Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

The receptor-mediated mechanisms for activation of phospholipase D (PLD) remain largely undefined. PLD can be activated in various types of cells by pharmacologic stimulants of protein kinase C and calcium-mobilizing agents (Exton, 1997), but there has been no clear demonstration that the enzyme can be activated directly through receptor-regulated trimeric G proteins or tyrosine kinases, as is the case for phospholipase C (PLC). Ongoing studies in this laboratory have shown that the PLD activity in RBL-2H3 mast cell line is regulated by calcium, protein kinase C, and unidentified receptor-mediated signal or signals. Although the inhibition of calcium influx and protein kinase C abrogates stimulation of PLD by thapsigargin and phorbol-12-myristate-13-acetate, respectively, a significant fraction of the PLD response to receptor agonists is resistant to such inhibition. A common feature, whether PLD is activated by pharmacologic stimulants or receptor, is that this activation is markedly synergized by treatment of RBL-2H3 cells with cholera toxin in a cAMP-independent manner (Cissel et al., 1998; P. F. Fraundorfer, W. A. Patton, J. Moss, and M. A. Beaven, submitted for publication).

Studies with recombinant PLD in vitro show that PLD1, which exists as alternatively spliced variants 1a and 1b, is synergistically activated by various small GTPases and protein kinase C in the presence of phosphatidylinositol 4,5-bisphosphate (Hammond et al., 1997). A PLD2 has also been cloned (Colley et al., 1997). This PLD, in contrast to PLD1, is constitutively active in the presence of phosphatidylinositol 4,5-bisphosphate, and this activity is not affected by the GTPases and protein kinase C, either alone or in combination (Colley et al., 1997). Except for the presence of a pleckstrin homology-like domain, no other recognizable sequence motifs have been described to account for the activation of PLD by these stimulants (Steed et al., 1998; Holbrook et al., 1999).

To investigate the possibility that receptor- or cholera toxin-mediated release of -subunits from trimeric G proteins (subunits of G proteins are denoted as G, G, G, and G with isoform undesignated or designated) provides an additional signal for the activation of PLD, we undertook studies with the G protein stimulant compound 48/80. This agent, which was originally described as a mast cell secretagogue, directly stimulates G_i and G_o subfamilies of G proteins to promote GDP-GTP exchange and dissociation into their constituent - and -subunits (Tomita et al., 1991; Tanaka et al., 1998). In mast cells, compound 48/80 partially penetrates the plasma membrane to stimulate membrane GTPase activity (Mousli et al., 1990a, and citations therein) and stimulates PLC-mediated events (Nakamura and Ui, 1985; Senyshyn et al., 1998, and citations therein). The present study was conducted with RBL-2H3 cells because the expression of G_i proteins is enhanced more than 7-fold in these cells after treatment with quercetin (Senyshyn et al., 1998). This treatment also transforms the cell from an unresponsive to a compound 48/80-responsive phenotype, thus providing a useful negative control for our experiments (Senyshyn et al., 1998). As described here, compound 48/80 activates PLD in a pertussis toxin-sensitive manner. Our findings support the notion that PLD is activated by G protein -subunits in addition to signals transduced via calcium and protein kinase C.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Reagents. Quercetin, compound 48/80, L--phosphatidic acid, isobutanol, and p-nitrophenyl-N-acetyl--D-glucosamide were obtained from Sigma Chemical. Ro31-7549 was obtained from Alexis Biochemicals (San Diego, CA). n-Butanol was purchased from Mallinckrodt (St. Louis, MO). Phosphatidylethanol and phosphatidylbutanol were purchased from Avanti Polar Lipids (Pelham, AL). Radiolabeled compounds and assay kits for the measurement of inositol 1,4,5-trisphosphate were obtained from DuPont-New England Nuclear (Boston, MA). Recombinant G_i-3 was obtained from Calbiochem (La Jolla, CA). Anti-G_i3, -G, and -G antibodies were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Anti-human PLD1 antibody was purchased from Upstate Biotechnologies (Lake Placid, NY). Phosphatase-labeled goat anti-rabbit IgG was obtained from Kirkegaard & Perry Laboratories Inc. (Gaithersburg, MD). Bacterial toxins were purchased from List Biologicals (Campbell, CA). Cell culture reagents and protein expression systems were obtained from Gibco/Life Technologies (Gaithersburg, MD). Silica TLC plates were obtained from EM Sciences (Gibbstown, NJ). Ni-NTA resin was purchased from Qiagen (Valencia, CA). The antibodies against rat PLD1 (epitope, residues 526-542) and rat PLD2 (epitope, residues 28-42) were custom prepared (anti-PLD1 was obtained from Lofstrand, Gaithersburg, MD; anti-PLD2 was obtained from Genosys, The Woodlands, TX) on the basis of published sequences (Nakashima et al., 1997). Sequences were verified by reverse transcription-polymerase chain reaction for the PLD isoforms in RBL-2H3 cells in our laboratory (P. G. Holbrook, A. Vaid, and M. A. Beaven, unpublished data). Myristoylated ADP-ribosylation factor (mARF-1) was kindly supplied by Dr. Joel Moss (National Heart, Lung, and Blood Institute, Bethesda, MD). All other reagents were obtained from sources listed elsewhere (Ali et al., 1996; Senyshyn et al., 1998).

