Cloning and Characterization of Phospholipase D from Rat Brain*

The regulation of phospholipase D cloned from rat brain (rPLD) was examined in vivo andin vitro. The enzyme was a shorter splice variant of human phospholipase D 1 (Hammond, S. M., Altshuller, Y. M., Sung, T.-C., Rudge, S. M., Rose, K., Engebrecht, J. A., Morris, A. J., and Frohman, M. A. (1995) J. Biol. Chem. 270, 29640–29643). Its expression in COS-7 cells led to increased phospholipase D (PLD) activity that was further stimulated by constitutively active V14RhoA. V14RhoA had no effect on the endogenous PLD of the COS-7 cells, but constitutively active L71ARF3 increased its activity. In contrast, L71ARF3 did not activate rPLD expressed in the cells. Addition of phorbol ester markedly increased the endogenous PLD activity of COS-7 cells, and there was a further increase in the cells expressing rPLD. In membranes from COS-7 cells expressing rPLD, addition of myristoylated ADP-ribosylation factor (ARF) and RhoA in vitro stimulated PLD activity. The effect of ARF was greater than that of RhoA, although the concentrations for half-maximal stimulation (0.08–0.2 μm) were similar. Membranes isolated from cells expressing rPLD plus L71ARF3 and/or V14RhoA also showed higher PLD activity but no synergism between the two G proteins. Addition of phorbol ester and protein kinase C α (PKCα) also stimulated PLD activity in membranes from COS-7 cells expressing rPLD, but it had no effect on the activity in control (vector) membranes and did not enhance the effects of constitutively active ARF or Rho. The stimulation by PKCα did not require ATP and was not increased by addition of this nucleotide. No synergism between ARF and Rho and between these and PKCα on PLD activity was observed when these were added to membranes from cells expressing rPLD. Oleate inhibited the PLD activity of membranes from both control and rPLD-expressing cells. In summary, these results indicate that in vitro, rPLD is stimulated by ARF, RhoA, and PKCα and inhibited by oleate. However, in intact COS-7 cells, ARF activates endogenous PLD but not rPLD, whereas the reverse is true for RhoA. In addition, the effects of phorbol ester are much greater in the intact cells. It is concluded that the regulation of rPLD in intact COS-7 cells differs significantly from that seen in vitro; possible reasons for this are discussed.

Phospholipase D (PLD) 1 is believed to play a role in signal transduction in many cell types because it is a ubiquitous enzyme that is regulated by a great variety of hormones, neurotransmitters, growth factors, cytokines, and other molecules involved in extracellular communication (Ref. 1 and references therein). The product of PLD, phosphatidic acid (PA) has been proposed to function in mitogenesis in fibroblasts, stimulation of respiratory burst in neutrophils, regulation of secretion, and activation of specific protein kinases and other proteins. Diacylglycerol, which can be formed from PA by phosphatidic acid phosphatase, is a major regulator of protein kinase C (PKC). Lysophosphatidic acid, which can be produced from PA through hydrolysis by a specific phospholipase, A 2 , is now recognized as an important extracellular signal.
Studies employing PKC inhibitors or down-regulation of the enzyme indicate the involvement of PKC in agonist regulation of PLD in many cell types (1). The stimulation of PLD by PKC may occur by phosphorylation-dependent (2) or phosphorylation-independent (3)(4)(5)(6) mechanisms. On the other hand, PKCindependent control of PLD by agonists has been observed in some studies (Ref. 1

and references therein).
Early studies indicated that the PLD activity of neutrophils or HL60 cells was enhanced by GTP␥S in combination with a cytosolic protein, which was subsequently identified as ARF (7,8). ARF was originally discovered as a factor required for the ADP-ribosylation of G s␣ by cholera toxin and was later shown to be involved in protein trafficking in the Golgi apparatus (9). Stimulation of PLD activity by ARF has now been observed using recombinant PLD and partially purified preparations of the enzyme from various tissues (Ref. 1 and references therein). ARF translocation to the membrane fraction has also been correlated with PLD activation after treatment of HL60 cells with phorbol ester or formyl-methionyl-leucyl-phenylalanine (10). Similar findings have been reported by Rü menapp et al. (11) in HEK cells expressing muscarinic receptors and treated with carbachol.
Cytosolic proteins have been reported to stimulate PLD syn-* Supported in part by National Institutes of Health Grant DK 47448. 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.
