas/FasL-dependent apoptosis upon activation. A variant of KMIs-8.3.5 bearing a mutation in TDAG51 expresses FasL, TNFR-1, and IL-2, but not Fas (17). Transfection of the active gene restored activation-induced Fas expression in this mutant. Thus, TDAG51 may play an essential role in the induction of apoptosis by coupling TCR stimulation to Fas expression. Our data show that TDAG51 expression could also be activated solely by PKC; it is very likely that the effect of TDAG51 acts downstream of PKC. How TDAG51 regulates Fas expression is currently not known. The transfection of TDAG51 does not appear to induce Fas expression (17). Thus, it is possible that other factors are required. Studies are in progress to determine how PKC regulates TDAG51 and Fas expression.

We thank Drs. David W. Scott, Wendy Davidson, Achsah D. Keegan, and Juliann G. Kiang for critical discussions. This is publication number 46 of the Department of Immunology, Holland Laboratory.