Regulated Exocytosis in Chromaffin Cells

The ADP-ribosylation factor (ARF) GTP-binding proteins have been implicated in a wide range of vesicle transport and fusion steps along the secretory pathway. In chromaffin cells, ARF6 is specifically associated with the membrane of secretory chromaffin granules. Since ARF6 is an established regulator of phospholipase D (PLD), we have examined the intracellular distribution of ARF6 and PLD activity in resting and stimulated chromaffin cells. We found that stimulation of intact chromaffin cells or direct elevation of cytosolic calcium in permeabilized cells triggered the rapid translocation of ARF6 from secretory granules to the plasma membrane and the concomitant activation of PLD in the plasma membrane. To probe the existence of an ARF6-dependent PLD in chromaffin cells, we measured the PLD activity in purified plasma membranes. PLD could be activated by a nonhydrolyzable analogue of GTP and by recombinant myristoylated ARF6 and inhibited by specific anti-ARF6 antibodies. Furthermore, a synthetic myristoylated peptide corresponding to the N-terminal domain of ARF6 inhibited both PLD activity and catecholamine secretion in calcium-stimulated chromaffin cells. The possibility that ARF6 participates in the exocytotic reaction by controlling a plasma membrane-bound PLD and thereby generating fusogenic lipids at the exocytotic sites is discussed.

The ADP-ribosylation factor (ARF) GTP-binding proteins have been implicated in a wide range of vesicle transport and fusion steps along the secretory pathway. In chromaffin cells, ARF6 is specifically associated with the membrane of secretory chromaffin granules. Since ARF6 is an established regulator of phospholipase D (PLD), we have examined the intracellular distribution of ARF6 and PLD activity in resting and stimulated chromaffin cells. We found that stimulation of intact chromaffin cells or direct elevation of cytosolic calcium in permeabilized cells triggered the rapid translocation of ARF6 from secretory granules to the plasma membrane and the concomitant activation of PLD in the plasma membrane. To probe the existence of an ARF6dependent PLD in chromaffin cells, we measured the PLD activity in purified plasma membranes. PLD could be activated by a nonhydrolyzable analogue of GTP and by recombinant myristoylated ARF6 and inhibited by specific anti-ARF6 antibodies. Furthermore, a synthetic myristoylated peptide corresponding to the N-terminal domain of ARF6 inhibited both PLD activity and catecholamine secretion in calcium-stimulated chromaffin cells. The possibility that ARF6 participates in the exocytotic reaction by controlling a plasma membranebound PLD and thereby generating fusogenic lipids at the exocytotic sites is discussed.
ADP-ribosylation factors (ARFs) 1 comprise a family of 20-kDa monomeric GTP-binding proteins that were discovered as one of several cofactors required in the cholera toxin-catalyzed ADP-ribosylation of the trimeric G s proteins (1). Six mammalian family members have been identified which have been classified into three groups according to their size and sequence homology. ARF1, ARF2, and ARF3 form class I, ARF4 and ARF5 form class II, and ARF6 forms class III (1). Members of the ARF family are subjected to myristoylation at the N-terminal glycine residue, a lipid co-translational modification that appears essential for functional activity (2). ARFs are ubiquitous among eukaryotes with an amino acid sequence that is highly conserved across diverse species, suggesting a fundamen-tal role in cellular physiology. Indeed, ARF proteins have been implicated in a wide range of vesicle transport and fusion steps along the secretory pathway (3)(4)(5). These include budding, transport, and fusion steps in the Golgi complex, in the endoplasmic reticulum and in the endocytotic and exocytotic pathways.
