Regulation of T Cell Receptor-induced Activation of the Ras-ERK Pathway by Diacylglycerol Kinase (cid:1) *

T cell development in the thymus and activation of mature T cells in the periphery depend on signals stimulated by engagement of the T cell antigen receptor (TCR). Among the second messenger cascades initiated by TCR ligation include the phosphatidylinositol pathway where the membrane phospholipid, phosphatidylinositol 4,5-bisphosphate, is hydrolyzed to inositol 1,4,5-trisphosphate and diacylglycerol (DAG). Inositol 1,4,5-trisphosphate signals a rise in intracellular free calcium, leading to translocation of nuclear factor of activated T cells into the nucleus. DAG activates RasGRP and protein kinase C (cid:2) . Because both RasGRP and protein kinase C (cid:2) are essential for thymocyte and T cell function, it is critical to understand how DAG is regulated. In this report, we demonstrate expression of DAG kinase (cid:1) (DGK (cid:1) , the enzyme that catalyzes the conversion of DAG

During T cell development, precursor cells transit from the CD4 Ϫ CD8 Ϫ double negative (DN) 1 to the CD4 ϩ CD8 ϩ double positive (DP) stage and then become either CD4 ϩ or CD8 ϩ single positive (SP) mature T cells. Each of these transitions requires appropriate signals delivered by the T cell antigen receptor (TCR) or, in the earliest stages of development, the pre-TCR. Defects in expression of these receptors or in their signaling pathways can result in developmental blockade or skewing of the T cell repertoire (1)(2)(3)(4). In peripheral lymphoid organs, engagement of the TCR by antigenic peptide presented by major histocompatibility complex molecules on antigen presenting cells initiates a signaling cascade that is required for T cell proliferation and effector function (1,5). Abnormalities in signals delivered by the TCR may result in either hypo-or hyperactivation leading to undesirable outcomes such as immune deficiency (6 -9) or autoimmunity (10 -12).
Engagement of the TCR initiates numerous second messenger cascades. The most proximal known biochemical signal is stimulation of protein tyrosine kinases with subsequent tyrosine phosphorylation of multiple substrates (5,13,14). Among these substrates are phospholipase C␥1 (PLC␥1) (15)(16)(17) and critical adapters including linker of activated T cells (18) and SLP-76 (SH2 domain containing leukocyte phosphoprotein of 76 kDa) (19). These proteins in association with another adapter protein, Grb2-related adapter molecule downstream of Shc (20 -22), are part of a larger multimolecular complex required for PLC␥1 to act efficiently on its substrate (23)(24)(25)(26)(27)(28). Activated PLC␥1 hydrolyzes phosphatidylinositol 4,5-bisphosphate to generate two second messengers, diacylglycerol (DAG) and inositol 1,4,5-trisphosphate (29). Inositol 1,4,5-trisphosphate induces an influx of Ca 2ϩ that activates the calcineurinnuclear factor of activated T cells (NFAT) signaling pathway (30). DAG allosterically activates both RasGRP (31,32), a nucleotide exchange factor for Ras, and PKC (33), a serine (threonine) kinase, through associating with the C1 domains of both molecules. Activated RasGRP (34,35) and PKC (36,37) in turn stimulate the Ras-ERK-AP-1 and NFB pathways, respectively. Propagation of these signaling cascades promotes the transcription of numerous genes including those encoding cytokines critical for T cell development, activation, and proliferation. Deficiencies of RasGRP (35) and PKC (37) cause defects in positive selection in developing thymocytes and inefficient activation of peripheral T cells. Similarly, PLC␥1 deficiency (with consequent defects in phosphatidylinositol 4,5bisphosphate-derived second messenger production) abrogates TCR-mediated cellular activation in the Jurkat T cell * 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.
¶ Supported by a fellowship from the Cancer Research Institute, New York.
DAG is phosphorylated by members of a family of kinases, the DAG kinases (DGKs), resulting in conversion of DAG to phosphatidic acid (39,40). Although certain DGKs have been implicated in the regulation of receptor signaling in various cellular contexts, the importance of DGKs as regulators of T cell activation is less clear. In the experiments described in this work, we focused on DGK (41)(42)(43), whose mRNA is expressed at high levels in the thymus (41). We documented its pattern of protein expression in the thymus and peripheral lymphoid organs and, by using the Jurkat T cell line as a model for TCR signaling, demonstrated that overexpression of DGK potently inhibits TCR-induced activation of the Ras-ERK pathway without affecting Ca 2ϩ influx. Concordant with these observations, DGK blocks TCR-induced expression of CD69 and activation of an AP-1 reporter construct. Structure/function analysis demonstrated that both the kinase activity and DAG binding domains of DGK are essential for its ability to interfere with TCR signaling, whereas the ankyrin repeats are dispensable for such inhibition. Collectively, these observations suggest DGK can function as a negative regulator of T cell activation, presumably by terminating signaling via DAG.

