Phosphorylation-dependent Regulation of Phospholipase D2 by Protein Kinase C (cid:1) in Rat Pheochromocytoma PC12 Cells*

Many studies have shown that protein kinase C (PKC) is an important physiological regulator of phospholipase D (PLD). However, the role of PKC in agonist- induced PLD activation has been mainly investigated with a focus on the PLD1, which is one of the two PLD isoenzymes (PLD1 and PLD2) cloned to date. Since the expression of PLD2 significantly enhanced phorbol 12- myristate 13-acetate (PMA)- or bradykinin-induced PLD activity in rat pheochromocytoma PC12 cells, we inves- tigated the regulatory mechanism of PLD2 in PC12 cells. Two different PKC inhibitors, GF109203X and Ro-31– 8220, completely blocked PMA-induced PLD2 activation. In addition, specific inhibition of PKC (cid:1) by rottlerin pre- vented PLD2 activation in PMA-stimulated PC12 cells. Concomitant with PLD2 activation, results

It has been suggested that PKC is an important regulator of PLD. PKC inhibitors and the down-regulation of PKC expression blocked PMA-induced PLD activation (24). Various agonists such as platelet-derived growth factor, epidermal growth factor, bradykinin, angiotensin II, thrombin, and carbachol activate PLD in many cell types, whereas PKC mediates agonist-induced PLD activation (24 -27). PLD1 was directly associated with, and activated by PKC␣ in the presence of PMA in NIH 3T3 cells and in PLD1-transfected COS-7 cells (28). In previous studies, we described the phosphorylation-dependent activation of PLD1 by PKC␣ in 3Y1 fibroblast cells and in PLD1-transfected COS-7 cells (29) and PKC␣-mediated PLD1 phosphorylation with the activation occurring in caveolin-enriched microdomains within the plasma membrane (30). However, the details of the regulation of PLD2 by PKC are not well understood, although recently, it was reported that PLD2 could be activated in different cell types in response to PMA (31)(32)(33).
The aim of this study was to examine the involvement of PKC in PMA-induced PLD2 stimulation in PC12 cells and to investigate whether PKC-mediated PLD2 activation is phosphorylation-dependent. In this report, we show that PLD2 can be activated in PC12 cells by PMA treatment. Moreover, for the first time, we were able to demonstrate that PLD2 becomes phosphorylated in response to PMA treatment and that PKC␦ mediates the phosphorylation-dependent activation of PLD2 in PC12 cells.
Infection with Recombinant PKC Adenovirus-Adenovirus expression vectors for wild type and the dominant-negative type of PKC␦ have been described previously (34,35). Subconfluent PC12 cells in 6-well plates were infected with wild type (WT-PKC␦ AdV) or dominant negative PKC␦ adenovirus (DN-PKC␦ AdV) for 12 h at different multiplicities of infection (m.o.i.) in 0.1% serum containing DMEM. After removing the virus, cells were cultured for an additional 24 h in DMEM supplemented with 10% heat-inactivated equine serum and 5% fetal calf serum. Cells were then incubated for 12 h in 0.5% fetal calf serum containing DMEM.
In Vivo Assay of PLD-PLD activity was assayed by measuring the formation of phosphatidylbutanol (PBtOH), the product of PLD-mediated transphosphatidylation, in the presence of 1-butanol as previously described (4). PC12 cells were subcultured in 6-well tissue culture plates at 1 ϫ 10 6 cells/well in the presence or absence of 0.5 g/ml tetracycline. The cells were loaded with [ 3 H]myristic acid (3 Ci/ml) for 3 h and then treated with 100 nM PMA in the presence of 0.4% 1-butanol (v/v) for the indicated times at 37°C. To inhibit PKC, 5 M GF109203X, 5 M Ro-31-8220, 0.5 M Go6976, or 15 M rottlerin were administered for 15 min before the PMA treatment. After the incubation, the medium was aspirated, and 0.4 ml of ice-cold methanol was added to the cells. The cell debris was scraped into an Eppendorf tube, and chloroform and 0.1 N HCl were added, resulting in a final chloroform, methanol, 0.1 N HCl ratio of 1/1/1 (v/v/v). After vortexing, the tubes were centrifuged at 15,000 ϫ g for 1 min, and the organic phase was harvested, dried, and spotted onto a Silica Gel 60 TLC plate which was then developed with chloroform/methanol/acetic acid (9/1/1, v/v/v). The amounts of labeled PBtOH and total lipids were determined using a Fuji BAS-2000 image analyzer (Tokyo, Japan).