The recent discovery that some members of the ARF family are effective activators of phospholipase D (PLD) has raised the possibility that a novel signal transduction pathway may regulate intracellular membrane traffic (6,7). PLD is an enzyme that catalyzes the hydrolysis of phosphatidylcholine to produce membrane-localized phosphatidic acid (PA) and soluble choline (8). In the presence of a primary alcohol, the enzyme can also catalyze a transphosphatidylation reaction that exchanges the polar headgroup of the phospholipid substrate with the given alcohol to form the corresponding phosphatidyl-alcohol (9). This unique and very useful property of PLD has been used to reveal PLD activation following agonist stimulation in many types of cells and tissues (10). Biochemical evidences suggest that multiple PLD isoenzymes with diverse mechanisms of activation occur in mammalian cells (11). An integral membrane-bound PLD that is highly specific for phosphatidylcholine as substrate and is activated by sodium oleate was recently purified (12). In addition, several forms of small G protein-dependent PLDs including ARF-sensitive (6, 7) and RhoA-sensitive (13) isoenzymes have been described. Phosphatidylinositol 4,5-bisphosphate (PIP 2 ), another important activator of PLD, seems to be generally required for the small G protein-dependent PLDs (6, 7) but not for the oleate-dependent PLD (12).
Protein kinase C appears also as a major regulator of PLD since phorbol esters are among the most effective stimuli of PLD reported in many cell types (11). To date, several mammalian PLDs have been cloned and sequenced (11,14).
Studies in neutrophils (15), pancreatic B cells (16), and pheochromocytoma PC12 cells (17) have suggested a role for PLD in exocytosis. In chromaffin cells, however, the occurrence of an agonist-regulated PLD activity remains a controversial issue (18,19). We recently described a secretory granule-associated ARF6 protein that may represent a key component of the exocytotic pathway in chromaffin cells (20). Since ARF6 is an established regulator of PLD (21), we examine here the PLD activity in resting and stimulated chromaffin cells. We found that stimulation of chromaffin cells triggered the rapid translocation of ARF6 from secretory granules to the plasma membrane and the concomitant activation of PLD in the plasma membrane. Both calcium-evoked PLD activation and calciuminduced catecholamine secretion could be inhibited by a synthetic peptide corresponding to the N-terminal domain of myristoylated ARF6. We propose that ARF6 may participate in the exocytotic reaction by controlling a plasma membranebound PLD and thereby contributing to the generation of fusogenic lipids at the exocytotic sites.

EXPERIMENTAL PROCEDURES
Culture of Chromaffin Cells-Chromaffin cells were isolated from fresh bovine adrenal glands and maintained in primary culture, essentially as described previously (22). Cells were usually cultured as monolayers either on 24 multiple 16-mm Costar plates (Cambridge, MA) at a density of 2.5 ϫ 10 5 cells/well or on 100-mm Costar plates at a density of 5 ϫ 10 6 cells/plate. To trigger exocytosis, chromaffin cells were washed twice with Locke's solution (140 mM NaCl, 4.7 mM KCl, 2.5 mM CaCl 2 , 1.2 mM KH 2 PO 4 , 1.2 mM MgSO 4 , 11 mM glucose, 0.56 mM ascorbic acid, and 15 mM Hepes, pH 7.2) and then stimulated 10 min with Locke's solution containing either 10 M nicotine or 59 mM K ϩ (made by decreasing NaCl isosmotically). Experiments were carried out at 37°C on 3-7-day-old cultures.
Permeabilization with Streptolysin O (SLO) and Cell Stimulation-Cultured chromaffin cells were washed four times with Locke's solution and twice with Ca 2ϩ -free Locke's solution (containing 1 mM EGTA). Permeabilization was performed with 15 units/ml SLO (Institut Pasteur, Paris, France) for 2 min in 200 l/16-mm well or in 5 ml/100-mm plate Ca 2ϩ -free KG medium (150 mM K ϩ -glutamate, 10 mM PIPES, pH 7.0, 5 mM nitrilotriacetic acid, 0.5 mM EGTA, 5 mM MgATP, 4.5 mM magnesium acetate, 0.2% bovine serum albumin). Cells were subsequently stimulated for 10 min in the presence of the compound to be tested in KG medium containing 20 M free Ca 2ϩ and 1 mM free Mg 2ϩ (22).