Preparation of Cell Cultures for Experiments. RBL-2H3 cells were maintained as monolayer cultures in modified Eagle's medium (minimum essential medium supplemented with Earle's salts) supplemented with 13% fetal calf serum, 1 mM glutamine, and 1% antibiotic-antimycotic solution (Gibco/Life Technologies). Trypsinized cells in culture dishes or 24-well multiwell cluster plates (3 × 10⁵ cells/0.4 ml medium/well) were incubated overnight in complete growth medium, 0.5 µg/ml O-dinitrophenol-specific IgE when required for stimulation with antigen (dinitrophenylated bovine serum antigen), radiolabeled reagents (Maeyama et al., 1986), and 30 µM quercetin as required (Senyshyn et al., 1998). Where indicated, pertussis toxin (0.2 µg/ml for 3 h), cholera toxin (1 µg/ml for 4 h), or [³H]myristic acid (2 µCi/ml for 90 min) was added to the cultures during the final period of incubation. For each experiment, cells were washed to remove quercetin, which would otherwise suppress intracellular kinases and cell responses (Senyshyn et al., 1998), and the medium was replaced with a glucose-saline, piperazine-N,N'-bis(2-ethanesulfonic acid) (PIPES)-buffered medium (25 mM PIPES, pH 7.4, 119 mM NaCl, 5 mM KCl, 0.4 mM MgCl₂, 1.0 mM CaCl₂, 5.6 mM glucose). In some experiments, calcium-free PIPES-buffered medium was prepared by substituting 0.1 mM EGTA for CaCl₂. Also where indicated, a 10 µM concentration of the protein kinase C inhibitor Ro31-7549 or 50 mM concentration of the PLD inhibitor butanol was added 10 min before the addition of stimulant. After stimulation and collection of the supernatant medium, cells were lysed in 0.1% Triton X-100.

Cell Permeabilization. Studies were performed with a genetically altered subline (RBL-2H3-m1) of RBL-2H3 cells made to express muscarinic m1 receptors and thus respond to carbachol as well as to antigen (Choi et al., 1993). Quercetin-treated RBL-2H3-m1 cells, labeled with [³H]myristic acid as described earlier, were permeabilized with streptolysin-O exactly as described (Pinxteren et al., 1998) except that a different source and concentration of streptolysin-O were used (300 U/ml; Sigma Chemical Co.). Experiments were performed as described by Pinxteren et al. (1998).

Measurement of Hexosaminidase, Inositol Phosphates, and Inositol 1,4,5-Trisphosphate. Secretion was determined by measurement of release of the granule marker hexosaminidase, which hydrolyses p-nitrophenyl-N-acetyl--D-glucosamide to the chromophore p-nitrophenol. A colorimetric assay based on this reaction was used to measure hexosaminidase in 10-µl aliquots of medium and cell lysate as described elsewhere (Choi et al., 1993). Values (mean ± S.E.) were expressed as the percentage of intracellular hexosaminidase released into the medium after correction for spontaneous release (2-4%). For measurement of total inositol phosphates, cells were incubated overnight in 24-well cluster plates in the presence of myo-[³H]inositol (4 µCi/ml). The next morning, the cultures were washed twice with PIPES-buffered medium before the final addition of the same buffer containing 10 mM LiCl. The cultures were incubated at 37°C for 10 min before the addition of compound 48/80. The reactions were terminated by placing the cultures on ice, and water- and chloroform-soluble [³H]inositol metabolites were assayed exactly as described previously (Maeyama et al., 1986). The amounts of water-soluble [³H]inositol phosphates formed were expressed as percent of chloroform-soluble [³H]inositol phospholipids in unstimulated cells. Values (mean ± S.E.) were corrected for spontaneous formation of [³H]inositol phosphates in unstimulated cells (1-3%). Inositol 1,4,5-trisphosphate was assayed in cell extracts by use of a receptor-binding assay kit (DuPont-New England Nuclear). The assay was performed according to the manufacturer's instructions, except that the final aqueous extract was passed through small ultrafiltration units to exclude proteoglycans that may interfere with the assay (Hide and Beaven, 1991).

