Attenuation of sn- 1,2-Diacylglycerol Second Messengers by Diacylglycerol Kinase INHIBITION BY DIACYLGLYCEROL ANALOGS IN VITRO AND IN HUMAN PLATELETS*

Diacylglycerol kinase is thought to play a central role in the metabolism of diacylglycerol second messengers in agonist-stimulated cells. A series of diacylglyc- erol analogs were tested for their ability to act as substrates or inhibitors of diacylglycerol kinase with the goal of determining the substrate specificity of the enzyme, and of discovering inhibitors. Screening of these compounds was performed using a partially pu- rified diacylglycerol kinase from pig brain. Modified assays for this enzyme using co-sonicated mixtures of diacylglycerol and anionic phospholipids were devel-oped. This enzyme was found to be quite specific for sn-1,2-diacylglycerol (KM 24 PM for dioctanoyl-glycerol). Among the analogs investigated, only 1,2-dioctanoyl-2-amino-l,3-propaned~ol was utilized at a significant rate. Two analogs, dioctanoylethylene glycol (Kz 58 PM) and 1-monooleoylglycerol (KI 91 pM), were potent inhibitors in vitro. These compounds were tested for effects on diacylglycerol formation and me- tabolism in thrombin-stimulated human platelets. Dioctanoylethylene glycol inhibited diacylglycerol phosphorylation in platelets (70-100% at 100 PM) leading to a longer-lived diacylglycerol signal. This compound may be a useful tool for studies of diacyl- glycerol kinase in other cell types. 1-Monooleoylglyc-erol treatment elevated diacylglycerol levels up to 4- fold in unstimulated platelets and up to 10-fold in thrombin-stimulated platelets. The implications with regard to the pathways of diacylglycerol metabolism in human platelets are discussed. using cell-permeable


Attenuation of sn-1,2-Diacylglycerol Second Messengers by Diacylglycerol Kinase
Diacylglycerol kinase is thought to play a central role in the metabolism of diacylglycerol second messengers in agonist-stimulated cells. A series of diacylglycerol analogs were tested for their ability to act as substrates or inhibitors of diacylglycerol kinase with the goal of determining the substrate specificity of the enzyme, and of discovering inhibitors. Screening of these compounds was performed using a partially purified diacylglycerol kinase from pig brain. Modified assays for this enzyme using co-sonicated mixtures of diacylglycerol and anionic phospholipids were developed. This enzyme was found to be quite specific for sn-1,2-diacylglycerol (KM 24 PM for dioctanoylglycerol). Among the analogs investigated, only 1,2dioctanoyl-2-amino-l,3-propaned~ol was utilized at a significant rate. Two analogs, dioctanoylethylene glycol (Kz 58 PM) and 1-monooleoylglycerol (KI 91 pM), were potent inhibitors in vitro. These compounds were tested for effects on diacylglycerol formation and metabolism in thrombin-stimulated human platelets. Dioctanoylethylene glycol inhibited diacylglycerol phosphorylation in platelets (70-100% at 100 PM) leading to a longer-lived diacylglycerol signal. This compound may be a useful tool for studies of diacylglycerol kinase in other cell types. 1-Monooleoylglycerol treatment elevated diacylglycerol levels up to 4fold in unstimulated platelets and up to 10-fold in thrombin-stimulated platelets. The implications with regard to the pathways of diacylglycerol metabolism in human platelets are discussed.
Many cells respond to a variety of extracellular stimuli by activation of a phospholipase C which catalyzes the phosphodiesteratic cleavage of phosphatidylinositol 4,5-bisphosphate (1,2). The products of this reaction, inositol trisphosphate and sn-1,2-diacylglycerol, both function as intracellular second messengers. Inositol trisphosphate functions in the mobilization of intracellular Ca2+ stores (l), whereas diacylglycerol activates protein kinase C (3-6). Protein kinase C plays a central role in signal transduction, cellular regulation, tumor promotion, and perhaps oncogenesis (1,4, 5, 7). Activation of this kinase using cell-permeable diacylglycerols has been useful in further defining its role in a variety of cells (8)(9)(10)(11)(12)(13).
The diacylglycerol signal produced in response to extracellular stimuli is transient, and may be removed via several pathways. Diacylglycerol kinase has been suggested to plaJ an essential role in this process in platelets (14,15) and othel cell types (16,17). The phosphatidic acid thus formed is believed to be recycled back to phosphatidylinositol 4,5-bis. phosphate via a sequence of reactions known as the phospha. tidylinositol cycle (reviewed in Ref.

