Earlier work in our laboratory showed that arachidonic acid (sn20:4?6) is a multifunctional regulatory molecule in submandibular glands, controlling several physiological functions, such as protein synthesis, mucin secretion, calcium mobilization and ATP metabolism . A prime source of the fatty acid is the membrane phospholipid, phosphatidylcholine, in which arachidonic acid is esterified at the sn-2 position. Muscarinic cholinergic stimulation of submandibular cells causes the release of arachidonate via a series of reactions, in which the initial stage is the receptor-coupled cleavage of phosphatidylcholine by the enzyme, phospholipase D . This produces phosphatidic acid, which in turn generates diacylglycerol and arachidonic acid. All three of these lipid-derived metabolites have proposed intracellular signalling or regulatory roles . Thus, the activation of phospholipase D is a key central factor in the control of several physiological processes in the submandibular gland.It was further demonstrated that muscarinic stimulaton of phospholipase D was partially mediated by the agonist-coupled phospholipase C/PIP₂ pathway . In this system, phospholipase C cleaves the membrane inositol phospholipid, PIP₂, to produce inositol trisphosphate and diacylglycerol, which in turn, respectively, release Ca²⁺ from endoplasmic reticulum stores and activate the enzyme protein kinase C. Both Ca²⁺ and protein kinase C play a part in the activation of phospholipase D in the submandibular model . However, approx. 40% of the ligand-induced effect on phospholipase D remains after complete blockade of the phospholipase C/PIP₂ pathway, leading to the speculation that a second, phospholipase C-independent pathway is involved in the activation of phospholipase D in the submandibular model. This is consistent with the proposal that such a pathway, which is guanine nucleotide-dependent, but phospholipase C/Gq-independent may operate in neutrophils .

A component that has been implicated in GTP-associated pathways of phospholipase D activation is the ADP-ribosylation factor, ARF. ARF was first identified as the protein cofactor for the cholera toxin-induced ADP-ribosylation of the -subunit of Gs of the adenylyl cyclase/cAMP signalling pathway . The protein is classified as a member of the ras superfamily of low molecular-weight, GTP-binding proteins . ARF has another established role in the transport of intracellular vesicles , and so may have a central regulatory function in the secretory process of exocrine glands, including salivary acinar cells. A further proposed role for ARF is in the activation of phospholipase D in neutrophils and kidney cells , although the protein’s mechanism of action is here unclear, but no such function has so far been examined in salivary glands.

Therefore, by using the submandibular mucous acinar-cell model, we have now sought to investigate whether phospholipase D is activated in a guanine nucleotide-dependent signalling pathway that is mediated by ARF, and to determine if this pathway is discrete from the phosphoinositide signal-transduction system.

Purified collagenase, CLSPA grade, was obtained from Worthington, Freehold (NJ, U.S.A.). [5.6.8.9.11.12.14.15-³H] arachidonic acid, spec. act. 7.73 TBq/mmol, was a product of Amersham (Arlington Heights, IL, U.S.A.). Silica gel-coated, thin-layer chromatography plates were from Whatman International (Maidstone, Kent, U.K.). The phosphatidylethanol standard was supplied by Avanti Polar-Lipids (Alabaster, ALA, U.S.A.). Compound U₇₃₁₂₂ and brefeldin A were obtained from Biomol Research Laboratory (Plymouth Meeting, PA, U.S.A.). GTP?S was from Boehringer Mannheim. All other chemicals were obtained from the Sigma Chemical Co. (St. Louis, MO, U.S.A.).

Male, Sprague–Dawley rats of 250–300 g body wt were anaesthetized by an intraperitoneal injection of sodium pentobarbital (Nembutal, 50 mg/kg body wt), then exsanguinated via the vena cava. This procedure was approved by the Animal Care Committee of the University of Manitoba. The submandibular glands were removed, acinar cells were prepared by collagenase dissociation and radiolabelled for 90 min with 60 kBq [³H]arachidonic acid/ml suspension. In experiments where intact cells were used, preparations were maintained in modified Hank’s balanced salt solution. In experiments where permeabilized cells were used (e.g., GTP?S treatment) the cells were maintained in cytosolic buffer containing 10 mM KCl, 20 mM NaCl, 25 mM NaHCO₃, 0.96 mM NaH₂PO₄, 5 mM MgSO₄, 0.01% soybean trypsin inhibitor, 1.5 mM Mg–ATP, 2 ?M CaCl₂, 15 mM Hepes, pH 7.2.