Immunoblot Analysis and Immunoprecipitation-To detect PLD2 and PKC, the proteins were separated by SDS-PAGE on 8% acrylamide gels, transferred to nitrocellulose membranes, and blotted with anti-PLD (16) and anti-PKC antibodies.
In Vitro Binding Assay-Direct interaction between PLD2 and binding protein was analyzed by the method, as described previously (19) with slight modifications. Recombinant PLD2 was expressed in baculovirus-infected Sf9 cells and purified, as described previously (16). Purified PLD2 was incubated with anti-PLD antibody immobilized on protein A-Sepharose in incubation buffer (20 mM Tris/HCl, pH 7.5, 1 mM MgCl 2 , 10 mM NaCl, 1% Triton X-100, and 1% sodium cholate). After incubating the immune complex (100 ng of purified PLD2, coupled to immunoaffinity resin) with 50 ng of PKC␦ in the absence or presence of 100 nM PMA at 37°C under in vitro binding conditions (30 mM Tris/ HCl, pH 7.0, 6 mM MgCl 2 , 0.25 mM EGTA, 0.4 mM CaCl 2 , and 0.1% Triton X-100), the samples were pelleted and washed 5 times with washing buffer (20 mM Tris, pH 7.5, 150 mM NaCl, 1 mM EDTA, 1 mM EGTA, and 1% Triton X-100). The existence of PKC␦ in the precipitate was determined by immunoblotting with anti-PKC␦ antibody.
Phosphorylation of PLD2 in PC12 Cells after Treatment with PMA-Initially, 1 ϫ 10 7 PC12 Tet-Off cells/100-mm dish were incubated with 5 mCi of [ 32 P]orthophosphate in 5 ml of phosphate-free DMEM for 5 h at 37°C. After the cells had been washed twice with serum-free medium, they were treated with 100 nM PMA for 5 min. To inhibit PKC, the cells were pretreated with 5 M Ro-31-8220, 0.5 M Go6976, or 15 M rottlerin for 15 min followed by 100 nM PMA for 5 min. The cells were then washed with ice-cold hypotonic buffer containing phosphatase inhibitors (30 mM NaF, 1 mM Na 3 VO 4 , and 30 mM Na 4 O 7 P 2 ) and lysed in 1 ml of hypotonic buffer containing 1% Triton X-100, 1% sodium cholate, and phosphatase inhibitors (30 mM NaF, 1 mM Na 3 VO 4 , and 30 mM Na 4 O 7 P 2 ). After centrifugation (15,000 ϫ g for 15 min), equal amounts of soluble extract were incubated with 2 g of anti-PLD antibody and 30 l of immobilized protein A. The immunoprecipitated proteins were then separated by 8% SDS-PAGE, transferred to a nitrocellulose membrane, and exposed to a photographic film for autoradiography.
Phosphopeptide and Phosphoamino Acid Analysis-Two-dimensional phosphopeptide mapping was performed as described previously, with slight modification (29). Immunoprecipitates of in vivo or in vitro 32 P-labeled PLD2 were resolved by 8% SDS-PAGE and then transferred to a nitrocellulose membrane. The PLD2 was localized by autoradiography, excised from the membrane, and washed with deionized water. The piece of membrane corresponding to the PLD2 band was then incubated with 0.5% polyvinylpyrrolidone in 100 mM acetic acid at 37°C for 30 min, washed with water, and then with fresh 50 mM ammonium bicarbonate. Tryptic digestion was achieved by incubating the piece of membrane in 150 l of 50 mM ammonium bicarbonate with 10 g TPCK-treated trypsin at 37°C for 6 h, followed by the addition of another 10 g of TPCK-treated trypsin and a second 6-h incubation at 37°C. The tryptic digest was lyophilized, oxidized with performic acid, re-lyophilized, and then dissolved in 20 l of pH 1.9 buffer (glacial acetic acid, formic acid (88%), H 2 O; 78:25:8973 v/v/v). Phosphotryptic peptides were separated on cellulose TLC plates by electrophoresis at pH 1.9 in the first dimension. In the second dimension TLC was performed in n-butanol:pyridine:glacial acetic acid:H 2 O (75:50:15: For phosphoamino acid analysis, the piece of membrane corresponding to the PLD2 band was dissolved in 6 N HCl and hydrolyzed for 1 h at 110°C. The HCl was then removed by lyophilization, and the pellet dissolved in pH 3.5 buffer (glacial acetic acid:pyridine:H 2 O; 50:5:945, v/v/v). A mixture of phosphoserine, phosphothreonine, and phosphotyrosine (1 g of each) was then added. The 32 P-labeled phosphoamino acids were separated by electrophoresis on 20 ϫ 20-cm cellulose TLC plates. After electrophoresis the plates were dried, the phosphoamino acids were visualized by staining with 0.2% (w/v) ninhydrin in acetone, and the 32 P-labeled amino acids identified by autoradiography.