Measurement of [³H]Phosphatidic Acid, [³H]Phosphatidylethanol, and [³H]Phosphatidylbutanol. Cultures (in 24-well plates), labeled with [³H]myristic acid, were washed with the PIPES-buffered medium, as described above, and then incubated in 0.2 ml of the PIPES-buffered medium in the absence or presence of 172 mM (1%) ethanol or 55 mM (0.5%) n-butanol as indicated for 10 min before stimulation. In the presence of ethanol or n-butanol, a phosphatidylalcohol is formed at the expense of phosphatidic acid, the normal PLD product, via a PLD-specific transphosphatidylation reaction (Dennis et al., 1991). Radiolabeled phosphatidic acid and phosphatidylethanol (or phosphatidylbutanol) were isolated and quantified through minor modifications of previously described procedures (Ali et al., 1996). Reactions were terminated by the addition of 0.75 ml of a mixture of chloroform/methanol/4 N HCl (100:200:2 v/v/v) to each culture well. The resultant monophasic mixture was separated into two phases by addition of 0.25 ml of chloroform, which contained 30 µg of unlabeled phosphatidic acid and phosphatidylethanol (or phosphatidylbutanol), and 0.25 ml of 0.1 N HCl. From the lower chloroform phase, 0.5 ml was removed and evaporated to dryness under nitrogen. The residue was dissolved in 0.1 ml of a mixture of chloroform/methanol (2:1), and a 25-µl sample of this mixture was then subjected to thin-layer chromatography on silica-gel sheets by use of chloroform/methanol/glacial acetic acid (65:15:2 v/v/v; Tomhave et al., 1994). The sheet was air dried and then exposed to iodine vapor to visualize the phospholipids. Phosphatidic acid and phosphatidylalcohol were then cut from the sheets for assay of tritium. The total amount of [³H]phosphatidylalcohol formed was calculated as a percentage of ³H-phospholipid in Triton X-100 extracts of unstimulated cells from the same set of cultures. Values were either left uncorrected or corrected for formation of [³H]phosphatidylalcohol in the absence of stimulant (0.1-0.2%) as noted in the figure legends.

Measurement of GTPase Activity. GTPase activity was determined through minor modifications of previously described procedures (Chahdi et al., 1998b). Cells were incubated for 10 min on ice in buffer A (1 mM ATP, 2 mM MgCl₂, 0.1 mM EDTA, 1 mM dithiothreitol, 150 mM NaCl, 1 mM phenylmethylsulfonyl fluoride, 50 mM triethanolamine-HCl, pH 7.4). Cells (~10⁸ cells) were sonicated, and the mixture centrifuged for 10 min at 4000g. The supernatant fraction was removed, adjusted to 4 ml with buffer A, and then centrifuged at 38,000g for 30 min. The pellet was resuspended in 1 ml of buffer A, and after the determination of protein concentration (BCA assay kit; Pierce, Rockford, IL), GTPase activity was assayed in samples containing 20 µg of protein in the presence of the indicated amounts of compound 48/80 in a final volume of 80 µl in buffer A. The assay mixture was incubated for 10 min at 25°C before the addition of 20 µl of [-³²P]GTP (30 Ci/mmol; final concentration, 0.1 µM) to initiate the reaction. After further incubation for 15 min at 25°C, the reaction was terminated by the addition of 0.7 ml of an ice-cold suspension of 5% (w/v) charcoal (pH 7.4) to adsorb radiolabeled nucleotides. The suspensions were centrifuged for 15 min at 9000g at 4°C, and 0.4 ml of the supernatant fraction was mixed with 3.6 ml of scintillation cocktail for assay of free [³²P]phosphate.

Electrophoretic Separation and Immunoblotting of G Protein Subunits and PLD. The procedures for the preparation of whole-cell lysates and the soluble and membrane fractions, as well as the separation and detection of G_i-3 protein by SDS-polyacrylamide gel electrophoresis and Western blotting, were performed as described previously (Hirasawa et al., 1995). Gels were loaded with equivalent amounts of protein. G_i-3 was detected with anti-G_i-3 antibody and phosphatase-labeled goat anti-rabbit IgG as the secondary antibody. PLD isoforms were separated on 4 to 20% gradient Tris-glycine gels, and G and G subunits were separated on 10 and 18% Tris-glycine gels, respectively, and detected according to the Amersham enhanced chemiluminescence system (Arlington Heights, IL). Immunoblotting was performed with antibodies against PLD1, PLD2, G_i-3, G₁, G₂, G₁, and G₂. The relative amounts of protein were determined by densitometric scanning (ImageQuant).