There is much indirect evidence that PLD exists as several isozymes. This is based on differences in substrate specificity; in the regulation of the enzyme from different tissues and subcellular fractions by PKC, PIP 2 , fatty acids, Rho, and ARF; and in the effects of divalent cations, detergents, and pH (1,30,31). Although it is probable that the regulation of PLD is complex, the results strongly suggest the presence of several isozymes. Recently, human PLD (hPLD1) was cloned from HeLa cells and characterized in vitro (32,33). To determine whether this enzyme is similar to that in other mammalian tissues, we cloned rPLD from a rat brain cDNA library and determined its regulation by ARF, RhoA, and PKC in vitro and in vivo.

EXPERIMENTAL PROCEDURES
Materials-Phorbol 12-myristate 13-acetate (PMA), sodium oleate, and PA were purchased from Sigma. Dipalmitoylphosphatidylcholine (PC), phosphatidylethanolamine, phosphatidylethanol, and phosphatidylbutanol (PtdBut) were purchased from Avanti Polar Lipids Corp. GTP␥S was obtained from Boehringer Mannheim. [2-palmitoyl-9,10-3 H]Phosphatidylcholine and [9,10-3 H]myristic acid were purchased from NEN Life Science Products. PKC␣ was purchased from Panvera Corp. Phosphatidylinositol 4,5-bisphosphate (PIP 2 ), RhoA, and recombinant myristoylated ARF3 (mARF) were prepared as described previously (31). SDS-polyacrylamide gels were purchased from Novex. [␣-32 P]dCTP, hybridization solution, and random DNA labeling kit were from Amersham. Monoclonal antibody against RhoA from mouse was from Santa Cruz, and the cDNA for RhoA was generously provided by Dr. R. Cerione (Cornell University, Ithaca, NY). Anti-sARFII, polyclonal antibody from rabbit (which recognizes ARF1 and ARF3), and the cDNA for ARF3 were kind gifts of Dr. J. Moss (National Institutes of Health, Bethesda, MD). A polyclonal antibody to the carboxyl-terminal 12 residues of hPLD1 from rabbit was a kind gift of Dr. S. H. Ryu (Postech, Pohang, Korea). Polyclonal antibodies to hPLD1 (33) and PLD2 (34) were kindly supplied by Dr. A. J. Morris (State University of New York, Stonybrook, NY), and polyclonal antibodies to the aminoterminal and carboxyl-terminal 12 residues of rPLD were prepared by Dr. D.-S. Min (Vanderbilt University, Nashville, TN). Anti-rabbit and anti-mouse antibodies conjugated with horseradish peroxidase were from Vector Laboratories. Opti-MEM and LipofectAMINE were from Life Technologies, Inc. COS-7 cells were from the American Type Culture Collection. The ATP bioluminescence kit CLSII was from Boehringer Mannheim.
Cloning of PLD from Rat Brain-Plaques (10 6 ) of a rat brain cDNA library (Stratagene) were screened with probe labeled with [ 32 P]dCTP using a hPLD1 gene fragment (nucleotides 1476 -2494) amplified from HL60 DNA using polymerase chain reaction primers corresponding to conserved sequences (WAHHEK and IIGSANIN) in hPLD1 (32). Three clones that hybridized to the probe were selected and their 5Ј-terminal sequences were determined, but they were not full-length. Two fulllength cDNA clones were obtained by rescreening the cDNA library with a probe made using the DNA fragment corresponding to nucleotides 299 -1026 of hPLD1. Their sequences, determined by the dideoxy method, were found to be identical.
Construction of Expression Plasmids-Constitutively active mutant RhoA (V14RhoA) was made by site-directed mutagenesis (Promega) using the oligonucleotide TCCACAGGCTACATCACCAAC. Constitutively active mutant ARF3 (L71ARF3) was obtained from Dr. Hiroyuki Kanoh (Gifu University, Gifu, Japan). The full open reading frame of rPLD was cut by the restriction enzymes HindIII and ApaI. The cDNA were inserted into the pcDNA3 vector under the control of the cytomegalovirus promoter. Plasmids were purified for transfection using Qiagen kits.