[ 3  H]Noradrenaline release after stimulation was determined by measuring the radioactivity present in the incubation medium and in cells after precipitation with 10% (w/v) trichloroacetic acid. Release of [ 3 H]noradrenaline is expressed as a percentage of total radioactivity present in the cells before Ca 2ϩ -induced stimulation. Release experiments were performed in triplicate on at least two different cell preparations. In the figures that are representative of a typical experiment, data are given as the mean of triplicate determinations on the same cell preparation Ϯ S.E. Endogenous catecholamine release was estimated by reverse phase high performance liquid chromatography with electrochemical detection as described previously (23).
Subcellular Fractionation of Cultured Chromaffin Cells-Cultured chromaffin cells were collected in 0.32 M sucrose, Tris 10 mM, pH 7.4, homogenized, and then centrifuged at 800 ϫ g for 15 min. After centrifugation at 20,000 ϫ g for 20 min, the pellet containing the crude membrane fraction was resuspended in 0.32 M sucrose (10 mM Tris, pH 7.4), layered on a continuous sucrose density gradient (1-2.2 M sucrose, 10 mM Tris, pH 7.4), and centrifuged for 90 min at 100,000 ϫ g. Twelve 1-ml fractions were collected from the top to the bottom and analyzed for PLD activity and protein content by the Bradford procedure. The distribution of dopamine-␤-hydroxylase (D␤H; chromaffin granule marker) and Na ϩ /K ϩ ATPase (plasma membrane marker) in the fractions of the gradient was estimated as described previously (24).
Plasma membranes were purified from fractions 2 and 3, which contained the highest Na ϩ /K ϩ ATPase activity. Fractions were diluted 10 times in 10 mM Tris, pH 7.4, 1 mM EDTA, and 1 mM dithiothreitol and membranes were collected by centrifugation for 30 min at 100,000 ϫ g.
Assay for Phospholipase D Activity in Cultured Chromaffin Cells-Chromaffin cells were labeled with 1 Ci/ml [9,10-3 H]myristic acid for 24 h at 37°C. Labeled cells were then washed, stimulated with nicotine or 59 mM K ϩ in the presence of 1% ethanol or permeabilized with SLO, and then stimulated with 20 M free Ca 2ϩ in the presence of 1% ethanol. Cells were subsequently collected and lipids extracted and separated with CH 3 OH/CHCl 3 /0.1 N HCl (1:1:1, v/v/v) according to the method of Bligh and Dyer (25). The lower lipid-containing phase was collected, spiked with a mixture of standard lipids containing L-␣ phosphatidic acid (1,2-diacyl-sn-glycero-3-phosphate) (PA) and 1-palmitoyl-2-oleoylsn-3-phosphoethanol (PEt), dried under vacuum, and redissolved in 20 l of CHCl 3 /CH 3 OH (2:1, v/v). Lipids were separated on one-dimensional TLC 0.25-mm oxalate-coated silica gel plates in a solvent system composed of CHCl 3 /CH 3 OH/CH 3 COOH/H 2 O (75:45:3:1, v/v/v/v). Labeled phospholipids were visualized with tritium imaging plates using the FUJIX BAS1000 Bio-Imaging Analyzer (Fuji, Tokyo, Japan). Standard lipids were stained with iodine vapor. R F values were 0.31 for PA and 0.67 for PEt.
