Phosphatidic acid generation through interleukin 2 (IL-2)-induced alpha-diacylglycerol kinase activation is an essential step in IL-2-mediated lymphocyte proliferation.

Proliferation of T lymphocytes is triggered by the interaction of interleukin 2 (IL-2) with its high affinity specific receptor that is expressed on the cell surface following T lymphocyte activation. Significant advances have recently been made in identifying the multiple signals that follow IL-2 receptor occupancy, although the exact mechanism responsible for IL-2-induced proliferation remains an enigma. It has been shown previously that unique species of phosphatidic acid are rapidly produced in vivo following IL-2 binding. It was then suggested that, in contrast to other eukaryotic growth factor systems, phosphatidic acid was at least in part generated through IL-2-induced diacylglycerol (DG) kinase activation. In the present study we demonstrate IL-2-dependent activation of the alpha isoform of DG kinase. Confocal microscopy studies reveal that the enzyme is located in the cytosol and nuclei of resting T cells. Interleukin 2 stimulation induces translocation of the enzyme to the perinuclear region. Furthermore, our results indicate that inhibition of the alpha isoform of DG kinase has a profound effect on IL-2-induced T cell growth. Studies on cell cycle distribution demonstrate that the inhibition of IL-2-induced phosphatidic acid production induces arrest in late G1 phase of IL-2 dependent cells. Altogether, these results link previous observations of interleukin 2 and phosphatidic acid production to activation of an specific isoform of DG kinase and suggest that activation of this enzyme is part of a novel signaling cascade that utilizes phosphatidic acid as an effector molecule.

Activation of T lymphocytes by cell-bound antigens induces the expression of high affinity interleukin 2 (IL-2) 1 receptors on the cell surface. The binding of IL-2 to its high affinity receptor triggers a complex signaling program that ultimately results in cell proliferation. Although the transmembrane pathways activated upon IL-2 binding have been the subject of intensive studies, the exact mechanism responsible for the IL-2-induced progression through the cell cycle remains largely undefined. The functional high affinity IL-2 receptor is a heterotrimeric complex composed of at least three distinct polypeptide chains designated IL-2R␣ (p55), IL-2R␤ (p75), and IL-2R␥ (p64) (1)(2)(3). Although IL-2R␣ is capable of low affinity binding to IL-2, high affinity binding and biological response to IL-2 require association of the three proteins. All the studies performed to date indicate that the cytoplasmic domains of the ␤ and ␥ subunits are involved in transducing the IL-2 proliferative signal (4,5). The cytoplasmic domains of both proteins are devoid of any known intrinsic catalytic activity; thus, the early responses to IL-2 stimulation must be therefore transmitted by receptorassociated cytoplasmic enzymes. In this regard, activation of Src family and Janus family tyrosine kinases occupy a central role in the initiation of the IL-2-induced proliferative signal (6,7). As has been described for other receptors, tyrosine kinase activation would in turn induce association and activation of other signaling molecules such as phosphatidylinositol 3-kinase, p74 Raf kinase, or p21 ras (8 -10).
In our search for new molecules implicated in the transduction of the IL-2 proliferative signal, we have recently demonstrated that IL-2 induces the rapid activation of a DG kinase (DGK) in T lymphocytes following IL-2 binding to the high affinity receptor (11). DGK (ATP:1,2-diacylglycerol 3-phosphotransferase; EC 2.7.1.107) rapidly converts DG to phosphatidic acid (PA). This activity has been largely known as a regulator of the intracellular levels of DG, a second messenger that, in turn, regulates cellular events through activation of protein kinase C. In T cells, however, previous analysis indicates that proliferation upon IL-2 binding does not require activation of the classical and new protein kinase C family (12,13). In fact, previous studies failed to reproducibly detect phosphatidylinositol 4,5-bisphosphate hydrolysis or increases in Ca ϩ flux as components of the IL-2 proliferative response. Although IL-2 does not generate DG as a result of phosphatidylinositol hydrolysis by phospholipase C ␥, accumulation of DG and PA through hydrolysis of GPI molecules following IL-2 binding has been reported (14,15). Activation of DGK could therefore serve as a source of PA as an alternative mechanism to GPI-phospholipase D activation. Several findings indicate that, rather than being an intermediate in phosphatidylinositol turnover, the PA generated upon IL-2 binding may itself modulate signaling pathways responsible for the IL-2-induced lymphocyte proliferation. For instance, addition of exogenous PA to CTLL2 cells is able to mimic the IL-2 effect in proliferation as well as c-Myc induction (11,16). The consecutive activation of a GPI-PLC and a DGK would therefore constitute a novel pathway for the generation of the bioactive lipid PA.