In Vivo Assay of PLD-PLD activity was assayed by measuring the formation of PtdBut, the product of transphosphatidylation in the presence of n-butanol. COS-7 cells were cultured at 37°C in humidified 5% CO 2 in high glucose Dulbecco's modified Eagle's medium (Irvine Scientific) containing 10% fetal calf serum (Sigma). The cells were transfected using LipofectAMINE according to the manufacturer's protocol.
Briefly, each well of 6-well plates was seeded with 2 ϫ 10 5 cells and then cultured overnight in a 5% CO 2 incubator. For transfection, 2-3 g of DNA (from rPLD, V14RhoA, L71ARF3, and vector alone or in combination) and 6 l of LipofectAMINE were used. Twenty-four h later, the cells were labeled with 3 ml of Opti-MEM containing [ 3 H]myristic acid (1 Ci/ml) and 0.5% fetal calf serum for 12-13 h. Unincorporated [ 3 H]myristic acid was removed by washing with phosphate-buffered saline, and cells were incubated in 3 ml of Opti-MEM for 1 h. Butanol was then added to give a concentration of 40 mM, and the cells were incubated for 15 min. In the case of studies of the effect of PMA, 40 mM butanol and 100 nM PMA were used, and the cells were incubated for up to 60 min.
The reaction was terminated by adding 6 ml of ice-cold phosphatebuffered saline to each well. The phosphate-buffered saline was removed, and 1.5 ml of ice-cold methanol was added. [ 3 H]PtdBut was quantitated as described previously (14). Briefly, cells were scraped into 0.5 ml of chloroform and 0.75 ml of water, and 1 h later, 0.75 ml of 0.2 N HCl and 1 ml of chloroform were added; the mixture was kept overnight at room temperature. An aliquot was taken to measure total radioactivity in the lipids, and PtdBut was separated by thin layer chromatography using Whatman LK6D silica gel plates and a solvent system (ethyl acetate:isooctane:acetic acid:water, 55:25:10:50). PtdBut and PA were located by visualizing authentic standards with iodine vapor, and PtdBut was scraped and quantitated by liquid scintillation spectrometry.
Western Blotting-COS-7 cells were washed with ice-cold phosphatebuffered saline, scraped in homogenization buffer (10 mM HEPES, pH 7.4, 1 mM EDTA, 1 mM dithiothreitol, 250 mM sucrose, 2 g/ml aprotinin, 2 mM phenylmethylsulfonyl fluoride), homogenized by 20 passes through a 27 gauge needle, and centrifuged at 500 ϫ g for 10 min. The supernatant was then centrifuged at 100,000 ϫ g for 1 h to obtain membrane and cytosol fractions. These were mixed with SDS-polyacrylamide gel electrophoresis sample buffer, subjected to SDS-polyacrylamide gel electrophoresis, and transferred to polyvinylidene difluoride membranes. After blocking overnight with 5% fat free milk, the membranes were incubated with anti-RhoA monoclonal antibodies, anti-ARF polyclonal antibodies, or anti-PLD polyclonal antibodies. Visualization of bands was by enhanced chemiluminescence (Amersham).
Northern Hybridization-Preblotted poly(A) ϩ RNA from various rat tissues (CLONTECH) was hybridized with a probe excised from rPLD in pBluescript SK by NotI in the multiple cloning site of the vector and Nsi1 in rPLD. It was labeled with [ 32 P]dCTP by random priming and contained 1080 base pairs of the rPLD coding sequence, corresponding to amino acid residues 231-590. Hybridization was performed at 42°C for 16 h in the presence of 40% formamide and 10% dextran sulfate. After hybridization, the membrane was washed at 50°C in the presence of 0.1 ϫ SSC buffer and 0.1% SDS. Visualization was by exposure to either a phosphor storage screen (Molecular Dynamics) or Kodak Biomax MS film. When hybridization was performed on RNA from cell lines, total RNA was extracted using Trizo (Life Technologies, Inc.) according to the manufacturer's instructions. Poly(A) ϩ RNA was purified by annealing with biotinylated oligo(dT) and selection with magnetic beads. The RNA was separated by agarose gel electrophoresis and transferred to Nytran membranes (Schleicher & Schuell).
