Selective Up-Regulation of Phosphatidylinositol 3′-Kinase Activity in Th2 Cells Inhibits Caspase-8 Cleavage at the Death-Inducing Complex: A Mechanism for Th2 Resistance from Fas-Mediated Apoptosis¹

Mature CD4⁺ T cells can be separated into Th1- and Th2-type effector cells based on the pattern of cytokines they produce upon stimulation (1, 2, 3). The Th1 subset produces IFN-γ and IL-2 and induces a cell-mediated immune response to Ag, whereas the Th2 subset produces IL-4, IL-5, and IL-10 and is necessary for generating a humoral immune response against Ag. The biological relevance of these two subsets in infection is underscored by the ability of Th1 and Th2 cells to influence disease outcome (4). For example, protection against most intracellular pathogens requires cell-mediated immune responses generated by Th1 cells, whereas biasing of the host immune response toward a Th2 type promotes humoral immunity, resulting in disease.

Although the impact on disease of a restricted cytokine production pattern is well established, a clear understanding of the mechanisms leading to dominance of a particular T cell subset in disease is just emerging (5). Understanding both the factors that modulate the decision of a naive CD4⁺ T cell to develop into a Th1 or Th2 cell in response to an Ag and the mechanisms regulating responsiveness to Ag of differentiated subsets may lead to rational approaches for generating appropriate immune responses to a pathogen. Ag dose (6), APCs (7, 8) with their associated costimulatory signals (9, 10) and cytokines (11, 12) all have varying effects on the differentiation of T cells and the final outcome of a disease state. Emerging evidence suggests that the Fas/Fas ligand (FasL)³ apoptotic pathway, required for maintenance of T cell homeostasis and self tolerance (13), in addition, may also contribute to certain disease states through preferential induction of Th1 cell apoptosis (14, 15, 16).

Fas (APO-1, CD95) receptor is a cell death receptor that is activated by FasL, a trimeric transmembrane protein (17). Receptor oligomerization results in the recruitment of the death domain protein FADD to the FasR (18, 19, 20, 21, 22), followed by the IL-1β-converting enzyme (ICE)-like protease, caspase-8, to form a functional death-inducing signaling complex (DISC) (23). Upon triggering of the FasR, caspase-8 bound to the DISC gets proteolytically cleaved to generate two active fragments (24). The active fragments are released into the cytosol where they initiate activation of a cascade of ICE-like proteases (25) and subsequent death. Following recruitment to the DISC, active caspase-8 that is generated can directly mediate activation of caspase-3 and other downstream caspases, leading to apoptosis (26). Alternatively, in some cell types, active caspase-8 cleaves BH3-interacting domain death agonist (Bid), which then acts on mitochondria to release cytochrome c. Cytochrome c, in association with Apaf-1 and procaspase-9, activates caspase-3 and other downstream caspases (26, 27).

There are two other pathways that may also be engaged following FasR ligation. Recruitment of an adaptor molecule Daxx to the receptor activates a c-Jun amino-terminal kinase (JNK) pathway that leads to gene transcription and subsequent death (28, 29). FasR ligation also catalyzes the hydrolysis of membrane sphingomyelin, which results in a rapid increase in endogenous levels of ceramides (30, 31, 32), which in turn can either directly activate caspases or intersect with a ras-dependent death pathway. Recent studies utilizing FADD (33), caspase-8, (34), caspase-9 (35, 36), and acid sphingomyelinase-knockout mice (37), however, indicate that the FADD/caspase-8 signaling pathway is central and probably nonredundant in inducing death by the FasR. Therefore, the role of other pathways in FasR signaling, besides FADD/caspase-8, remain debatable.

Previously, we demonstrated that mature Th1 and Th2 cells can be differentially regulated to undergo Fas-mediated apoptosis upon CD3 ligation (14), implying that this form of apoptosis may play a significant role in suppressing Th1 cells. We showed that activated Th1 cells are susceptible to Fas-mediated apoptosis upon CD3 ligation. In contrast, prior activation through CD3 of Th2 cells induced resistance to Fas-mediated apoptosis. However, direct ligation of the FasR induced apoptosis of both Th1- and Th2-type cells. Thus, differences in sensitivity to apoptosis of Th1 and Th2 cells is not due to an intrinsic defect in the Fas signal transduction pathway in Th2 cells, as reported by others (16, 38, 39), but due to interception of this pathway from signals generated through CD3.