Studies with G Proteins in Permeabilized Cells. Recombinant histidine-tagged 22-subunits, prepared from human G₂ and G₂ cDNA (a gift from Dr. Narasimhan Gautam, Washington University School of Medicine, St. Louis, MO), were expressed using the Bac-to-Bac baculovirus Sf9-insect cell system (Gibco/Life Technologies). Sf9 cells (30 × 10⁶/ml) were suspended in fresh medium and incubated simultaneously with G₂- and G₂-containing baculovirus. The cells were then diluted (3 × 10⁶/ml) with additional medium and transferred to a 250-ml Erlenmeyer flask. Cells were maintained at 27°C and stirred at a rate of 90 rpm for 72 h. Cells were harvested by centrifugation (800g for 5 min at 4°C) and placed in lysis buffer [20 mM HEPES, 150 mM NaCl, 6.5 mM 3-[(3-cholamidopropyl)dimethylammonio]propanesulfonate (CHAPS), 1 mM phenylmethylsulfonyl fluoride, pH 8.0] for 30 min at 0°C. The cell lysate was centrifuged (1000g for 10 min) to remove cell debris. The supernatant fraction was further clarified through centrifugation (100,000g for 1 h) before the addition of 20 mM imidazole and 1 ml Ni-NTA resin. The mixture was stirred gently at 4°C for 1 h to permit binding of the histidine-tagged proteins to the resin, after which the resin mixture was poured into 0.8 × 4-cm polypropylene filter columns (Bio-Rad, Hercules, CA). The columns were washed extensively with a washing buffer (20 mM HEPES, 3 mM MgCl₂, 500 mM NaCl₂, 0.05% CHAPS, 40 mM imidazole, pH 8.0) before elution of proteins with 2 ml of an elution buffer (40 mM HEPES, 3 mM MgCl₂, 50 mM NaCl, 300 mM imidazole, 0.025% CHAPS, pH 8.0). The eluted proteins (>80% G by Western blot) were concentrated by the use of Nanosep ultrafiltration centrifugal devices (3-kDa exclusion; Gelman Sciences, Ann Arbor, MI) and dialyzed against 0.3% CHAPS exactly as described previously (Pinxteren et al., 1998). Permeabilized cells were incubated with the dialyzed preparation of 22-subunits or [AlF₄]-preactivated G_i-3 (Pinxteren et al., 1998) at the indicated concentrations for 10 min. Cells were stimulated with compound 48/80 in the presence of 50 mM n-butanol for the measurement of PLD activity as described above.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Compound 48/80 Stimulates Membrane GTPase Activity, PLC, PLD, and Secretion in Quercetin-Treated Cells. As in previous studies (Senyshyn et al., 1998), overnight treatment of RBL-2H3 cells with quercetin resulted in increased expression of G_i3 (Fig. 5A). Such cells responded to compound 48/80 with enhanced GTPase activity (data not shown), phospholipid metabolism (Fig. 1, A and B), and secretion of the granule marker hexosaminidase (Fig. 1C). The stimulation of GTPase activity was dependent on dose of compound 48/80 (up to 70% increase with 100 µg/ml compound 48/80; data not shown) and was not apparent in untreated cells. Untreated cells also showed no detectable stimulation of lipid metabolism or secretion in response to compound 48/80 (data not shown, but see Senyshyn et al., 1998). Stimulation of phospholipid metabolism was apparent from transient increases in levels of inositol 1,4,5-trisphosphate and, in cells previously labeled with ³H-myristic acid, sustained increases in levels of [³H]phosphatidic acid that continued to increase over the course of 15 min well beyond the point at which production of inositol 1,4,5-trisphosphate had ceased (Fig. 1A). The transient increase in levels of inositol 1,4,5-trisphosphate was in accord with the transient calcium signal that is induced by compound 48/80 in quercetin-treated RBL-2H3 cells (Senyshyn et al., 1998).