) .
The best-characterized mammalian diacylglycerol kinase: are those from pig brain (18) and rat liver (19). Kanoh et a1 (18,20) have described a membrane bound and a soluble forn of the enzyme in both tissues. These two forms of the enzymc have similar properties (20). The soluble pig brain enzymc has been purified to homogeneity and studied in vitro (18 21).
We have investigated the ability of a number of diacylglyc. erol analogs to act as substrates or inhibitors of diacylglycero: kinase. These analogs were constructed with constant acy: chain length (Cs), whereas the headgroup region was variec to explore the specificity of the kinase. This acyl chain lengtk was chosen because these compounds are cell-permeable (8)(9)(10)(11)(12)(13). These same analogs have been tested as effectors oj protein kinase C (22). For screening of these compounds we chose a partially purified preparation of the soluble pig brair enzyme. We report here a modified assay for this enzymr employing diacylglycerol substrates co-sonicated with variouf phospholipids. Several of'the diacylglycerol analogs were substrates and several were inhibitors. The effects of diacylglyc. erol kinase inhibitors on diacylglycerol formation and metab. olism in human platelets were investigated.

EXPERIMENTAL PROCEDURES
Materials--sn-l,2-Dioleoylglycerol (diC1kl)l was prepared by phos pholipase C digestion of dioleoylphosphatidylcholine as previouslJ described (23). All phospholipids and diC8 were obtained from Avant: Polar Lipids. 1-Monooleoylglycerol and 2-monooleoylglycerol were from Serdary Research Laboratories. DiCs analogs were prepared as previously reported (22). All other reagents were of the highest qualit3 commercially available.
Purification of Pig Brain Diacylglycerol Kinase-Diacylglycerol kinase was purified through G-150 column chromatography (Step 4) as described by Kanoh et al. (18). Purification was monitored using the deoxycholate assay of Kanoh et al. (18) except assays were performet at 23 "C and products were extracted by the method of Bligh and Dyer (24) using 1% HClO, as upper phase. Under these conditions, activity was proportional with the amount of protein employed, a n d with time over 15 min. This extraction procedure produced results equivalent to those obtained using butanol.
Assay of Diacylglycerol Kinuse-Enzyme assays were performed essentially as described (18) with the modifications noted above. Assays were usually performed for 3 min using 6.5 pg of protein. Assays in the presence of phospholipids were performed by mixing the diacylglycerol substrate with the appropriate phospholipid in CHC13, drying under N,, and resuspension in 2 X assay buffer by vortexing and sonication. Diacylglycerol analogs were also added in CHCL prior to drying. Phorbol esters, 40-phorbol 120-myristate 13aacetate and phorbol 12,13-dibutyrate, were added to the reaction mixture as dimethyl sulfoxide or ethanol solutions, respectively. pzP]Phosphatidic Acid Production in Thrombin-stimulated Platelets-Human platelets were prepared from freshly drawn blood essentially as described by Siess et al. (25), and were resuspended to 6.25 X 10' platelets/ml in modified Tyrode's buffer (25). ["PIPi was added to 0.6 mCi/ml, and labeling was allowed to proceed for 75 min at 37 "C. Platelets were pelleted at 600 X g for 10 min and resuspended in the same volume of Tyrode's buffer, and samples were taken for pretreatments. Diaeylglycerol analogs were added as ethanol solutions (0.5% ethanol, final) and platelets were preincubated for 10 min at 23 "C. Thrombin (100 units/ml in 50% glycerol) was then added to 1 unit/ml and, at the indicated times, samples (6.25 X 10' platelets) were extracted by dilution into 3 ml of CHCla/methanol (12). Extractions were performed by the method of Bligh and Dyer (24); the CHCl, phase was washed twice with 2 ml of 1% HClO,. Diacylglycerol Production in Thrombin-stimulated Platelets-Platelets were prepared and incubated (37 "C, 75 min) as described above, without 32P-labeling. Diacylglycerol analogs were added, platelets were preincubated for 10 min, and thrombin was added. Samples (5 X lo8 platelets) were taken at the indicated times and extracted as described above, except 1 M NaCl was used as upper phase to prevent acid-catalyzed isomerization of sn-1,2-diacylglycerols? Samples (1.5 ml) of the final CHCl, phase (1.8 ml total) were dried under Nz and redissolved in 20 pl of 7.5% 0-octyl glucoside, 25 mM dioleoylphosphatidylglycerol, and were then incubated with 1 mM [T-~'P]ATP and Escherichia coli diacylglycerol kinase as described by Preiss et aZ.* for 30 min at 23 "C to allow quantitative conversion of sn-1,2diacylglycerol to [32P]PA. Samples containing known amounts of sn-1,2-diCl&. were run simultaneously to ensure that the reactions were proceeding to completion. Following incubations, samples were extracted (24) using 1% HC10, as upper phase. The final lower phase was dried under N, and redissolved in 0.1 ml of CHC13,20-@1 aliquots were spotted on Merck Silica Gel 60 plates, and thin layer chromatography was performed in CHCl3/methanol/glacia1 acetic acid (65: 15:5).' [32P]PA was localized by autoradiography, scraped, and counted. With known amounts of sn-l,2-diCl~l the per cent diacylglycerol detected was 96.3 +-5.1. In addition, diacylglycerol levels measured from individual platelet preparations agreed closely as described in detail elsewhere?
Protein Phosphorylation by Platelets-Platelets (6.25 X 108/ml) were labeled with 0.1 mCi/ml ["PIP, (75 min, 37 "C) as described above, followed by a 10-min preincubation with the various analogs. Samples (0.1 ml), taken before and after addition of thrombin (1 unit/ml), were diluted into an equal volume of 2 X sample buffer (28) and boiled for 5 min. 0.1 ml was loaded to 10% sodium dodecyl sulfate-polyacrylamide gels, and electrophoresis was performed according to Laemmli (28). Gels were fixed in water/methanol/glacial acetic acid (60:3010), dried, and subjected to autoradiography.
Protein Determinations-Protein was determined using a modification of the method of Lowry et al. (29,30).

