Differences in signal transduction between platelet-derived growth factor (PDGF) alpha and beta receptors in vascular smooth muscle cells. PDGF-BB is a potent mitogen, but PDGF-AA promotes only protein synthesis without activation of DNA synthesis.

Cultured vascular smooth muscle cells (VSMC) from spontaneously hypertensive rats express both alpha and beta isoforms of the platelet-derived growth factor (PDGF) receptors at high levels (100,000 and 240,000 sites/cell, respectively). In this cell type, PDGF-BB elicited a mitogenic response; however, PDGF-AA increased only protein synthesis without activating DNA synthesis. Protein kinase C (PKC) was activated by PDGF-AA as well as PDGF-BB with concomitant translocation from cytosol to membrane fractions. However, the hypertrophic effect of PDGF-AA was not affected by depletion of cellular PKC, whereas the mitogenic action of PDGF-BB was partially attenuated by the depletion. Following incubation with PDGF-AA or -BB, phospholipase C-gamma 1 (PLC-gamma 1) and phosphatidylinositol 3-kinase were tyrosine phosphorylated; however, the phosphorylation of Ras-GTPase-activating protein was induced only by PDGF-BB. Both PDGF isoforms resulted in a prompt and transient increase in the level of 1,2-diacylglycerol (DAG), presumably through the action of PLC-gamma 1. After returning to basal levels, the rate of DAG synthesis steadily increased for at least 15 min due to activation of phosphatidylcholine-hydrolyzing phospholipase C (PC-PLC). Incubation with PDGF-BB-activated phospholipase D (PLD) in a PKC-dependent manner resulting in the formation of phosphatidic acid (PA). PA was also formed by the sequential reactions of PC-PLC and DAG kinase in the PDGF-BB-stimulated VSMC, and these sequential reactions were not affected by PKC depletion. In contrast, PDGF-AA stimulation did not result in increased PA synthesis as neither PLD nor DAG kinase activities were affected. PA may be a significant second messenger in the activation of DNA synthesis by PDGF-BB. These results indicate that signaling mechanisms of the PDGF-alpha and -beta receptors in VSMC are distinctly different in signal transduction in VSMC and that the alpha receptor promotes cellular hypertrophy (but not hyperplasia), whereas a mitogenic response is mediated only through the beta receptor.

1 To whom correspondence should be addressed.
PDGF' is found in three isoforms consisting of disulfidelinked homo-and heterodimers of two homologous polypeptide chains designated A and B. These polypeptides are encoded by two distinct genes (1)(2)(3)(4). The dimeric ligands (PDGF-AA, -A B , and -BB) exert their biological effects by binding to two monomeric units of the receptors, PDGFR-a or -p, each with distinct binding properties (5-8). These receptors are structurally similar glycoproteins. The extracellular structures of these receptors are composed of five immunoglobulin-like domains and their intracellular region contains a tyrosine-kinase domain. Each subunit of the PDGF dimer binds t o one receptor molecule, therefore two molecules of receptors are necessary in order to accommodate the PDGF dimer upon binding the ligand (g-ll), these two receptor moieties form a noncovalent dimer. PDGFR-a shows a high affinity for both the PDGF-A and -B chains, whereas PDGFR-p preferentially binds the B chain. Receptor dimerization is closely related to the activation of the intrinsic protein tyrosine kinase (12)(13)(14). The tyrosine phosphorylation of various proteins in the signaling pathway appears to be involved in the mitogenic response stimulated by PDGF (11,15). Functional differences between PDGFR-a and -p have been reported in actin organization of porcine aortic endothelial cells (16), in chemotaxis on human foreskin fibroblasts (171, and in transformation of NIW3T3 fibroblasts (18).
However, it is generally thought that both PDGFR-a and -p are almost equally capable of triggering DNA synthesis in any types of cells, including those cited above (10,(16)(17)(18), and if there was a difference in mitogenic sensitivity to PDGF isoforms, it has been thought to be due to differences in expression levels of the two receptor isoforms which vary depending on cell type (10).
