Regulation of Membrane-bound Phospholipase D by Protein Kinase C in HL60 Cells

In HL60 cells, the membrane-bound phospholipase D (PLD) was stimulated by 4 (cid:98) -phorbol 12-myristate 13-acetate (PMA) in the presence of the cytosolic fraction from HL60 cells or rat brain. The cytosolic factor for this PMA-induced PLD activation was subjected to purifica- tion from rat brain by sequential chromatographies. The PLD stimulating activity was found in protein ki- nase C (PKC) fraction containing (cid:97) , (cid:98) I, (cid:98) II, and (cid:103) isozymes. PKC isozymes were further separated by hy- droxylapatite chromatography. PKC (cid:97) and - (cid:98) , but not (cid:103) , isozymes were found to activate membrane-bound PLD. PKC (cid:97) was much more effective than PKC (cid:98) for PLD activation. Millimolar concentrations of MgATP were re- quired for the PKC-mediated PLD activation in HL60 membranes. MgATP is utilized to maintain the levels of phosphatidylinositol 4,5-bisphosphate (PIP 2 ) under these assay conditions. The PKC-mediated PLD activation was completely inhibited by neomycin, a high affin- ity ligand for PIP 2 , and this suppression was recovered by the addition of exogenous PIP 2 . Thus, these results suggest that PIP 2 is supposed to play a key role in PKC- mediated PLD activity in HL60 membranes. Further-more, analysis. Electrophoresis Western Blot Analysis— Proteins were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis on 8 or 12% polyacrylamide gels for PKC or small GTP-binding protein, respec- tively, according to the method by Laemmli (27). Proteins were electro-phoretically transferred onto nitrocellulose membrane (28). Blocking was performed in Tris-buffered saline containing 5% skimmed milk powder and 0.05% Tween 20. Western blot analysis specific antibodies leading to PLD activation to be disclosed. In the present study, the mechanism of PMA-induced PLD in HL60 membranes. cell-free isolated membranes HL60 cells, PMA cytosolic fractions HL60 cytosolic factors for partially purified by sequential chromatographies and identified PKCs. Our previous (20) that the PKC-mediated PLD activation was most effectively induced in the presence of Ca 2 (cid:49) . In addition, PKC (cid:97) , - (cid:98) I, - (cid:98) II, and lesser amounts of - (cid:122) isozyme were observed, but PKC (cid:103) , - (cid:100) , - (cid:101) , - (cid:104) , and - (cid:117) isozymes were undetectable in HL60 cells by Western blot analysis (data not shown). These results suggest that conven-tional PKC (cPKC) isozymes may play a role in PMA-stimu-lated PLD activation in HL60 membranes. We have further examined PKC isozymes involved in the regulation of mem-brane-bound PLD. The results indicate that PKC (cid:97) and - (cid:98) activate the membrane-bound PLD and also that PKC (cid:97) is much more effective in its activation in HL60 cells. The study of the regulation of PLD in membranes isolated from Chinese ham-ster lung (CCL39) fibroblasts addition purified PKC (cid:97) and - (cid:98) from rat brain could activate PLD (8). In

In HL60 cells, the membrane-bound phospholipase D (PLD) was stimulated by 4␤-phorbol 12-myristate 13acetate (PMA) in the presence of the cytosolic fraction from HL60 cells or rat brain. The cytosolic factor for this PMA-induced PLD activation was subjected to purification from rat brain by sequential chromatographies. The PLD stimulating activity was found in protein kinase C (PKC) fraction containing ␣, ␤I, ␤II, and ␥ isozymes. PKC isozymes were further separated by hydroxylapatite chromatography. PKC␣ and -␤, but not ␥, isozymes were found to activate membrane-bound PLD. PKC␣ was much more effective than PKC␤ for PLD activation. Millimolar concentrations of MgATP were required for the PKC-mediated PLD activation in HL60 membranes. MgATP is utilized to maintain the levels of phosphatidylinositol 4,5-bisphosphate (PIP 2 ) under these assay conditions. The PKC-mediated PLD activation was completely inhibited by neomycin, a high affinity ligand for PIP 2 , and this suppression was recovered by the addition of exogenous PIP 2 . Thus, these results suggest that PIP 2 is supposed to play a key role in PKCmediated PLD activity in HL60 membranes. Furthermore, PKC␣-mediated PLD activation was potentiated by the addition of recombinant RhoA protein in the presence of guanosine 5-O-(3-thiotriphosphate) (GTP␥S). The results obtained here indicate that PKC␣ and RhoA (GTP form) exert a synergistic action in the membrane-bound PLD activation in HL60 cells.