Although increases in DGK activity following IL-2 binding were previously described, the identity of the enzyme responsible for this activity was not known. Several mammalian DGKs have been purified from various tissues so far, and some of them have already been cloned. At the present time the best characterized is the one having a mass of 80 kDa (␣DGK) originally purified from porcine brain cytosol (17). This isoform, which is highly tissue-specific, is only present in brain and lymphoid tissue. This enzyme is extremely abundant in the cytosol of lymphocytes and comprises more than 0,2% of the total soluble protein in T cell-enriched pig splenocytes (18). The aa sequences described for this DGK isozyme from porcine, human, and rat tissue are more than 80% identical to each other. The primary structure contains two sets of EF-hand motifs and a cysteine-rich zinc finger-like sequence (19). Although lymphocytes contain very high levels of DGK ␣ isoform, the role of this enzyme on the regulation of the immune response has never been addressed. In the present study we have investigated the activation of the ␣DGK isoform upon IL-2 binding on T lymphocytes. Immunoprecipitation studies using a specific antibody against ␣ DGK demonstrate IL-2-dependent activation of this enzyme. By using inhibitors of this enzyme, we demonstrate that this isoform is responsible for the majority of PA generated in response to IL-2 through DGK activation. Confocal microscopy studies reveal that the enzyme, located in the cytosol and nuclei of resting cells, translocates to the perinuclear space in response to IL-2 binding. Using ␣DGK inhibitors we demonstrate the essential role of the PA generated through ␣DGK activation on T cell growth. Finally, cell cycle analysis demonstrates that inhibition of PA production prevents the cells from entering S-phase. These results further demonstrate that activation of the ␣DGK isoform and the subsequent accumulation of PA play a significant role in IL-2-dependent control of cell cycle progression in T lymphocytes.

EXPERIMENTAL PROCEDURES
Cells-The natural killer-like cell line YT was a gift from Dr. Kendall Smith (Dartmouth University). YT cells were maintained in RPMI 1640 medium (Life Technologies, Inc.) supplemented with 10% (v/v) fetal bovine serum, 4 mM L-glutamine, 50 M 2-mercaptoethanol, and 10 mM HEPES. CTLL-2 cells (obtained from the American Type Culture Collection) were maintained in the same medium supplemented with 10% (v/v) conditioned medium from concanavalin A-stimulated rat splenocytes. Before stimulation with IL-2, the cells were washed and resuspended in serum-and IL-2-free medium for 2 to 4 h. This incubation time was the minimum period necessary to arrest cell growth without loss of cell viability.
[␥-32 P]ATP, FITC-conjugated anti-rabbit monoclonal antibody, and ECL Western blotting detection system were from Amersham Corp. Silica Gel G-60 TLC plates were obtained from Whatman. All the organic solvents were from Merck. The preparation and characterization of rabbit antibody against ␣DGK has been described previously (17).
Measurement of DGK Activity-15 ϫ 10 6 quiescent cells were stimulated with either buffer or 100 units/ml recombinant IL-2. After the periods of time indicated in the figure legends, the cells were harvested, washed two times with ice-cold phosphate-buffered saline, and frozen at Ϫ70°C. The cells were thawed and lysed by nitrogen cavitation (10 min at 500 p.s.i., 4°C) in a buffer containing 10 mM Tris-HCl, pH 7.4, 1 mM EDTA, 50 mM NaF, 2 mM Na 3 VO 5 , 1 mM PMSF, 10 g/ml leupeptin, and 10 g/ml aprotinin. The lysates were then centrifuged at 800 ϫ g, and the supernatant was used to determinate DG kinase activity. Enzyme activity was determined by measuring the incorporation of radioactive phosphate into PA in the presence of DG. A 25-l aliquot of the cell lysate (20 -40 g of protein) was the source of the enzyme. Other assay constituents were 0.2 mg/ml DG (18:1/18:1), 5 mM MgCl 2 , and 20 M [␥-32 P]ATP (20 Ci/mmol). The stock solution of pure DG (18:1/18:1) in chloroform was dried under nitrogen, resuspended in a small volume of 10 mM Tris, and sonicated before being added to the phosphorylation mix. Standard phosphorylation assays were carried out for 15 min at room temperature in a final volume of 50 l. When inhibition studies were performed, DGK inhibitors were added to lysates or immunopellets 10 min before the DG. At the end of phosphorylation assay, lipids were extracted by subsequent addition of 200 l of CHCl 3 /MeOH (2:1), 50 l of CHCl 3 , and 50 l of 0.1 N HCl. After centrifugation at 500 ϫ g, the organic layers were recovered, dried under a stream of nitrogen, taken up in 20 l of CHCl 3 /MeOH (2:1), and applied to Silica Gel G-60 plates along with dioleoyl PA standard. Plates were developed with a solvent system of CHCl 3 , MeOH, 4 M NH 4 OH (9:7:2, v/v/v). Dried plates were subjected to autoradiography, and the bands corresponding to PA were quantified by scanning of the autoradiograms.