View larger version (31K): [in this window] [in a new window]   Fig. 1.   Stimulation of PLC, PLD, and secretion by compound 48/80 in quercetin-treated RBL-2H3 cells. RBL-2H3 cells were incubated with 30 µM quercetin for 24 h and then washed. Cells were labeled with [3H]myristic acid for 90 min, washed, and then stimulated with compound 48/80 (25 µg/ml for the indicated times in A and B or for 5 min at the indicated concentrations in C) before the measurement of inositol 1,4,5-trisphosphate and [3H]phosphatidic acid (A), [3H]phosphatidic acid and [3H]phosphatidylethanol (B), or [3H]phosphatidylethanol and secretion of hexosaminidase (C). B and C, cells were stimulated in the presence of 172 mM ethanol. Data are expressed as pmol of inositol 1,4,5-trisphosphate in 106 cells (A), percent of total 3H-lipid recovered as [3H]phosphatidic acid or [3H]phosphatidylethanol (A-C), and percent of intracellular hexosaminidase that was released into the medium after correction for spontaneous release (2 ± 1%) in unstimulated cells (C). Data points are mean ± S.E. of values from three similar experiments. Error bars have been omitted for clarity in some panels.

Sensitivity of PLC- and PLD-Mediated Reactions to Cholera and Pertussis Toxins. Studies in cholera toxin-treated RBL-2H3 cells indicated that although compound 48/80-stimulated production of [³H]inositol phosphates was modestly enhanced (Fig. 2A), productions of [³H]phosphatidic acid (Fig. 2B) and, in the presence of ethanol, [³H]phosphatidylethanol (Fig. 2C) were substantially enhanced by this toxin. These data were consistent with those obtained with thapsigargin (Cissel et al., 1998) and other stimulants (P. F. Fraundorfer, W. A. Patton, J. Moss, and M. A. Beaven, submitted for publication) in which the activation of PLD was substantially enhanced in cholera toxin-treated RBL-2H3 cells, whereas the activation of PLC and PLA₂ was unaffected.

View larger version (39K): [in this window] [in a new window]   Fig. 2.   Enhanced production of phospholipid metabolites in cholera toxin (CTx)-treated RBL-2H3 cells and suppressed production in pertussis toxin (PTx)-treated cells. Quercetin-treated RBL-2H3 cells were incubated in the absence or presence of cholera toxin or pertussis toxin and labeled with [3H]inositol (A) or [3H]myristate (B-E) as described in Materials and Methods. Cells were stimulated with 25 µg/ml compound 48/80 for the indicated times in the absence (A) or presence (C-E) of 172 mM ethanol. The amounts of [3H]inositol phosphates, [3H]phosphatidic acid, and [3H]phosphatidylethanol were expressed as percent of total 3H-phospholipid in unstimulated cells. Data are mean ± S.E. of values from three experiments. Data were corrected for values in unstimulated cells (A, 1-2%; B-E, <0.04%). The differences between levels of [3H]inositol phosphates (P < .05), [3H]phosphatidic acid (P < .001), and [3H]phosphatidylethanol (P < .001) in untreated and toxin-treated cells were significant (paired t test) for all experiments except for the 10-min point in C.

Calcium/Protein Kinase C-Dependent and -Independent Responses of PLD to Compound 48/80. The activation of PLD by compound 48/80 was partially blocked by the removal of external calcium with EGTA or by the addition of 10 µM Ro31-7549, a selective inhibitor of protein kinase C catalytic activity (Ozawa et al., 1993; Wilkinson et al., 1993). Even the combination of these two treatments failed to completely suppress the production of [³H]phosphatidic acid (Fig. 3A) and, in the presence of ethanol, [³H]phosphatidylethanol (Fig. 3B) to reveal a substantial component of PLD activation that was calcium and protein kinase C independent. As in previous studies (Senyshyn et al., 1998), the same treatments, individually or in combination, totally suppressed compound 48/80-induced secretion (data not shown). These results suggested that PLD was activated by calcium/protein kinase C-dependent and -independent mechanisms.

View larger version (26K): [in this window] [in a new window]   Fig. 3.   Incomplete suppression of PLD responses by calcium deprivation and inhibition of protein kinase C. Quercetin-treated RBL-2H3 cells, labeled with [3H]myristic acid, were incubated for 10 min in normal medium; in the presence of 10 µM Ro31-7549, a protein kinase C inhibitor; in calcium-free medium with 0.1 mM EGTA; or in calcium-free medium with 0.1 mM EGTA plus 10 µM Ro31-7549. The cells were then stimulated with 25 µg/ml compound 48/80 for the indicated times. The amounts of [3H]phosphatidic acid and [3H]phosphatidylethanol (formed in the presence of ethanol) were expressed as percent of total 3H-lipid in unstimulated cells. Data are mean ± S.E. of values from three experiments and were uncorrected for values in unstimulated cells.