RESULTS AND DISCUSSION
Rationale-The objective of this work was to screen a series of diacylglycerol analogs for their ability to act as substrates or inhibitors of diacylglycerol kinase. This should facilitate the identification of biologically useful inhibitors of diacylglycerol kinase which will further understanding of diacylglycerol metabolism in uiuo. To accomplish this it was necessary to develop suitable in uitro assays for this enzyme. These assays were then employed for preliminary screening. Analogs that were found to inhibit pig brain diacylglycerol kinase in uitro were then tested in platelets where they had marked effects on diacylglycerol metabolism. Fig. 1 shows the structures of several diacylglycerol analogs which were substrates or inhibitors of the kinase.
In Vitro Assays of Pig %rain Diacylglycerol Kinase-Cosonicated mixtures of diacylglycerol and various phospholipids were good substrates for the partially purified pig brain kinase in uitro (Table I). These results are unlike those of Kanoh et al. (18). When diC1,, and phospholipid were sonicated separately and mixed prior to assay, little or no activity was detected (Table I). This is likely due to inefficient dispersal of the diCIB:l, since separate sonication of the more water-soluble compound diCs yielded good activity. A variety of anionic phospholipids were activators of this enzyme, as were mixtures of phospholipids containing anionic species, such as PE/PG/CL (6l:l) ( Table I). Activities greater than observed with the deoxycholate assay were obtained with PS alone, PG alone, CL alone, or a mixture of PE/PG/CL.
In contrast, little activity was observed with the zwitterionic phospholipids, phosphatidylcholine or PE. These results differ somewhat from those of Kanoh et al. (18) who found optimal stimulation by phosphatidylcholine, and little or no stimulation by PG and CL. These differences do not appear to be due to the state of purity of the enzyme used, since similar results were obtained by Kanoh et al. (18) using homogeneous or partially purified (Step 2) enzyme. Cofactor dependence may be modulated by degree of sonic dispersion, temperature, or other factors.
The properties of the diacylglycerol kinase assay were further characterized using phospholipid co-sonicated with diC1,,. or diCs as substrate ( Table 11)   With all activators examined, diCs was a better substrate, as evidenced by the lower apparent K , and the higher reaction velocities (Table 11). The apparent K, for diCs was 1.5-2-fold lower with phospholipid activators than with deoxycholate.
To understand further the role of phospholipids in the activation of diacylglycerol kinase, the amount of phospholipid present was varied at fixed concentrations of diC, (Fig.  2). When PS was used as activator, optimal activity occurred with a molar ratio of 0.5-1 PS/1 diCs. This was true at 30 pM diCs and 500 &M die, (Fig. 2 A ) . When a greater molar excess of PS was used, a decline in activity occurred.
These results suggest that optimal activity of diacylglycerol kinase is obtained over a narrow phospholipid/diacylglycerol range, and that 1 mol of anionic Iipid/mol of diacylglycerol is sufficient for maximal activity. The physical properties of is well documented (32,33). To circumvent these problems, we attempted to develop a mixed micellar assay for diacylglycerol kinase using a variety of detergent and phospholipid mixtures. None of these supported kinase activity. The inability to measure activity of the rat liver enzyme in nonionic detergents has been previously reported (19). Therefore, investigations on the effect of diacylglycerol analogs on kinase activity were performed using both PS and PE/PG/CL as activators.
Phosphorylation of Diacylglycerol Analogs by Pig Brain Diacylglycerol Kinase-All of the diacylglycerol analogs previously described (22) which contain a free OH group were tested as substrates of pig brain diacylglycerol kinase. Among the analogs designed to test the positional requirement of the hydroxyl group, neither diC8-butanetriol nor hexanetriol which displace the hydroxyl group 1 and 3 methylenes, respectively, showed any activity. 1,3-DiC8 had about 5% of the activity seen with 1,2-diC8. This activity is undoubtedly due to acyl group migration and therefore, contamination of the 1,3 isomer with small amounts of 1,2. Specificity for the sn-1,2 isomer has been reported for the rat liver enzyme (20).
Among the analogs designed to test the requirement for 0ester carbonyls, most were not substrates, including diC,ether, diC8-thioether, monoCs-propanediol, and monoC8-ethylene glycol. Only the diCs-%amide was a good substrate for the kinase (Fig. 3). DiCs-2-amide had an apparent K , of 90 pM, and a maximal velocity of 31 nmol/min/mg (compared to 58.5 nmol/min/mg for diCs). Although~the kinase tolerated an amide linkage in the 2 position, only slight activity was observed using diC,-1-amide. 2-Monooleoylglycerol was a substrate for diacylglycerol kinase, with an apparent K , of 82 p M and a V,,, of 9.6 nmol/ min/mg (Fig. 3). 1-Monooleoylglycerol was not a substrate. In addition, ceramide and phorbol esters (40-phorbol 12pmyristate 1301-acetate and phorbo112,13-&butyrate) were not substrates. These results indicate that diacylglycerol kinase from pig brain is quite specific for sn-1,2-diacylglycerol, both in regard to the 0-ester carbonyl linkage and in regard to the position of the hydroxyl group relative to the esters.