PDGF is one of the major mitogens in serum and is responsible for proliferation of certain cell types, including VSMC (1,11). We have previously studied the differential cellular response to PDGF isoforms in cultured rat VSMC and found that the PDGF-BB isoform was a potent mitogen, whereas PDGF-AA was essentially inactive (19). We attributed a lack of mitogenic activity of PDGF-AA to the absence of expression of PDGFR-a (20). In addition, PDGF-AB, which had been thought to require PDGFR-a for binding to its target cells (9-111, bound to PDGFR-p alone, forming a homodimer in VSMC which usu-PDGFR, platelet-derived growth factor receptor; VSMC, vascular The abbreviations used are: PDGF, platelet-derived growth factor; smooth muscle cells; SHR, spontaneously hypertensive rats; WKY, Wistar-Kyoto rats; PA, phosphatidic acid; PLD, phospholipase D; DAG, 1,2-diacylglycerol; PLC, phospholipase C; GAP, Ras-GTPase-activating protein; PI3K, phosphatidylinositol 3-kinase; BSA, bovine serum albumin; PAGE, polyacrylamide gel electrophoresis; PKC, protein kinase C; TPA, phorbol 12-myristate 13-acetate; PEt, phosphatidylethanol; PIP,, phosphatidylinositol 4,5-bisphosphate; PC, phosphatidylcholine. ally lacks . In this cell type, the binding of PDGF-AI3 induced DNA synthesis to almost the same extent as In a preliminary report (21), we proposed that PDGFR-a was expressed at a level comparable with PDGFR-P in cultured VSMC from a genetically hypertensive rat strain (SHR). However, in VSMC from a normotensive rat strain (WKY), PDGFR-a was suppressed almost completely. "his view was further supporting by the observation that PDGF-AA induced the tyrosine phosphorylation of a group of proteins almost to the same extent as PDGF-BB in the SHR cells (but not in the WKY cells). In the SHR cells, protein synthesis was promoted markedly by PDGF-AA, but not DNA synthesis. These observations suggest that PDGFR-a mediates cellular hypertrophy rather than hyperplasia in VSMC, whereas mitogenic responses are mediated only through PDGFR-p. This is in contrast to other types of cells which express PDGFR-a and their DNA synthesis is induced by both PDGF-AA and -BB (10). As a result, the SHR-derived VSMC provide us with a unique system for comparing the differential signaling mechanisms of PDGFR-a and -p. In addition, differences in signaling pathways between hyperplastic and hypertrophic responses can also be examined.
Here we report observations of marked differences in the signaling mechanism by PDGFR-a and -p in cultured VSMC isolated from SHR. In particular, we have observed that synthesis of PA is induced in response to PDGF-BB through PDGFR-p, due to activation of PLD and DAG kinase, whereas PDGF-AA (which binds only to PDGFR-a) does not result in enhanced PA synthesis. The observed differences may account, at least in part, for the differential cellular response to PDGF-AA(hypertrophy) and -BB (hyperplasia) in this cell type.