In view of the role that apoptosis may play in the loss of Th1-dependent effector functions in disease, it becomes necessary to establish the molecular mechanisms controlling T cell subset-specific apoptosis. In this work, we therefore investigate the mechanism of differential susceptibility to apoptosis of Th1 and Th2 cells and examine CD3-generated signals that regulate this apoptotic pathway. We report that the difference in Fas sensitivity of CD3-activated Th1 and Th2 cells is not at the level of recruitment of caspase-8 to the DISC, but at the level of caspase-8 cleavage to its active fragments. We identify phosphatidylinositol 3′-kinase (PI3′-K) as the key signaling molecule generated through CD3 engagement on Th2 cells that regulates the resistance to Fas-induced apoptosis by preventing caspase-8 cleavage to its active fragments. These findings describe a novel mechanism that T cells utilize to resist FasR-induced apoptosis. It also provides the first evidence that PI3′-K interferes with caspase-8 activation.

Tetanus-toxoid and purified protein derivative-reactive clones were established by limiting dilution technique, as previously described (3). Cytokine profile was determined by stimulating cells with plate-bound anti-CD3 Abs and by testing 24-h supernatants for the presence of TNF-α, IFN-γ, and IL-4 by ELISA. To expand cell numbers, clones were maintained in IL-2, with biweekly stimulation of Ag and autologous APCs. Cytokine profile of the clones was routinely reconfirmed after several rounds of stimulation. We have observed that, even after several passages in vitro, the cytokine profile and Ag reactivity are unchanged. As previously described, Th1 and Th0 type 1 clones undergo apoptosis when anti-CD3 stimulated; for sake of clarity, herein Th1 and Th0 type 1 clones will be referred to as apoptosis-susceptible clones. Conversely, Th2 and Th0 type 2 clones will be referred to as apoptosis-resistant clones. A total of nine individual T cell clones of both the apoptosis-susceptible (CG2, JC2, HG17, and PS13) and -resistant (HG3, HG5, HG6, HG10, and SPU) phenotype were included in this study.

Anti-CD3 mAb was obtained from Biosource International (Camarillo, CA). The anti-APO-1 mAb (IgG3,κ) is an agonistic Ab recognizing an extracellular epitope of CD95 (40). The three anti-caspase-8 Abs utilized recognize the three functional domains of caspase-8. N2 recognizes an epitope adjoining the tandem death effector domain containing pro-domain. C15 detects the catalytically active p18 fragment, and C5 binds to an epitope within the p10 fragment (41). The NF6 anti-cellular FLICE-inhibiting protein (c-FLIP) mAb (IgG1) recognizes the N-terminal part of human c-FLIP (42). IETD-fmk is a cell permeable caspase-8-specific inhibitor (Clontech, Palo Alto, CA) and was used at 25 μM. Wild-type (WT) and dominant-negative phosphatidylinositol 3′-kinase p85 regulatory subunit containing plasmids (43) were provided by Tomas Mustlin (La Jolla Institute of Allergy and Immunology, San Diego, CA). Successfully transfected T cells were selected on the basis of their expression of a murine MHC class I molecule encoded by the cotransfected plasmid, pMACS Kk (Miltenyi Biotec, Gladbach, Germany). Wortmannin was purchased from Sigma (St. Louis, MO).

Apoptosis of cells was determined using a sandwich ELISA (44). The ELISA relies on the detection of cytoplasmic nucleosomal fragments released from the nucleus during the apoptotic process. To obtain cytoplasmic nucleosomes, cells were lysed in ice cold buffer (1% Nonidet P-40, 20 mM EDTA, 50 mM Tris-HCl (pH 7.5)), for 0.5 h, in a volume sufficient for a final concentration of 1.25 × 10⁵ cell equivalents/ml. Nuclei were pelleted by centrifugation at 13,000 × g for 10 min, following which the supernatant was carefully removed and used for analysis. Lysates from 5 × 10³ cell equivalents were added to each well, with each sample tested in triplicate. Nucleosomal fragments were captured with an anti-histone mAb (LG11-2A) and detected with a biotinylated secondary mAb (PL2-3) that recognizes a DNA-histone complex. Bound Ab/Ag complexes were then reacted with alkaline phosphatase-conjugated streptavidin (Southern Biotechnology, Birmingham, AL), to which substrate was then added. Color development was measured at 405 nm. The percentage of apoptosis is expressed in arbitrary units, calculated as follows using cells maintained in IL-2 as control: 100 − [(absorbance values at OD₄₀₅ of IL-2 culture/absorbance value at OD₄₀₅ of treated cells) × 100].