EXPERIMENTAL PROCEDURES
Cell Lines-The human Jurkat leukemia T cell line and the J14 SLP-76-deficient Jurkat variant (kind gift of A. Weiss, University of California, San Francisco) (26) were grown in RPMI 1640 medium supplemented with 10% fetal bovine serum (FBS), 100 units/ml penicillin G, 100 units/ml streptomycin, and 292 g/ml of L-glutamine (RPMI-10).
DNA Constructs-The pEF-IRES-NGFR bi-cistronic expression vector was generated by replacing the Myc-His from pEF-Myc-His vector (Invitrogen, Carlsbad, CA) with an IRES-NGFR cassette (kindly provided by Dr. Warren Pear, University of Pennsylvania). FLAG-tagged wild type (WT), the N-terminal (⌬NT), and C-terminal (⌬CT) truncation mutants of DGK were generated by PCR with Pfu-Turbo (Stratagene, La Jolla, CA) using pcDNA-DGK WT (44) as a template and using hDGKz89F (5Ј GAA GAT CTA TGG AGC CGC GGG ACG GTA) and hDGKz2900R (5Ј CGG AAT TCC CTC CTG CTG CCC GTG GGC), hDGKz842F (GAA GAT CTC AGA ATA CTC TGA AAG CAA G) and hDGKz2900R, or hDGKz89F and hDGKz2449R (CGG AAT TCA TGC AGC ATC CCC TTG CAG T) primer pairs. The kinase dead (KD) mutant of DGK was generated using pcDNA-DGK KD (44) as the template, and hDGKz89F and hDGKz2900R as primers. The PCR products were digested and cloned in-frame with a FLAG tag coding sequence at the N terminus into pEF-IRES-NGFR. An N-terminal hemagglutinin (HA)-tagged human H-Ras (HA-Ras) construct was generated by reverse transcriptase-PCR using total RNA from Jurkat cells as template and HRas5ЈKpn (5Ј CGG GGT ACC ACG GAA TAT AAG CTG GTG G) and HRas3ЈRI (5Ј CGG AAT TCT CAG GAG AGC ACA CAC TTG) as primers. Amplified Ras cDNA was digested with KpnI and EcoRI and ligated into the corresponding restriction sites in pHM6 (Roche Diagnostics) for HA tag and mammalian expression. Sequences of all cloned cDNAs were confirmed using automated sequence analysis.
Transfection, Stimulation, and Luciferase Assay-Jurkat cells were harvested and washed once with phosphate-buffered saline (PBS) and once with cytomix (25 mM HEPES, 120 mM KCl, 0.15 mM CaCl 2 , 10 mM K 2 HPO 4 , 10 mM KH 2 PO 4 , 2 mM EGTA, 5 mM MgCl 2 , pH 7.6) (45). Cell pellets were resuspended in cytomix spiked with ATP (2 M) and glutathione (5 M) at a concentration of 5 ϫ 10 7 cells/ml. For each transfection, 40 g of test plasmids and 20 g of the AP-1-Luc reporter (46) were added to 0.4 ml of cells in a 4-mm gap cuvette. Cells were electroporated at 250 V and 950 microfarads, incubated on ice for 10 min, transferred into 15 ml of RPMI-10, and incubated at 37°C with 5% CO 2 . Cells were harvested 18 h later and seeded in 96-well plates in triplicate at 1 or 2 ϫ 10 5 cells/well in 200 l of RPMI-10 with various stimuli; the final concentrations are as follows: 1:1000 of C305 ascites (a monoclonal IgM antibody specific for the TCR on Jurkat cells) (47), 1 g/ml monoclonal anti-human CD28 antibody (Caltag Laboratories, Burlingame, CA), 50 ng/ml PMA (Sigma), and 1 g/ml ionomycin (Sigma). Eight hours after stimulation, cell pellets from each well were lysed in 110 l of luciferase lysis buffer (1% Triton, 120 mM K 2 PO 4 , 15 mM KH 2 PO 4 , pH 7.8, with freshly added 5 mM dithiothreitol). To measure luciferase activity, 50 l of each lysate was mixed with 50 l of luciferase assay buffer (176 mM K 2 PO 4 , 24 mM KH 2 PO 4 , 10 mM MgCl 2 , 10 mM ATP, pH 7.8) in duplicate in a 96-well white polystyrene assay plate (Corning Incorporated, Acton, MA). Fifty l of 1 mM luciferin (Sigma) was injected into each well, and luciferase activity was measured for 10 s after a 2-s delay by fluorimetry.