Inhibition of Diacylglycerol Kinase by Diacylglycerol Ana-
1ogs"Diacylglycerol analogs were tested as inhibitors of pig brain diacylglycero1 kinase, and these results are summarized in Table 111. In contrast to the specificity observed when analogs were tested as substrates, virtually all of the analogs were inhibitory in assays using PE/PG/CL as activator. With the exception of dice-ethylene glycol and l-monooleoylglycerol, only slight inhibitions were observed when analogs were present at 0.1 mM, whereas at 0.5 mM many analogs exerted considerable inhibitory effects. Among the 3-hydroxy analogs, diCs-ethylene glycol, diCs-glyceramide, and diCs-butanetriol were the most potent inhibitors. Among the 0-ester analogs, 1-monooleoylglycerol and d1C8-ether were most potent.
Analogs were also tested in assays using PS as activator (Table 111). At the lower concentration of PS (0.1 mM) the extent of inhibition was similar to that seen in the PE/PG/ CL assays. However, when the PS concentration was raised to 0.5 mM, inhibition by a number of analogs (e.g. dice-ether, diCs-butanetriol, and deoxydiCs) was alleviated. This suggests that inhibition by these analogs may have been due to changes in the physical state of the diacylglycerol/analog/phospholipid mixture. In contrast, inhibition by the most potent inhibitors, diCa-ethylene glycol and 1-monooleoylglycerol was not affected by elevating the PS level. We chose, therefore, to focus our attention on these compounds.
Double-reciprocal plots indicated that inhibition by diCaethylene glycol or 1-monooleoylglycerol was competitive with respect to diCB (Fig. 4). The K r values observed were 58 pM for diCs-ethylene glycol and 91 pM for 1-monooleoylglycerol. A similar extent of inhibition occurred when diC,,,, was employed as substrate (data not shown).

Effects of Diacylglycerol Analogs on Diacylglycerol Formation and Turnover in Thrombin-stimulated Human Platelets-To extend the in vitro results, diacylglycerol kinase inhibitors
were tested for their effects on diacylglycerol production and