Association of caspase-8 with the DISC of unactivated and CD3-activated apoptosis-resistant and -susceptible clones was determined after directly ligating FasR. T cells (10⁷) of the resistant and susceptible clone were either treated with CD3 Abs for 30 min or left untreated, following which the cells received 500 ng/ml anti-APO-1 Abs for 5 min at 37°C. Control cells were stimulated with anti-CD3 Abs alone; however, the lysates of these cells received anti-APO-1 Abs. Cells were harvested, washed in cold buffer, pelleted, and lysed in ice cold lysis buffer (30 mM Tris-HCl (pH 7.5), 150 mM NaCl, 1% Nonidet P-40, 1 mM PMSF, and small peptide inhibitors, as described in Ref. 45 , for 30 min. Cell debris was pelleted by centrifugation for 10 min at 13,000 × g. To the lysates, 50 μl of rabbit anti-mouse (Cappel, Cochranville, PA)-coated staphylococcus A was added. DISCs were allowed to immunoprecipitate in the cold for at least 1 h, then were washed in cold TNE buffer. After washing 6×, immune complexes and staphylococcus A were separated by 15 min incubation in 2-ME containing sample buffer, staphylococcus A was spun out, and the supernatants containing the immunoprecipitates were boiled. For Western blotting, lysates were separated on 12% SDS-PAGE, transferred to supported nitrocellulose membrane (Bio-Rad, Hercules, CA), and blocked with nonfat milk-containing TTBS. Blots were incubated with primary Ab diluted in TTBS overnight at 4°C, washed 3 to 4 times for 10 min each with TTBS, then detected with HRP-conjugated secondary Ab diluted 1:3000 in TTBS + 5% nonfat milk, and developed using the enhanced chemiluminescence method (ECL Plus, Amersham, Arlington Heights, IL) following the manufacturer’s protocol.

Cells (10 × 10⁶) of an apoptosis-resistant clone (SPU) were suspended in 250 μl serum free RPMI 1640 (Mediatech, Herndon, VA), and transferred to an electroporation cuvette specific for eukaryotic cells (Bio-Rad Gene Pulser Cuvette, 0.4-cm electrode, gap 50). To the cuvette, 20 μg total plasmid DNA was added, with the p85WT or p85 MUT DNA in a 1:1 ratio with the cotransfected pMAX plasmid, and allowed to incubate for 10 min at room temperature. Electroporation was conducted at 250 V, 960 μF. Immediately after electroporation, cells were placed at 37°C for 10 min, then cultured in IMDM (Life Technologies, Grand Island, NY) supplemented with 10% AB serum (BioReclamation, Long Island, NY) and 20 U/ml IL-2 (National Cancer Institute, Frederick, MD). Following 72 h incubation, dead cells were removed by separation using Ficoll-Paque (Pharmacia, Uppsala, Sweden) density gradient centrifugation, with the remaining cells resuspended in PBS + 0.5% BSA. To magnetically label the suspension, MACSelect Kk microbeads were added and allowed to bind the T cells at 4°C for 15 min, following which volume was adjusted with PBS, and the cells were applied to the magnetic selection column. Cells not expressing the murine class I molecule were not retained in the column, while transfectants were eluted and collected by removing the column from the magnetic field and flushing the column with additional buffer. Recovery of cells after positive selection was 2% of the input number.

The activity of immunoprecipitated PI3′-K was determined by the phosphorylation of the substrate phosphatidylinositol to phosphatidylinositol 3-phosphate with and without a 10-min anti-CD3 stimulation. Lipid substrate was chloroform extracted from the kinase reaction buffer, then separated by TLC and analyzed by autoradiography. Quantitative assessment of kinase activity was obtained using phosphorimaging (30).