Immunoprecipitation and Western Blot for ERK Activation-J14 cells were transfected with a control vector, pEF-hSLP-76 (48), or pEF-hSLP-76 plus the DGK expression vector. Thirty-six hours later, cells were harvested and washed twice with PBS. Cells were resuspended in PBS at a concentration of 1 ϫ 10 7 cells/ml, aliquoted at 0.5 ml/tube, rested at 37°C for 30 min, and then stimulated with 1 l of C305 ascites at 37°C for 0, 1, 5, 15, and 30 min or stimulated with PMA (50 ng/ml) at 37°C for 5 min. Stimulated cells were pelleted and lysed in 1% Nonidet P-40 lysis buffer (1% Nonidet P-40, 150 mM NaCl, 50 mM Tris, pH 7.4) with freshly added protease inhibitors and phosphatase inhibitors. Proteins in cell lysates were resolved by SDS-PAGE (12%) and transferred to Trans-Blot Nitrocellulose membrane (Bio-Rad) and analyzed for ERK activation by probing the membrane with an anti-phospho-ERK antibody (Cell Signaling Technology, Beverly, MA). Levels of transfected SLP-76 and DGK were assessed with anti-FLAG antibody. The membrane was stripped and reprobed with an anti-ERK antibody (Santa Cruz Biotechnology, Inc., Santa Cruz, CA) for loading control.
In some experiments, ERK activation was analyzed in cotransfection experiments where a vector encoding Myc-tagged ERK was used. To assess the effect of DGK on the activation of cotransfected ERK, cell lysates were immunoprecipitated with a rabbit polyclonal anti-Myc antibody (Upstate Biotechnology, Inc., Lake Placid, NY) coupled to Gammabind TM Plus Sepharose beads (Amersham Biosciences) at 4°C for 2 h. The beads were washed three times with high salt buffer (1% Nonidet P-40, 450 mM NaCl, 50 mM Tris, pH 7.4) and twice with 1% Nonidet P-40 lysis buffer. Proteins on the beads were denatured, resolved by SDS-PAGE (12%), and transferred as described above. The membrane was blotted with an anti-phospho-ERK antibody, stripped, and reprobed with monoclonal anti-Myc (Upstate Biotechnology, Inc.) to control loading.
GST-Raf-RBD "Pull-down" Assay for Ras Activation-J14 cells were transfected with 40 g of test plasmids, 20 g of pEF-hSLP-76, and 20 g of pHM6-Ras and stimulated as described above. Stimulated cells were pelleted and lysed in 400 l of ice-cold MLB (25 mM HEPES, 150 mM NaCl, 1% (v/v) Nonidet P-40, 0.25% (w/v) sodium deoxycholate, 10 mM MgCl 2 , 1 mM EDTA, 10% (v/v) glycerol, pH 7.5) with freshly added phenylmethylsulfonyl fluoride (Sigma) and protease inhibitor mixture (Sigma) on ice for 15 min. Cell lysates were spun at 13,000 rpm at 4°C for 10 min. Supernatants were mixed with 30 g of GST-Raf-RBD fusion protein (plasmid kindly provided by Dr. Jeff Pessin, University of Iowa) coupled to glutathione-Sepharose 4B beads (Amersham Biosciences) and rocked at 4°C for 30 min. After three washes with 0.6 ml of cold MLB, the beads were mixed with 40 l of 2ϫ SDS reducing buffer and boiled for 5 min. Proteins were resolved by SDS-PAGE (13%), transferred nitrocellulose membrane, and blotted with an anti-HA antibody (Roche Diagnostics). Protein levels of FLAG-DGK, FLAG-SLP-76, and HA-Ras in each sample before GST-Raf-RBD pull-down were determined by Western blot with anti-FLAG (Upstate Biotechnology, Inc.) and anti-HA antibodies.