Immunocytochemical Analysis-Immunocytochemical analysis was performed as described previously (30). Briefly, PC12 cells were grown on coverslips. After treatment with 100 nM PMA, cells were fixed in 3.7% (w/v) paraformaldehyde for 10 min at 37°C, washed with phosphate-buffered saline, and then incubated in blocking buffer (1% goat serum in phosphate-buffered saline containing 0.2% Triton X-100) at 4°C for 1 h. Subsequently cells were incubated with 2 g/ml anti-PLD antibody and 2 g/ml anti-PKC␦ antibody overnight at 4°C. After washing with phosphate-buffered saline containing 0.05% Triton X-100, cells were incubated with secondary antibodies: TRITC-conjugated anti-mouse antibody and fluorescein isothiocyanate-conjugated anti-rabbit antibody. After washing with phosphate-buffered saline containing 0.05% Triton X-100, the slides were mounted and examined under a fluorescence microscope (Nikon, Inc., Melville, NY).

PMA-induced PLD2 Activation in PC12 Cells-It has been
reported that PMA, an activator of PKC, stimulated PLD activity in PC12 cells (37) and that PLD2 was prominently expressed in PC12 cells (33). We confirmed the presence of PLD2 in PC12 cells using anti-PLD antibody (data not shown). Therefore, to investigate whether PLD2 is activated in PC12 cells upon PMA treatment, we used a PLD2-inducible PC12 Tet-Off cell line (4). In this case, recombinant human PLD2 protein is expressed under the control of an inducible tetracycline-regulated promoter. The removal of tetracycline from the culture medium thus results in increased expression of PLD2, which was verified by Western blot analysis using anti-PLD antibody, as shown in Fig. 1C. Treatment of the cells with 100 nM PMA stimulated endogenous PLD activity (ϩTet) in PC12 cells, while the induction of heterologous PLD2 expression (ϪTet) led to a further increase in PMA-induced PLD activity, indicating that the PLD2 activity had increased in response to PMA. The effect of PMA on the PLD2 activity was both time-and concentration-dependent. Activation of PLD2 by 100 nM PMA occurred within 5 min, and the PLD activity was saturated at 100 nM PMA in the presence or absence of tetracycline (Fig. 1, A and B). These data show that PLD2 can be activated by PMA in PC12 cells in time-and concentration-dependent manner. To confirm that PLD1 is also activated by PMA treatment, we created an inducible PC12 Tet-Off cell line, which overexpressed PLD1 in the absence of tetracycline. On checking the PMA-induced PLD1 activity in these PC12 cells, we found that the activity of PLD1 was also enhanced by the PMA treatment (data not shown).