Cell Culture-VSMC (passage 7-12) derived from SHR (21) were seeded in a dish (lo4 cells/cm2) and grown in Dulbecco's modified Eagle's medium with 10% fetal calf serum. After reaching confluence, the cells were transferred to the above medium (without serum) containing insulin (10 pg/ml), transferrin (10 pg/ml), and sodium selenite (10 ng/ml). These cells were cultured for an additional 2 days (serum-free culture) and then used for the following determinations. Measurement of DNA and Protein Synthesis-Cells in a 13-mm well were incubated with or without PDGF (100 ng/ml) in 1 ml of a 1:l mixture (v/v) of Dulbecco's modified Eagle's medium and Ham's F-12 medium containing insulin (10 pg/ml), transferrin (10 pg/ml), sodium selenite (10 ng/ml), and BSA(1 mg/ml) for 2 h at 37 "C. After washing, incubation was continued in the absence of PDGF but in the presence of [U-"Clleucine (0.5 pCi/well) for an additional 18 h. During the final 4 h of incubation [methyl-3Hlthymidine was added to the medium at 2 pCi/ well (final). The cells were washed with BufferA(20 IIU~ HEPES-NaOH buffer, pH 7.4, containing 130 nm NaCl, 5 m M KCI, 1 m M MgCl, and 1.5 m M CaCI,), and treated with 10% (w/w) trichloroacetic acid. After washing with a mixture of ethanol and ethyl ether (2:1, v/v), the trichloroacetic acid-insoluble fraction containing DNA and proteins was solubilized with 0.5 N NaOH. DNA and protein synthesis were measured by radioactive incorporation by [rnethyl-3H]thymidine and [U-'4C]leucine, respectively, Binding Analysis-Competitive binding analysis was done by using "'I-PDGF-BB as a labeled ligand at 4 "C according to a published method (20). Determination ofPKCActiuity-Cells in a 100-mm dish were stimulated with PDGF (100 ng/ml) or TPA (100 m) in Buffer A containing 10 ~l l~ glucose and 1 mg/ml BSA for 5 min at 37 "C. After removing the medium, cells were quickly rinsed with ice-cold Buffer A and then scraped with Buffer B (25 m M Tris-HC1, pH 7.5, containing 2.5 m M EDTA, 0.5 IIU~ EGTA, 1 m M dithiothreitol, 100 pg/ml leupeptin, 25 pg/ml aprotinin, 1 nm phenylmethylsulfonyl fluoride, and 0.02% Triton X-100). Cells were homogenized by using a Dounce homogenizer, and the lysate was centrifuged at 100,000 x g for 30 min. Supernatant was used for the assay of PKC activity in the cytosolic fraction. PKC in membrane pellets were solubilized with Buffer B containing 1% Triton X-100 prior to assay. PKC activity was measured using a synthetic peptide from myelin basic protein as a substrate provided with an assay kit (Life Technologies, Inc.).
Immunoblotting-Cells in a 100-mm dish were stimulated with PDGF (100 ng/ml) in Buffer A containing 10 m M glucose and 1 mg/ml BSA for 10 min at 37 "C. After washing with ice-cold Buffer A containing 2 m M EGTA in place of CaCI,, cells were lysed with 20 m M HEPES-NaOH buffer, pH 7.4, containing 0.5% Nonidet P-40, 50 m M P-glycerophosphate, 0.1 m M orthovanadate, 10 molybdate, 1 nm phenylmethylsulfonyl fluoride, 0.1 mg/ml leupeptin, and 10 pg/ml aprotinin at 4 "C.
This lysate was centrifuged (14,000 xg, 5 min) to remove cellular debris and supernatant was retained as "cell lysate." To isolate tyrosine-phosphorylated proteins, cell lysate wag incubated with an anti-phosphotyrosine antibody in the presence of protein A-Sepharose, and the immunoprecipitated proteins were redissolved directly in SDS-PAGE buffer. SDS-PAGE (7.5% gel) and immunoblotting were done according to a published method (20). The protein concentration within cell lysate fraction was measured according to Lowry et al. (22) by comparison with a standard curve obtained with BSA.
Zkacer Experiments with PHlArachidonic Acid and [**C]Myristic Acid-Cells in a 35-mm dish were prelabeled for 2 days with [5,6,8,9,11,12,14,15-3Hlarachidonic acid and/or [l-'4Clmyristic acid (1 pCi eaclddish) in serum-free culture containing 2 mg/ml BSA. Following three washes with BufferA containing 10 m M glucose and 1 mg/ml BSA, cells were equilibrated in the same buffer for 30 min at 37 "C and then stimulated by the addition of PDGF (100 ng/ml) or TPA (100 m). "0 measure synthesis of PEt, ethanol was added to the equilibration buffer to 400 mM. Stimulation by PDGF was stopped by rapid removal of the culture medium, and 1 ml of a mixture of chloroform and methanol (1:2, v/v) was immediately added to extract total lipids from cells. Lipids were extracted and separated by thin-layer chromatography as described previously (19, 23). Radioactivity in each lipid class was measured by scintillation counting.

