(Received for publication, March 24, 1997)
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 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 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, 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.
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 (5-8). 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 phosphatidylcholine (PC) by PLD produces phosphatidic acid (PA). PA is proposed to play a role in signal transduction as a lipid mediator or mediator precursor (11-14). Some PLDs are regulated by the small GTP-binding proteins ARF and/or Rho (15-17). 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, Gi, 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 (
2AAR) have been
used to examine coupling of this receptor to its effectors (21). The
2AAR couples to the heterotrimeric GTP-binding protein
Gi, is widely expressed, and mediates the central
hypotensive effects of
2 agonists (22, 23).
Gi 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,
2AAR-expressing PC12 cells were used to examine the
potential role of Gi in regulating PLD activity.
PC12 cell lines were
maintained on Primaria plastic (Falcon) in Dulbecco's modified
Eagle's medium (DME) supplemented with either 10% fetal calf serum or
7.5% fetal calf serum plus 2.5% horse serum. PC12K (a wild-type cell
line) and RG20 (stably transfected with human 2A/DAR)
(21) were grown for 3-5 days in 35- or 100-mm dishes. Cytosolic and
membrane extracts were prepared as described (18). Briefly, cells were
incubated in culture medium at 37 °C, washed, resuspended in lysis
buffer (20 mM HEPES (pH 7.5), 80 mM
-glycerophosphate, 10 mM EGTA, 2 mM EDTA,
and 2 mM dithiothreitol), sonicated, and sedimented at
100,000 × g for 20 min at 4 °C. The supernatant
("cytosol") was used for ERK assays and the pellet ("membranes") for PLD assays. Protein concentrations in cell
extracts were determined using Coomassie reagent (Pierce).
PLD activity was measured in vitro using the fluorescent substrate BODIPY-phosphatidylcholine (BPC), 2-decanoyl-1-(O-(11-(4, 4-difluoro-5,7-dimethyl-4-bora-3a,4a-diaza-s-indacene-3-propionyl)amino)undecyl)sn-glycero-3-phosphocholine (Molecular Probes). BODIPY is a trademark of Molecular Probes, Inc. The assay was performed, as described previously (18-20), using 10 µg membrane protein and 1% butanol (for transphosphatidylation) with incubation for 60 min at 30 °C. BODIPY-labeled products were separated by thin-layer chromatography and quantitated by Molecular Dynamics FluorImager or Helena Laboratories scanning densitometer. Product formation was calculated as percent of total fluorescence.
PLD activity was assessed in intact cells as described previously (18-20). Briefly, cells were incubated overnight with [3H]palmitate or [3H]oleate, washed, and incubated in triplicate with agonists and 0.5% ethanol at 37 °C in DME, 10 mM HEPES. Products were separated by TLC; PEt and PA were quantitated by liquid scintillation spectrometry after scraping from the plate. Product formation was calculated as a percent of total radioactivity recovered from each lane.
ERK activity was assessed in vitro as described previously (10, 20), using myelin basic protein as substrate. Results were normalized for cytosolic protein.
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 activity and its regulation were therefore further characterized. PC12K membranes convert BPC to
PBt, the transphosphatidylation product generated by PLD in the
presence of butanol (Fig. 1A). Most of the
PLD activity was membrane-bound, with cytosol producing only 21% of
the PBt seen in membranes with equal amounts of cellular protein (data not shown). PBt can be metabolized to lyso-PBt by a calcium-independent PLA2 in membranes from mammalian cells (18), including
PC12K (Fig. 1A). PLD and PLA2 activities were
similar between undifferentiated and NGF-differentiated cells.
The phorbol ester PMA generally increases PLD in mammalian cells. PMA activated PLD in intact undifferentiated PC12K cells, as seen using either [3H]oleate (Fig. 1B) or [3H]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/DAR, a
Gi-coupled receptor, was examined in a PC12 cell line
overexpressing this receptor. Radioligand binding, using the ligand
[3H]RX 82102, confirmed that RG20 overexpressed the
2A/DAR (7585 fmol/mg of protein in RG20
versus 78 in PC12K) (data not shown). Basal membrane PLD
activity was substantially decreased in
2AAR-expressing cells, as shown for two clonal RG20 lines (Fig.
2A). This decrease was not due to enhanced
degradation of PBt by PLA2, since lyso-PBt production was
not greater in RG20 than in PC12K. Calcium-independent PLA2
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 2AR expression inhibited PLD
activity was explored. The
2AR is coupled to inhibition
of adenylate cyclase. However, dibutyryl cAMP (10 µM) 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
2AR expression was not due to decreased cAMP levels.
Forskolin (10 µM) and epinephrine (1 µM)
likewise had no effect on membrane PLD activity (data not shown). The
lack of effect of epinephrine suggested that the overexpressed
2AAR might be functionally "precoupled" to
Gi in RG20, as indicated previously for this cell line (27, 28). PC12K and RG20 were therefore incubated with rauwolscine, an
2AR antagonist, to uncouple the receptor from
Gi. 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
2AR
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
2AR 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 2AR. 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
2AR (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 Gi. Other forms
likely contribute to the activity measured in intact cells. Immunoblots
obtained using an anti-PLD antibody (34) suggest that expression of PLD
is similar in PC12K and RG20 (data not shown).
The mechanism of the novel negative interaction between Gi
and PLD is unknown. Gi2, a pertussin toxin-sensitive G
protein, can positively regulate PLA2 (35). Negative
regulation of PLD by Gi could involve cross-talk between
heterotrimeric Gi and small GTP-binding proteins (24). For
example, the
2AR can activate ras via a
Gi-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 Gi 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 Gi-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 (41-43).
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.
We thank Dr. John Hildebrandt for helpful suggestions. The advice and assistance of Drs. Stephen Lanier, Motohiko Satoh, Qing Yang, and Sebastien Saulnier-Blache are gratefully acknowledged.