Phospholipase D (PLD) 1 has been recognized to play an important role in signal transduction of many types of cells. PLD hydrolyzes phosphatidylcholine (PC) to generate phosphatidic acid (PA) and choline (1). PA and its dephosphorylated product diacylglycerol are important second messengers. PLD is activated in a variety cells in response to receptor agonists, phorbol ester and Ca 2ϩ ionophore (2). Recently, it has been pointed out that several factors are required for activation of PLD. In reconstitution experiments, activation of membrane-bound PLD induced by phorbol 12-myristate 13-acetate (PMA) or nonhydrolyzable guanine nucleotide (GTP␥S) was observed only when cytosol and membranes were present (3). Similar findings were obtained in permeabilized cell preparations in which leakage of cytosolic components resulted in reduction of PLD activity (4,5). These results imply that cytosolic factors for PLD activation are involved in protein kinase C (PKC) and GTP-binding proteins.
PKC has been reported to be implicated in PLD activation in various cell types. Evidence that PKC up-regulates PLD activity is supported by the observations that PKC inhibitors or PKC down-regulation by long-term exposure to PMA prevents the increase of PLD activity (2). Recently, a role for specific PKC isozymes in the regulation of PLD is presented by the studies overexpressing PKC␣ or -␤ isozymes in cells. Overexpression of PKC␤I enhances PMA-induced PLD activity in rat fibroblasts (6). Overexpression of PKC␣ in Swiss-3T3 fibroblasts (7) does not induce an acute stimulation of PLD by PMA, but rather it plays a role in the expression of PLD enzyme. Furthermore, in membranes isolated from CCL39 fibroblasts (8), only PKC␣ and -␤ are capable of activating PLD. However, exact molecular interactions between PLD and PKC have not yet been studied.
On the other hand, stimulation of PLD activity by GTP␥S in permeabilized cells and cell lysates indicates its regulation by GTP-binding proteins (2). Recent studies have demonstrated the implication of two small GTP-binding proteins, ADP-ribosylation factor (ARF) (9, 10) and Rho (11,12) in the regulation of PLD activity in several types of cells. Furthermore, in some types of cells (13)(14)(15)(16)(17)(18)(19)(20), PLD activation induced by both GTP␥S and PMA was greatly enhanced, compared with that caused by either stimulant alone. These findings suggest that PKC may play an important role in positively modulating GTP-binding protein-mediated PLD activity. However, the precise relationship between GTP-binding proteins and PKC has not yet been disclosed. Our recent study (20) has demonstrated that the partially purified PKC fraction from rat brain cytosol showed a synergistic stimulation of PLD activity of HL60 membranes by PMA and GTP␥S. Moreover, this synergistic activation of the membrane-bound PLD was prevented by pretreatment with Rho GDP dissociation inhibitor (RhoGDI), suggesting a potential role of RhoA in the PKC-mediated PLD activation.