Phospholipase D was assayed by measuring the formation of [³H]phosphatidylethanol, which is produced in the presence of ethanol by the phospholipase D-specific transphosphatidylation reaction., except that all reagent volumes were proportionally reduced to produce a final reaction mixture of 200 ?l. Radiolabelled phosphatidylethanol was extracted by thin-layer chromatography on silica-gel plates and quantitated by scintillation counting as before.

The effect of GTP?S on phospholipase D was examined in cells permeabilized by treatment with digitonin (50 ?g/ml) in a final concentration of 1% ethanol, which also provided the substrate for the phospholipase D/transphosphatidylation reaction. GTP?S was routinely added to cell suspensions at the same time as digitonin so that the nucleotide would react with cytosolic components before they were washed out of the cells. Preliminary studies on dose and time responses of phospholipase D activation were carried out to establish the optimal concentration of GTP?S and duration of exposure. In other, prepermeabilization experiments, cells in cytosolic medium were treated for 20 min with 50 ?l/ml digitonin in the absence of GTP?S and ethanol, followed by several washings with the same medium, to deplete cytosolic components before the addition of nucleotide and phospholipase D substrate. In add-back experiments, exogenous cytosol from rat submandibular glands or human platelets, or purified, recombinant ARF1, was added to these cytosol-depleted cells, and the transphosphatidylation reaction triggered by the addition of 1% ethanol.

In another series of experiments, the effect on phospholipase D activity of brefeldin A, an inhibitor of the GDP/GTP exchange proteins that regulate ARF activation , was investigated. In these studies, phospholipase D was activated with GTP?S, or with the muscarinic agonist carbachol, or with the heterotrimeric G-protein activator AlFn. Permeabilized cells were pretreated with 400 ?M brefeldin A for 10 min before the activation of phospholipase D. AlFn was generated by the addition of 10 mM NaF plus 50 ?M AlCl₃.

In further studies, the possible involvement of the phospholipase C/PIP₂ hydrolysis signal-transduction pathway was examined in the GTP?S-or AlFn stimulation of phospholipase D. In these experiments, cells were preincubated in the absence or presence of 10 ?M U₇₃₁₂₂, an inhibitor of phospholipase C, for 10 min before nucleotide or AlFn treatment. U₇₃₁₂₂ was solubilized in dimethylsulphoxide (final concentration 1%), which was also added to controls.

Rat submandibular-gland acini were suspended in cytosolic buffer plus 1 mM dithiothreitol, 1 mM EDTA, 1 mM EGTA, and 0.3 mM PMSF. The cells were disrupted by sonication for 2×15 sec, then centrifuged at 500 g for 10 min. Unbroken cells and nuclei (pellet) were discarded, and the supernatant fluid was further centrifuged at 100,000 g for 1 hr at 4°C. The supernatant was retained as the cytosolic fraction. Human platelets were lysed by three freeze–thaw cycles in liquid nitrogen, and the cytosolic fraction was prepared by centrifugation at 100,000 g for 1 hr at 4°C. The cytosolic fractions from both sources were stored at ?70°C for further use in the add-back phospholipase D assays.

The protein concentration in various samples was determined by using the Bio-Rad dye reagent assay, based on the method of with bovine serum albumin as standard.

The rARF1 used in this study was from two sources. In some experiments, purified rARF1 was supplied by Dr B. Geny, Institut Cochin de Genetique Moleculaire, Paris. In other experiments, rARF1 was purified in our own laboratory from Escherichia coli strain BL21(DE3), transformed with ARF-expression plasmid pOW12, provided by Dr R. Kahn (Emory University, Atlanta, GA, U.S.A.). The expression of rARF₁ was induced with isopropylthio-?-galactoside at a final concentration of 1 mM. BL21(DE3) cells transformed with pOW12 were grown in LB medium containing 50 ?g/ml ampicillin at 37°C. The bacteria were lysed as reported by , by incubation for 10 min in 0.2% Triton X-100, 50 mM Tris–HCl, 100 mM MgCl₂, pH 8.0, with an additional step of a brief sonication at the end of the lysis. The preparation was centrifuged at 100,000 g for 1 hr at 4°C and the supernatant used as a source of rARF1. rARF1 was purified by successive column-chromatographic techniques on DEAE–Sephacel and Ultrogel AcA-54. In the present study, a further purification step consisted of the elution of rARF-containing fractions from a hydroxyapatite column. The column was equilibrated in 20 mM Hepes, 1 mM EDTA, 100 mM NaCl, 1 mM dithiothreitol, 2 mM MgCl₂, pH 7.4. Bound proteins were eluted with a gradient of NaH₂PO₄/Na₂HPO₄ (0–250 mM). The fractions from each step of the purification were analysed by SDS–PAGE and gels stained with Coomassie blue. Fractions containing species around the 21-kDa range (the M_r of ARF) were combined in each case for the subsequent phase of purification. These fractions were also checked for rARF content by reactivity with an anti-ARF antibody (see below). Fractions from the hydroxyapatite column that showed a single, ARF-positive band were pooled, dialysed in 10 mM Hepes, 10% glycerol, pH 7.2, and concentrated (Centricon 10, Amicon) to a protein content of 0.8 mg/ml.