Measurement of PLD Activity in Membranes and Cytosol from Transfected COS Cells-COS-7 cells transfected with either vector or rPLD as described earlier were washed twice with Buffer A (20 mM HEPES (pH 7.2), 250 mM sucrose, 0.1 mM EDTA, 0.1 mM EGTA, 1.0 mM dithiothreitol, 0.5 mM phenylmethylsulfonyl fluoride, and 0.1 M leupeptin) and scraped into 2.7 ml of Buffer A per 6-well tissue culture plate. The cells were homogenized by 20 passes in a Dounce tissue grinder. Unbroken cells were removed by centrifugation at 500 ϫ g for 5 min. Crude membranes and cytosol were then isolated after centrifugation at 100,000 ϫ g for 1 h. The resulting pellet (crude membranes) was resuspended and washed in Buffer A. PLD activity was measured by the formation of [ 3 H]BtdBut from [ 3 H-palmitoyl]PC in the presence of 1% butanol as described previously (31). Phospholipid vesicles containing phosphatidylethanolamine/PIP 2 /[ 3 H]PC (16:1.4:1) were used as substrate (7).
Assay of ATP-The concentration of ATP in COS-7 cell membranes was assayed by luciferase bioluminescence (Boehringer Mannheim) according to the manufacturer's directions.

RESULTS
Cloning of Rat Brain PLD cDNA-Rat brain PLD cDNA was obtained from a rat brain cDNA library as described under "Experimental Procedures" using a DNA fragment of the hPLD1 gene from HL60 cells. Using this approach, two fulllength cDNA clones were obtained that had the translation initiation codon ATG, an in-frame stop codon, and a poly(A) tail. Both had identical sequences and were named rPLD. Analysis of the cDNA predicted an open reading frame encoding a 1036-amino acid protein with a calculated molecular mass of 119 kDa. The sequence showed 87% amino acid identity with hPLD1 (32) (Fig. 1) and had no recognizable motifs.
The most striking difference from hPLD1 was that rPLD had a 38-amino acid deletion corresponding to residues 585-622 of hPLD1. To determine whether this region was species-specific or not, polymerase chain reaction was performed on the cDNA of HL60 cells using a primer spanning this area. Two kinds of polymerase chain reaction fragment were amplified, i.e. one with the insertion present and one with it absent (data not shown). In addition, two cDNAs selected from a human pla-centa cDNA library (Stratagene) did not have the insertion (data not shown). These results indicate that the insertion is not species-specific and is probably a splicing variant, in agreement with a recent report (33).
Tissue Distribution of rPLD Transcript-Northern hydribidization using mRNA from various rat tissues revealed the presence of a 5.3-kilobase transcript in all tissues examined (Fig. 2). It was highest in lung, brain, and kidney and weakly present in testis (Fig. 2). Kidney contained an additional band of 3.8 kilobases. The transcript was not detected in COS-7 cells but was evident in C6 glioma cells, Rat-1 fibroblasts, and PC12 pheochromocytoma cells (data not shown).
Stimulation of rPLD by Constitutively Active ARF and Rho in COS-7 Cells-There are many reports that PLD is regulated by the small GTP-binding proteins ARF and Rho (1). To test the in vivo regulation of rPLD by these proteins, rPLD was expressed in COS-7 cells. A recombinant protein of approximately 120 kDa was detected in the membranes by an antibody raised to a peptide corresponding to the carboxyl terminus of hPLD1 ( Fig.  3) but not by a antibody raised to a sequence (residues 525-541) in hPLD1 that differs from that in rPLD (data not shown). Lesser amounts of the enzyme were also detected in the cytosol. 2 The enzyme had a molecular mass of 120 kDa as determined by SDS-polyacrylamide gel electrophoresis (not shown), which corresponds to the value of 119 kDa deduced from the sequence. The antibody cross-reacted with a 98-kDa protein that was endogenously present in the COS-7 cells (Fig. 3). Although this protein could represent an endogenous PLD, it did not cross-react with antibodies raised to PLD2 or to other sequences in hPLD1 or rPLD (data not shown).
The basal (vector alone) PLD activity of the cells was low and was enhanced 2-3-fold by transfection with rPLD (Fig. 4A). To study the regulation of rPLD in vivo, COS-7 cells were transfected with constitutively active forms of ARF and RhoA (L71ARF3 or V14RhoA) with and without rPLD. Western blotting showed increased expression of both of these small G proteins in the cytosol of the COS-7 cells (Fig. 3). V14RhoA was also detected in the membranes, but L71ARF3 was barely detectable in this fraction (Fig. 3). Cells transfected with only L71ARF3 showed a 2-fold increase of PLD activity, indicating that the endogenous PLD was responsive to ARF. In contrast, cells transfected with V14RhoA showed no significant increase of basal PLD activity. However, cells transfected with rPLD plus V14RhoA or rPLD plus L71ARF showed increases over basal PLD activity of 8.5-and 4.6-fold, respectively (Fig. 4A). There were no significant differences in the incorporation of [ 3 H]myristic acid into phospholipids in the variously transfected cells, indicating that [ 3 H]PtdBut production was a valid means of assaying PLD activity (data not shown).