The present study was designed to gain more insight into the mechanisms underlying the PMA-induced PLD activation in HL60 membranes. A cytosolic factor reconstituting PMA-induced PLD activity was resolved as PKC fraction from rat brain. Among PKC␣, -␤, and -␥ isozymes, PKC␣ was the most effective in activating membrane-bound PLD. PKC␣-mediated PLD activation was synergistically stimulated by RhoA in the presence of GTP␥S. Furthermore, MgATP and phosphatidylinositol 4,5-bisphosphate (PIP 2 ) were necessary for the membrane-bound PLD activation by the partially purified PKC fraction which was free from small GTP-binding proteins of brain cytosol. Cell Culture and Cell Labeling-HL60 cells were grown in RPMI 1640 medium supplemented with 10% fetal bovine serum, 100 units/ml penicillin, and 100 g/ml streptomycin. All cells were grown at a cell density of 0.2-1.0 ϫ 10 6 cells/ml in a humidified atmosphere containing 5% CO 2 at 37°C. For assay of PLD activity, cells were labeled with [ 3 H]oleic acid (0.5 Ci/ml) for 12-15 h. Under these conditions, 60 -65% of total radioactivity incorporated into cells was found in the phosphatidylcholine (PC) fraction. For the analysis of phosphoinositides, cells were labeled with myo-[ 3 H]inositol (2 Ci/ml) in inositol-free ␣-minimum essential medium supplemented with 2 mM glutamine and 3.5 mg/ml bovine serum albumin for 24 h.
Preparation of Membranes and Cytosol Fractions from HL60 Cells-Membranes and cytosol fractions were prepared by the method described by Olson et al. (3), with minor modifications. The labeled cells were washed twice with buffer A (25 mM HEPES, pH 7.4, 100 mM KCl, 3 mM NaCl, 5 mM MgCl 2 , 0.5 mM MgATP, 1 mM EGTA, 5 mM dithiothreitol, 0.5 mM phenylmethylsulfonyl fluoride (PMSF), and 10 g/ml leupeptin) and resuspended in buffer A. Cells were then disrupted by N 2 cavitation (600 p.s.i. at 4°C for 30 min). After unbroken cells were removed by centrifugation at 900 ϫ g for 5 min, membrane and cytosol fractions were separated by centrifugation at 100,000 ϫ g for 60 min. Membranes were washed once and resuspended in buffer A. Membranes were used within 12 h after isolation. Cytosol proteins were concentrated using Centricon 10 (Amicon) and stored at Ϫ80°C until use.
Assay of PLD Activity in HL60 Membranes-[ 3 H]Oleic acid-labeled HL60 membranes (50 g of protein/assay) and crude cytosol or purified PKC fractions were incubated in buffer A containing CaCl 2 to give a final free Ca 2ϩ concentration of 1 M (total 0.1 ml) and were stimulated with 100 nM PMA and/or 10 M GTP␥S at 37°C for 15 min in the presence of butanol (0.3%, v/v). Reactions were terminated by the addition of chloroform/methanol (1:2, v/v). Lipids were extracted according to the method of Bligh and Dyer (21) and separated on Silica Gel 60 TLC plates in a solvent system of the upper phase of ethyl acetate/2,2,4trimethylpentane/acetic acid/water (13:2:3:10, v/v) as described previously (22). The plates were exposed to iodine vapor, and [ 3 H]PBut was identified by comigration with PBut standard which was prepared by the partially purified cabbage PLD (23). The spots scraped off the plates were mixed with scintillation mixture, and the radioactivity was counted in a liquid scintillation counter (Beckman LS-6500). To measure PLD activity using exogenous substrate, assays were carried out with phosphatidylethanolamine (PE)/PIP 2 /[ 3 H]DPPC essentially according to the method of Brown et al. (9).
Analysis of Phosphoinositides in HL60 Membranes-myo-[ 3 H]Inositol-labeled HL60 membranes (50 g of protein/assay) were prepared as described above and were incubated under the same conditions of PLD assay except butanol. Reactions were terminated by the addition of chloroform/methanol/concentrated HCl (20:40:1, v/v). Lipids were extracted and separated on Silica Gel 60 TLC plates, impregnated with 1.2% potassium oxalate in a solvent system of chloroform/methanol/ 28% ammonia water/water (45:40:5:8, v/v) as described previously (24). The plates were exposed to iodine vapor, and [ 3 H]PIP 2 was identified by comigration with PIP 2 standard. The radioactivity of spots was counted as described above. In another set of experiments, HL60 membranes (50 g of protein) were incubated with [␥-32 P]ATP (10 Ci/ml) in the presence or absence of 0.5 mM ATP at 37°C for 15 min. [ 32 P]PIP 2 was analyzed as above or by autoradiography.