PMA-induced PLD2 Activation Is Mediated by PKC-Next, we determined whether the PMA-induced PLD2 activation in PC12 cells was PKC-dependent. We used the PKC inhibitors GF109203X and Ro-31-8220 to detect any effect on PLD activity after PMA stimulation. As shown in Fig. 2, in the PLD2overexpressing PC12 cells (ϪTet) basal activity had increased, and furthermore, maximal PMA-induced PLD activation had also significantly increased compared with the control PC12 cells (ϩTet), which do not overexpress PLD2. This indicated that the overexpressed PLD2 was responsive to PMA treatment. Inhibition of PKC by pretreating of the cells with GF109203X and Ro-31-8220, specific inhibitors of PKC, abrogated the PMA-induced PLD activity in the presence or absence of tetracycline. It has been reported that PC12 cells express PKC␣, ␤, ␥, ␦, , and (43), and that PMA activates all PKC isozymes except . To determine which isozyme of PKC was involved in the PMA-induced PLD2 activation in PC12 cells, we used two different PKC isozyme-specific inhibitors, classical PKC isozymes-specific Go6976 and PKC␦-specific rottlerin. As shown in Fig. 2, Go6976 had a slight inhibitory effect on the PMA-induced PLD2 activity (ϳ14%), while rottlerin strongly inhibited the PMA-induced PLD2 activity (ϳ69%), suggesting that PKC␦ might be mainly involved in the PMAinduced activation of PLD2 in PC12 cells.
PLD2 Becomes Phosphorylated in Response to PMA in PC12 Cells-To examine whether PLD2 is phosphorylated upon PMA treatment, we looked for PMA-dependent phosphoryla- tion of PLD2 in PC12 cells. Lysates from [ 32 P]orthophosphate-loaded control PC12 cells (ϩTet and ϪTet) and PC12 cells (ϩTet and ϪTet) treated with 100 nM PMA for 5 min were immunoprecipitated with anti-PLD antibody. As seen in Fig. 3A, in the PLD2-overexpressing PC12 (Ϫ Tet) cells, PLD2 became basally phosphorylated during the initial incubation with [ 32 P]orthophosphate, but more PLD2 became phosphorylated after PMA treatment. However, in (ϩTet) cells, the amount of endogenous PLD2 was negligible (Fig. 1C), we could not detect phosphorylation of endogenous PLD2 in (ϩTet) cells (data not shown). To determine whether PKC mediated the PMA-dependent PLD2 phosphorylation in PC12 cells, we in-hibited PKC activity by pretreating the cells with Ro-31-8220. We found that the intensity of the PMA-induced PLD2 phosphorylation was reduced to basal level by the Ro-31-8220 pretreatment, suggesting that PKC is indeed involved in the PMA-induced PLD2 phosphorylation. Next, we further checked which isozyme of PKC was involved in the PMA-induced PLD2 phosphorylation using the PKC isozyme-specific inhibitors Go6976 and rottlerin. As shown in Fig. 3A, Go6976 slightly attenuated, but rottlerin significantly inhibited the PMA-induced PLD2 phosphorylation, suggesting that the PMA-induced PLD2 phosphorylation may be mainly mediated by PKC␦. Direct analysis of PLD2 phosphoamino acids proved Protein extracts (2 mg) were then prepared with lysis buffer (10 mM Tris/HCl, pH 7.5, 1 mM EDTA, 0.5 mM EGTA, 10 mM NaCl, 1% Triton X-100, 1% sodium cholate, 0.5 mM phenylmethylsulfonyl fluoride, 1 g/ml leupeptin, 5 g/ml aprotinin, 30 mM NaF, 1 mM Na 3 VO 4 , and 30 mM Na 4 O 7 P 2 ) and immunoprecipitated with anti-PLD antibody. The immunoprecipitated proteins were separated by 8% SDS-PAGE, transferred to a nitrocellulose membrane, and exposed to a photographic film for autoradiography (upper panel). The amount of PLD2 bound to the immune complex was analyzed by immunoblotting with anti-PLD antibody (lower panel). B, the band of phosphorylated PLD2 described in A was excised and subjected to phosphoamino acid analysis. 32 P-Labeled PLD2 was hydrolyzed with 6 N HCl, and the phosphoamino acids were resolved as described under "Experimental Procedures." The positions of phosphoserine (P-Ser), phosphothreonine (P-Thr), and phosphotyrosine (P-Tyr) are indicated by arrows. C, the phosphorylated PLD2 was excised, digested with trypsin, and the tryptic digest was subjected to two-dimensional phosphopeptide mapping. D, PC12 cells were loaded with [ 32 P]orthophosphate (1 mCi/ml) for 5 h, pretreated with 15 M rottlerin or control Me 2 SO for 15 min, and treated with 10 nM bradykinin for 1 min. As a control, PC12 cells were treated with 100 nM PMA for 5 min. Cell lysates were immunoprecipitated with anti-PLD antibody, the resulting