Immunoprecipitation and Immunoblotting-15 ϫ 10 6 quiescent cells were stimulated with either buffer or 100 units/ml recombinant IL-2. After the periods of time indicated in the figure legends, the cells were harvested, washed two times with ice-cold phosphate-buffered saline, and lysed in ice-cold lysis buffer (1% Nonidet P-40, 10 mM Tris-HCl, pH 7.4, 1 mM EDTA, 50 mM NaF, 2 mM Na 3 VO 5 , 1 mM PMSF, 10 g/ml leupeptin, and 10 g/ml aprotinin). After 15-30 min at 4°C, the whole cell lysate was centrifuged at 15000 rpm and the supernatant precleared by treatment with 50 l/sample protein A-Sepharose. The precleared lysates were then subjected to immunoprecipitation with the ␣DGK antibody. The immune complexes were collected on protein A-Sepharose. After extensive washing immunopellets were ready to use for the DG kinase assay, performed as described above, or for immunoblotting. When immunoblotting was to be performed, proteins were eluted from protein A beads in Laemmli buffer, separated on reduced SDS-polyacrylamide gels, and then electrophoretically transferred for 1 h at 250 mA to nitrocellulose (Bio-Rad) in buffer containing 25 mM Tris, 192 mM glycine, 0,1% SDS, and 20% methanol. The membranes were then blocked by incubation with Tris-buffered saline (0.15 M NaCl, 20 mM Tris-HCl, pH 7.4) containing 5% dried nonfat milk. The specific proteins were detected by enhanced chemiluminescence (ECL system) and autoradiography following incubation of the blot with primary antibody and horseradish peroxidase-linked secondary antibody.
Indirect Immunofluorecence-1 ϫ 10 5 exponentially growing cells were washed, deprived of serum and IL-2 for 4 h, and then incubated either with buffer or with 100 units/ml IL-2 plus or minus R59022 (10 M final) for 10 min. Cells were harvested and centrifuged in a cytospin (Heraeus) for 10 min at 600 rpm over glass slides. Cells were fixed with methanol for 3 min at Ϫ20°C and permeabilized with 100 g/ml lysophosphatidylcholine in phosphate-buffered saline for 2 min at room temperature. Slides were washed twice for 2 min at room temperature with 1% BSA in phosphate-buffered saline. Rabbit polyclonal anti-␣DGK antibody diluted in PBS/BSA was added to the slides under a coverslip and incubated in a moist chamber for 1 h at room temperature. After washing once with PBS and twice with PBS/BSA, anti-rabbit FITC monoclonal antibody was added and incubated for 45 min in a moist chamber at room temperature. After washing three times with PBS/BSA and once with PBS, slides were mounted in Moviol (Sigma) with anti-fading. All confocal experiments were analyzed in a Zeiss laser scanning microscope. The images were acquired using software from Carl Zeiss.
DNA Synthesis-CTLL2 cells exponentially growing on IL-2-supplemented medium were washed free of IL-2 and resuspended in RPMI, 10% fetal calf serum. After 15 h of starving, the IL-2-deprived cells were restimulated with either IL-2 (50 units/ml) alone or IL-2 plus DGK inhibitors at the indicated concentrations. 10,000 cells were seeded in 96-well plates, and quadruplicate samples were pulsed at 2-h intervals with 1 mCi/well [ 3 H]thymidine. Cellular DNA was subsequently harvested on glass-fiber filters, and 3 H uptake was quantified by liquid scintillation counting.