A final concentration of 100 ?M GTP?S plus 50 ?g digitonin was added to a 1 ml acinar-cell suspension containing 200 ?l total cell protein, in cytosolic medium. After 20 min, 1 mM dithiothreitol, 1 mM EDTA, 1 mM EGTA, and 0.3 mM PMSF were added to the samples. Cells were disrupted by three 10-sec pulses of sonication. The suspension was subjected to ultracentrifugation at 100,000 g for 1 hr at 4°C. The retained pellet and supernatant were regarded as membrane and cytosol fractions, respectively. Samples containing the total amount of cytosolic and membrane extracts, representing 200 ?g of the original total-cell protein, were subjected to SDS–PAGE on 13% acrylamide. Separated proteins were transferred to nitrocellulose membranes by electroblotting in buffer containing 0.5% SDS . The blots were blocked with 5% milk powder for 1 hr at room temperature in a buffer containing 100 mM Tris, pH 7.5, 100 mM NaCl and 0.1% Tween 20. A polyclonal anti-ARF antibody (1:1000 dilution) was then added to the incubation mixture. The antibody was kindly provided by Drs J. Moss and M. Vaughan, NIH, Bethesda, MD. After 1 hr of incubation at 20°C, the blots were washed (5×5 min) in a buffer lacking milk power. The blots were reacted with goat anti-rabbit-IgG–horseradish peroxidase conjugate (1:3000 dilution, 1 hr, 20°C) followed by 5×5 min washing. The antigen–antibody complex was visualized on Hyperfilm-ECL by the enhanced chemiluminescence procedure (Amersam), as described .

Data were analysed statistically by two-way analysis of variance. Means were compared by Duncan’s multiple-range test or the student t-test. Values of p<0.05 were considered significant.

In a dose–response study, GTP?S, at an optimal concentration of 100 ?M, activated phospholipase D to approx. 5-fold control values . In time-course experiments, the activity of phospholipase D was maximal at 20-min incubation with GTP?S, and dropped only slightly over the following 30 min. The standard conditions of a 100-?M concentration of nucleotide and a 20-min incubation were therefore used in subsequent experiments, unless otherwise indicated. Permeabilization conditions proved to be crucial in demonstrating the stimulation of phospholipase D by GTP?S. When cells were permeabilized by digitonin for 15 min in the absence of nucleotide to allow depletion of cytoplasmic components (prepermeabilization), the subsequent addition of GTP?S had no effect on the activity of phospholipase D. When GTP?S was included with digitonin at the beginning of the permeabilization, the nucleotide did elevate that activity (.

This observation suggested a requirement for a cytosolic component in the stimulation of phospholipase D by GTP?S, which was investigated further in add-back experiments with cytosolic extracts of submandibular acinar cells and human platelets. Exogenous cytosol from both these sources (200 ?g protein) restored the guanine nucleotide-activating effect on phospholipase D in prepermeabilized, cytosol-depleted cells to 2.4-fold (submandibular-cell cytosol) and 6.8-fold (platelet cytosol), the levels found with GTP?S alone (p<0.01; data not shown).

In a further series of add-back experiments, the effect of exogenous rARF on the nucleotide stimulation of phospholipase D was investigated. rARF was purified to a single, 21-kDa species that reacted with an anti-ARF antibody . The purified protein, at a concentration of 220 ?g/ml, significantly enhanced the GTP?S elevation of phospholipase D activity by approx. 2.3-fold . rARF had no effect on enzyme activity in the absence of nucleotide.