Evidence of GTP-binding protein regulation of phospholipase A2 activity in isolated human platelet membranes.

G protein regulation of human platelet membrane phospholipase A2 activity was investigated at pH 8.0 and 9.0 by studying the effects of the nonhydrolyzable GTP analogue, guanosine 5'-O-(3-thiotriphosphate) (GTP gamma S), and of F-/Al3+ ions on arachidonic acid (AA) release. The membrane acted as the source of the enzyme, the substrate, and the G protein. At pH 8.0, 10 and 100 microM GTP gamma S stimulated AA mobilization at least 6-fold. Optimum AA release conditions required 1 mM Ca2+ and 5 mM Mg2+. Nonspecific nucleotide effect was excluded since similar stimulatory effects on AA release were not observed by ATP, GTP, ADP, and NADP. Although at pH 9.0 the GTP gamma S-stimulated AA release was greater than at pH 8.0, it constituted only 26% of the total. At both pH values the effect of F- (10 mM) in the presence of Al3+ (2 microM) was similar to that of GTP gamma S. The G protein inhibitor, guanosine 5'-O-(2-thiodiphosphate), inhibited the GTP gamma S-stimulated AA release by about 80% at pH 8.0 and by 100% at pH 9.0. To determine a possible contribution to AA mobilization by the phospholipase C and diacylglycerol lipase pathway, the effects of neomycin, a phospholipase C inhibitor, were investigated. 100 microM neomycin did not inhibit the GTP gamma S-stimulated AA release at pH 8.0 and only slightly so (17%) at pH 9.0. At pH 8.0 in the presence of Ca2+ the released fatty acids consisted mainly of arachidonic and docosahexaenoic acids (80 and 8%, respectively). GTP gamma S had no effect on the fatty acid profile but only on their quantity. These results provide evidence of G protein regulation of phospholipase A2 activity in isolated platelet membranes.

sitide-specific PLC leading to the formation of AA-rich diacylglycerol which may be further metabolized by diacylglyceride lipases to yield AA (3,4). In view of the selectivity with which stimulated cells release AA, the apparent lack of specificity for AA by the soluble PLA2s posed a serious objection to the PLAZ pathway (5). This objection, however, seemed to have been alleviated with the isolation of a PLA, from macrophage cell line RAW 264.7 which possessed a considerable preference for AA (6). However, the relative importance of the two pathways has yet to be resolved ( 5 ) .
The mechanisms of signal transduction leading to AA mobilization are not yet well understood. There is considerable evidence in platelets of G protein regulation of PLC activity (7-10). Some evidence of G protein regulation of PLA2 activity has also been reported in platelets (10, 11) as well as in other cells (12)(13)(14)(15)(16). In platelets (10) and in FRTL5 thyroid cells (13), different G proteins were responsible for the regulation of the respective PLC and PLA2 activities. Although Crouch and Lapetina (17) reported that the adenylate cyclase inhibitory G protein (GJ was not involved in platelet PLAz regulation, their results did not rule out the possibility of its regulation by another G protein. We have elected to study the possibility of G protein regulation of the previously reported Ca2+-dependent PLA, in isolated platelet membranes which required for optimum activity pH 9.5 and 10 mM Ca2+ (18). This enzyme was linked to AA mobilization in intact cells by the findings that the release of AA in isolated membranes and in thrombin-and collagen-stimulated platelets was inhibitable in a similar manner by the sulfhydryl-blocking reagent, 5,5'-dithiobis(2-nitrobenzoic acid) (19). By means of the G protein activators, GTPyS and F-/A13' (20), and utilizing an assay in which the membrane acts as the source of the enzyme, the substrate (19), and the G protein, we report evidence of G protein activation of the platelet membrane PLA,.