Cell Cycle Analysis-Cells were harvested by centrifugation and washed in phosphate-buffered saline buffer. After centrifugation the cells were resuspended in permeabilization solution (0.1% sodium citrate, 0.05% Nonidet P-40). The samples were treated with 50 g/ml RNase A for 30 min at room temperature, and then propidium iodine was added to a final concentration of 20 g/ml. After 20 min, the fluorescence of the propidium iodine-stained DNA was quantitated on a per cell basis with an EPICS-XL flow cytofluorometer (Coulter).

IL-2-induced Production of PA Is Inhibited by the ␣DGK
Inhibitor R59022-Previous studies have demonstrated that DGK activity was markedly increased following addition of IL-2 to lymphoid cells (11). This indicates that IL-2-dependent increases in PA can, at least in part, be attributed to DGK-dependent conversion of DG to PA. As a first step to demonstrate the potential involvement of the ␣DGK isoform in the generation of PA by IL-2, we have utilized a specific inhibitor of this enzyme R59022. This inhibitor has been demonstrated to abolish activity of purified ␣DGK, whereas other DGK species do not (20). For these experiments we have used the human natural killer cell line YT, which expresses high levels of high affinity IL-2 receptor. Cells harvested at different times after addition of IL-2 were lysed by nitrogen cavitation. The assay was performed by quantifying the in vitro phosphorylation of exogenously added DG to yield [ 32 P]PA as described under "Experimental Procedures." Fig. 1 shows the transient increase in phosphorylated PA obtained after addition of IL-2 (upper panel). The rate of PA formation, and thus DGK activation, was maximum at 5 min, approximately 6-fold higher in IL-2-treated cells than in control cells. When R59022 at a concentration of 25 M final was added to the phosphorylation reaction, the rate of PA formation diminished by 70% of the initial value (lower panel). This inhibitor concentration has been shown previously to be enough to abolish ␣DGK activity. Similar results were obtained with another DGK inhibitor, R59949 (21), added at a final concentration of 10 M (data not shown). The autoradiograms were scanned, and a quantification of four different assays results is represented.
IL-2-stimulated DGK Activity Is Immunoprecipitated by Anti-DGK␣ Antibodies-Although known to be very abundant in T lymphocytes, no agonist-induced activation of the ␣DGK isoform has been demonstrated. Using a polyclonal antibody, we immunoprecipitated the enzyme from detergent cell lysates of quiescent and IL-2-stimulated YT cells. To verify the specificity of the antibody in immunoprecipitating this DGK isoform, immunoblot analysis of the immunoprecipitates was performed using the same antibody. As shown in Fig. 2A, a protein with molecular weight corresponding to the ␣DGK is recognized by the anti-␣DGK antibody in the immunoprecipitates. The next step was to demonstrate activation of the ␣DGK upon IL-2 binding. To this end, immunoprecipitated samples were resuspended in phosphorylation buffer and assayed for DGK activity as described under "Experimental Procedures." The samples were analyzed by TLC and autoradiography. As is shown in Fig. 2B, the ratio of phosphorylation of exogenously added DG is increased after 2 min of IL-2 treatment, reaching the maximum level of phosphorylation at 5 min and decrease thereafter. These results clearly indicate IL-2-induced activation of ␣DGK in YT cells. To further demonstrate that the IL-2-stimulated DGK activity present in the immunoprecipitates corresponds to the ␣ isoform of the DGK, the R59022 inhibitor was added to the phosphorylation reaction. When the immunoprecipitates of ␣DGK were assayed for DGK activity in the presence of the inhibitor, the activity was completely abolished. Addition of R59949 to the immunoprecipitates had the same effect on the phosphorylation reaction (data not shown).
IL-2 Induces Translocation of ␣DGK to the Perinuclear Region-DGK activation has been detected previously on IL-2stimulated T lymphocytes in both soluble and membranebound subcellular fraction (11). Whether both activities arose from the same enzyme species was at that time unknown. Previous studies have suggested that ␣DGK, located mainly on the cytosol, could translocate to the plasma membrane upon activation. Immunofluorescence analysis was performed to determine the distribution pattern of ␣DGK after IL-2 stimulation. YT cells were stained with a polyclonal antibody against ␣DGK and the primary antibody detected by using anti-rabbit antibody coupled to FITC. Confocal images from the middle of the cell were obtained. Three different fields under each condition are shown to guarantee the consistency of the results. In quiescent cells, staining with ␣DGK antibody resulted in a diffuse homogeneous pattern (Fig. 3A). This indicates that, aside from a cytosolic localization, a considerable portion of ␣DGK is present in the nucleus. After 10 min of IL-2 stimulation, the pattern observed was distinctly different (Fig. 3B). Staining disappeared from inner nucleus and also diminished in the cell periphery and tended to concentrate in the perinu-

FIG. 1. ␣DGK inhibitor prevents IL-2-induced DGK activation.