When the cells were transfected with wild-type forms of ARF3 and RhoA with or without rPLD, the pattern of PLD 2 The amount of enzyme in the cytosol was variable. Compare Fig. 3 with Fig. 4C.

FIG. 1. Comparison of human and rat brain PLD sequences.
The upper sequence (hPLD1) is of the enzyme cloned from HeLa cells (32). The lower sequence (rPLD) is that of the PLD cloned in the present study. Nonidentical residues are shown in black boxes.
FIG. 2. Tissue distribution of rPLD transcript. mRNA from various rat tissues was analyzed as described under "Experimental Procedures." activity ( Fig. 4B) was similar to that seen with the constitutively active forms of the small G proteins (Fig. 4A). As discussed below, the stimulatory effects of the transfected wildtype G proteins can be attributed to factors present in the serum, because the stimulations were much less in serum-free medium (data not shown). When both ARF3 and RhoA were transfected together with rPLD, no synergistic interaction between the two G proteins was observed (Fig. 4B). In these experiments, rPLD expression in the co-transfections was similar to rPLD expression in the transfections with rPLD alone (Fig. 4C). 3 The expression of the small G proteins was similar to that seen in Fig. 3 except that ARF was not detectable in the membranes.
Stimulation of rPLD activity by PMA in COS-7 Cells-The involvement of PKC in agonist stimulation of PLD activity differs among cell types (1). To see whether rPLD responds to the PKC activator PMA, COS-7 cells transfected with vector or rPLD were treated with 100 nM PMA. PMA produced a very large stimulation of the PLD activity of COS-7 cells, which was enhanced in the cells expressing rPLD (Fig. 5). The changes with PMA in both types of cell were substantial at 5 min and maximal at 30 min. These findings suggest that both the endogenous PLD of COS-7 cells and rPLD are activated by PKC in vivo.
Stimulation of rPLD in COS-7 Cell Membranes and Cytosol by ARF and RhoA in Vitro-To further investigate the regulation of rPLD by RhoA and ARF, COS-7 cells were transiently transfected with either vector alone or rPLD, and cell fractions were prepared and assayed for PLD activity as described under "Experimental Procedures." In both membranes and cytosol, the basal PLD activity of rPLD-transfected cells was at least 2-fold higher than that of control cells (Fig. 6) and was increased by addition of GTP␥S. When ARF was included with GTP␥S in the assay, the PLD activity of both fractions was increased in both control and rPLD-transfected cells. However, the rPLD-transfected cells showed a much greater stimulation than the control cells (Fig. 6). In comparison to ARF, exogenous RhoA stimulated the PLD activity of membranes from rPLDtransfected cells but not that of control membranes (Fig. 6A). The stimulation by RhoA was consistently less than by ARF, and no stimulation by added RhoA was evident in cytosol from either control or rPLD-transfected cells (Fig. 6B). Preloading RhoA with GTP␥S prior to the PLD assay did not increase the magnitude of its effect (data not shown).
To determine the concentration dependence of activation of 3 The higher expression of membrane rPLD in the co-transfection with ARF3 (Fig. 4C) was not routinely observed. Compare Fig. 3 with Fig. 8B.   FIG. 3. Expression of rPLD, L71ARF3, and V14RhoA in COS-7 cells. Membrane and cytosol fractions of COS-7 cells transfected with vector control or cDNAs for rPLD, L71ARF3, and V14RhoA were analyzed by Western blotting as described under "Experimental Procedures." For PLD, antibodies raised to the carboxyl-terminal 12 residues of hPLD1 were employed. The data are representative of two analyses.

FIG. 4. Effects of constitutively active (A) and wild-type (B) ARF and Rho on rPLD activity in COS-7 cells.