Up-regulation of CD69 after TCR Stimulation-Jurkat cells were transfected with WT FLAG-DGK or the control expression vector as described above. Twenty-four hours after transfection, live cells were separated from dead cells via Ficoll-Paque TM Plus (Amersham Biosciences) gradient centrifugation (2500 rpm, 20 min, room temperature). Live cells were then washed three times with RPMI-10 and resuspended at 0.5 ϫ 10 6 cells/ml. 1.0 ml of cells was transferred to individual wells of a 24-well plate pre-coated with different dilutions of C305 and stimulated overnight. Stimulated cells were harvested, stained with allophycocyanin (APC)-conjugated anti-CD69 and phycoerythrin (PE)-conjugated anti-NGFR antibodies (BD PharMingen), and analyzed on a FACSCalibur (BD PharMingen) with Cellquest software.
Ca 2ϩ Influx-Calcium flux was measured by flow cytometry using a fluorescent calcium indicator. Briefly, 24 -30 h after transfection, Jurkat cells were labeled with Indo-1 (Molecular Probes, Eugene, OR) and PE-conjugated anti-NGFR antibody to distinguish transfected from nontransfected cells. After establishing the base line of the ratio of FL5 to FL4, cells were stimulated with C305, and calcium flux was meas-ured on a BD-LSR (BD PharMingen), and analyzed based on the change of ratio of FL5 to FL4.
Protein Lysates from Murine Primary Cells-Thymocytes, splenocytes, and lymph node cells were harvested from 8-week-old C57B/6 mice according to standard procedures. Single cell suspensions (5 ϫ 10 7 cells/ml) in PBS with 5% FBS and 10% normal rat serum were stained with various combinations of PE-or APC-conjugated antibodies (BD PharMingen) at 4°C for 45 min and washed three times with PBS, 2% FBS by centrifugation. Different populations of cells were sorted using a MoFlo (Cytomation, Fort Collins, CO) and lysed in 1% Nonidet P-40 lysis buffer with protease inhibitors. DGK expression in individual populations was determined by Western blotting of lysates from 3 ϫ 10 6 cells with anti-DGK antibody (42). (41,42) have demonstrated expression of DGK transcripts in the thymus and brain as well as two isoforms of DGK in skeletal muscle. The smaller transcript includes the first coding exon spliced to the third exon, whereas the larger form is generated by eliminating the first coding exon but utilizing the second (49). We investigated expression of DGK protein in the lymphoid system by Western analysis using a DGK-specific antibody (Fig. 1A), and we found that two isoforms of DGK protein are expressed in the thymus, spleen, and lymph node (apparent molecular masses of 115 and 130 kDa). Interestingly, in the thymus the 130-kDa form is expressed at a higher level, whereas in the spleen and lymph node the smaller (115 kDa) form predominates, suggesting developmental regulation of the two isoforms.

DGK Is Expressed at High Levels in Primary Murine T Cells-Previous reports
To analyze further the pattern of DGK expression in hematopoietic lineages, T cells of different developmental stages, B cells, and myeloid cells were purified from mouse spleen and thymus by fluorescence-activated cell sorting (FACS). Lysates were prepared from each population and examined for DGK expression by Western analysis. As shown in Fig. 1B, we found that DGK protein is preferentially expressed in T cells, with severalfold lower expression in CD11b-positive (myeloid) cells and B220-positive (Fig. 1B) cells. Within the T cell compartment, DGK is expressed throughout all developmental stages with both the 115-and 130-kDa isoforms easily detected. In addition, the two isoforms of DGK appear to be differentially regulated in the thymus with the 130-kDa form preferentially expressed in immature DN and DP thymocytes and the 115-kDa form in CD4 ϩ and CD8 ϩ single positive mature thymocytes. Furthermore, in the thymus and spleen, it appears that CD4 ϩ cells express more DGK than CD8 ϩ cells. Although found at a lower level, both the 115-and 130-kDa forms of DGK are expressed in B cells and myeloid cells with the 115-kDa species predominant. In myeloid cells, the 130-kDa form is barely detectable. Interestingly, another smaller band (Fig. 1B, arrowhead) which migrates slightly faster than the 115-kDa form exists in both B and myeloid cells but is not detectable in T cells, suggesting the possibility that an additional splice variant of DGK may be present in specific hematopoietic lineages. Alternatively, this could be because of differential post-translational modifications of the smaller form of DGK in these subsets of cells. Further experiments will distinguish these possibilities.