Separation of PKC Isozymes and Small GTP-binding Proteins from Rat Brain Cytosol-Rat brains (approximately 7.0 g) were homogenized in buffer B (25 mM Tris-HCl, pH 7.5, 0.25 M sucrose, 2 mM EGTA, 1 mM PMSF, 50 mM 2-mercaptoethanol, and 10 g/ml leupeptin) with Poly-tron homogenizer (Kinematica). The homogenate was centrifuged at 100,000 ϫ g for 60 min to obtain cytosolic fraction. The supernatant was loaded onto a Mono Q column equilibrated with buffer C (20 mM Tris-HCl, pH 7.5, 0.5 mM EGTA, 0.5 mM EDTA, and 10 mM 2-mercaptoethanol). Proteins were eluted with linear gradient of NaCl (0 -0.7 M) in buffer C using first protein liquid chromatography (Pharmacia). The PKC activity peak eluted with 0.2-0.3 M NaCl was then applied to a Superose 12 column equilibrated with buffer C containing 150 mM NaCl and eluted with the same buffer. The PKC activity peak was pooled and then was applied to a hydroxylapatite column equilibrated with buffer D (20 mM KPO 4 , pH 7.5, 0.5 mM EGTA, 0.5 mM EDTA, 10% glycerol, and 10 mM 2-mercaptoethanol). Proteins were eluted with linear gradient of potassium phosphate (20 -300 mM) in buffer D. Individual fractions obtained from each chromatography were subjected to Western blot analysis using antibodies against PKC␣, -␤I, -␤II, and -␥ isozymes and small GTP-binding proteins.
Assay of Protein Kinase C Activity-The PKC activity was assayed as described previously (25) by using myelin basic protein (MBP) as a substrate (26). The reaction mixture (50 l) contained 50 mM Tris-HCl, pH 7.5, 10 mM MgCl 2 , 2 mM CaCl 2 , 10 nM PMA, 50 g/ml phosphatidylserine, 10 M [␥-32 P]ATP, and 50 g/ml MBP. After incubation at 30°C for 5 min, reactions were terminated with 50 l of 20 mM ATP. A 50-l aliquot was spotted on P81 phosphocellulose paper, and then the paper was washed with 75 mM phosphoric acid. The radioactivity retained on the paper was determined by Cerenkov counting. One unit of PKC was expressed as 1 pmol of [␥-32 P]ATP incorporated into MBP/5 min.
Translocation of PKCs and Small GTP-binding Proteins to Membrane-After incubation under the same conditions as the PLD assay, the reaction mixture was centrifuged at 100,000 ϫ g for 30 min to obtain the membrane pellet. Membranes were washed once in buffer E (20 mM Tris-HCl, pH 7.4, 10 mM EGTA, 2 mM EDTA, 5 mM dithiothreitol, 1 mM PMSF, and 10 g/ml leupeptin) and resuspended in buffer E containing 1% Triton X-100. After incubation at 4°C for 60 min, the suspension was centrifuged at 100,000 ϫ g for 60 min to obtain the membrane extract. Aliquots mixed with Laemmli's sample buffer (27) were subjected to electrophoresis and Western blot analysis.
Electrophoresis and Western Blot Analysis-Proteins were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis on 8 or 12% polyacrylamide gels for PKC or small GTP-binding protein, respectively, according to the method by Laemmli (27). Proteins were electrophoretically transferred onto nitrocellulose membrane (28). Blocking was performed in Tris-buffered saline containing 5% skimmed milk powder and 0.05% Tween 20. Western blot analysis using specific antibodies was performed as described previously (29).