ARF6-dependent Phospholipase D Activity in Purified Plasma Membranes-ARF-dependent phosphatidylcholine hydrolysis in purified plasma membranes was determined according to the procedure previ-ously described by Brown et al. (6) in the presence of MgCl 2 and 0.4 M NaCl to favor the nucleotide exchange on ARF. The reaction was carried out in a final volume of 60 l. All assays contained 6.25 g of plasma membrane proteins in buffer A (50 mM Na-Hepes, pH 7.5, 3 mM EGTA, 80 mM KCl, 2 mM MgATP, 400 mM NaCl, 4.5 mM MgCl 2 , 3 mM CaCl 2 , and 1 mM dithiothreitol). When indicated, 30 M GTP␥S and 1 M recombinant myristoylated ARF6 were included. The membranes and the above constituents in a volume of 34 l were preincubated for 30 min at 37°C. The reaction was subsequently started by the addition of 24 l of lipid substrate and 1% ethanol. The final concentration of lipids in the assay were 135 M PE, 12 M PIP 2 , 8 M phosphatidylcholine (PC), and 1 Ci of L-␣-dipalmitoylphosphatidylcholine (2-palmitoyl-9,10-3 H-labeled). Lipids were previously dried and sonicated in buffer A without MgCl 2 and CaCl 2 . The incubation was performed for 30 min at 37°C and stopped by the addition of 350 l of 1 M HCl, 5 mM EGTA, and 1 ml of CHCl 3 /CH 3 OH/HCL (50:50:0.3, v/v/v). After vortexing and centrifugation (2000 ϫ g for 5 min), 400 l of the aqueous phase was extracted and analyzed as described above.
Recombinant myristoylated ARF6 protein (rARF6) was produced in a bacterial expression system containing the N-myristoyltransferase in the plasmid pACYC177/ET3d, and purified according to the procedure previously described by Haun et al. (27).
Protein Determination, Electrophoresis, and Immunoblotting-Protein concentration was routinely determined using the Bradford procedure with Bio-Rad dye reagent and bovine serum albumin as standard. One dimensional SDS-polyacrylamide gel electrophoresis was performed on 12% acrylamide gel in Tris-glycine buffer (24). Proteins were transferred to nitrocellulose sheets at a constant current of 120 mV for 1 h. Blots were developed with secondary antibodies coupled to horseradish peroxidase (Amersham, Les Ulis, France), and immunoreactive bands were detected with the ECL system (Amersham).

Activation of Phospholipase D in Response to Chromaffin Cell
Stimulation-To measure PLD activity, we made use of the specific transphosphatidylation reaction in which the phospholipid headgroup is exchanged for ethanol, producing PEt at the expense of PA. Therefore, cultured chromaffin cells were labeled with [ 3 H]myristic acid. Preliminary time-course experiments indicated that maximal incorporation of the radioactivity in total lipids occurred after 20 h of incubation. [ 3 H]Myristic acid was predominantly incorporated into phosphatidylcholine (data not shown), which represents the major PLD substrate (10 (Table I). Both secretagogues elicited a similar increase in PA and PEt levels, suggesting that an agonist-stimulated phospholipase D activity was present in chromaffin cells.
To investigate the mechanisms of activation of PLD, chromaffin cells were permeabilized with SLO to clamp cytosolic Ca 2ϩ at known values and to control the cytosolic levels of nucleotides. Permeabilized cells were maintained under resting conditions or stimulated with 20 M Ca 2ϩ in the presence of 5 mM MgATP and 1% ethanol. We found that calcium strongly stimulated the formation of both [ 3 H]PEt (Fig. 1) and [ 3 H]PA (data not shown), suggesting that PLD activation might be a component of the calcium signaling cascade in stimulated chromaffin cells. PLD can be activated by protein kinase C in many cell types (11). We looked, therefore, at the effect of phorbol 12-myristate 13-acetate (PMA) at 100 nM, a concentration known to give maximal stimulation of protein kinase C in chromaffin cells (28). However, as seen in Fig. 1, PMA was unable to significantly modify the basal or the calcium-evoked formation of [ 3 H]PEt in SLO-permeabilized chromaffin cells, indicating that PLD is not modulated by a protein kinase C-dependent pathway. In agreement, Purkiss et al. (18) previously reported that neither bradykinin nor 12-O-tetradecanoylphorbol-13-acetate stimulated PLD in intact chromaffin cells.