The effect of the inhibitor R59022 on total cellular DGK activity was determined. Resting cells were stimulated with recombinant IL-2 at the times indicated, and DGK activity in cell lysates was assayed by measuring the incorporation of 32 P into PA in the absence (upper panel) or presence (lower panel) of the ␣DGK inhibitor R59022. The formation of PA was further determined by ascending thin layer chromatography and autoradiography. Autoradiograms were scanned, and a quantification of four independent assays is represented. IL-2, open symbol; IL-2 ϩ R59022, closed symbol.

FIG. 2. IL-2-stimulated DGK activity is immunoprecipitated by anti-␣DGK antibodies.
Quiescent cells were stimulated with IL-2 for the indicated times, and, after lysis in detergent-containing buffer, ␣DGK was immunoprecipitated with an specific anti-␣DGK polyclonal antibody. A, immunoblot analysis of immunoprecipitates. B, autoradiograms of the thin layer chromatography separation of the [ 32 P]PA formed when the immunoprecipitates are subjected to the phosphorylation assay performed in the absence or presence of the ␣DGK inhibitor R59022. Autoradiograms of three independent phosphorylation experiments were scanned and represented. IL-2, open circles; IL-2 ϩ R59022, closed circles. clear region. The localization of the signal suggested a translocation of both nuclear and cytoplasmic ␣DGK to the perinuclear region. Ten minutes of pretreatment with the ␣DGK inhibitor R59022 prior to IL-2 stimulation resulted in a granular pattern in the cytoplasmic region, with no effect in the nucleus (Fig. 3C). When the cells were incubated with R59022 alone, the pattern was the same that the one observed in control cells (data not shown). These results suggest that R59022 inhibits cytosolic ␣DGK, seizing it and not allowing the enzyme to reach its substrate.
␣DGK Inhibition Prevents IL-2-induced DNA Synthesis-Earlier findings have suggested that PA could mediate the growth-promoting effects of IL-2. The previous experiments using ␣DGK immunoprecipitates indicated that both R59022 and R59949 inhibit IL-2-induced ␣DGK activation. To determine if the PA accumulated in response to IL-2 through ␣DGK activation was responsible to the same extent of IL-2-dependent cell growth, we determined the effect of the two inhibitors of ␣DGK, R59022 and R59949, on cell proliferation. To further assess the specificity of DGK inhibition on cell proliferation, we also tested the effect of a structurally different DGK inhibitor on the IL-2-induced proliferation assay. We used ethylene glycol dioctanoate, a competitive DGK inhibitor with a very different structure from R59022 and R59949, which has been shown previously to inhibit DGK activity both in vitro and in vivo (22). In our system, and as shown previously for R59022, addition of ethylene glycol dioctanoate markedly reduces the activity of the immunoprecipitated ␣DGK (data not shown). For these experiments, we used the IL-2 growth-dependent cell line CTLL2. This cell line has an absolute requirement of IL-2 to grow and has been utilized previously to demonstrate the IL-2 mimetic effect on growth of PA (11,16). As is shown in Fig.  4, addition of the different inhibitors reduced IL-2-stimulated DNA synthesis, as quantitated by measuring [ 3 H]thymidine incorporation, in a dose-dependent manner. Cell integrity at all the concentrations tested was determined by trypan blue exclusion.