Dependence of Compound 48/80-Induced Secretion on PLD. The role of PLD in secretion was tested by the use of the PLD inhibitor n-butanol. This primary alcohol, but not its isomer, isobutanol, serves as an efficient donor for the PLD-catalyzed transphosphatidylation reaction and thereby suppresses the normal formation of phosphatidic acid by PLD in a variety of cells (Billah, 1993), including RBL-2H3 cells (Cissel et al., 1998). As shown in Fig. 4, 50 mM butanol inhibited compound 48/80-induced production of [³H]phosphatidic acid and secretion to the same extent (by 70-76%; P < .01), whereas the same concentration of isobutanol had much less effect on these responses, suggesting that suppression of secretion was due to specific rather than nonspecific actions of butanol.

View larger version (20K): [in this window] [in a new window]   Fig. 4.   Inhibition of phosphatidic acid formation and secretion by n-butanol in compound 48/80-stimulated cells. Quercetin-treated [3H]myristate-labeled RBL-2H3 cells were stimulated with 25 µg/ml compound 48/80 for 10 min in the absence or presence of 50 mM n-butanol or isobutanol as indicated for the measurement of secretion (release of hexosaminidase) and [3H]phosphatidic acid. Values are mean ± S.E. of data from three similar experiments.

Effect of Quercetin Treatment on Expression of G_i-3 and PLD in RBL-2H3 Cells. The increased expression of G_i-3 in quercetin-treated cells was not accompanied by an increased expression of PLD2 or PLD enzyme activity. Treatment with quercetin caused a 10-fold increase in G_i-3 and a modest increase in G₂ and G₂ but no increase in PLD2 in the membrane fraction (Fig. 5A). We were unable to detect PLD1 protein with the available antibodies (see Materials and Methods), although RBL-2H3 cells contain relatively small amounts of mRNA for PLD1b compared with mRNA for PLD2 (P. G. Holbrook, A. Vaid, and M. A. Beaven, unpublished data). It was unlikely, however, that the effects of quercetin were attributable to increased expression of PLD1. This PLD isoform is activated by guanosine-5'-O-(3-thio)triphosphate (GTPS) in the presence of ARF (Hammond et al., 1997). The extent of PLD activation by GTPS/ARF was the same for quercetin-treated and untreated cells in permeabilized cells, as was the case for other stimulants, such as antigen and carbachol in intact cells (Fig. 5B). Previous studies have shown that antigen and carbachol stimulate an ARF-insensitive and cholera toxin-sensitive form of PLD distinct from that stimulated by the combination of GTPS/ARF (P. F. Fraundorfer, W. A. Patton, J. Moss, and M. A. Beaven, submitted for publication).

View larger version (57K): [in this window] [in a new window]   Fig. 5.   Increased expression of Gi-3, G2, and G2 but not PLD2 or PLD activity in quercetin-treated RBL-2H3 cells. Cells were incubated overnight (18 h) with vehicle or 30 µM quercetin. A, cell lysates were separated into soluble and membrane fractions for electrophoretic separation and immunoblotting of proteins as described in Materials and Methods. Only blots for Gi-3, G2, G2, and PLD2 (as identified on right) in the membrane fractions are shown. These proteins were undetectable in the soluble fractions. G1, G1, and PLD1 were not detected in either fraction (not shown). The blots were from one of three similar experiments. B, untreated (open columns) and quercetin-treated (filled columns) RBL-2H3-m1 cells (see Materials and Methods for description of RBL-2H3-m1 cells) were labeled with [3H]myristic acid and then stimulated with 1 mM carbachol (CBC) or 20 ng/ml antigen (Ag, dinitrophenylated BSA) or permeabilized before stimulation with 100 µM GTPS and 1 µM myristoylated ARF-1 (see Materials and Methods for further details). Cells were stimulated for 10 min in the presence of 55 mM n-butanol for measurement of [3H]phosphatidylbutanol, which is expressed as percent of total 3H-phospholipids. Values are mean ± S.E. of three experiments and are corrected for values in unstimulated cells (0.05%).

Potentiation of Compound 48/80-Induced PLD Activation by G Subunits. In permeabilized quercetin-treated RBL-2H3 cells, the provision of recombinant G₂₂ subunits resulted in a modest increase in basal PLD activity as measured by the formation of [³H]phosphatidylbutanol (Fig. 6A). In addition, the presence of G₂₂ subunits markedly synergized the PLD response to compound 48/80. Maximal synergy was observed at concentrations of 500 to 1000 nM G₂₂ and were not apparent with heat-inactivated (100°C, 10 min) G₂₂ subunits (data not shown).