EXPERIMENTAL PROCEDURES
Materials-All reagents were purchased from Sigma, except GTPyS and GDPpS from Boehringer Mannheim, solvents from Mallinckrodt Chemical Works, and AA and heneicosanoic acid (21:O) standards from Alltech Associates. Outdated human platelet concentrates were from the New York Blood Center. Solutions and Buffers-Buffers in the PLA? assays were: at pH 8.0, 0.133 M Tris, 0.067 M KC1, and 0.53% bovine serum albumin (essentially fatty acid free); at pH 9.0 and 9.5, 0.133 M glycine was substituted for Tris. The concentrations of other reagents used in the assay are reported under "Results." Membrane Preparation-Within 3 days following expiration, platelet concentrates were washed and disrupted by nitrogen decompression as previously described (19). Briefly, the homogenate was layered on a 30% (w/v) sucrose cushion and ultracentrifuged in an SW 50.1 rotor at 300,000 X g for 60 min. The membranes, from buffer-sucrose interface, were washed and collected by ultracentrifugation in a type 65 fixed angle rotor at 217,000 X g for 2 h. The membranes, were stored at -80 "C and used within 6 months.
PLA, Assay-PLA2 activity was measured by the extent of AA release as previously described (19) except 40-60 j~l of membrane suspension containing 0.3-0.5 mg of protein were added to the incubation solution containing 300 ~1 of the assay buffer and any additional reagents indicated under "Results." The final reaction volume was 400 ~1 .
Following incubation, the suspension was acidified, and the free fatty acids were extracted with ethyl acetate, methylated with diazomethane, and analyzed on a Varian 4600 gas-liquid chromato-graph equipped with a flame ionization detector and a 180-220 "C temperature program. Heneicosanoic acid was used as the internal standard. Initial identification of AA was carried out on a Hewlett-Packard 5985 gas chromatograph-mass spectrograph. All experiments were done in duplicate.

RESULTS
To determine possible G protein regulation of the platelet membrane PLA2, the effects on AA mobilization of both the nonhydrolyzable GTP analogue, GTPyS, and of F-/A13+ were investigated (Table I). At pH 8.0 in the presence of 1 mM Ca2+ and 10 mM Mg2+, both GTPyS and F-/A13+ stimulated the membrane PLAz activity 6-fold. 10 mM Ca2+ alone had an effect on AA release comparable with that of 100 p~ GTPyS at a 10-fold lower Ca2+ concentration. Under those conditions the GTPyS effect was not observed (results not shown). At pH 9.0 ( Table I)  utable to GTPyS activation was substantially reduced. The effects of F-/AP+ were similar to those of GTPyS. In the presence of 10 mM EDTA and no added Ca2+ and M$+, AA release was very low at pH 8.0 and 9.5 and no GTPyS stimulation was observed. We also found that 2 mM GSH increased GTPyS-stimulated AA mobilization by about 16% but had no effect on control values (results not shown). Time courses of the reactions at pH 8.0 and 9.0 are shown in Fig. 1. The increase in GTPyS-stimulated AA accumulation was linear during the first hour of a 3-h incubation period. Approximately 80% of the AA was released during the first hour of incubation at pH 8.0 and 67% at pH 9.0. Because the ratio of the GTPyS-stimulated AA release to that obtained under control conditions was about the same, valid comparisons could be made between the results obtained from 1-or 3-h incubation periods.
The possibility of a nonspecific nucleotide effect on the PLAz activity was also investigated. The effects of ATP, GTP, ADP, and NADP were compared with those of GTPyS (Table  11). No significant difference in AA mobilization was observed when GTPyS concentration was reduced from 100 to 10 PM. AA mobilization was not stimulated either by 10 or 100 p~ ADP or NADP, or by 10 p~ GTP or ATP. However, 100 p~ ATP appeared to stimulate AA release about %fold.
Events mediated by G proteins are known to be inhibitable by nonhydrolyzable GDP analogs (14,20). The effects of GDPDS on GTPyS-stimulated PLA2 activity are shown in Table 111. 524 p~ GDPDS inhibited 100 p~ GTPyS-stimulated AA release by about 80% at pH 8.0 and by 100% at pH 9.0 but had no significant effect on the control values.
Reports of the presence in platelets of a membrane-bound PLC (7, 9) raised the possibility that we might be observing the results of sequential actions of a G protein-activated PLC followed by a diacylglycerol lipase. We therefore used neomycin, a PLC inhibitor (lo), to determine a possible contribution to the AA mobilization by the PLC/diacylglycerol lipase pathway. At pH 8.0,O.l mM neomycin appeared to have no significant inhibitory effect. However, 1.0 mM neomycin inhibited GTPyS-stimulated AA release by about 37% (Fig.   -0.0 0.5 1.0 1.5 2.0 2.5 3.0 0.0 0.5 1.0 1.5 2.0 2   No significant G T P r S effect was observed in the absence of added Ca2+ or in the presence of 1.0 mM EGTA. At a constant 1.0 mM Ca2+ concentration, the GTPyS-stimulated AA release was maximal at 5.0 mM Mgz+. Although omission of M$+ from the incubation buffer had no inhibitory effect on the GTP+stimulated AA release, Mg2+ was required for optimum effect (Fig. 3). AA release of the control experiments was low and varied minimally with Ca2+ and Mg'+ concentrations.
At pH 8.0, the released fatty acids consisted mostly of arachidonic (80%) and docosahexaenoic (22:6, 8%) acids. No significant variation in the fatty acid profile was observed in the presence or absence of GTPyS and neomycin or when Ca2+ concentration was varied between 0.05 and 1 mM.