DGK␣ Inhibition Prevents the Entry of IL-2-stimulated T Cells into S-phase-Inhibition of DGK activity resulted in impairment of IL-2-induced cell growth. We questioned whether this inhibition was related to the blocking of IL-2-induced G 1phase progression. CTLL-2 cells were stained with propidium iodine, and cell cycle distributions were analyzed by flow cytofluorometry. When exponentially growing CTLL2 cells were washed free of IL-2 and cultured for 12 h in medium without IL-2 (basal medium), a marked increase in the percentage of G1-phase cells was observed (Fig. 5A). Upon addition of IL-2, a typical histogram from CTLL2 cells cultured under exponential growth is shown in Fig. 5B. The addition of either R59022 or R59949 prevented the IL-2 effect, as determined by propidium iodine staining (Fig. 5, C and D). As a control of late G 1 arrest, starving CTLL2 cells were restimulated with IL-2 plus rapamycin (Fig. 5E). In Fig. 6, it is shown how the growth factordeprived CTLL2 cells entered S-phase in a synchronous fashion after 10 -12 h of restimulation with IL-2. However, the addition of IL-2 plus either R59022 or R59949 to the IL-2-starved CTLL2 cells completely inhibited IL-2-induced S-phase entry, as determined by [ 3 H]thymidine incorporation (see Fig. 6). DISCUSSION This study shows what our understanding is the first evidence of agonist activation of ␣DGK. Following three different approaches: immunoprecipitation, inhibition studies, and confocal microscopy analysis, the data presented in this study indicate for the first time that IL-2 activates ␣DGK. Indirect evidence, such as the lack of selectivity for different DG substrates as well as phosphatidylserine-mediated activation of this DGK, also indicates that ␣DGK is the isoform activated by IL-2 (data not shown). Moreover, our studies indicate that IL-2 activation of this isoform of DGK is the main source of PA production through this mechanism. Finally, another important observation derived from the present work is the finding that, in the IL-2 signaling system, PA and not DG is the lipid second messenger with mitogenic activity; PA generation is a key event for the correct proliferative function of this cytokine.
The results presented here demonstrate that the formation of [ 32 P]PA and thus DGK activation takes place over a similar time course after IL-2 addition in both total cell lysates and specific immunoprecipitates of ␣DGK. While DGK activity is completely abolished when R59022 is added to the immunoprecipitates, there is some activity still present if the assay is performed on cell lysates. These data suggest that another DGK isoforms could also be activated by IL-2. In this regard we have previously reported the presence of DGK activity in the membrane fraction that is enhanced in response to IL-2 (11). This subtype of DGK could be similar to the arachidonoyl-DGK identified by Glomset and others as a membrane-bound DGK (23).
Initial studies demonstrated that ␣DGK is mostly present in the cytosolic fraction, being the content in the particulate fraction very scarce (24). By analogy with other DG-binding proteins, translocation from the cytosol to the plasma membrane has been postulated. To determine if IL-2 induces translocation of ␣DGK, confocal studies were performed. The confocal immunofluorescence assays indicate that, in resting cells, ␣DGK is present not only in the cytosol but also in the nucleus and is perhaps partly associated with the cytoskeleton. This is not surprising since DGK activity has been previously described in both the nuclei (25) and associated to the cytoskeleton (26 -28). Our results here show that, upon IL-2 stimulation of the cells, the enzyme does not translocate to the plasma membrane and is instead accumulated mostly in the perinuclear space. In view of these results, it must be postulated that at least some if not most of the PA being generated in response to IL-2 binding is produced in inner compartments and not in the plasma membrane.
Previous studies by Sakane and co-workers (29) suggested that the translocation of ␣DGK to membranes occurs in a Ca 2ϩ -dependent manner. Furthermore, the existence of EFhand motifs in the ␣DGK sequence suggested that the enzyme was functionally linked to Ca 2ϩ signals (30). IL-2 fails to induce Ca 2ϩ mobilization, but we have shown here that ␣DGK is successfully translocated in an IL-2-dependent manner. In this regard other groups have described the Ca 2ϩ -independent activation and translocation of this DGK isoform (31,32). Interestingly, Kanoh and co-workers (33) have reported that an ␣DGK mutant lacking EF-hands lost Ca 2ϩ binding activity, but could be fully activated by phosphatidylserine or deoxycholate in the absence of Ca 2ϩ . Recently another DGK isoform containing EF-hands in its sequence has been cloned, and its activity resulted to be Ca 2ϩ -independent (33). All these experiments suggest that a Ca 2ϩ -independent mechanism may be responsible for the IL-2-induced ␣DGK translocation, and therefore some other unidentified roles could be attributed to the EFhand motif. As for other possible mechanisms responsible for the enzyme translocation, Besterman and co-workers (32) have reported that the presence of DG and phosphatidylserine is enough to induce translocation of the DGK enzyme from cytosol to membranes. The same mechanism where substrate concentration is responsible for the translocation has been also described for protein kinase C-depleted 3T3 cells when cell-permeable dioctanoylglycerol is added (34).
