Phospholipase D Activity in PC12 Cells

PC12 neuronal cells express a membrane phospholipase D (PLD) activity that is detected at similar levels in undifferentiated or differentiated cells. The regulation of this activity by agonists was explored. Membrane phospholipase D activity was increased by treatment of cells with the phorbol ester phorbol 12-myristate 13-acetate (PMA) or with nerve growth factor. The ability of PMA to activate PLD was confirmed in intact PC12 cells. Basal activity of PLD in membranes was reduced in RG20, a PC12 cell line overexpressing the human α2A-adrenergic receptor. PMA did not increase PLD activity in RG20 cells, as assessed both in membrane preparations and in intact cells. Cyclic AMP levels did not regulate phospholipase D activity in either cell type. However, incubation of RG20 cells with the α2-adrenergic antagonist rauwolscine or with pertussis toxin increased membrane PLD activity and restored activation of PLD by PMA. These data suggest that the effects of the overexpressed α2A-adrenergic receptor on PLD activity are mediated by precoupling of the receptor to the heterotrimeric GTP-binding protein, Gi, but are independent of adenylate cyclase regulation. The results of this study suggest that membrane phospholipase D activity can be negatively regulated via Gi in PC12 cells.


From the Department of Cell and Molecular Pharmacology, Medical University of South Carolina, Charleston, South Carolina 29425
PC12 neuronal cells express a membrane phospholipase D (PLD) activity that is detected at similar levels in undifferentiated or differentiated cells. The regulation of this activity by agonists was explored. Membrane phospholipase D activity was increased by treatment of cells with the phorbol ester phorbol 12-myristate 13acetate (PMA) or with nerve growth factor. The ability of PMA to activate PLD was confirmed in intact PC12 cells. Basal activity of PLD in membranes was reduced in RG20, a PC12 cell line overexpressing the human ␣ 2A -adrenergic receptor. PMA did not increase PLD activity in RG20 cells, as assessed both in membrane preparations and in intact cells. Cyclic AMP levels did not regulate phospholipase D activity in either cell type. However, incubation of RG20 cells with the ␣ 2 -adrenergic antagonist rauwolscine or with pertussis toxin increased membrane PLD activity and restored activation of PLD by PMA. These data suggest that the effects of the overexpressed ␣ 2A -adrenergic receptor on PLD activity are mediated by precoupling of the receptor to the heterotrimeric GTP-binding protein, G i , but are independent of adenylate cyclase regulation. The results of this study suggest that membrane phospholipase D activity can be negatively regulated via G i in PC12 cells.
Phospholipase D (PLD) 1 isoforms have been sequenced from yeast (1-3) and mammalian (4) cells. It appears that more than one form of PLD is expressed in mammals (5)(6)(7)(8). PLDs can be activated in response to extracellular signals such as growth factors, hormones, and neurotransmitters (9) and by phorbol esters that stimulate protein kinase C (10). Hydrolysis of phos-phatidylcholine (PC) by PLD produces phosphatidic acid (PA). PA is proposed to play a role in signal transduction as a lipid mediator or mediator precursor (11)(12)(13)(14). Some PLDs are regulated by the small GTP-binding proteins ARF and/or Rho (15)(16)(17). However, the full range of effectors and mediators regulating different isoforms of PLD remains to be elucidated. This laboratory has characterized regulated PLDs in yeast and mammalian cells (10, 18 -20). In this study, we explore the role of a heterotrimeric GTP-binding protein, G i , in the regulation of PLD activity in a neuronal cell line.
Signal transduction pathways have been extensively studied in PC12, a rat phaeochromocytoma cell line that can be induced to differentiate to a neuronal phenotype. PC12 transfected with the ␣ 2A -adrenergic receptor (␣ 2A AR) have been used to examine coupling of this receptor to its effectors (21). The ␣ 2A AR couples to the heterotrimeric GTP-binding protein G i , is widely expressed, and mediates the central hypotensive effects of ␣ 2 agonists (22,23). G i proteins are heterotrimeric GTP-binding proteins containing an ␣ i subunit. They are coupled to inhibition of adenylate cyclase as well as to pathways involving additional effectors, such as small GTP-binding proteins (24). In this study, ␣ 2A AR-expressing PC12 cells were used to examine the potential role of G i in regulating PLD activity.
ERK activity was assessed in vitro as described previously (10,20), using myelin basic protein as substrate. Results were normalized for cytosolic protein.