DNA and Protein Synthesis Induced by PDGF-AA and -BB in
VSMC from SHR-Cultured VSMC isolated from SHR were exposed to PDGF-AA or -BB for 2 h, and effects on DNA and protein synthesis were examined (Fig. 1). Incorporation of L3H1thymidine into DNA increased approximately 10-fold following exposure to PDGF-BB. However, incubation in the pres- Binding competition by unlabeled PDGF-AA and -BB for Iz5I-PDGF-BB. Cells (3.5 x lo5 cells/well) in a 13-mm well were incubated with '"'I-PDGF-BB (1 ng, 69,000 cpm) in the presence of unlabeled PDGF-AA (0) or -BB (A) in 1 ml of Dulbecco's modified Eagle's medium containing 10 mM HEPES-NaOH buffer, pH 7.4, and 5 mg/ml BSA a t 4 "C for 3 h. After washing five times with the same medium, cells were lysed directly with 20 mM HEPES-NaOH buffer, pH 7.4, containing 1% Triton X-I00 and 10% glycerol, and radioactivity associated with cells was counted. Nonspecific binding was determined in the presence of unlabeled PDGF-BB a t 500 ng/ml. Specific binding in the absence of unlabeled ligand (9,788 cpdwell) was set at 100%. Data are presented as mean of four determinations (S.E. < 10%). ence of PDGF-AA did not produce a significant increase in DNA synthesis over cells incubated in the absence of PDGF. In contrast to DNA synthesis, protein synthesis, as determined by [U-14C]leucine uptake, was substantially increased (1.7-fold) following exposure to PDGF-AA, although again the increase was smaller than that seen after stimulation by PDGF-BB (2.5-fold). Mitogenic stimulation by PDGF-AB was comparable with PDGF-BB (data not shown). These results are consistent with preliminary studies involving long term exposure (20 h) to PDGF (21). The data also suggest that binding of PDGF-AA to PDGFR-a in VSMC results in cellular hypertrophy rather than hyperplasia, whereas mitogenesis occurs only when PDGFR-/3 is activated by either PDGF-AB or -BB.
To clarify the role of PKC in the induction of DNA and protein synthesis, cells were preincubated with TPAfor 24 h in order to deplete cellular PKC activity (23) prior to incubation with PDGF-AA or -BB. An increase (approximately 6-fold) in DNA synthesis levels over control cells by PDGF-BB was found in PKC-depleted cells, although the increase was lower than that seen in the PKC-intact cells. In contrast, PKC depletion did not have a significant effect on the induction of protein synthesis by PDGF-AA or -BB.
Expression Levels of PDGFR-a and -&To determine the levels of PDGFR-a and -P expressed in SHR-derived VSMC, competitive binding analysis was done to displace the binding of lz51-PDGF-BB by unlabeled PDGF-AA or -BB a t 4 "C ( Fig. 2).
Binding of the labeled ligand decreased by increasing the concentration of unlabeled PDGF-BB. Unlabeled PDGF-AA also displaced the binding of 1251-PDGF-BB in a concentration-dependent manner, but the extent of the displacement was only about 30%, even at a concentration as high as 100 ng/ml. On the basis of a binding model for PDGF isoforms proposed previously (9-ll), the numbers of PDGFR-a and -P expressed in this cell type are estimated to be approximately 100,000 and 240,000 sites/cell, respectively. Protein Tyrosine Phosphorylation-Cells were incubated with PDGF-AA or -BB for 10 min, and tyrosine-phosphorylated proteins were detected by SDS-PAGE followed by immunoblotting with an anti-phosphotyrosine antibody. As shown in  PLC-yl (24, 25), GAP (26)(27)(28)(29), and PI3K (85-kDa subunit) (29-31) have all been considered to be direct substrates for the intrinsic tyrosine kinase action of PDGFR-P. In the VSMC system examined here, all three proteins were phosphorylated on their tyrosine residues following incubation with PDGF-BB (Fig. 4, lanes 3 ) . Incubation with PDGF-AA (lanes 2 ) resulted in phosphorylation of PLC-y, and PI3K a t levels comparable with PDGF-BB-induced phosphorylation. In contrast, the phosphorylation of GAP was markedly dependent on which PDGF isoform was present, with phosphorylation in the presence of PDGF-AA substantially less (10%) than that in the presence of PKC Activation-PKC, an important second messenger for transmembrane signaling, is generally activated in response to stimulation by growth factors and undergoes translocation from cytosol to plasma membranes (32). The translocation from cytosol to the membrane fraction was examined to see if PKC is activated by PDGF-AA in this VSMC system. As shown in Fig.  5, PKC activity associated in the membrane fraction was increased by PDGF-AA to an extent comparable with that attained with TPA or PDGF-BB.