Current evidence supports the existence of two types of PLD activities: an oleate-dependent form and a form modulated by monomeric GTP-binding proteins (21). To probe the involvement of a GTP-binding protein in the calcium-evoked PLD activation, we introduced a nonhydrolyzable analogue of GTP into SLO-permeabilized chromaffin cells. Fig. 1 shows that GTP␥S at 20 M produced a large increase in the Ca 2ϩ -dependent PLD activity, revealing that a G protein probably regulates PLD in chromaffin cells. In contrast, AlF 4 Ϫ , which selectively activates trimeric G proteins (29), inhibited the Ca 2ϩ -dependent PLD activity in SLO-permeabilized cells by more than 80% (Fig. 1). Taken together, these findings suggest that the Ca 2ϩdependent PLD activity in chromaffin cells might be regulated by both monomeric and heterotrimeric G proteins.
Subcellular Localization of the Phospholipase D Activity in Stimulated Chromaffin Cells-To determine the intracellular localization of the PLD activity in stimulated chromaffin cells, we analyzed the phospholipid content of subcellular fractions collected from a sucrose density gradient layered with a crude membrane preparation. Chromaffin cells labeled with [ 3 H]myristic acid were permeabilized with SLO and subsequently exposed to 1% ethanol and GTP␥S in the presence or absence of 20 M free Ca 2ϩ . Cells were then collected and processed for subcellular fractionation. As expected, the [ 3 H]PEt levels detected in the fractions obtained from resting cells remained negligible ( Fig. 2A). Interestingly, radioactive [ 3 H]PEt was essentially collected in fractions 2 and 3 in gradients prepared from stimulated cells (Fig. 2A). These fractions contain plasma membranes as estimated by the Na ϩ /K ϩ -ATPase activity. It is noteworthy that fractions 11 and 12, enriched in chromaffin granules revealed by the peak of D␤H, contain very little radioactive [ 3 H]PEt in both resting and stimulated cells. In other words, PLD activity in chromaffin cells is not associated with secretory granules. However, a rise in cytosolic calcium triggers the activation of a G protein-regulated PLD activity, essentially in the plasma membrane of chromaffin cells.  Cells were subsequently collected and processed for subcellular fractionation. Fractions from sucrose density gradients (10 g of protein) were subjected to gel electrophoresis and immunodetection on nitrocellulose sheets using anti-ARF6 antibodies. In resting cells, ARF6 protein is essentially present in fractions 10 and 11 containing chromaffin granules. Stimulation with nicotine triggers the translocation of ARF6 to fraction 3 enriched in plasma membranes.

Translocation of ARF6 from Secretory Chromaffin Granules to the Plasma Membrane in Stimulated Chromaffin Cells-We
previously reported that ARF6 is specifically associated with secretory granule membranes in chromaffin cells, most likely through an interaction with the ␤ subunit of a granule-bound trimeric G protein (20). Stimulation of chromaffin cells with nicotine or direct elevation of cytosolic Ca 2ϩ in permeabilized cells triggered the rapid dissociation of ARF6 from secretory granules (20). To identify the target membrane to which ARF6 translocates in stimulated cells, we compared here the distribution of ARF6 in membrane fractions collected from resting or nicotine-stimulated chromaffin cells. Fig. 2B illustrates an immunoreplica analysis, using an anti-ARF6 antibody, of fractions collected from a sucrose density gradient layered with chromaffin cell crude membranes. The parallel distribution of ARF6 immunoreactivity and D␤H activity (Fig. 2, compare A  and B) confirms the specific association of ARF6 with chromaffin granule membranes in resting cells, in agreement with our previous observations (20). Stimulation of chromaffin cells with 10 M nicotine for 5 min completely modified the distribution of ARF6 in the sucrose gradient since the highest immunosignal for ARF6 was then detected in fraction 3 containing the plasma membrane (Fig. 2B). Similar results were obtained in gradients prepared from SLO-permeabilized cells stimulated by a rise in cytosolic calcium (Fig. 4). These findings strongly suggest that ARF6 translocates from secretory granules to the plasma membrane upon chromaffin cell stimulation.