RESULTS AND DISCUSSION
A fluorescent in vitro assay using a BODIPY-labeled substrate, BPC, was used to screen for PLD activity in membranes prepared from a variety of cell types. PC12K cells were observed to express a very high level of membrane PLD activity as compared with other cell types (data not shown). This activ-* This work was supported by National Institutes of Health Grants CA58640-04 and HL07260, by the University Research Committee of the Medical University of South Carolina, and by National Science Foundation Grant EPS-9630167. 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.
The phorbol ester PMA generally increases PLD in mammalian cells. PMA activated PLD in intact undifferentiated PC12K cells, as seen using either [ 3 H]oleate (Fig. 1B) or [ 3 H]palmitate (data not shown) as metabolic label. The greatest accumulation of PEt occurred in the first 15 min after PMA addition. PEt did not accumulate in cells incubated without PMA (data not shown).
The potential influence of the ␣ 2A/D AR, a G i -coupled receptor, was examined in a PC12 cell line overexpressing this receptor. Radioligand binding, using the ligand [ 3 H]RX 82102, confirmed that RG20 overexpressed the ␣ 2A/D AR (7585 fmol/mg of protein in RG20 versus 78 in PC12K) (data not shown). Basal mem-brane PLD activity was substantially decreased in ␣ 2A AR-expressing cells, as shown for two clonal RG20 lines ( Fig. 2A). This decrease was not due to enhanced degradation of PBt by PLA 2 , since lyso-PBt production was not greater in RG20 than in PC12K. Calcium-independent PLA 2 activity, as measured by lyso-PC production, was similar in both cell lines (Fig. 2A).
The abilities of PC12 and RG20 cells to activate PLD were compared. In PC12K, but not RG20, treatment with PMA or NGF increased membrane PLD activity (Fig. 2B). PMA treatment has been shown to activate ERK mitogen-activated protein kinases in PC12 cells (25). The lack of response to PMA was not due to an inability of RG20 cells to respond to phorbol ester, since a 15-min treatment with PMA activated cytosolic ERK mitogen-activated protein kinases to a similar extent in PC12K and RG20, as assessed by an in vitro assay using myelin basic protein as substrate (data not shown). The reduced ability of PMA to activate PLD in RG20 was also apparent in assays using intact cells (Fig. 2C). The intact cell assay could not be used to assess basal PLD activity, since radioactivity co-migrating with PEt was not significantly different for untreated cells incubated in the absence or presence of ethanol (data not shown). The ability of the in vitro assay to quantitate basal PLD activity is thus advantageous.
The mechanism by which by ␣ 2 AR expression inhibited PLD activity was explored. The ␣ 2 AR is coupled to inhibition of adenylate cyclase. However, dibutyryl cAMP (10 M) had no effect on membrane or cytosolic PLD activity in PC12K or RG20 cells at times from 15 to 60 min (data not shown), suggesting that the effect of ␣ 2 AR expression was not due to decreased cAMP levels. Forskolin (10 M) and epinephrine (1 M) likewise had no effect on membrane PLD activity (data not shown). The lack of effect of epinephrine suggested that the overexpressed ␣ 2A AR might be functionally "precoupled" to G i in RG20, as indicated previously for this cell line (27,28). PC12K and RG20 were therefore incubated with rauwolscine, an ␣ 2 AR antagonist, to uncouple the receptor from G i . Rauwolscine increased membrane PLD activity in RG20, but had no significant effect in PC12K (Fig. 3A). One explanation for the fact that rauwolscine did not restore PLD levels in RG20 cells to that seen in PC12K cells is that a portion of the precoupled ␣ 2 AR receptors in RG20 cells are rauwolscine-resistant, as suggested previously (28). Incubation with pertussis toxin, an inhibitor of G␣ i -mediated signaling, likewise increased membrane PLD activity in RG20 (Fig. 3B). Rauwolscine and pertussis toxin had no effect on PLD activity in PC12K. These findings were confirmed by PLD assays using intact cells (Fig.  3C). Pertussis toxin alone slightly increased basal PLD activity in both PC12K and RG20. Neither pertussis toxin nor rauwolscine significantly increased activation of PLD in response to PMA in PC12K (Fig. 3C, left). In contrast, both pertussis toxin and rauwolscine enhanced PMA-induced PLD activation in RG2 (Fig. 3C, right). Rauwolscine alone had no effect on PLD activity in intact PC12K or RG20 cells (data not shown). These results support the hypothesis that the effect of ␣ 2 AR overexpression on PLD activity is mediated by precoupling of the receptor to G␣ i .
In summary, membrane PLD activity in PC12 cells is positively regulated by PMA and NGF and negatively regulated by the ␣ 2 AR. Previous reports have suggested that ␣ 2 -agonists, with protein kinase C co-activation, can stimulate PLD activity in myristate-labeled intact or broken PC12 cells transfected with the ␣ 2 AR (29,30). We observed that an ␣ 2 -selective agonist did not significantly enhance the ability of PMA to activate PLD in intact oleate-labeled RG20 cells (data not shown). More than one form of PLD is expressed by mammalian cells (31,32). In one study, PMA-activated PLD was detected in fibroblasts isotopically labeled with either a fatty acid or an alkyl-lyso-PC precursor, while v-Src-activated PLD could be detected only using the fatty acid precursor (14). In Madin-Darby canine kidney cells, PMA-activated PLD preferentially utilizes alkyl-PCs (33). The alkyl-PC substrate used in our in vitro assays may thus detect a PLD preferentially regulated by G i . Other forms likely contribute to the activity measured in intact cells. Immunoblots obtained using an anti-PLD antibody (34) sug-gest that expression of PLD is similar in PC12K and RG20 (data not shown).
The mechanism of the novel negative interaction between G i and PLD is unknown. G␣ i2 , a pertussin toxin-sensitive G protein, can positively regulate PLA 2 (35). Negative regulation of PLD by G i could involve cross-talk between heterotrimeric G i and small GTP-binding proteins (24). For example, the ␣ 2 AR can activate ras via a G i -mediated pathway when transfected into fibroblasts (36). However, activation of rho is induced via the pertussis-toxin insensitive proteins G␣ 12 and G␣ 13 , but not by G␣ i2 or G␣ q , in fibroblasts (37). G␣ 12 can also activate signaling mediated by Ras and Rac (38). Since rho activates some forms of PLD, a previously unidentified negative influence of a G i component on the function of Rho (or another small GTP-binding protein) could potentially inhibit PLD activity. However, it should be noted that some agonists (e.g. LPA) bind to a G i -coupled receptor that activates both rho (39) and PLD (40). The mechanism by which protein kinase C isoforms activate PLD remain to be defined, but may involve protein-protein interactions rather than phosphorylation (41)(42)(43). Interestingly, the effects of PMA on PLD activity in fibroblasts were recently shown to be independent of Rac (44), but partially dependent on Rho (45). The roles of heterotrimeric G-proteins in PLD regulation appear worthy of further study.