Requirement of Cytosolic Components for Activation of Membrane-bound PLD in HL60
Cells-Recent reconstitution studies in the cell-free system and cytosol-depleted permeabilized cells have shown that several cytosolic factors are required for activation of PLD by GTP␥S and/or PMA (2). We examined the effects of cytosolic fractions separated from HL60 cells or rat brain on the membrane-bound PLD activity from HL60 cells in response to PMA and/or GTP␥S (Fig. 1). Weak PLD activity was detected in both cytosolic fractions, when measured using the PE/PIP 2 /[ 3 H]DPPC (16:1.4:1) substrate system as described by Brown et al. (9). However, the cytosolic PLD activity was stimulated at most 2-fold with both 100 nM PMA and 10 M GTP␥S (data not shown). In the absence of cytosolic fraction, little membrane-bound PLD activity was detected in response to either PMA or GTP␥S. However, when the cytosolic fractions from HL60 cells or rat brain were included in the reaction mixture, nearly 5-fold enhancement of PLD activity was seen in response to either 100 nM PMA or 10 M GTP␥S. Furthermore, in the presence of both PMA and GTP␥S, the enhancement (more than 15-fold) of membrane-bound PLD activity was much greater than that observed in response to either stimulant alone.
The formation of phosphatidylalcohol such as PBut is commonly utilized to monitor PLD activity (1, 2). Although [ 3 H] PBut has been reported to be metabolically stable, its degradation in various assay conditions is not clearly known. We Resolution of PMA-dependent PLD Activating Fractions from Rat Brain Cytosol-The PMA-or GTP␥S-induced PLD activation in HL60 membranes was dependent on rat brain cytosol ( Fig. 2A). To separate the PMA-and GTP␥S-dependent cytosolic factors responsible for membrane-bound PLD activation, the brain cytosol was subjected to Mono Q anion exchange chromatography (Fig. 3A). The fractions enhancing PMA-and/or GTP␥S-induced PLD activity in HL60 membranes showed a broad peak. Unadsorbed fractions were ineffective in stimulating PLD activity. The active peak enhancing the PMA-induced PLD activity (fractions 35-55) was found to overlap with the peak of PKC activity (Fig. 3B). However, column fractions 32-41 also could augment the GTP␥S-induced PLD activity (Fig. 3A). A small GTP-binding protein RhoA was detected in these fractions by Western blot analysis (Fig. 3A, inset). The fractions 42-55 (PKC-rich fractions) were pooled and concentrated. However, this PKC-rich fraction (Mono Q-PKC) still had an ability to increase GTP␥S-induced PLD activity (Fig.  2B), suggesting the coexistence of small GTP-binding proteins.
In order to separate PKCs from small GTP-binding proteins, the Mono Q-PKC fraction was subjected to Superose 12 gel filtration chromatography. The PKC fraction eluted at a position of about 80 kDa was separated from the fraction containing small GTP-binding proteins (less than 80 kDa) and concentrated. This PKC fraction (Superose-PKC) contained ␣, ␤I, ␤II, and ␥ isozymes, but Rho family small GTP-binding proteins, RhoA, Rac1/Rac2, and Cdc42Hs were undetectable by Western blot analysis (data not shown). The Superose-PKC fraction stimulated PLD activity in response to PMA but failed to fulfill the stimulatory effect of GTP␥S (Fig. 2C). PLD activity was not detectable in this PKC fraction using the PE/PIP 2 /[ 3 H]DPPC (16:1.4:1) substrate system (data not shown) as described by Brown et al. (9), although crude rat brain cytosol contained weak PLD activity as described above.