Presence of an ARF6-regulated Phospholipase D Activity in Chromaffin Cell Plasma Membranes-The recent identification of PLD as a possible effector of ARF proteins (6, 7) led us to investigate whether PLD might be activated by an ARF6-dependent pathway in chromaffin cell plasma membranes.
To probe the existence of an ARF6-dependent PLD in chro-maffin cells, we first examined the effect of recombinant ARF6 on PLD activity associated with purified plasma membranes. Plasma membranes recovered from a sucrose density gradient were assayed for PLD activity using lipid vesicles labeled with [ 3 H]dipalmitoyl phosphatidylcholine. As illustrated in Fig. 3 These results indicate that the plasma membrane-associated PLD can be activated by myrARF6 in chromaffin cells.
To further correlate the PLD activity with the presence of ARF6 in the plasma membrane, we compared the PLD activity in plasma membranes purified from either resting or stimulated SLO-permeabilized chromaffin cells. The amount of ARF6 in the two membrane preparations was quantified by immunodetection on nitrocellulose sheets using specific anti-ARF6 antibodies. As shown in Fig. 4A, stimulation with 20 M free Ca 2ϩ strongly enhanced the quantity of ARF6 associated with the plasma membrane. We then compared the PLD activity in the two plasma membrane fractions by measuring the GTP␥S-dependent formation of [ 3 H]PEt (Fig. 4B). We found that the translocation of ARF6 in stimulated cells was accompanied by a significant increase in the GTP-dependent PLD activity in the plasma membrane (Fig. 4B), suggesting that endogenous ARF6 was able to activate the plasma membraneassociated PLD. Accordingly, the GTP-dependent PLD activity measured in plasma membranes from stimulated cells could be totally reversed by anti-ARF6 antibodies (Fig. 4). These findings are in line with the idea that, in stimulated chromaffin cells, ARF6 translocates from secretory granules to the plasma membrane to activate a plasma membrane-bound PLD.
Correlation between Catecholamine Secretion and ARF6-Phospholipase D Activation in Chromaffin Cells- Fig. 5 illustrates the time course and calcium dose-response curve for secretion and PLD activation in SLO-permeabilized chromaffin cells. Secretion was estimated by measuring the release of endogenous catecholamines, and PLD activity was detected by measuring the formation of labeled PEt and PA. We found a strong similarity between the calcium sensitivity for PLD activation and the calcium concentration required for the exocytotic reaction in permeabilized cells (Fig. 5A). Furthermore, time-course experiments revealed that catecholamine secretion was always accompanied by the formation of PEt (Fig. 5B), in agreement with the hypothesis that PLD activation represents an important event in the pathway of regulated exocytosis. We previously reported that translocation of ARF6 was maximal within 1 min of stimulation (20). By comparison, maximal PLD activation was observed after 3 min of stimulation (Fig. 5B), an observation that correlates well with the idea that PLD stimulation requires the translocation of ARF6 to the plasma membrane.
Since PLD selectively requires primary alcohols to catalyze the transphosphatidylation reaction (9), we compared the effects of ethanol, 1-butanol, and 2-butanol on Ca 2ϩ -evoked catecholamine secretion in permeabilized chromaffin cells. We found that, at 1.5%, both ethanol and 1-butanol significantly reduced the secretory response whereas 2-butanol had little effect (Table II). Since ethanol and 1-butanol can divert the PLD production of PA to the formation of phosphatidylethanol and phosphatidylbutanol, these results suggest that the PA produced by PLD may play an essential role in the regulation of exocytotic events in chromaffin cells.