Our previous data demonstrated that, upon CD3 stimulation, Th1 cells undergo Fas-mediated apoptosis. In contrast, Th2 cells do not undergo apoptosis upon CD3 stimulation (14). We also showed that Th0 cells consist of two subsets, one that behaves like Th1 cells and undergoes Fas-mediated apoptosis following CD3 stimulation (Th0 type 1), and the other subset is similar to Th2 cells and is resistant to apoptosis (Th0 type 2). Paradoxically, all clones were susceptible to apoptosis when Fas receptor was directly ligated by anti-Fas Abs. However, prior activation of Th2 and Th0 type 2 clones by mAbs to CD3/TCR complex induced resistance against Fas-mediated apoptosis (14). In Table I, we have listed the sensitivity to both CD3 and direct FasR ligation-mediated apoptosis of the clones utilized in the present study. As shown in Table I, and as previously reported (14), clones susceptible to CD3-mediated apoptosis are Th1 and Th0 type 1, and clones resistant to CD3 stimulation are Th2 and Th0 type 2. Studies described in this article are designed to provide a mechanism for our previous observation that anti-CD3 rapidly protects Th2 cells from death caused by direct FasR ligation with agonistic Abs (14).

To determine whether the FADD/caspase-8 pathway is being differentially regulated in the Th1 and Th2 clones, we utilized the inhibitor IETD-fmk, which interrupts the FADD/caspase-8 apoptotic pathway by functioning as a noncleavable substrate for active caspase-8 (20). CD3 ligation induces apoptosis in all three susceptible clones (Fig. 1). However, addition of IETD-fmk during CD3-activation fully protects these cells from Fas-mediated death (Fig. 1). We therefore further investigated the FADD/caspase-8-associated death pathway used in activated Th1- and Th2-type cells.

Recently, a caspase-8-like protein without enzymatic activity, c-FLIP, was cloned (46). In stable transfectants, c-FLIP was shown to interfere with the activation of caspase-8 at the DISC (42). Thus, we wished to determine whether preferential recruitment of c-FLIP to the DISC in Th2-type cells could be responsible for its resistance to Fas. An apoptosis-resistant and an apoptosis-susceptible clone were treated with anti-CD3 Abs for 30 min before ligating their Fas receptors with anti-APO-1 Abs. FasR was immunoprecipitated, and the DISCs were Western blotted and probed with anti-c-FLIP Abs. As shown in Fig. 2, in anti-CD3-treated cells of both resistant (lane 1) and susceptible clones (lane 4), no recruitment of c-FLIP to the nonactivated Fas receptor is observed. However, detectable c-FLIP recruitment to the DISC is observed in both apoptosis-resistant and -susceptible clones, with (lanes 3 and 6) or without (lanes 2 and 4) prior CD3 treatment. This suggests that additional regulatory mechanisms, such as an improper DISC assembly or inappropriate caspase-8 cleavage at the DISC, are conferring Th2 resistance to Fas-induced apoptosis. Thus, we next tested the ability of Th2 cells to recruit and process caspase-8.

Caspase-8, as with all known caspases, exists in the cytoplasm in a pro-form (20). Previous studies have described that, on triggering of the FasR, caspase-8 gets recruited to the DISC where it is proteolytically activated (24). Initial cleavage of DISC-associated caspase-8 generates a p43 and a p12 fragment. The latter gets further processed to a p10 fragment. Subsequently, the DISC-associated p43 gets cleaved to release the active site-containing fragment p18 from the prodomain p26 (24) (Fig. 3, A and B). We therefore examined by immunoprecipitation and Western blotting the DISCs of Th1 and Th2 cells activated under different conditions. FasR on a susceptible (JC-2) clone and a resistant (SP-U) clone was directly ligated with anti-APO-1 (anti-Fas) Abs for 5 min, with or without prior stimulation with anti-CD3 Abs. FasR was immunoprecipitated, separated on 12% SDS-PAGE gel, and probed with anti-caspase-8 Ab C15, which recognizes full-length caspase-8, and the first cleavage intermediate of caspase-8, the p43/41 fragment (Fig. 3, A and B). As seen in Fig. 3,C, 56-kDa pro-caspase-8 is associated with the FasR under conditions where the cells are undergoing apoptosis (APO-1 Abs alone, lane 2) and also under conditions where the cells are resistant to apoptosis (anti-CD3 + anti-APO-1 Abs, lane 3). A 43-kDa band indicative of the first cleavage fragment of caspase-8 is also observed in the resistant clone (Fig. 3,C, lanes 2 and 3), suggesting that cleavage has been initiated at the DISC of the apoptosis-resistant clone. In the susceptible JC-2 clone, as expected, both in the presence of anti-APO-1 Abs alone (lane 5) and in the presence of anti-APO-1 + anti-CD3 Abs (lane 6), caspase-8 is recruited to the DISC, and the presence of p43/41 is observed (Fig. 3,C, lanes 5 and 6). Both apoptosis-resistant and -susceptible clones stimulated with CD3 Abs alone do not demonstrate any caspase-8 recruitment (Fig. 3 C, lanes 1 and 4).