Inhibition of TCR-induced ERK Activation by DGK-Jurkat T cells have been used extensively as a model to examine the biology of TCR signaling. As expected from our studies of primary cells, DGK protein is expressed in Jurkat cells by Western analysis (data not shown). To study the role of DGK in modulating TCR signaling, we first tested whether overexpression of DGK affects TCR-induced activation of the Ras-ERK pathway by examining TCR-inducible ERK phosphorylation. The 115-kDa isoform of DGK was used in this study because previous experiments (50) have indicated that the 115-kDa form but not the 130-kDa form of DGK can inhibit RasGRP activation when overexpressed in HEK293, a human embryonic kidney cell line. Because transfection efficiency in Jurkat cells is not 100%, we made use of J14, a mutant variant of Jurkat which lacks the adapter protein, SLP-76. Due to absence of functional SLP-76, stimulation of the TCR on J14 cells is deficient in activation of the Ras-ERK and calcium signaling pathways. Reconstitution of SLP-76 expression by gene transfer rescues the TCR signaling defect (26,27). Thus, when using J14 cells, only cells successfully transfected with SLP-76 (and cotransfected with other genes of interest) will signal efficiently after TCR ligation, whereas nontransfected cells should not contribute to the TCR-mediated response.
J14 cells were transfected with a control vector, FLAGtagged SLP-76, or SLP-76 plus FLAG-tagged DGK. Transfectants were left unstimulated, stimulated via the TCR for various times, or stimulated with PMA. ERK activation was analyzed by Western blot using an antibody specific for phosphorylation of threonine and tyrosine residues (Thr 202 /Tyr 204 ) (51), markers of active ERK. As shown in Fig. 2A, we observed a low but reproducible degree of TCR-induced ERK phosphorylation in J14 cells transfected with the control vector (similar to that seen in non-transfected cells, data not shown). When SLP-76 was reintroduced into the cells, TCR-induced ERK phosphorylation was markedly increased. However, when DGK was cotransfected with SLP-76 into J14 cells, ERK phosphorylation was decreased to a level approximating that seen in cells transfected with the control vector, suggesting an inhibitory effect of DGK on TCR-induced ERK activation. Stripping the blot and reprobing with anti-ERK documented that the decreased ERK phosphorylation caused by DGK overexpression was not due to a decrease in ERK protein expression (Fig. 2B). Similar levels of SLP-76 expression were detected in both transfectants (Fig. 2C), indicating that the failure of ERK to be inducibly phosphorylated in the cells overexpressing DGK was not due to poor expression of SLP-76.
Although J14 cells have a severe defect in TCR-mediated signaling resulting in impaired second messenger production, the TCR is not completely uncoupled from its signaling machinery (note the small amount of TCR-inducible ERK activation in Fig. 2A consistent with results from reference 26). Because this low level signaling could potentially confound our results, we continued our studies using vectors that express epitope-tagged signaling molecules. This allowed us to study ectopically expressed proteins present only in those cells that were successfully transfected. As a first experiment, J14 cells were cotransfected with SLP-76 along with Myc-tagged ERK plus DGK or control vector. As for the experiments described in Fig. 2, transfectants were left unstimulated or stimulated via the TCR for various times or with PMA. Ectopically expressed ERK was immunoprecipitated from cell lysates with an anti-Myc antibody and its activation analyzed by Western blot with the anti-phospho-ERK antibody. Concordant with the results examining endogenous ERK, TCR-induced Myc-ERK phosphorylation is greatly decreased when DGK is overexpressed compared with vector control (Fig. 3A). Equal levels of Myc-ERK were observed in all lanes after stripping and reprobing with anti-Myc antibody (Fig. 3B). In addition, similar levels of SLP-76 were detected in cell lysates from all cells (Fig.  3C). Importantly, a similar level of ERK activation was observed when cells were stimulated with PMA, a phorbol ester derivative that functions as a homologue of DAG but is not a substrate for DGK activity.