To further probe the importance of ARF6-regulated PLD in the exocytotic process, we used various synthetic peptides corresponding in sequence to the N-terminal region of ARF proteins and investigated their effects on catecholamine secretion and PLD activation in chromaffin cells. Fig. 6 compares the effects of the myristoylated myrARF6-(2-13) peptide, the nonmyristoylated ARF6-(2-13) peptide, and the myristoylated myrARF1-(2-17) peptide on calcium-evoked [ 3 H]noradrenaline release and [ 3 H]PEt formation in SLO-permeabilized chromaffin cells. MyrARF6-(2-13) produced a dose-dependent inhibition of Ca 2ϩ -dependent catecholamine secretion with a mean inhibitory dose, ID 50 , of approximately 50 M (Fig. 6A). In contrast, neither ARF6-(2-13) nor myrARF1-(2-17) affected noradrenaline release in the range of concentrations tested (Fig. 6A). We then examined the effects of these peptides on PLD activity in stimulated chromaffin cells. At 50 M, myrARF6-(2-13) inhibited both the calcium-stimulated PLD activity and the calcium-evoked noradrenaline release by 50% (Fig. 6B). MyrARF1-(2-17) did not modify the formation of [ 3 H]PEt, and ARF6-(2-13) reduced only slightly the Ca 2ϩ -dependent PLD activity in chromaffin cells (Fig. 6B). Thus, there is a close relationship between the effects of the various N-  terminal ARF peptides on catecholamine secretion and their ability to inhibit PLD in stimulated chromaffin cells. Our results suggest that myristoylated ARF6 may be implicated in regulated exocytosis in chromaffin cells most likely by stimulating the plasma membrane-associated PLD to produce PA.

DISCUSSION
Dramatic advances have recently been made in our understanding of the protein machinery responsible for the formation, targeting, and fusion of vesicles along the secretory pathway. Most of the emerging models emphasize the convergence in protein molecules and mechanisms underlying the multiple steps of intracellular vesicular transport (30,31). Since ARF proteins belong to the molecules that have been implicated as ubiquitous regulators in membrane traffic (3)(4)(5), we recently investigated the possible function of ARF in calcium-regulated exocytosis in chromaffin cells (20). We found that ARF6 is specifically associated with the membrane of secretory chromaffin granules, most likely through an interaction with ␤␥ subunits of a trimeric G protein. Interestingly, nicotine-in-duced stimulation of intact cells or direct elevation of cytosolic calcium in permeabilized cells triggered the rapid dissociation of ARF6 from secretory granules (20). Although ARF proteins are generally believed to cycle on and off the membrane in a manner that is tightly coupled to the binding and hydrolysis of GTP, we could not detect ARF6 in the cytosol in subcellular fractionation experiments (20) or among the cytosolic proteins released through the pores created in the plasma membrane of SLO-permeabilized cells. 2 Thus, we postulated that ARF6 translocated from chromaffin granules to an unknown membrane-bound compartment upon cell stimulation. Interestingly, the subcellular distribution of human ARF proteins has recently been examined in detail in Chinese hamster ovary cells, and ARF6 was the only isoform that could not be detected in the cytosol (32). It is also noteworthy that both wild-type and mutant forms of ARF6 were exclusively localized in membrane compartments when overexpressed in fibroblasts (33). Thus, ARF6 seems to behave quite distinctly from other ARFs, at least regarding its membrane binding activity and intracellular localization. Based on subcellular fractionation techniques and immunological detection, we report here that agonist stimulation triggers the translocation of ARF6 from secretory granules to the plasma membrane in chromaffin cells. Moreover, we found a close correlation between the presence of ARF6 in the plasma membrane and the activation of a GTP-dependent PLD activity in the plasma membrane, suggesting that PLD may be a possible effector for ARF6 in the exocytotic pathway.