Activation of Membrane-bound PLD of HL60 Cells by Partially Purified PKC from Rat Brain-To examine the mechanism of PKC-mediated PLD activation, isolated HL60 membranes and the Superose-PKC fraction obtained after gel filtration were used. In the presence of the Superose-PKC fraction, PLD activation by PMA was concentration-dependent and reached the maximal level at around 100 nM PMA (Fig. 4A). An inactive phorbol ester, 4␣-PDD failed to activate PLD (Fig. 4A). The PKC-mediated PLD activation was timedependent and reached a plateau at 10 -15 min after addition FIG. 3. Mono Q anion exchange chromatography of rat brain cytosol. Rat brain cytosol was subjected to Mono Q anion exchange chromatography, and each fraction was assayed for its ability to reconstitute HL60 membrane PLD activity and PKC activity as described under "Experimental Procedures." A, [ 3 H]oleic acid-labeled HL60 membranes (50 g of protein) and 10 l of each fraction were incubated with 100 nM PMA, 10 M GTP␥S, or both PMA and GTP␥S at 37°C for 15 min in the presence of 0.3% butanol. Western blot analysis of RhoA is shown in the inset. B, PKC activity of each fraction was measured using MBP as a substrate in the presence of 2 mM CaCl 2 , 10 nM PMA, and 50 g/ml phosphatidylserine. of the PKC fraction (Fig. 4B). Incubation of the HL60 membranes with the PKC fraction in the presence of 100 nM PMA caused translocation of PKC␣, -␤I, and -␤II isozymes to membranes as inferred by Western blot analysis (Fig. 4C). The time course of PKC translocation (Fig. 4D) showed a good correlation with PMA-induced PLD activation (Fig. 4B).
Effect of MgATP and Phosphatidylinositol 4,5-Bisphosphate on PKC-mediated PLD Activity-The enhancement by MgATP of PLD activation in response to GTP␥S was demonstrated in many types of cells (15, 30 -34). The effect of MgATP on PKCmediated PLD activation was examined in HL60 membranes. Although no stimulatory effect on the PKC-mediated PLD activity was observed at concentrations less than 50 M, MgATP greatly potentiated PKC-mediated PLD activity, and the max-imal effect was obtained at 0.5 mM (Fig. 5A). Recent reports (9, 35) have indicated evidence that PIP 2 functions as a cofactor for PLD activity. We examined the effect of MgATP on the level of [ 3 Fig. 5B, the level of [ 3 H]PIP 2 decreased in the absence of MgATP. In contrast, 0.5 mM MgATP caused an increase in the level of [ 3 H]PIP 2 . PMA alone or both PMA and partially purified PKC fraction had no stimulatory effects on the level of [ 3 H]PIP 2 . Additionally, the incubation of HL60 membranes with [␥-32 P]ATP in the presence of 0.5 mM MgATP led to incorporation of 32 P into PIP 2 (data not shown). Similar observations have been observed in rat brain membranes (36) and permeabilized U937 cells (37). These results suggest that the effect of MgATP is likely due to its ability to maintain the level of PIP 2 .

H]PIP 2 in membranes prepared from myo-[ 3 H]inositol-labeled HL60 cells. As shown in
In order to further assess the involvement of PIP 2 in PMAmediated PLD activation, we examined the effect of neomycin, which binds polyphosphoinositides with high affinity. However, neomycin is reported to inhibit PKC activity at higher concentrations (more than 2 mM) (38). Therefore, the HL60 membrane fraction was treated with neomycin (less than 1 mM) and then excess neomycin was washed out prior to stimulation of partially purified PKC fraction and PMA. The PKC-mediated PLD activity in HL60 membranes was suppressed in parallel to increasing concentrations of neomycin (Fig. 6). 1 mM neomycin caused a complete inhibition (approximately 95%) of the PLD activity. However, the suppressed PKC-mediated PLD activity by 1 mM neomycin was restored by addition of PIP 2 in a concentration-dependent manner (Fig. 6). In these experiments, PIP 2 was mixed with phosphatidylethanolamine (PIP 2 /PE, 1:5 mol/mol), because of effective incorporation into membranes. PE alone had no effect on restoring PLD activity in neomycintreated membranes (data not shown).