In Fig. 3,D, the blot was reprobed with the N2 Ab, which recognizes p26 fragment of caspase-8, (Fig. 3, A and B). As observed previously with C15 Ab, p43/41 is found in the resistant clone, both under apoptosing and nonapoptosing conditions. p26, the cleavage product of p43/41, is observed in the resistant clone when treated with anti-APO-1 Abs (Fig. 3,D, lane 2). However, when cells are stimulated with anti-CD3 Abs before addition of anti-APO-1 Abs, conditions that protect these cells from apoptosis, no p26 is detected (Fig. 3,D, lane 3). In the susceptible clone, again as expected, p26 was observed both when cells are treated with anti-APO-1 Abs (Fig. 3,D, lane 5) and also when pretreated with CD3 before addition of anti-APO-1 Abs (Fig. 3,D, lane 6). In lane 4 of Fig. 3 D, the susceptible clones do not show any caspase-8 recruitment, since they are treated with CD3 for only 30 min, which is insufficient length of time to induce endogenous FasL. However, 30 min of CD3 engagement is sufficient to generate protective signals in resistant clones.

Lysates from Th1 and Th2 clones were further probed with C15 and C5 Abs (Fig. 3, E and F) to analyze caspase-8 cleavage products released from the DISC into the cytosol. C15 Ab detects the p18 fragment, and C5 Ab detects the p10 fragment (Fig. 3,A). The presence of active site-containing p18 (Fig. 3,E) and p10 (Fig. 3,F) fragments is observed only in the apoptosis-sensitive clones (lanes 5 and 6). The presence of the p12 fragment, a cleavage intermediate of full length caspase-8 (p56 cleavage giving rise to p43 and p12 products), suggests that caspase-8 cleavage is initiated in both the resistant and susceptible clones. However, further processing of p12 to the active p10 fragment and cleavage of p43 to p26 and the active p18 fragment do not occur in the resistant Th2 clone (Fig. 3, E and F). This cleavage process is being specifically prevented in the Th2 clone by signals generated upon CD3 ligation.

The immunoprecipitation and Western blot results indicate that Th2 cells utilize a novel strategy to protect themselves from Fas-mediated death, by actively disrupting the DISC. Since this protection was only obvious after CD3 ligation, it was important to uncover which signals generated by CD3 intersected with the FADD/caspase-8 pathway. Keeping in mind that PI3′-K protects from apoptosis in other systems, we used wortmannin, an inhibitor of PI3′-K, and examined induction of apoptosis in five individual Th2 clones. As previously observed, anti-APO-1 (Fas) Abs induce apoptosis of Th2 cells that can be abrogated if the cells are stimulated with CD3 Abs before FasR ligation. However, CD3 stimulation of wortmannin-treated cells followed by FasR ligation leads to a dramatic increase in apoptosis of the normally resistant Th2 clones (Fig. 4 A). Pretreatment of Th2 clones with wortmannin before CD3 stimulation does not induce death in the T cell clones, confirming that the wortmannin is not toxic to the cells.

The wortmannin results suggest that, upon CD3 stimulation, PI3′-K becomes active and, acting on downstream substrates, is able to interfere with an orderly apoptotic cascade. To further confirm the role of PI3′-K in preventing Fas-mediated apoptosis, WT and a dominant negative mutant of the p85 regulatory subunit of PI3′-K were transiently transfected into a typical apoptosis-resistant Th2-type clone (SPU). Successful transfectants were selected on the basis of their expression of a murine MHC class I molecule, which had been cotransfected with the p85 subunit encoding plasmid. Following 8 h of CD3 stimulation, cells were harvested and prepared for apoptosis ELISA. Cells receiving the WT p85 behaved as normal untransfected counterparts, i.e., resistant to Fas-mediated death (Fig. 4,B). However, the cells transfected with the dominant-negative mutant become susceptible to apoptosis (Fig. 4 B).