DAG Formation-Cells were simultaneously prelabeled with both [3H]arachidonic acid and [l4C]rnyristic acid, then stimulated by addition of either PDGF-AA or -BB. Stimulation with either isoform resulted in an immediate and transient increase in cellular DAG levels (within the first 0.5 rnin), which returned to base-line values in 1 min (Fig. 6). Following this brief peak, the rate of DAG production rose steadily for a t least 15 min. 3H-Labeled DAG species (arachidonate-derived) were dominant in the first transient synthesis, whereas the subsequent sustained increase resulted in I4C-labeled DAG species (myristate derived), suggesting that the precursor for DAG was different during the two phases of synthesis. When cells were simultaneously prelabeled with [3H]arachidonic acid and ['"C]myristic acid, the ratio of [3H] to [14C] in PIP, was high (4.91, but the ratio in PC was low (0.9) compared with other phospholipid classes (data not shown). Therefore, PLC-y,, which is activated by the phosphorylation of tyrosine residues and catalyzes hydrolysis of PIP, (331, appears to be responsible for the initial phase of DAG synthesis seen in Fig. 6. Indeed, in PDGF-BB.  WKY-derived VSMC, which express only PDGFR-P, we have reported that PLC-y, can induce DAG formation immediately following a challenge with PDGF-BB (19). In contrast, the gradual increase of predominantly 14C-labeled DAG species found after 1 min in these SHR-derived VSMC may be due to the activation of PC-hydrolyzing PLC. Importantly, the second phase of DAG synthesis induced by both PDGF-AA and -BB was almost completely suppressed by D 609, a specific inhibitor for PC-PLC (34) (Fig. 7, solid bars). PKC depletion, through preincubation with TPA, did not alter the rate of DAG formation (Fig. 7, shaded bars), indicating that PC-PLC is activated by both PDGF-AA and -BB in a PKC-independent manner.
PA Formation-In contrast to DAG levels, which were equally affected by both isoforms, levels of PA did not significantly increase following stimulation of VSMC with PDGF-AA but did increase following stimulation with PDGF-BB (Fig. 8). requirement for the formation of PA by the sequential reactions of PC-PLC and DAG kinase. DISCUSSION The PDGFR isoforms, a and P, have different ligand specificities; the PDGF-A chain selectively binds and activates PDGFR-a, whereas the PDGF-B chain stimulates both PDGFR-a and -0 (9)(10)(11). Recently, Heidaran et al. (18) found that tyrosine kinase activities of PDGFR-a and -P also showed differing substrate specificities, and the difference may account for the greater ability of PDGFR-fl over PDGFR-a in mediating PDGF-induced transformation of NIW3T3 fibroblasts. In addition, Eriksson et al. (16) reported a significant difference between PDGFR-a and -0 in the ability to induce actin organization in porcine aortic endothelial cells. However, it has been generally found that, if expressed, both PDGFR-a and -P are equally capable of mediating a mitogenic response in any cell types (including the two cell lines above) (10,16,18,38).
The data obtained by a competitive binding analysis (Fig. 2) showed that both PDGFR-a and -P were expressed at high levels (100,000 and 240,000 sitedcell, respectively) in cultured VSMC derived from the genetically hypertensive rat strain, SHR. The results are quite different from those obtained in cultured VSMC from the normotensive rat strain, WKY, in which PDGFR-a is suppressed almost completely, whereas The expression level of PDGFR-a observed in the SHR-derived cells is comparable with or higher than those reported in other cell types in which PDGF-AA can act as a mitogen (10,39,40).