If IL-2 induces translocation of the enzyme, could the increases on activity be due to an increase on the level of protein? In the immunoprecipitation studies, the content of ␣DGK in the immunoprecipitates is the same in resting and activated cells. This implies that the increase in the DGK activity is not related to an increase in the ␣DGK protein level and suggests that the activation occurs independently and previous to translocation. In fact, maximum translocation measured by microscopy confocal analysis is obtained about 10 min after IL-2 addition, several minutes after the maximum activation is obtained in the in vitro studies. It could be hypothesized that a covalent modification occurs. In this context, Schaap and coworkers (35) have reported that ␣DGK can be phosphorylated by Ser/Thr kinases and in Tyr by epidermal growth factor receptor. However, we have not been able to demonstrate any increase on tyrosine phosphorylation of the enzyme in response to IL-2 treatment (data not shown). Experiments are currently under way to study if any other covalent modification of the ␣DGK takes place in response to IL-2.
The main biological function of DGK has largely been considered to modulate the levels of DG that, in turn, regulate the activation of classical and new PKCs. Our experiments support a new role for this enzymatic activity in the generation of PA, a lipid with a potential central role in IL-2-induced proliferation. When DGK is inhibited by using specific inhibitors in IL-2-stimulated cells, IL-2-induced proliferation is impaired. Cell cycle analysis demonstrates that, upon DGK inhibition, the cells become arrested in the late G 1 phase. These results indicate that, in contrast to what has been described for other systems, it is not DG but the ␣DGK product PA that is the effector molecule in this system. We must therefore consider the novel hypothesis where accumulation of certain ␣DGKderived PA species following IL-2 stimulation constitutes an essential step for the T cells to reach the restriction point and enter S-phase, progressing afterward through the cell cycle.
Several observations have previously suggested that PA may act as a second messenger. Specific species of PA that contain polyunsaturated fatty acids have been shown to be increased in ras-transformed fibroblasts (36). PA has also been shown to induce invasion of tumor cells in vitro (37). When added exogenously to cells, PA exhibits growth factor-like activity (11, 16, 38 -40) and can induce actin polymerization (41). Moreover, intracellular accumulation of PA, rather than an increase in DG, correlates well with mitogenesis in growth factor-stimulated fibroblasts (42), similarly to what has been described here. Therefore, it appears that specific PA species may be associated with proliferation and a high level of this bioactive lipid could be present in tumor cells. Cellular targets of PA action have not been identified to date, although activation of cellular kinases has been suggested (43). Limatola and coworkers (44), using a cell-free assay system, have demonstrated a strong PA-dependent activation of the DG-insensitive isotype of PKC. Interestingly, PA-induced activation of PKC is inhibited by Ca 2ϩ . This would imply that, in intact stimulated cells, PKC can only be activated by PA when Ca 2ϩ remains at basal levels as it happens upon IL-2 binding. On the other hand, Gomez and co-workers (45) have recently reported, using antisense techniques, that the isotype of PKC could be implicated in the control of IL-2 mediated proliferation. Nevertheless, further studies will be necessary to establish the exact role of PKC on IL-2 signaling processes and its activation mechanism by PA. Future work must also contemplate the identification of other possible targets of the PA action as well as the characterization of the fatty acid composition of the effector PA species generated in response to IL-2.
In the last few years, a great number of studies have emerged describing the role of the sphingomyelin cycle in signal transduction (46). Ceramide, the product of this pathway, has been implicated as a mediator of programmed cell death (47)(48)(49), cell cycle arrest (50), and differentiation (51). All of these mechanisms imply a stop in cell cycle progression. Thus, ceramide could be considered an antimitogenic lipid. Interestingly, ceramide inhibits the activity of a DGK isoform (52). It also inhibits the activation of phosphatidylcholine-phospholipase D by several agonists (53). Therefore, the generation of ceramide would inhibit the production of PA either by DGK or phosphatidylcholine-phospholipase D activation. Furthermore, it has been reported that an inverse relationship exists between the cellular concentrations of ceramide and the proliferative capacity of human T-lymphocytes after IL-2 stimulation (47). Thus, we can envisage the exciting hypothesis that PA (a mitogenic lipid) and ceramide (an antimitogenic lipid) act as biosensors of the cellular state determining, together with other factors, if the cells are going to die, proliferate, or remain quiescent.