A, COS-7 cells were transfected with cDNAs for L71ARF3, V14RhoA, and rPLD (or vector) and assayed for PLD activity as described under "Experimental Procedures." The data are means Ϯ S.E. of three experiments performed in duplicate. B, the experimental conditions were the same as for A, except that the cells were transfected with cDNAs for wild-type ARF and RhoA instead of the constitutively active forms. Data are means Ϯ S.E. of three experiments performed in duplicate. C, membranes and cytosol from the COS-7 cells used in the experiments shown in B were analyzed by Western blotting as described under "Experimental Procedures." rPLD by RhoA and ARF, transfected membranes were incubated with GTP␥S and increasing concentrations of either RhoA or mARF. Both proteins stimulated rPLD in a dosedependent manner, but ARF gave a much greater (8-fold) stimulation of rPLD than RhoA (2.5-fold) at maximal concentrations (compared with GTP␥S alone). The half-maximally effective concentrations of RhoA and ARF were 80 and 200 nM, respectively.
Stimulation of rPLD in COS-7 Cell Membranes by PKC␣ and Phorbol Ester in Vitro-Membranes from both control and rPLD-transfected COS-7 cells were incubated with various combinations of PKC␣, PMA, and ATP (Fig. 7). In the control (vector) cells, addition of PMA and/or PKC␣ had no effect. However, in the rPLD-transfected cells, PMA increased PLD activity 2-fold over basal. The PMA stimulation of rPLD was seen immediately after addition, was linear for 20 min, and continued to increase up to 60 min (data not shown). PLD stimulation by PMA alone indicates that these membranes contained a PKC isozyme(s). When purified PKC␣ was incubated with membranes from rPLD-transfected cells, a 3-fold increase in PLD was seen, and when both PMA and PKC␣ were added together, the PLD activity was no higher than with PKC␣ alone (Fig. 7), suggesting that the membranes contained activators of PKC. Surprisingly, the PKC␣ effect was independent of ATP because measurement of this nucleotide in the incubations using a bioluminescence luciferase assay showed a concentration of 0.3 nM, which is far below the K m of the enzyme for ATP (6 M) (35). Furthermore, there was no further stimulation when ATP or its nonhydrolyzable analogs were added along with PKC␣ and PMA ( Fig. 7 and data not shown).
Membranes from COS-7 cells that had been transfected with rPLD, L71ARF3, or V14RhoA were assayed for PLD activity using the in vitro assay system to see whether the results agreed with those observed in vivo. The membranes from the rPLD-transfected cells showed higher PLD activity than vector controls (Fig. 8A), although the magnitude of the increase was not as great as that seen in vivo (Fig. 4A). Co-transfection of L71ARF or V14RhoA resulted in small but reproducible further increases in activity, which were not synergistic (Fig. 9). The addition of GTP␥S slightly enhanced the effects of the constitutively active small G proteins (data not shown). The addition of PKC␣ plus PMA produced a further stimulation of PLD activity in the rPLD-transfected cells (Fig. 8A), but no synergism with L71ARF or V14RhoA was observed. Western blotting showed that co-expression with the small G proteins did not alter rPLD levels in the cells (Fig. 8B).
To further test for synergism, membranes from COS-7 cells transfected with rPLD were incubated with GTP␥S, ARF, RhoA, and PKC␣ alone or in various combinations (Fig. 9). Fig.  9 shows that the combination of ARF and RhoA with GTP␥S produced an additive effect on PLD activity, but not synergism. Likewise, the combination of PKC␣ with either ARF or RhoA plus GTP␥S did not result in a synergistic effect. In fact, inhibition was consistently observed when all three agents were combined (compare GTP␥S plus RhoA, ARF, and PKC␣ with GTP␥S plus RhoA and ARF).
Oleate Inhibition of rPLD-Sodium oleate is known to both activate and inhibit PLD activity depending on the tissue source of the enzyme (1,30). Therefore, the effect of oleate on rPLD activity was investigated. As seen in Fig. 10, oleate inhibited both control and ARF-stimulated rPLD activity. Oleate inhibition of basal and ARF-stimulated PLD activity was also observed in control (vector) membranes, but the changes were of much lower magnitude.