precipitates were subjected to SDS-PAGE, and the autoradiography was visualized or quantified using anti-PLD antibody. The data are representative of three separate experiments. that serine and threonine residues were phosphorylated. Pretreatment of rottlerin greatly reduced the PMA-induced serine/ threonine phosphorylation of PLD2 (Fig. 3B). Next, we eluted 32 P-labeled PLD2 species, digested it with trypsin, and subjected it to two-dimensional phosphopeptide mapping (Fig. 3C). Multiple phosphopeptides were resolved from PLD2 before and after PMA treatment. Seventeen phosphopeptides (1-17) were either enhanced or created by the PMA treatment and pretreatment with rottlerin significantly reduced PMA-induced phosphopeptides, suggesting that 17 phosphopeptides from PLD2 after PMA treatment were the exclusive result of PKC␦ kinase activity. We previously demonstrated that bradykinin activates PLD2 via PKC␦ (44). Therefore, we also examined whether PLD2 is also phosphorylated by bradykinin treatment. As shown in Fig. 3D, PLD2 became phosphorylated after treatment of 10 nM bradykinin for 1 min and the intensity of the bradykinin-induced PLD2 phosphorylation was reduced to basal level by the pretreatment of rottlerin, suggesting that PKC␦ mediates PLD2 activation and phosphorylation in PC12 cells with bradykinin as well as PMA. (Fig. 4A). The endogenous level of PKC␣ protein was unchanged by infection with WT-PKC␦ or DN-PKC␦ adenovirus. The expression of kinase-deficient-, DN-PKC␦ in PC12 cells inhibited the basal and PMA-induced phosphorylation of PLD2, as assessed by autoradiogram. Infection of PC12 cells with DN-PKC␦ adenovirus also inhibited basal and PMA-induced activation of PLD2 in the presence or absence of tetracycline (Fig. 4C). Interestingly, the overexpression of WT-PKC␦ further enhanced PMA-induced PLD2 phosphorylation, and increased basal and PMA-induced PLD2 activities (Fig. 4, B and C). These results strongly suggest that the activity of PLD2 is regulated by its phosphorylation status as mediated by PKC␦.

Effect of Wild Type or Kinase-deficient Mutant of PKC␦ on PMA-induced PLD2 Activation and Phosphorylation-To further confirm that the PKC␦ isoform is involved in PMA-induced PLD2 activation in PC12 cells and that PLD2 is regulated in a phosphorylation-dependent manner, wild type (WT) or the dominant negative (DN) type of PKC␦ was expressed using adenovirus expression vector. Infection of PC12 cells with WT-PKC␦ or DN-PKC␦ adenovirus resulted in dose-dependent increases in the amounts of PKC␦, as assessed by immunoblot analysis. The amount of PKC␦ proteins in cells infected at an multiplicity of infection (pfu/cell) of 100 was about 20 times that of endogenous PKC␦
PMA-induced PLD2 Phosphorylation Is Directly Mediated by PKC␦-To examine whether PKC␦ is directly involved in PMAinduced PLD2 phosphorylation, we looked for an in vivo and in vitro association between PKC␦ and PLD2. As shown in Fig.  5A, PKC␦ and PLD2 could be co-immunoprecipitated from PLD2-overexpressing PC12 cells (ϪTet) treated with 100 nM PMA for 5 min, but could not be co-immunoprecipitated in the absence of PMA, which suggests that the association of PLD2 with PKC␦ in PC12 cells is PMA-dependent. Since we could not exclude the possibility that the association between PLD2 and PKC␦ was indirect, we examined the directness of the PMAdependent association between PLD2 and PKC␦ in vitro. Purified PLD2 was bound to an anti-PLD antibody-coupled protein A-Sepharose resin, and the immune complex was then incubated with PKC␦ in the presence or absence of PMA. As shown in Fig. 5B, PKC␦ did not interact with PLD2 in the absence of PMA. On the other hand, in the presence of PMA the association between PLD2 and PKC␦ was greatly enhanced. This indicated that PLD2 directly associates with PKC␦ in vitro in a PMA-dependent manner. We also investigated whether PLD2 is directly phosphorylated by PKC␦ in a PMA-dependent manner in vitro. Under in vitro binding assay conditions, [␥-32 P]ATP was added to the PLD2 immune complexes, and found that PLD2 was negligibly phosphorylated in the absence of PMA. However, in the presence of PKC␦ and PMA, PKC␦- dependent PLD2 phosphorylation and PKC␦ autophosphorylation were evident (Fig. 5C), indicating that PLD2 can become directly associated with and phosphorylated by PKC␦ in vitro in a PMA-dependent manner.