PKC Isozymes Responsible for Activation of Membranebound PLD-The Superose-PKC fraction from rat brain cytosol capable of activating PLD contained ␣, ␤I, ␤II, and ␥ isozymes. In order to determine which isozyme was involved in PLD activation, the PKC fraction was further subjected to hydroxylapatite column chromatography; the PKC fraction after Su- perose 12 gel filtration was applied to a hydroxylapatite column and eluted by a linear concentration gradient of potassium phosphate. As shown in Fig. 7A, two major fractions capable of enhancing PMA-induced PLD activity were separated. Western blot analysis revealed that the first peak (peak 1) contained PKC␤ (␤I plus ␤II) and PKC␣ was present in the second peak (peak 2). PKC␥ was found in fractions 5-15 (Fig. 7B). Each PKC isozyme fraction; PKC␥ (fractions 5-15), PKC␤ (fractions 20 -35), and PKC␣ (fractions 50 -70) was separately pooled and concentrated.
The effects of these purified PKC isozymes on HL60 membrane PLD activity were examined. In the presence of PMA (100 nM), PLD activity was enhanced by PKC␣ or -␤ in a concentration-dependent manner (Fig. 8, A and B). PKC␣ was the most effective with a maximal effect obtained at about 10 units/assay. The maximal PLD activation obtained at around 10 units/assay of PKC␤ was almost half of that induced by PKC␣. PKC␥ had no effect on PLD activation in HL60 membranes (data not shown). The PLD activity stimulated by PKC␣ (2.5 units/assay) plus ␤ (2.5 units/assay) was the same as that obtained with PKC␣ alone (Fig. 8C). GTP␥S in the absence of PMA did not stimulate PLD activity. However, in the presence of both PMA and GTP␥S, the PKC-mediated PLD activity was synergistically potentiated (Fig. 8). Additionally, recombinant PKC␣ stimulated the membrane-bound PLD activity in HL60 cells in a quite similar manner as observed with the purified PKC␣ fraction (data not shown).
Effect of Small GTP-binding Protein RhoA on PKC␣-mediated PLD Activation-Recently, we have demonstrated possible evidence that PKC-mediated PLD activity in HL60 membranes could be enhanced by RhoA in the presence of GTP␥S (20). The PLD activity of HL60 membranes was stimulated by recombinant GTP␥S-bound RhoA alone in a concentration-dependent manner (Fig. 9A). RhoA at 20 nM caused a remarkable activation of the PKC␣-mediated PLD of HL60 membranes in the presence of both PMA and GTP␥S (Fig. 9B). Incubation of HL60 membranes with HL60 cytosol in the presence of PMA or GTP␥S caused translocation of PKC␣ and RhoA to membranes, respectively, as inferred by Western blot analysis (Fig. 10). Cdc42Hs also was translocated to membranes although the extent was much less than that of RhoA (data not shown). Another Rho family member, Rac, was undetectable in either cytosol or membrane of HL60 cells as mentioned previously by Siddiqi et al. (39). DISCUSSION Several factors have been implicated in the regulation of PLD activity, such as Ca 2ϩ , PKC, protein-tyrosine kinase, and GTP-binding proteins (2). However, their detailed mechanisms are not fully understood. Recently, PLD assay systems using permeabilized cells or cell-free preparations have been developed, and cytosolic factors including ARF and Rho protein are identified as regulatory factors for PLD activity. PMA, know as a PKC activator, activates PLD in many types of cells. Although the effect of PMA is assumed to be mediated through PKC, the pathway leading to PLD activation remains to be disclosed. In the present study, the mechanism of PMA-induced PLD activation was investigated in HL60 membranes.
In the cell-free system using isolated membranes from HL60 cells, the membrane-bound PLD activity was stimulated by PMA in the presence of the cytosolic fractions from HL60 cells or rat brain. The cytosolic factors required for this activity were partially purified by sequential chromatographies and identified as PKCs. Our previous report (20) has shown that the PKC-mediated PLD activation was most effectively induced in the presence of Ca 2ϩ . In addition, PKC␣, -␤I, -␤II, and lesser amounts ofisozyme were observed, but PKC␥, -␦, -⑀, -, and isozymes were undetectable in HL60 cells by Western blot analysis (data not shown). These results suggest that conventional PKC (cPKC) isozymes may play a role in PMA-stimulated PLD activation in HL60 membranes. We have further examined PKC isozymes involved in the regulation of membrane-bound PLD. The results indicate that PKC␣ and -␤ activate the membrane-bound PLD and also that PKC␣ is much more effective in its activation in HL60 cells. The study of the regulation of PLD in membranes isolated from Chinese hamster lung (CCL39) fibroblasts demonstrated that addition of purified PKC␣ and -␤ from rat brain could activate PLD (8). In our study, PKC␣ and -␤ did not additively activate the PLD in HL60 membranes (Fig. 8C), suggesting that both PKC isozymes act at the same step for the PLD activation.