DISCUSSION
The PLD cloned from rat brain (rPLD) is closely related to hPLD1 cloned from HeLa cells. It shows 87% amino acid sequence identity with the human enzyme; the major difference is a 38-amino acid deletion (residues 585-624). When this deletion is omitted, the sequence identity rises to 91%, and when homologous replacements are also included, the proteins are 95% similar. The deletion is not species-specific, because it was also detected in human placenta. Furthermore, a very recently published report (33) has shown the existence of a variant of hPLD1 that has the same deletion and shows 90% amino acid identity with rPLD. The short and long forms of the enzymes probably arise from the alternative splicing of exons, although efforts to demonstrate this by Southern analysis were inconclusive (data not shown).
Northern hybridization indicated a 5.3-kilobase transcript for rPLD that was present in all tissues examined, although it was only weakly detectable in liver and testis (Fig. 2). Kidney contained an additional 3.8-kilobase transcript, which probably corresponds to a closely related PLD. Western blotting of tissues with antibody to the carboxyl terminus of hPLD1 was not performed because, although the antibody recognizes rPLD (Fig. 3), its interaction with other PLD isozymes is unknown.
In agreement with Hammond et al. (32,33) who studied two alternatively spliced forms of hPLD1, rPLD was stimulated by ARF, RhoA, and PKC␣ in vitro (Figs. 6 and 7). 4 The enzyme was also responsive to constitutively active V14RhoA and phorbol ester in intact COS-7 cells (Figs. 4A and 5). Surprisingly, although there was evidence that constitutively active L71ARF3 activated the endogenous PLD of COS-7 cells in vivo (Fig. 4A), an unequivocal effect of this small G protein on rPLD in the intact cells could not be demonstrated (i.e. the increase in PLD activity in the cells transfected with L71ARF3 plus rPLD 4 Although addition of RhoA stimulated rPLD in the membranes assayed with exogenous substrate (Fig. 6A), it did not do so in the cytosol (Fig. 6B). There are many possible explanations for this, including the fact that COS-7 cell cytosol contains a high level of endogenous RhoA, whereas it is barely detectable in membranes (Figs. 3 and 4C). Although cytosolic RhoA is probably mainly complexed to the GDP dissociation inhibitor of Rho, it can be activated by exogenous GTP␥S (36,37). This explanation is supported by the observation that addition of GTP␥S alone to the cytosol produces a large stimulation of PLD activity (Fig. 6B). Another explanation is the possible presence of inhibitors (or absence of stimulators) of PLD activation by Rho in the cytosol. could be attributed to the sum of their effects when expressed alone) (Fig. 4A). This cannot be ascribed to a failure of active ARF to be expressed because the protein was detected by Western blotting (Fig. 3) and stimulated the endogenous PLD in vivo (Fig. 4A). However, the possibility that L71ARF3 and rPLD were not localized in the same subcellular compartment remains, particularly because L71ARF3 was barely detectable in the membranes (Fig. 3). In this regard, membranes from cells transfected with rPLD, L71ARF3, and V14RhoA did show stimulation of PLD activity by the constitutively active forms of ARF3 and RhoA, although the magnitude of the effects was smaller than observed in vivo (Fig. 8A). Other explanations include an intrinsic insensitivity of rPLD to stimulation by ARF compared with endogenous (COS-7) PLD, a greater susceptibility of rPLD to inhibitory proteins possibly present in COS-7 cells (25)(26)(27)(28)(29), and a lack of responsiveness of rPLD to proteins that magnify the effect of ARF (5, 15, 20 -24). In relation to these possibilities, it should be noted that the en-dogenous PLD of COS-7 cells differs in several respects from rPLD. For example, it does not respond to RhoA in vivo or in vitro, and it does not respond to PKC␣ in vitro. Furthermore, it is not recognized by several antisera raised to sequences in rPLD, hPLD1, and PLD2.
In COS-7 cells expressing rPLD in combination with wildtype ARF3 or RhoA, stimulation of PLD activity by the G protein was observed. This is probably due to presence of stimulatory agonists (e.g. lysophosphatidic acid) present in the serum added to the incubation medium. Lysophosphatidic acid induces translocation of RhoA, consistent with its activation (38). The postulated role of serum was supported by the decrease in PLD activation in serum-deprived cells (data not shown).
The observation that rPLD responds to RhoA and PKC but not to ARF in the intact cells raises questions about the physiological significance of in vitro findings with PLD preparations. For example, it is not certain how the concentrations of ARF and Rho employed in such studies relate to those in the cell, and the effects of proteins that enhance or inhibit the effects of these G proteins in vitro are also difficult to translate to the in vivo situation. However, in vitro studies do allow PLD isozymes to be characterized and differentiated and provide better definition of regulatory mechanisms.