As expected, PDGF-AA caused tyrosine phosphorylation of a group of proteins, including PDGFR-a in the SHR-derived cells (Fig. 3). However, DNA synthesis, as measured by L3H]thymidine incorporation, was not affected by PDGF-AA, although protein synthesis (measured by [14C]leucine incorporation) was significantly stimulated (Fig. 1). In contrast, the presence of PDGF-BB, which activates both PDGFR-a and -P, induced marked activation of both DNA and protein synthesis. The difference in the effects of PDGF-AA and -BB on DNA and protein synthesis indicates that in VSMC PDGFR-a mediates cellular hypertrophy (but not hyperplasia), whereas PDGFR-P mediates a mitogenic response. These results suggest, therefore, that in this cell type the signaling mechanism of PDGFR-a is distinct from that of PDGFR-P. Thickening of the vascular wall due to an increase in the number and/or size of VSMC is commonly involved in chronic hypertension regardless of the underlying mechanism causing elevated blood pressure (41-43). PDGF-AA is synthesized and secreted from VSMC (44). The data presented here suggest that PDGF-AA could cause cellular hypertrophy in an autocrine fashion within VSMC of SHR (a genetically hypertensive rat strain used as a model for human essential hypertension (45)).
Therefore, expression of PDGFR-a in VSMC may be considered a pathophysiologically important factor which may induce vascular hypertrophy during hypertension. The mechanism by which PDGFR-a is abnormally expressed in VSMC of SHR is still unclear.
When tyrosine kinase of PDGFR-/3 is activated by its ligand, receptor autophosphorylation ensues which is followed by the tyrosine phosphorylation of, among other proteins, PLC-y,, GAP, and PI3K. The present studies with the VSMC system confirm that PDGF-BB induced the phosphorylation of these three proteins (Fig. 4). The activation of PDGFR-a by PDGF-AA also induced the phosphorylation of PLC-yl and PI3K in VSMC; however, phosphorylation of GAP in the presence of PDGF-AA was minimal. The relatively low level of GAP phosphorylation by PDGF-AAin VSMC is consistent with other studies that show GAP is a poor substrate for the intrinsic kinase activity of PDGFR-a (18,381. Ras, a key protein involved in mitogenic signaling pathways, is activated by binding GTP and is thought to be an important mediator of PDGF-stimulated responses (46)(47)(48). GAP stimulates the hydrolysis of Rasbound-GTP and the activity of GAP is reduced when the protein is tyrosine-phosphorylated (49,501. As found here, and in other systems, low levels of GAP phosphorylation seem t o be a general property of PDGFR-a-mediated responses. Therefore, in cells in which PDGF-AA can act as a potent mitogen, it is reasonable to consider that tyrosine phosphorylation of GAP is not required for the induction of DNA synthesis. Recently, Bazenet and Kazlauskas (38) reported that PDGFR-a! activated Ras and triggered DNA synthesis without the participation of GAP in mouse fibroblasts, PhB. Results of experiments using a series of mutated chimeric PDGFR-p, each of which lacks tyrosine phosphorylating ability for a specific protein, indicate that the tyrosine phosphorylation of PLC-y, andor PI3K is essential for mitogenic signaling by PDGF (25, 29, 31). Valius and Kazlauskas (15) demonstrated that DNA synthesis was activated to near maximal levels when either PLC-y, or P13K was tyrosine-phosphorylated by the action of PDGFR-p in a human hepatoma cell line (HepG2) and a canine kidney epithelial cell line (TRAMP). Furthermore, Satoh et a2. (51) reported that PDGFR-p-mediated Ras activation might be correlated with the tyrosine phosphorylation of PI3K (but not GAP) in Chinese hamster ovary cells (CHO cells). However, the present results (Figs. 1 and 4) clearly indicate that in VSMC, in contrast t o other types of cells, the tyrosine phosphorylation of PLC-y, and PI3K is not sufficient (but perhaps essential) for the activation of DNA synthesis. The tyrosine phosphorylation of GAP, which is not induced by the action of PDGFR-a, may be one of essential steps involved in mitogenic signaling triggered by PDGFR-p in this cell type.