Although PLD2 may be directly phosphorylated by PKC␦ in vitro, we could not exclude the possibility that PLD2 might be indirectly regulated by PKC␦ in PC12 cells. To examine whether PKC␦ directly mediated PMA-induced PLD2 phospho-rylation in PC12 cells, the two-dimensional tryptic phosphopeptide map of PLD2 phosphorylated by PKC␦ in vitro was compared with that of PLD2 phosphorylated in vivo. As seen in Figs. 3C and 6, 17 distinct phosphopeptides were generated from the in vivo phosphorylated PLD2 (designated as 1-17) and 22 phosphopeptides were generated from the in vitro phosphorylated PLD2 (designated as a-v). Among these, only seven phosphopeptides overlapped (1/a, 2/b, 7/h, 8/i, 9/f, 13/m, and 14/l). Therefore, our results indicate that these seven common phosphopeptides are the direct result of PKC␦ phosphorylation in PC12 cells.
PMA-induced Co-localization of PLD2 and PKC␦-PLD activities have been detected from several region of the cell, including the plasma membrane (38,39), nucleus (40), Golgi (41), and endoplasmic reticulum (42). Therefore, we examined the subcellular localization of PLD2 by immunocytochemical analysis in PC12 cells. As shown in Fig. 7, we found that PLD2 localizes primarily to the plasma membrane in PLD2-overexpressing PC12 cells (ϪTet). In the presence of tetracycline (ϩTet), PLD2 was not overexpressed (Fig. 1C) and not stained in PC12 cells (Fig. 7), which indicated that the localization is due to PLD2 itself. The subcellular localization of PLD2 was not changed by PMA treatment in PC12 cells. The localization of endogenous PKC␦ was also monitored in PC12 cells by immunocytochemical analysis. In the presence and absence of tetracycline (ϩTet and ϪTet), endogenous PKC␦ was stained predominantly in the cytoplasm in quiescent PC12 cells, and not merged with PLD2. However, after treatment with PMA, PKC␦ translocated from the cytosol to the plasma membrane, and co-localized with PLD2 in PLD2-overexpressing PC12 cells (ϪTet). These results suggest that phosphorylation and activation of PLD2 by PKC␦ occurs in the plasma membrane in PC12 cells.

FIG. 5. PMA-dependent association between PLD2 and PKC␦ in vivo and in vitro.
A, in the absence of tetracycline (ϪTet), PC12 cells were treated with 100 nM PMA for 5 min and then disrupted in cold lysis buffer (20 mM Tris/HCl, pH 7.5, 1 mM MgCl 2 , 150 mM NaCl, 0.5 mM phenylmethylsulfonyl fluoride, 1 g/ml leupeptin, and 5 g/ml aprotinin) by sonication. The membrane fraction was solubilized with lysis buffer containing 1% cholic acid. The extract (1 mg) was immunoprecipitated using control IgG or anti-PLD antibody and then washed three times with washing buffer (20 mM Tris/HCl, pH 7.5, 1 mM MgCl 2 , 150 mM NaCl, and 0.1% Triton X-100). The immune complex was subjected to SDS-PAGE and immunoblotted with anti-PLD antibody and anti-PKC␦ antibody. B, after incubating the immune complex (PLD2, coupled to immunoaffinity resin) with 50 ng of PKC␦ in the absence or presence of 100 nM PMA at 37°C, the samples were pelleted. The existence of PKC␦ in the precipitate was determined by immunoblotting with anti-PKC␦ antibody, and the amount of PLD bound to the immune complex was analyzed by immunoblotting with anti-PLD antibody. The results shown are representative of three separate experiments. C, immune complex (PLD2, coupled to immunoaffinity resin) and 50 ng of PKC␦ were incubated in phosphorylation buffer (30 mM Tris/HCl, pH 7.0, 6 mM MgCl 2 , 0.25 mM EGTA, 0.4 mM CaCl 2 , 0.1% Triton X-100, 0.12 mM ATP, 2 Ci of [␥-32 P]ATP (3,000 Ci/mmol)) in the presence of 100 nM PMA for 15 min. The reaction mixture was then electrophoresed through 8% SDS-PAGE, transferred to a nitrocellulose membrane, and exposed to photographic film for autoradiography. The results shown are representative of three separate experiments.