DISCUSSION
The previously demonstrated PLA2 activity in isolated platelet membranes required high pH and high Ca2+ concentration (pH 9.5, 10 mM Ca2+) for optimum activity (18). In view of such nonphysiological activating conditions and the recently emerging evidence of G protein regulation of a number of PLA2s (10-16), we investigated the existence of a similar regulating mechanism in platelet membranes. The membranes acted as the source of PLA,, the enzyme substrate, and a regulatory G protein. In this assay the use of labeled exogenous phospholipids was avoided and all of the released AA was quantitated. In view of reports that PLAP activities depend on the nature and physical state of the substrates (21,22) the preservation of the substrate-enzyme relationship in the isolated membranes seemed most desirable.
At pH 8.0, our results were consistent with G protein activation of PLA, (20). The enzyme was activated by both GTPyS and by F-/A13+ ions (Table I). Activation by GTPyS was specific. None of the other nucleotides used, with an apparent exception of 100 PM ATP, seemed to have any PLA, stimulatory effect (Table 11). GTPyS activation of the PLA, was significantly inhibited by GDPpS (Table 111). The lack of inhibition of AA release by 0.1 mM neomycin (Fig. 2) as well as the reported (9) low pH optimum (6.5-7.0) of the membrane-bound PLC argued against the possibility that AA was released by the PLC/diacylglycerol lipase pathway. In platelets, the reported stimulation of AA release by neomycin at concentrations greater than 0.1 mM indicated still another function of this reagent (11). Therefore, in our experiments, the partial inhibition of the GTPyS stimulation of AA release by 1 mM neomycin may be due to effects other than on PLC. The above results provided evidence that at pH 8.0, G proteinactivated PLA' was responsible for the released AA.
We found at pH 8.0 a disproportional release of arachidonic (80%) and docosahexaenoic acids (8%) to their reported concentrations in membrane phospholipids (29% and less than 2%, respectively)' (23). Docosahexaenoic acid was reported to be almost exclusively acylated in phosphatidylethanolamine which contained also 48% of all of the platelet membrane AA. These results are indicative of an enzymatic process where at least part of the AA would be expected to be mobilized from phosphatidylethanolamine. The findings that GTPyS in the M. J. Broekman, personal communication. presence of Ca2+ affected only the quantity of the released fatty acids without altering their relative concentrations and that high Ca2+ concentration activated the membrane PLA, to the same extent as G protein effectors at lower Ca2+ concentrations were also suggestive of a common enzymatic mechanism. These results were consistent with known guanine nucleotide effects found to lower Ca2+ requirements for enzymatic activities (20), such as in platelets, where Ca2+ alone was found to stimulate membrane PLC to the same extent as GTPyS at lower Ca2+ concentrations (9).
At pH 9.0, two apparently independent Ca2+-requiring mechanisms contributed to AA mobilization: 1) G proteinmediated and 2) high pH-facilitated ( Table I). As at pH 8.0, PLA, activity was responsible for the G protein-mediated fraction of the released AA. AA release was activated by GTPyS and by F-/A13+. The GTPyS-stimulated AA release was inhibitable by GDPBS (Table 111) but only moderately so (17%) by 100 PM neomycin. The high pH-facilitated AA release had an absolute requirement for Ca2+ indicating an enzymatic nature of the mechanism. There is some experimental evidence that negatively charged substances introduced into the substrate environment increased substrate availability and thereby increased PLA2 activity (6, 21). At pH 9.0 an increase in the membrane negative charges may thus effect PLA2 activity. Whether in our experiments the pH-facilitated and the G protein-activated AA release was due to the activation of the same enzyme remains to be elucidated.