Pharmacological and gene transfer experiments imply that the apoptosis-resistant and -susceptible clones differ in their ability to activate PI3′-K. To explore this possibility, we compared the activity of PI3′-K in four resistant and four susceptible clones, before and after CD3 ligation. T cell clones were rested for 8 h in the absence of IL-2 (this length of time does not affect the viability of the cells) to reduce background PI3′-K activity. Cells (5 × 10⁶) each were stimulated with anti-CD3 Abs, or the same number of cells was left untreated. Cell lysates were prepared and subjected to immunoprecipitation with Abs to the p85 regulatory subunit of PI3′-K. Kinase activity was measured using [γ-³²P]ATP and phosphatidylinositol (PI) as the substrate. Radiolabeled lipids were separated by TLC and quantitated by phosphorimaging. Radiolabeled lipids were identified by comparison of rf to known standards (Fig. 5). The four resistant clones SPU, HG3, HG5, and HG15 exhibited percentage increases in kinase activity of 156, 1000, 285, and 251%, respectively. In contrast, the susceptible clones had significantly lower activity. PS13, 59.5, JC6, and JC1 exhibited percentage increases of kinase activity of only 79, 41, 26, and 16%, respectively (Fig. 5).

To test how PI3′-K was regulating apoptosis in Th2 cells, we next determined whether blockade of PI3′-K permitted the generation of active caspase-8 fragments. We tested caspase-8 cleavage pattern in an apoptosis-resistant Th2-type clone (SPU) that was stimulated with anti-CD3 Abs in the presence or absence of wortmannin, followed by FasR ligation. FasR was immunoprecipitated, separated on 12% SDS-PAGE gel, and probed with the anti-caspase-8 Ab C15, which recognizes full-length caspase-8, and the first cleavage intermediate of caspase-8, the p43/41 fragment. Both full-length p56 caspase-8 and partially cleaved p43/p41 are detected with C15 Abs when cells are treated with anti-APO-1 alone (Fig. 6,A, lane 3), Abs to APO-1 and CD3 (Fig. 6,A, lane 4), or Abs to APO-1 and CD3 in the presence of wortmannin (Fig. 6,A, lane 5). C15 Abs were also used to probe the lysates for the presence of p18. Ligation of FasR alone leads to the generation of the catalytically active p18 in the resistant clone (Fig. 6,B, lane 3), which can be blocked by CD3 stimulation (Fig. 6,B, lane 4). However, in the presence of wortmannin, complete cleavage of caspase-8 to p18 fragment is observed (Fig. 6 B, lane 5). These results demonstrate that inhibiting PI3′-K allows complete cleavage of caspase-8, which is inhibited in CD3-activated Th2 cells that resist apoptosis.

In this report we characterize the Fas signal transduction pathway in activated Th1- and Th2-type cells and identify PI3′-K as a critical molecule in mediating protection from Fas-induced death in activated Th2 cells. We present a novel mechanism of how Th2 cells resist Fas-mediated apoptosis.

The search for mechanisms that permit T cells to resist apoptosis has led to the discovery of a number of molecules that block the FADD/caspase-8 apoptotic pathway (46, 47, 48). Previous work indicates that bringing caspase-8 into close proximity with another caspase molecule at the DISC is sufficient to permit autoproteolysis of the caspase to occur (24). Therefore, the most efficient way to inhibit the FADD/caspase-8 apoptotic pathway is to prevent recruitment of caspase-8, resulting in a nonfunctional DISC. Indeed, recently such a protein, termed c-FLIP, has been cloned (46) that can compete with caspase-8 for binding to FADD, and thereby blocking recruitment of caspase-8 to the DISC (42). That this strategy may be utilized by cells that undergo IL-2-dependent Fas-mediated apoptosis has been suggested (49). However, recent studies examining the inability of short-term activated peripheral T cells to recruit caspase-8 (45) indicates that c-FLIP is not involved in preventing caspase-8 recruitment to the DISC, unless it is overexpressed (42). The present data also support this finding by showing that resistance of Th2 cells is not at the level of c-FLIP. The amount of endogenous c-FLIP present in T cells is insufficient to prevent caspase-8 recruitment and apoptosis induction (present data and Ref. 42). c-FLIP, only when overexpressed, can compete with caspase-8 for binding to FADD, and thereby halt caspase-8 recruitment and activation.