FIG. 6. Comparison of the two-dimensional tryptic phosphopeptide maps of PLD2 phosphorylated in vivo and in vitro.
The phosphorylated PLD2 obtained from PC12 cells by 100 nM PMA treatment for 5 min and in vitro phosphorylated PLD2 were digested with TPCK-trypsin. Each tryptic digest and the mixture of the tryptic digests were subjected to two-dimensional phosphopeptide mapping.
The arrows indicate common overlapping phosphopeptides. The data shown are representative of three separate experiments.

DISCUSSION
Although PKC is an important regulator of PLD, the regulation mechanism in vivo remains controversial. In the case of PLD1, several studies have reported the phosphorylation-independent regulation of PLD1 by PKC. Purified PLD1 could be activated by PKC␣ and PMA further increased PKC␣-stimulated PLD1 activity in reconstitution assays (6,28). The regulatory domain of PKC␣ was found to be a critical region for PLD1 activation, suggesting that PKC␣ modulates PLD1 activity by physical interaction, rather than by a phosphorylationdependent process (15,51). In neutrophil and HL60 cells, PMAdependent activation of PLD required ATP in a cell-free system (52). In addition, it has been reported that Ro-31-8220 or staurosporin, competitive inhibitors of the ATP-binding domain of PKC, inhibited PMA-induced PLD activation, suggesting that PLD regulation is phosphorylation-dependent. In previous studies, we described the phosphorylation-dependent activation of PLD1 by PKC␣ in 3Y1 fibroblast cells and in PLD1-transfected COS-7 cells (29) and PKC␣-mediated PLD1 phosphorylation with the activation occurring in caveolin-enriched microdomains within the plasma membrane (30). In the present study, the PMA-induced PLD2 phosphorylation and activation was blocked by Ro-31-8220 and rottlerin (Figs. 2  and 3). Moreover, DN-PKC␦ completely inhibited, but WT-PKC␦ greatly enhanced PMA-induced PLD2 activation and PLD2 phosphorylation in PC12 cells (Fig. 4, B and C). DN-PKC␦ adenovirus encodes a mutant of PKC␦ (K376R), in which conserved lysine residues in the ATP-binding domain are replaced. This mutant yields enzymatically inactive trans-dominant proteins (47). Therefore, our results strongly indicate that PLD2 is regulated by PKC␦ in a phosphorylation-dependent manner.
Although PLD2 overexpressed in HeLa cells could be phosphorylated by okadaic acid, a Ser/Thr phosphatase inhibitor (45), there has been no proof that PLD2 is phosphorylated at serine or threonine in any cell type upon treatment with PMA or agonists. For the first time, by phosphoamino acid analysis, PLD2 was found to be basally phosphorylated at Ser/Thr amino acids and further phosphorylated by PMA at Ser/Thr amino acids (Fig. 3B). Previously, it has been reported that mouse PLD2 is constitutively associated with epidermal growth factor receptor and becomes phosphorylated on tyrosine 11 after stimulation with epidermal growth factor, but the mutation of tyrosine 11 to phenylalanine has no effect on epidermal growth factor-induced PLD2 activity (46). In our results, phospho-Tyr was not detected, indicating that tyrosine kinases were not involved in PMA-induced PLD2 phosphorylation.