DNA synthesis was activated by PDGF-BB in VSMC, even when the cellular PKC level was depleted, although the extent of the induction was substantially lower than that in PKCintact cells (Fig. 1). This observation would imply that both PKC-dependent and -independent signaling pathways are involved in the activation of DNA synthesis in this cell type in response to PDGF-BB. In contrast, PKC was not essential for the promotion of protein synthesis by PDGF-AA. However, PKC should be activated both by PDGF-AAand -BB, since both of the PDGF isoforms promoted translocation of PKC from cytosol to plasma membranes (Fig. 5). In addition, the level of DAG, a well known activator for PKC (32), was increased either by PDGF-AA or -BB (Fig. 6). Thus, PKC activation itself may not be a critical factor in determining differences in cellular response mediated by PDGFR-a and -p.
In VSMC from SHR, the synthesis of PA was increased by PDGF-BB, but not by PDGF-AA (Fig. 8). PA is thought to be an important second messenger for mitogenic signaling (52), and it has been reported that PA triggers DNA synthesis when added to culture medium of human A 431 carcinoma cells (53) and Balb/c 3T3 cells (52). We have shown that exogenously added bacterial (Streptomyces chromofuscus) PLD, which hydrolyzes PC in cell membranes to form PA, mimicked the mitogenic effects of PDGF-BB in VSMC (23). Therefore, the inability of PDGF-AA to enhance DNA synthesis in VSMC from SHR may be due at least in part to its inability to enhance PA synthesis. Furthermore, when PA synthesis by DAG kinase action was blocked by R 59022, the level of PDGF-BB-induced DNA synthesis was attenuated in a dose-dependent manner correlating to decrease in PA production in this cell type. 2 PDGF-BB activated PLD in VSMC, and the activation was largely but not completely abolished when PKC was depleted lished data.
* H. Inui, T. Kondo, F. Konishi, Y. Kitami, and T. Inagami, unpub-( Fig. 9). Furthermore, TPA itself increased the PLD activity significantly, and the activation was diminished by PKC depletion ( Fig. 9 and Refs. 23 and 54). These results suggest that PKC activation automatically leads to the activation of PLD. However, PDGF-AA failed to activate PLD (Fig. 9), even if PKC was activated by either PDGF-AA or -BB to comparable levels (Fig. 5). Analogous to PDGF-AA, angiotensin I1 does not significantly activate PLD in VSMC (541, even if PKC should be activated as this peptide induces DAG production (55). These results lead to the postulate that although PKC is necessary for the activation of PLD in VSMC, it is not sufficient and an additional unknown factor is involved. This additional mechanism (or factor) is activated by PDGF-BB or TPA but not by PDGF-AA or angiotensin 11. Recently, Brown et al. (56) reported that the ADP-ribosylation factor subfamily of small GTP-binding proteins is a critical factor for PLD activation. Our data also suggest that TPA elicits PLD activation not only by the direct activation of PKC but also by a PKC-independent mechanism in VSMC. This notion is supported by previous studies suggesting that phorbol esters act in part through a PKC-independent mechanism in certain cells (57,58).
In addition to PLD, DAG kinase also contributed t o the synthesis of PA by phosphorylation of DAG which was produced by the action of PC-PLC in PDGF-BB-stimulated VSMC (Fig. 10). PC-PLC was activated by PDGF-AA as well as PDGF-BB, resulting in increased DAG levels (Figs. 6 and 71, but PA levels were not affected by incubating with PDGF-AA (Fig. 8). These results indicate that DAG kinase activity is low a t a basal level in VSMC, and it is stimulated specifically by PDGF-BB (but not by PDGF-AA). Recently, van der Bend et al. (59) reported that PA was actively synthesized by DAG kinase action in bradykinin-treated human foreskin fibroblasts and endothelin-ltreated Rat-1 fibroblasts. However, DAG generated artificially by exogenously added PLC was not converted to PA by basal DAG kinase activity in these cells. Furthermore, Ohanian and Heagerty (60) observed that noradrenaline increased membrane-associated DAG kinase activity transiently in the artery, whereas angiotensin I1 did not alter the DAG kinase activity in this fraction. Perhaps the inability of PDGF-AA to activate PLD and DAG kinase may be ascribed to a mechanism related to the inability of PDGFR-a! to phosphorylate GAP; however, details still remain to be clarified.