Inhibition of TCR-induced Ras Activation by DGK-Downstream of the TCR but upstream of ERK stimulation is the activation of Ras. To evaluate the effects of DGK overexpression on TCR-induced Ras stimulation, FLAG-tagged SLP-76 and HA-tagged Ras were cotransfected into J14 cells along with FLAG-tagged DGK or control vector. Transfected cells were left unstimulated or stimulated via their TCR for various times. Lysates were prepared and analyzed for Ras activation using a GST-Raf-RBD fusion protein pull-down assay based on high affinity of the active GTP-bound form of Ras for the Ras binding domain (RBD) of Raf (52,53). The amount of activated Ras was then detected by Western blot with an anti-HA antibody. As shown in Fig. 4A, when control vector was transfected with SLP-76 and HA-Ras, Ras activation was observed at 1 min following TCR stimulation and reached a maximum at 5 min. However, when DGK was cotransfected with SLP-76, the TCR-induced increase of GTP-bound Ras was virtually ablated. Equal levels of FLAG-SLP-76 and HA-Ras were detected in lysates from both transfections (Fig. 4, B and C), ruling out the possibility that the inhibitory effect is due to differential expression of these proteins. Consistent with the ERK activation experiments, PMA overcomes the inhibitory effect of DGK on TCR-induced Ras activation. Collectively, these results indicate DGK inhibits TCR-induced activation of the Ras-ERK pathway.
Inhibition of TCR-induced AP-1 Activity and CD69 Expression by DGK-TCR-induced activation of the Ras-ERK pathway leads to activation of the AP-1 transcription factor thereby promoting the transcription of genes encoding cytokines and other molecules important for T cell function (54). When DGK was cotransfected into Jurkat cells along with an AP-1 luciferase reporter (46), TCR-induced luciferase activity in lysates from transfected cells was greatly inhibited compared with luciferase activity in cells transfected with the control vector. Moreover, although ligation of the costimulatory CD28 synergizes with the TCR to induce more potent AP-1 activity, engagement of this coreceptor failed to overcome the inhibitory effect of DGK (Fig. 5). Similar results were observed when using the NFAT-AP-1-luciferase reporter, in which activation of both NFAT and AP-1 transcription factors is required to activate transcription of the luciferase reporter (data not shown).
CD69 is a surface antigen up-regulated following TCR stimulation in an AP-1-dependent manner (55,56). We therefore used CD69 expression as a marker for more distal events in T cell activation which could potentially be modulated by DGK. For these experiments, we made use of an expression vector containing an internal ribosomal entry site, thus allowing DGK and a truncated form of the NGFR to be translated from the same bi-cistronic transcript. Analysis for NGFR expression by cell-surface staining demonstrates that 8 -15% of Jurkat cells are successfully transfected using our protocol (Fig. 6A). For the experiments shown in Fig. 6, B and C, cells were transfected with the NGFR expressing control vector or a vector encoding both NGFR and DGK. Twenty-four hours later, cells were left unstimulated or were treated with various concentrations of anti-TCR antibody ranging from 1:1000 to 1:16000 dilutions of ascites for an additional 15 h. Up-regulation of CD69 expression was evaluated by flow cytometry. As can be appreciated from Fig. 5B, similar levels of TCR-induced CD69 expression are found on the non-transfected (NGFR Ϫ ) cells regardless of the plasmid used. As shown also (FACS plots in Fig. 6B and calculated values for the full dose range in Fig.  6C), the ratio of CD69 ϩ to CD69 Ϫ cells in the non-transfected and NGFR only transfected groups becomes greater as the strength of TCR stimulation increases. In contrast, when examining the NGFR ϩ (successfully transfected) cells, which received the DGK construct, there is a marked abrogation of TCR-induced CD69 expression. As expected, those cells brightest for NGFR (and presumably expressing the greatest amount of DGK) demonstrate the most marked inhibition of CD69 expression.

DGK Overexpression Does Not Affect TCR-induced Ca 2ϩ
Influx-To ensure that overexpression of DGK was modulating specific signaling pathways and not merely globally interfering with TCR-mediated activation events, we examined TCR-induced intracellular calcium flux, a response that is dependent on PLC␥1 activity and independent of DAG (5). Jurkat cells were transfected with vector encoding DGK/NGFR or vector encoding NGFR alone. Twenty-four hours later, the cells were labeled with the calcium-sensitive dye, Indo-1, and stained with anti-NGFR antibody. Cells were analyzed for intracellular free calcium concentration in real time by flow cytometry following engagement of the TCR. As shown in Fig. 7, TCR stimulation induces a similar calcium response in cells transfected with either the control vector or DGK. These data indicate that DGK selectively regulates DAG-dependent signaling pathways and does not cause a global nonspecific perturbation of TCR signaling.