View larger version (28K): [in this window] [in a new window]   Fig. 6.   Effects of G22, Gi-3, and calcium ions on compound 48/80-induced stimulation of PLD in permeabilized RBL-2H3 cells. [3H]Myristate-labeled cells were permeabilized with streptolysin-O, incubated with the indicated amount of G22 (or 1000 nM G22 in C) or [AlF4]-preactivated Gi-3 for 10 min as described by Pinxteren et al. (1998), and then stimulated with 25 µg/ml compound 48/80 (+ Cpd. 48/80) or left unstimulated ( Cpd. 48/80) for 10 min in the presence of 55 mM isobutanol. A, heat-inactivated (100°C for 5 min) G22 was tested as a control (). Medium was buffered to give 1 µM free calcium (A and B) or the indicated concentrations of free calcium (C). The amount of [3H]phosphatidylbutanol formed was calculated as percent of total 3H-lipid in unstimulated cells. The data are mean ± S.E. from three experiments.

i-3

i-3

i-3

22

22

22

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Compound 48/80 (a mixture of polymers derived from N-methyl-p-methoxy-phenylethylamine), mastoparan, and certain polybasic neuropeptides are members of a family of polybasic mast cell secretagogues that are known to activate trimeric G proteins, primarily those of the G_i and G_o categories. Studies with rat peritoneal mast cells demonstrate that these compounds act at the cell surface, not by binding to discrete receptors but rather by directly stimulating pertussis toxin-sensitive membrane-associated GTPase activity (Mousli et al., 1990c; Chahdi et al., 1998a,b). Other pertussis toxin-sensitive events include a transient hydrolysis of inositol phospholipids (Nakamura and Ui, 1985) and a modest increase in levels of 1,2-syn-diacylglycerol, derived largely from phosphatidylcholine (Kennerly, 1990) to indicate possible activation of PLC and PLD. As shown in this and a previous study (Senyshyn et al., 1998), quercetin-treated RBL-2H3 cells respond similarly to rat peritoneal mast cells. This study establishes that both PLC and PLD are activated in a pertussis toxin-sensitive manner via the measurement of enzyme-specific products, namely inositol 1,4,5-trisphosphate, phosphatidylethanol, and phosphatidylbutanol. Furthermore, the synergistic actions of cholera toxin and recombinant G subunits noted here suggest that the release of G subunits may provide a signal for the activation of PLD in RBL-2H3 cells.

Compound 48/80 promotes GDP-GTP exchange and GTPase activity of purified preparations of G_i and G_o in vitro (Mousli et al., 1990b; Tomita et al., 1991), possibly by decreasing the ellipticity of the -helices in the receptor-binding domains of the G protein -subunits (Tanaka et al., 1998). The mast cell secretagogues target primarily G_i and G_o, although studies with mastoparan in reconstituted systems indicate weak stimulation of G_s as well (Higashijima et al., 1988). These reactions are thought to promote dissociation of the G proteins into their constituent - and -subunits (Tomita et al., 1991) and the subsequent activation of effector enzymes by either the -subunits (Mousli et al., 1990c) or, as suggested here, -subunits.

It is unclear why compound 48/80 activates only certain subtypes of mast cells, but both RBL-2H3 cells and mouse bone marrow-derived mast cells can be made to respond to compound 48/80 and other polybasic mast cell secretagogues by coculture with fibroblasts (Sakaguchi et al., 1992; Swieter et al., 1993; Ogasawara et al., 1997) or prolonged exposure to quercetin (Senyshyn et al., 1998). In the latter case, the exposure of RBL-2H3 cells to quercetin leads to cell maturation (Trnovsky et al., 1993) and substantial increases in the expression of G_i2, G_i3, and G (Senyshyn et al., 1998; current study) but not, apparently, in PLD2 (current study). The effects of quercetin treatment on PLD1 are uncertain because of difficulties in its detection. However, it is unlikely that the acquired sensitivity to compound 48/80 occurred through increased levels of PLD activity (see Fig. 5) but rather occurred through increased expression of G_i-2 and G_i-3. This sensitivity, however, is unmasked only on the removal of quercetin, which normally inhibits a variety of protein kinases and secretion in RBL-2H3 cells (Senyshyn et al., 1998). As far as we could determine, no residual effects of quercetin were apparent in washed cells. The secretory response to stimulants other than compound 48/80 (Senyshyn et al., 1998), as well as the activation of PLC (J. Senyshyn and M. A. Beaven, unpublished data) and PLD (current study; Fig. 5B), was the same in untreated and quercetin-treated cells.