Our results demonstrate that the differences in caspase-8 processing at the DISC of Th1 and Th2 cells is due to differential PI3′-K activation in the two subsets following CD3 stimulation. PI3′-K has been shown previously to protect cells from a variety of apoptotic stimuli (50, 51, 52, 53). More recently, it has been clearly demonstrated that serine threonine kinase AKT/protein kinase B, a target of PI3′-K, is an essential signaling intermediate in this antiapoptotic pathway. AKT can catalyze BAD phosphorylation and thereby sequester it from bcl-xL/bcl-2, thus preventing apoptosis (51, 54). AKT can also phosphorylate caspase-9 and prevent it from activating caspase-3 (66). Both these targets of AKT are components of the mitochondrial-dependent pathway of Fas-mediated apoptosis. This pathway is characteristic of cells that recruit low amounts of caspase-8 to the DISC and are therefore not efficient at directly activating caspase-3 (26). In contrast, cells that recruit large amounts of caspase-8 to the DISC can directly activate caspase-3 and bypass the mitochondria. These cells therefore may have developed other PI3′-K/AKT targets whose phosphorylation would abrogate apoptosis. Cells described in these studies recruit large amounts of caspase-8 to the DISC and thus provide a model system to address in the future how PI3′-K directly or through AKT regulates caspase-8 cleavage and subsequent caspase-3 activation in the mitochondrial-independent pathway of Fas signaling.

Engagement of the TCR/CD3 complex initiates the activation of a number of protein tyrosine kinases, followed by recruitment and activation of ZAP-70 (55, 56). One of the key substrates for ZAP-70 is linker for activation of T cells (LAT). Phosphorylated LAT associates with the p85 subunit of PI3′-K directly and also via tyrosine phosphorylated cbl (57). TCR-interacting molecule (TRIM) (58) is another linker protein containing several tyrosine-based signaling motifs that, upon phosphorylation, can also associate with p85 (59). Thus, both TRIM and LAT function as linker molecules to target PI3′-K to the membrane where it can potentially interact with its lipid substrates and other activators, such as Ras (60, 61). The data reported here reveal that Th1 and Th2 cells regulate PI3′-K activation differently upon CD3 activation leading to differences in their ability to resist Fas-mediated apoptosis. Future studies will include a systematic comparative analysis of CD3 signaling in the two subsets to gain a better insight into the mechanisms of why CD3 signaling in Th1 cells results in poor activation of PI3′-K. Indeed, previous studies have indicated that Th1 and Th2 cells signal differently upon engagement of their TCRs. Fitch and colleagues studying murine Th1 and Th2 clones have shown that stimulation of Th1, but not Th2, clones with anti-TCR Abs leads to elevated [Ca²⁺] (9, 62). Similar observations were made in another study where calcium ion signaling was measured in T cells as they differentiated from naive T cells to mature Th1 and Th2 effector populations (63). Naive T cells engaged the Ca²⁺ pathway, which was further enhanced if the cells were biased to differentiate into Th1 cells. In contrast, under Th2-differentiating conditions, the naive T cells selectively lost their ability to use this signaling pathway. PKC-dependent signaling leading to preferential up-regulation of FasL and subsequent susceptibility to apoptosis occurring only in Th1 cells, is yet another instance of differences between Th1 and Th2 cells in CD3 signaling (39). Thus, distinct signaling pathways are activated in Th1 and Th2 cells upon ligation of their CD3/TCR that can lead to differences in their effector functions and activities.

The biological significance of the differential susceptibility of Th1 and Th2 cells to undergo apoptosis when activated in the absence of costimulation become important during receptor triggering in response to Ag. During Ag presentation, and in a milieu that provides T cells with necessary costimulatory signals, Th1 cells will be capable of proliferating, and thus activating cellular immune responses. However, if an intracellular pathogen has disabled much of the costimulatory machinery (64, 65), Ag triggering of T cells will lead to skewing of the immune response by default to a Th2 type. In conclusion, this study provides a paradigm to further investigate how PI3′-K regulates caspase-8 activation, and also to examine whether preferential Fas-mediated apoptosis of Th1 cells is an additional modality for immune deviation.

We thank Tomas Mustelin for his generous gift of the PI3′-K constructs, Danny Dhanashekaran for help with the PI3′-K assays, and the National Cancer Institute for IL-2. We also thank Abul Abbas for helpful discussions.