When VSMC were colabeled with L3H]arachidonic acid and ['4Clmyristic acid and then stimulated by PDGF-BB for 15 min, 14C-labeled DAG species were the major products whereas 3Hlabeled species were dominant in the increased PA (Figs. 6 and 8). This may appear to be incompatible with the current view that PC is the source of PA and DAG. However, it has been reported that PA is synthesized by the action of a membranebound DAG kinase in PDGF-BB-stimulated Swiss 3T3 cells, and the membrane-bound DAG kinase shows high substrate specificity for arachidonate-containing DAG species (36,61). It is thus reasonable to consider that arachidonate-containing DAG species (but not myristate-containing ones) which were produced by PC-PLC were selectively converted to PA by DAG kinase in VSMC stimulated by PDGF-BB, and the difference in the distribution of 3H and 14C observed between DAG and PA may be attributed at least in part to the selective conversion of DAG to PA by DAG kinase. The ratio of 3H to '*C in PEt produced by PLD in the presence of ethanol was 1.3 in VSMC stimulated by either PDGF-BB or TPA (Fig. 9). Interestingly, this value is substantially higher (about 1.4-fold) than that of its major substrate PC (0.9). To further clarify the characteristics of the PLD action in VSMC, cells were prelabeled with a combination of L3H1arachidonic acid and [14C]palmitic acid, or of [3Hloleic acid and ['4Clmyristic acid, and then stimulated by PDGF-BB in the presence of ethanol. In both cases, the ratio of 3H t o 14C in PEt produced was higher (1.3-to 1.5-fold) than that in PC (data not shown). These observations suggest that PLD has a relative preference for PCs containing unsaturated fatty acids to those containing saturated fatty acids. Alternatively, other phospholipid classes such as phosphatidylethanolamine, in addition to PC, might also be substrates for PLD. The ratio of 3H to I4C in phosphatidylethanolamine was over 3.0 in VSMC colabeled with [3H]arachidonic acid and [14C]myristic acid (data not shown), and this phospholipid has been reported to be hydrolyzed by PLD in NIW3T3 fibroblasts (62).
Larrodera et al. (63) report that exogenous bacterial PC-PLC can induce DNA synthesis in Swiss 3T3 cells, and the authors propose that the activation of PC-PLC resulting in formation of DAG is an important step in the mitogenic signaling by PDGF. Here, PKC-independent activation of PC-PLC was observed in response to stimulation with both PDGF-AA and -BB in VSMC (Fig. 7). However, despite the activation of PC-PLC, PDGF-AA did not show any effect on DNA synthesis in this cell type. The discrepancy may be due to differences in cell types. Analogous to our results, Fukami and Takenawa (52) reported that in Balblc 3T3 cells the mitogenic activities of exogenously added PLC and DAG were significantly lower than those of PLD, PA, and PDGF. In VSMC stimulated by PDGF-AA, the activation of PC-PLC resulting in increase in cellular DAG level may be correlated with increased protein synthesis. This notion is supported by the following observations. Angiotensin I1 can activate PC-PLC (but not PLD) (54,551 and causes cellular hypertrophy in VSMC (64, 65). However, detailed mechanisms of signaling involved in the cellular hypertrophy induced by PDGF-AA and angiotensin I1 still remain unclear.
In conclusion, in VSMC derived from SHR, PDGF-BB acts as a potent mitogen, whereas PDGF-AA induces only protein synthesis without activating DNA synthesis despite that PDGFR-a as well as PDGFR-p is expressed at a high level. PDGFR-a and -p expressed in the SHR-derived VSMC, consistent with other known sources, show differential substrate specificities as tyrosine kinases, that is, GAP tyrosine phosphorylation occurs only when PDGFR-p is activated by PDGF-BB. The activation of PLD and DAG kinase, both of which participate in the synthesis of PA in response to PDGF-BB, is clearly not involved in signal transduction by PDGFR-a in this cell type. These differences observed in signal transduction by PDGFR-a and -p may account for the differential cellular response to PDGF-AA and -BB in the SHR-derived VSMC.