As reported here, quercetin-treated RBL-2H3 cells responded to compound 48/80 with a transient activation of PLC and a somewhat more sustained activation of PLD. Both responses were inhibited by pertussis toxin and enhanced by cholera toxin, thus implicating G_i and G_s (Senyshyn et al., 1998; current study). These responses were accompanied by a transient release of calcium from inositol 1,4,5-trisphosphate-sensitive stores and rapid secretion of granules in a calcium/protein kinase C-dependent manner (Senyshyn et al., 1998). The present study reveals features that were not apparent from previous work with rat peritoneal mast cells; these include the activation of PLD by compound 48/80, the effects of pertussis and cholera toxins on this activation, and the inhibition of PLD activation and secretion by n-butanol. Based on these findings, we hypothesized that PLD was stimulated through the release of -subunits from trimeric G proteins and that PLD contributed a significant signal for secretion in compound 48/80-stimulated cells. In support of this hypothesis, recombinant G₂₂ subunits, but not G_i-3 subunits (Fig. 6), markedly synergized PLD activation by compound 48/80 in permeabilized RBL-2H3 cells. In addition, the overexpression of G₂₂, but not a constitutively active form of G_s, in hamster kidney cells (BHK-21) substantially enhanced the activation of PLD by pharmacologic stimulants and receptor ligands and further enhanced the potentiating effects of cholera toxin (P. F. Fraundorfer, J. Rivera, and M. A. Beaven, unpublished data).

The transient activation of PLC by compound 48/80 most likely results from the activation of PLC₃ through the release of -subunits from G_i as well as from G_s in cholera toxin-treated cells. PLC₃, the only G protein-sensitive PLC isoform found in RBL-2H3 cells (Ali et al., 1997; Senyshyn et al., 1998), can be activated by G subunits but not by -subunits of G_i and G_s (Noh et al., 1995). Furthermore, multiple isoforms of G and G are present in RBL-2H3 cells in combinations that permit functional coupling between muscarinic receptors and PLC (Dippel et al., 1996). If, as noted earlier, the G protein stimulants activate G_s weakly (Higashijima et al., 1988), treatment with cholera toxin might be expected to enhance release of -subunits from G_s and the activation of PLC₃. The release of additional -subunits from G_s may also increase the efficacy of compound 48/80 in activating G_i. Stimulation of G_i is significantly increased in the presence of excess -subunits (Higashijima et al., 1990).

We suggest similar scenarios for the stimulation of PLD by compound 48/80. The notion that PLD, like PLC, is activated by subunits is consistent with the observed enhancement of PLD activation by cholera toxin and G₂₂ subunits in compound 48/80-stimulated cells. Cholera toxin most likely synergizes activation of PLD at the level of the membrane rather than through soluble messengers such as calcium and cAMP as the effects of the toxin are still apparent in washed permeabilized-cells and plasma membrane vesicles (Cissel et al., 1998; P. F. Fraundorfer, unpublished data). Collectively, these studies suggest that PLD can be activated by G subunits as well as by calcium and protein kinase C. In contrast, PLC₃ is negatively regulated by protein kinase C in RBL-2H3 cells (Ali et al., 1997) which might account for the short-lived activation of PLC by compound 48/80. Interestingly, stimulation of RBL-2H3 cells via adenosine A₃ receptors also results in a pertussis toxin-sensitive activation of PLD that is sustained long after PLC-mediated events have decayed (Ali et al., 1996). As with compound 48/80, PLD appears to be activated via G_i independent of PLC-mediated events. Recently and consistent with our results, G subunits have been implicated in the activation of PLD via the angiotensin II receptor based on the inhibitory effects of anti-G antibody on PLD activation (Ushio-Fukai et al., 1999).

In conclusion, our results show that free G subunits and cholera toxin enhance the activation of PLD by compound 48/80 and that this activation may provide a necessary signal for secretion. We suggest that compound 48/80 interacts with PLD through release of G subunits from G_i and, in cholera toxin-treated cells, from G_s as well. Stimulation of PLC activity by compound 48/80 is minimally enhanced by cholera toxin, possibly because unlike PLD, this enzyme is negatively regulated by protein kinase C. A potential but untested target for the -subunits is the recently identified pleckstrin homology-like domain in PLD (Steed et al., 1998; Holbrook et al., 1999), although it is possible that -subunits act indirectly through activation of phosphatidylinositol 3'-kinase (Vanhaesebroeck et al., 1997), which, in turn, may stimulate PLD (Kozawa et al., 1997).