The participation of ARF in exocytosis in endocrine and neuroendocrine cells has been previously postulated (34 -36). To probe the importance of ARF6 in agonist-stimulated PLD activity and exocytotic response in chromaffin cells, we used here synthetic N-terminal ARF peptides described to block ARF activities in various cellular processes (37,38). We found that myrARF6-(2-13), a peptide corresponding in sequence to the myristoylated N-terminal domain of ARF6, specifically inhibited calcium-evoked catecholamine release and PLD activation in stimulated chromaffin cells. By comparison, the nonmyristoylated ARF6-(2-13) peptide had little effect, an observation that may be related to the presumed importance of the myristoyl group in the binding of ARF6 to membranes (2,33). Dose-response experiments indicated that myrARF6-(2-13) was able to block almost completely the calcium-evoked secretory response in permeabilized chromaffin cells (Х90% inhibition). Thus, ARF6 activation of PLD may represent a key event in the exocytotic pathway in neuroendocrine cells.
Interestingly, we found that AlF 4 Ϫ , which activates specifically heterotrimeric G proteins (29), inhibited the calciuminduced PLD activity in permeabilized chromaffin cells. This observation correlates well with our previous findings that AlF 4 Ϫ can prevent the calcium-induced uncoupling of ARF6-G␤␥ on the secretory granule membrane (20) and strongly reduce the calcium-evoked exocytotic activity in stimulated cells (22). The regulation of ARF activities in the Golgi complex (39) and in the endocytotic pathway (40) by trimeric G proteins has already been reported. In chromaffin cells, activation of the secretory granule-associated Go inhibits the ATP-dependent priming step of exocytosis (22,23). This suggests that activated G o blocks the exocytotic machinery when the ␣ o subunit is dissociated from ␤␥. Our data support a model in which exocytosis requires the inactivation of the granule-bound G o , leading to the reassociation of ␣ o with ␤␥. G␤␥ interacting with G␣ o is then unable to retain ARF6, which translocates to the plasma membrane and activates PLD. We recently identified another putative effector of G o in the exocytotic pathway, namely the monomeric GTP-binding protein Rho, which seems to regulate the peripheral actin network (41). Interestingly, several reports describe a reciprocal regulatory relationship between actin reorganization and PLD activity (42)(43)(44). Thus, an attractive speculation is that the granule-bound G o plays a double control of the plasma membrane-associated PLD in the exocytotic pathway: through ␤␥ and ARF6, which directly activate the enzyme and through ␣ o , and Rho, which may modulate PLD by a specific cytoskeletal reorganization.
PLD hydrolyzes PC to generate PA and choline. PA seems to be important for the exocytotic reaction since primary alcohols, which divert the production of PA to phosphatidylalcohol, inhibited calcium-evoked catecholamine secretion in permeabilized cells. How does PA relate to our current understanding of the protein machinery responsible for regulated exocytosis? In principle, PA can be rapidly converted into diacylglycerol, and one of the functions of the PLD pathway might be, therefore, the provision of diacylglycerol with the consequent activation of protein kinase C. PA is also a known stimulator of phosphatidylinositol 4-phosphate 5-kinase (45), an enzyme that has been implicated in the priming of exocytosis in chromaffin and PC12 cells (46,47). Perhaps the most exciting speculation relates, however, to the changes in the lipid bilayer properties created by the activation of PLD. The conversion of PC to PA in the plasma membrane replaces a nonfusogenic phospholipid with a fusogenic one (48), a potentially positive effect for the exocytotic reaction. The local elevation of PA may also favor interactions between membranes and annexins (49). Annexin II is a Ca 2ϩdependent phospholipid-binding protein that forms cross-links between secretory granules and plasma membranes in stimulated chromaffin cells (50). Furthermore, the translocation of annexin II from the cytosol to the subplasmalemmal region in stimulated cells seems to be an essential event for catecholamine secretion (51). In view of a recent report describing that PLD lowers the calcium concentration required for annexin-induced liposome aggregation by increasing the PA composition (52), it is tempting to postulate that the activation of PLD and the production of PA may facilitate annexin II-mediated membrane-membrane apposition in the early stages of the exocytotic pathway in neuroendocrine cells. Since the role played by the lipid bilayer is almost completely lacking in the recent models for calcium-regulated secretion, it will be quite interesting to prove or refute any of these possibilities.