It has been known that PC12 cells express PLD1 and PLD2 endogenously as confirmed by Northern blot and Western blot analyses (9,33) and that PKC␣ and PKC␤ are major regulators of PLD1 (53). As shown in Fig. 2, PMA treatment increased PLD response in the presence of tetracycline (ϩTet), and that GF109203X or Ro-31-8220 completely inhibited this activity, which means that PMA-induced endogenous PLD activity is also mediated by PKC in PC12 cells. However, Go6976 and rottlerin partially inhibited PMA-induced PLD response in the presence of tetracycline (ϩTet). Therefore, it seems that Go6976 inhibited only PMA-induced endogenous PLD1 activity and had no effect on the PMA-induced endogenous PLD2 activity. Conversely, the PKC␦ inhibitor rottlerin inhibited PMAinduced PLD2 activity, but had no effect on the PMA-induced endogenous PLD1 activity. DN-PKC␦ completely inhibited PMA-induced PLD2 phosphorylation and activity (Fig. 4, B and C), but the fact that PMA still slightly increased PLD response despite the overexpression of DN-PKC␦ shows that although the PKC␦/PLD2 pathway is blocked, endogenous PLD1 may be activated by PMA treatment.
The two-dimensional tryptic phosphopeptide map of PLD2 showed that the phosphopeptide spots could be categorized into two groups. The first (1/a, 2/b, 8/i, 7/h, 9/f, 13/m, and 14/l) contained common phosphopeptides generated in vivo and in vitro. These spots thus represented peptide fragments directly phosphorylated by PKC␦ in vivo and in vitro. The phosphorylation of these peptide fragments was increased after PMA treatment and decreased upon pretreatment with PKC␦ inhibitor. The second group of spots represented phosphopeptides that became further phosphorylated after PMA treatment in a rottlerin-sensitive manner, and did not overlap the phosphopeptide spots generated by PKC␦ in vitro. These phosphopeptides may have been phosphorylated by other PKC␦-dependent kinase(s). Therefore, PLD2 may be linked to diverse regulation through phosphorylation in PC12 cells.
PLD activities have been detected from several regions of the cell, including plasma membrane, nucleus, Golgi, and endoplasmic reticulum (38 -42). The subcellular localization of PLD2 was determined by immunocytochemical analysis. PLD2 positive signals were primarily detected at the plasma membrane in the absence of tetracycline (ϪTet) and no signals were detected in the presence of tetracycline (ϩTet). PMA treatment did not induce a change in the subcellular localization of PLD2. We also followed the translocation of endogenous PKC␦ induced by PMA. Immunocytochemical study showed an almost homogeneous cytoplasmic distribution of endogenous PKC␦ in unstimulated cells, translocation to the plasma membrane in response to PMA, and finally co-localization with PLD2. These results suggest that the plasma membrane is used as a site for the phosphorylation and activation of PLD2 by active PKC␦ in PC12 cells and support the proposition that PKC␦ directly regulate PLD2.
The mechanism of PLD2 activation by PKC␦ phosphorylation may be explained either by direct activation of PLD2 catalytic activity by a phosphorylation-induced conformational change or by some indirect mechanism, such as the release of PLD2 from inhibitor(s) or the association between PLD2 and another activator(s) by phosphorylation. When we measured the PKC␦-dependent PLD2 activity in vitro, no significant change was observed by phosphorylation (data not shown), excluding the possibility of the first case. Sequence-specific phosphoserine/threonine-binding domains, such as the WW domains, WD40 domains, FHA domains, and leucine-rich repeats are known (48, 49). Unidentified PLD2 stimulators carrying these domains may exist and may be involved in PLD2 activation following PKC␦ phosphorylation. Recently, we reported that a cytoskeletal protein, ␣-actinin, bound and inhibited PLD2 and that this inhibition could be reversed by ADP-ribosylation factor (19). Similarly, there is also a possibility that PKC␦-mediated phosphorylation of PLD2 overcomes the inhibitory effect of ␣-actinin. Recently, it was reported that PMAstimulated phosphatidylcholine hydrolysis requires the overexpression of myristoylated alanine-rich protein kinase C substrate (MARCKS) as well as PKC␣ in SK-N-MC cells, while PKC␣ alone was insufficient (50). Therefore, there exists a possibility that PKC␦-mediated PLD2 phosphorylation and activation may require another factor(s) in cells.
In this study, we show for the first time that the activity of PLD2 is directly up-regulated by PKC␦ in a phosphorylationdependent manner. Accordingly, these results contribute to the understanding of the regulatory mechanism and the role of PLD2 in cellular signaling pathways.