Previously, Tettenborn and Mueller (40) demonstrated that the PLD activation by PMA was dependent on the presence of ATP in HL60 cell lysates. Olson et al. (3) also reported that PMA-induced PLD activation was dependent on ATP in the neutrophil cell-free system. The PKC-mediated PLD activation in HL60 membranes required MgATP at millimolar concentrations (Fig. 5). Our previous study (20) has shown that the PKC-mediated PLD activation in HL60 membranes was not suppressed by Ro31-8425, a potent PKC inhibitor. Similar findings were reported by Conricode et al. (41) showing that membrane-bound PLD in CCL39 fibroblasts could be activated by PKC in a phosphorylation-independent mechanism and that the PLD activation by PKC was observed even in the absence of ATP, suggesting that PKC may activate PLD by an allosteric mechanism without ATP-dependent phosphorylation. On the other hand, it was reported more recently that the effect of ATP on PKC-mediated PLD activation is mediated by phosphorylation in human neutrophils (42). Although this discrepancy could reflect difference in cell type, our present data cannot exclude this possibility.
Several recent studies have provided evidence that PIP 2 may act as an important cofactor for PLD activity. Brown et al. (9) developed a reconstitution system for solubilized PLD activity from HL60 cells in which the substrate PC was present in the form of mixed phospholipid micelles including PIP 2 and demonstrated the requirement of PIP 2 in the ARF-mediated PLD activation. Liscovitch et al. (35) have shown that the activity of partially purified PLD from brain membranes was stimulated considerably by PIP 2 . In permeabilized U937 cells (37), GTP␥S was observed to elevate the levels of polyphosphoinositides in the presence of MgATP, and either GTP␥S-or PMA-induced PLD activation was prevented by the antibody against phosphatidylinositol 4-kinase. Furthermore, neomycin, a high affinity ligand for PIP 2 , inhibited the activity of purified PLD from brain membranes (35) and GTP␥S-induced PLD activation in permeabilized HL60 cells (15), human neutrophils (17), and U937 cells (37). The results obtained in the present study also indicated that PIP 2 was required for PKC-mediated PLD activation in HL60 membranes. In fact, the level of PIP 2 in HL60 membranes was increased by incubation with MgATP (Fig.  5B). Therefore, MgATP is supposed to play a key role in maintaining the level of PIP 2 for PLD activity. The effect of PIP 2 might be partly explained by the fact that PIP 2 (100 M) caused little enhancement (approximately 1.4-fold) of PKC activity (data not shown). However, at present, detailed information for the site of action of PIP 2 and involvement of protein phosphorylation is not available and should be obtained by additional experiments which are currently in progress in our laboratory.
The PKC-mediated PLD activation in HL60 membranes was potentiated by the addition of GTP␥S (Fig. 8). This finding led us to assume that translocated PKC␣ or -␤ interacts with membrane-associated GTP-binding proteins, resulting in a synergistic activation of PLD. Several lines of evidence are present to indicate that activation of PLD is mediated by small GTP-binding proteins which interact directly with the solubilized PLD (9,39,(43)(44)(45). In addition, it was demonstrated that PIP 2 synthesis is regulated by Rho family small GTP-binding proteins (46,47). Our previous study (20) has shown that RhoGDI, Rho GDP dissociation inhibitor which extracts Rho proteins from membranes, prevented the synergistic effect by GTP␥S in PKC-mediated PLD activation in HL60 membranes, and also that this suppressed PLD activation was restored by the addition of recombinant RhoA. These results suggest the