Structural Requirements for the Inhibitory Effect of DGK on TCR Signaling-DGK contains two cysteine-rich (C1) domains at its N terminus, a kinase domain and four ankyrin repeats at its C terminus (Fig. 8A) (41,42). To determine the structural features of DGK important for its inhibitory effect on TCR signaling, mutant variants of DGK tagged with the FLAG epitope were cloned into the pEF-IRES-NGFR expression vector. In addition to wild type (WT) DGK, we generated a "kinase dead" (KD) variant containing a point mutation in the ATP-binding site (G355D) (44), a mutant lacking the N-terminal C1 domains by deleting amino acids 1-251 (⌬NT), and a mutant lacking the C-terminal ankyrin repeats by deleting amino acids 787-928 (⌬CT).
We first determined whether the kinase activity of DGK is required for its inhibitory effect on TCR signaling (Fig. 8B). Cells cotransfected with the AP-1 luciferase reporter plus WT or KD mutant DGK were analyzed for luciferase activity following engagement of the TCR in the absence or presence of costimulation with anti-CD28. Whereas WT DKG demonstrates a potent inhibitory effect on AP-1 function, no such inhibition is seen with the KD variant, even though high levels of the construct were expressed (Fig. 8B, inset). These data indicate that the kinase activity of DGK is essential for its inhibitory effect on TCR signaling. Consistent with this, PMA plus TCR stimulation can overcome the inhibition of AP-1 activation by DGK.
In separate experiments (Fig. 8C), we tested the ability of the ⌬NT and ⌬CT mutants to interfere with TCR signaling. Dele-tion of the C-terminal 142 amino acid of DGK has no obvious effect on the ability of the enzyme to inhibit AP-1 activation, indicating that, at least in this assay system, the ankyrin repeats are not essential for its function. In contrast, deletion of the N-terminal 252 amino acids of DGK results in a complete loss of the inhibitory effect of DGK, suggesting that the C1 domains are required for optimal DGK function.

DISCUSSION
In this report, we have shown that DGK is expressed in multiple lymphoid tissues, with highest expression in the thymus and mature T cell compartment. Interestingly, there appears to be differential expression of DGK isoforms based on stage of T cell development, an observation that we are currently pursuing further. In Jurkat T cells, we found that overexpression of DGK inhibits TCR-induced Ras and ERK activation. AP-1 activation, a downstream effector of the Ras-ERK signal pathway, is also inhibited, as is expression of the activation antigen, CD69. This inhibition requires enzymatic function of DGK as a kinase-inactive mutant does not affect AP-1 activation, suggesting that DGK interferes with TCRinduced activation of the Ras-ERK pathway through the phos-phorylation of its substrate DAG. This is consistent with our observation that PMA, which functions as DAG homologue but is not a substrate of DGKs (39,40), can overcome the inhibitory effect of DGK on Ras, ERK, and AP-1 activation. The inhibition of TCR signaling by DGK is selective to pathways downstream of DAG because TCR-induced Ca 2ϩ influx is not affected by overexpressed DGK. These data suggest that DGK-mediated regulation of DAG function is required for controlled signaling via the TCR.
Although previous experiments showed that overexpression of a kinase-inactive mutant of DGK can prolong TCR-induced Ras activation in Jurkat cells (50), in the current experiments we find no evidence of a dominant effect of this mutant on AP-1 activation. This may reflect differences in the sensitivity of these two assays measuring TCR function or possible differences in the levels of expression of transfected constructs achieved in the various studies. Future experiments will complement these current studies by generating genetically manipulated mice to modulate DGK expression and assess its importance in the regulation of TCR signaling in primary T cells and during T cell development.
The mechanism by which DGK interacts with and thereby regulates DAG following TCR engagement is not yet well understood. DGK has been shown to interact with several other proteins including ␥ 1 -syntrophin, the long form of the leptin receptor (Ob-Rb) and RasGRP (50,57,58). It is possible that DGK is directed to the correct subcellular location via these (or perhaps other) intermolecular interactions. Supporting this possibility are recent data demonstrating that in a glioblastoma cell line expressing high levels of DGK and RasGRP, both proteins colocalize in peripheral cellular extensions. Concordant with this observation is the finding that, when ectopically expressed in HEK293 cells, DGK can inhibit the ability of RasGRP to activate Ras (50). Although a direct interaction between DGK and RasGRP has yet to be confirmed in T cells, it is possible that this association is important in bringing DGK to its site of action. Precise mapping of the amino acid residues which mediate the interaction between DGK and RasGRP will be required to further investigate the functional importance of this inter-molecular interaction.