PKC␣ produced a small (2-fold) but reproducible stimulation of PLD activity in COS-7 membranes (Fig. 7). The findings produced some surprises. For example, phorbol ester and PKC␣ alone stimulated the enzyme, implying the presence of a PKC isozyme(s) and a PKC activator(s) in the membranes. However, PKC is known to associate with membranes, and diacylglycerol can be generated in membranes through phospholipase C and D action. The greatest surprise was the fact that the effects of PMA and PKC␣ were observed in the virtual absence of ATP, as determined by assay, and were not enhanced by addition of the nucleotide or its nonhydrolyzable analogs (Fig. 7). Phosphorylation-independent activation of PLD by PKC isozymes has previously been reported (3-6), but there is also evidence that the activation requires ATP (2). It should be noted that the observation that PKC can activate PLD in vitro without phosphorylation does not preclude an additional ATP-dependent regulatory mechanism in vivo. Indeed, phorbol ester produced a large increase in PLD activity in vector-transfected COS-7 cells in vivo (Fig. 5), but neither PMA nor PKC␣ could activate the endogenous PLD of these cells in vitro. Furthermore, the fold changes induced by PMA in the intact cells (Fig. 5) were much greater than those observed in vitro (Fig. 7). These findings indicate the existence of other mechanisms by which PMA activates PLD in intact cells.
The issue of the subcellular localization of small G proteins and PLD has been alluded to above. The present study just examined cytosol and a crude membrane fraction of COS-7 cells, but rPLD was consistently found in the membranes, although it was also variably present in the cytosol (Figs. 3 and   4C). 5 The presence of some rPLD in the cytosol was also indicated by the activity measurements shown in Fig. 6B. However, it should be cautioned that these findings with overexpressed enzyme may not reflect the distribution of PLD in normal cells, where the enzyme may be largely confined to membranes. Concerning the subcellular location of the expressed small G proteins, ARF and L71ARF3 were predominantly in the cytosol (Fig. 3), and endogenous RhoA was also almost exclusively in the cytosol, whereas there was some V14RhoA in the membranes. Because the natural substrate for PLD is membrane phospholipid, these observations suggest that membrane translocation of one or more of these G proteins would be required for activation of PLD in vivo. Agonist-induced translocation of ARF and Rho family proteins has been observed in some cell types (10,11,36), and there is some evidence that this is an important component of the mechanism(s) by which these proteins act (10,11,38,39).
Although several studies have shown that ARF, RhoA, and PKC interact synergistically to stimulate PLD activity in enzyme and membrane fractions from several tissues or cell types (5,6,14,22,40) and these proteins produce striking synergism on pure preparations of hPLD1 (33), we failed to see such synergism in intact COS-7 cells expressing these proteins (Fig.  4B) or in membranes from cells expressing rPLD and incubated with ARF, RhoA, and PKC␣ in vitro (Fig. 9). Although the synergism observed in previous studies with membrane fractions and partially purified PLD could be explained by the presence of other isozymes of PLD, the reason(s) for the difference between our results and those of Hammond et al. (33), who studied the human homologue of rPLD, is not clear. Comparison of the two sequences indicates substantial amino acid differences between residues 507 and 574. It could be speculated that residues in this sequence could be involved the synergistic interactions observed with hPLD1.
A surprising observation of the present study, in agreement with the findings of Hammond et al. (32,33), was that a single PLD isozyme could be regulated in vitro by ARF, RhoA, and PKC␣. Conventional wisdom, based on findings with phosphoinositide phospholipase C and previous biochemical data (1,30,31), suggested that these regulatory agents would act on sep- 5 Almost all the membrane PLD activity could be released by sonication, consistent with it being a peripheral rather than integral membrane protein. arate PLD isozymes. As described above, selective regulation of PLD by these agents could occur as a result of their selective activation by agonists or their translocation to membranes enriched in certain PLD isozymes. The possibility also exists that selective inhibitors or stimulators of their actions on PLD could be very important in regulation of the enzymes. These regulatory proteins could also show specificity for certain PLD isozymes. In conclusion, the present findings indicate that the in vivo regulation of PLD may be more complex than originally envisaged.