Our data demonstrate that the C1 domains of DGK are critical for it to function as a negative regulator of TCR signaling. This may be due to the requirement of these domains for optimal enzymatic activity and/or for the interaction between DGK and DAG. In this regard, the generation of DAG by TCR engagement may serve to recruit DGK to its substrate via the C1 domains. This possibility is supported by the observation that interactions between the C1 domains of PKCs and Ras-GRP with DAG are involved in the membrane localization of these proteins (31,32,59).
Although it has been clear for many years that engagement of the TCR stimulates multiple signaling cascades, most work has explored how activation of these signaling pathways is regulated. In the last few years, however, evidence has accumulated that negative regulation of TCR signaling is also tightly controlled. The importance of these inhibitory mechanisms for overall function of the immune system is underscored by a number of recent studies. These include evaluation of mice genetically deficient in putative negative regulators such as the lipid phosphatase Pten (10) and the ubiquitin ligase Cbl-b (60). Both Pten and Cbl-b null mice exhibit hyper-reponsive T cells leading to severe autoimmunity (11,12,61,62). The identification of DGK as a negative regulator of TCR signaling suggests a potential role for this enzyme in the maintenance of immune system homeostasis. By inactivating DAG, DGK may be responsible for down-regulating the Ras-ERK and other effector pathways downstream of DAG, thereby terminating an appropriate immune response or perhaps playing a role in the maintenance of T cell tolerance to self-antigen.
In addition to the RasGRP-Ras-ERK pathway, DAG regulates many other signaling pathways after TCR engagement. The activation and membrane localization of PKC, actin rearrangement, integrin-mediated adhesion, and cell surface downregulation of TCR expression are among the many responses to TCR engagement or PMA treatment. Future experiments will investigate whether and how these signaling events may be regulated DGK. As one example, PKC has been shown to be required for activation of NFB in mature T cells (36,37). Additionally, it is known that the association of DAG with the C1 domain of PKC is important for PKC activation (33,63). Thus, overexpression of DGK with consequent conversion of DAG to phosphatidic acid could potentially blunt PKC activation leading to decreased nuclear translocation of NFB in the transfected Jurkat cells. Another mechanism by which overexpressed DGK could impact T cell activation events could involve modulation of TCR expression. Others (64 -66) have shown that stimulation with PMA or direct activation of PKC results in TCR down-regulation. Although we have not found a significant impact on TCR levels in the DGK-overexpressing cells, it is possible that dysregulation of this signaling pathway could alter the subcellular localization of the TCR or its associated signaling complex.
It is important to note that in addition to DGK, other isoforms of the DGK family are expressed in T cells (67,68). Previous studies have suggested that DGK␣ may be involved in IL-2 receptor signaling. In these reports, it was shown that IL-2 stimulation induces rapid relocalization of DGK␣ in an IL-2dependent cell line. Furthermore, inhibition of DGK␣ activity has been reported to interfere with IL-2-induced phosphatidic acid production and correlates with cell cycle arrest at the G 1 phase (69,70). More recently, it was reported that TCR/CD28 stimulation can induce translocation of DGK␣ to the plasma membrane in Jurkat T cells expressing a GFP-DGK␣ fusion protein (71). However, in contrast to our results with DGK, overexpression of wild type DGK␣ did not impact TCR-induced CD69 expression, although overexpression of a constitutively active form revealed some inhibition of this marker of T cell activation (71), and a kinase dead mutant of DGK␣ enhanced ERK activation following TCR engagement (72). To date, there are no data available addressing the effect of wild type DGK␣ on TCR-induced activation of the Ras-ERK pathway. Thus, it remains to be determined whether there is DGK isoform specificity in the modulation of TCR signaling events.
The work described in this report, together with previous studies from a number of other laboratories, has demonstrated that members of the DGK family may serve critical roles as physiological regulators of several signaling pathways in T cells. The presence of multiple isoforms of DGKs in T cells, which may work in distinct, overlapping, or redundant pathways, will certainly add to the complexity as efforts are made to understand the exact role of each in T cell development and function. Clues will come from additional studies of endogenous DGK isoforms examining basal and induced enzymatic activity and subcellular localization in resting and stimulated cells. Additionally, development of genetically altered mice where expression of one or several DKG isoforms is eliminated should provide important insights into how these proteins regulate different aspects of immune cell function.