TABLE I11
Effect of diacylglycerol analogs on diucylglyeerot kinase activity Assays were performed for 3 min using 30 @M dice and 6.5 pg of protein. Activators were present at 1 mM PE, 0.16 mM PG, 0.16 mM CL, 0.1 mM PS, or 0.5 mM PS, and the indicated concentration of analog was employed. Although substantial inhibition was seen using high concentrations of 4P-phorboI 12P-myristate 13a-acetate, this was not pursued in uiuo, since physiological effects of this compound are observed at nanomolar concentrations (10)(11)(12)  Addition of thrombin to platelets caused a rapid, transient rise in diacylglycerol levels, as determined by mass measurements (2-3fold within 15 sec)' (e.g. Fig. 5R). [32P]PA levels also increased upon thrombin stimulation (4-5-fold, maximal by 2 min) (e.g. Fig. 5A; Refs. 15,34), reflecting successive action of phospholipase C and diacylglycerol kinase. The effects of analogs on diacylglycerol formation and 13'P]PA production were examined.
Proximal portions of the thrombin-stimulated phosphatidylinositol cycle (Le. phospholipase C activation) were not affected by diC,-ethylene glycol treatment, since a rapid rise in diacylglycerol levels occurred similar to that seen in control platelets (Fig. 5B). The diacylglycerol signal generated in diCs-ethylene glycol-treated platelets was longer-lived than in controls. 5 min after stimulation, diacylglycerol levels had declined markedly, but basal levels were not achieved. Even though inhibition of the kinase was apparent, the diacylglycerol generated was still metabolized. The small elevation of basal diacylglycerol levels in diCs-ethylene glycol-treated platelets may be due to a decreased ability to metabolize diacylglycerol via the kinase pathway. Other
In the case of the 3-hydroxy analogs tested, [3ZP]PA levels were found to decline more rapidly than in control platelets. This was observed with the deoxy-, chloro-, and diCs-Omethyl ether analogs. The reason for this is not clear, but the possibility that these compounds stimulate some other metabolic fate of PA cannot be excluded. As shown in Fig. 7B, pretreatment with 0.5 m M deoxydiC, had no effect on diacylglycerol production.
1 -Monooleoylglycerol-Despite in vitro inhibition of pig brain diacylglycerol kinase by 1-monooleoylglycerol, this com-The results in platelets following treatment with diC8-glyceramide were variable. Considerable (up to 75%) inhibition of [32P]PA formation was observed in two experiments; however, this was not consistently seen.
The effect of 1-monooleoylglycerol on diacylglycerol levels in platelets was also examined (Fig. 8B). Basal levels of diacylglycerol were elevated following a 10-min preincubation with 1-monooleoylglycerol (42 pmol/5 X lo8 platelets in controls, 125 pmol following treatment with 30 p~, and 155 pmol following treatment with 400 p~) .
These elevated levels were stable over the 5-min time course examined (Fig. 8B). Diacylglycerol accumulation could be due to inhibition of diacylglycerol utilization or possibly conversion of exogenous monoacylglycerol to diacylglycerol via the action of a monoacylglycerol acyltransferase.
The diacylglycerol levels reached following l-monooleoylglycerol treatment are comparable to those seen following thrombin stimulation of control platelets. This finding led us to investigate the effect of 1-monooleoylglycerol treatment on protein kinase C activation, as measured by phosphorylation of a 40-kDa protein (35). 1-Monooleoylglycerol treatment (400 p~) only weakly stimulated 40-kDa phosphorylation in comparison to stimulation with thrombin or diCs (Fig. 9). This suggests that much of the diacylglycerol formed in response to monoacylglycerol treatment is in a compartment where it cannot activate protein kinase C. Importantly, l-monooleoylglycerol treatment did not impair the ability of platelets to phosphorylate 40 kDa in response to thrombin (Fig. 9). Pretreatment of platelets with diCs-ethylene glycol or diCs-glyceramide did not affect 40-kDa phosphorylation (Fig. 9).
When 1-monooleoylglycerol-treated platelets were stimulated with thrombin, the level of diacylglycerol rose even further (Fig. 8B). Platelets treated with 30 p~ l-monooleoyl- glycerol demonstrated a rapid rise (1.6-fold) in diacylglycerol upon stimulation, and this level remained elevated for 5 min. Diacylglycerol levels in platelets treated with 400 p~ l-monooleoylglycerol continued to rise for 5 min following stimulation (to 405 pmol/5 X lo8 platelets). The level reached is nearly 10 times the basal level in control platelets. These persistent elevated levels of diacylglycerol suggest an inability of monooleoylglycerol-treated platelets to metabolize the diacylglycerol signal. It is possible that this reflects an inhibition of diacylglycerol lipase (see "Discussion") by monoacylglycerol. It is also likely that inhibition of diacylglycerol kinase (as measured by ["PIPA production) was overcome by the elevated levels of diacylglycerol in these platelets.

CONCLUDING DISCUSSION
I n Vitro Properties of Diacylglycerol Kinase-The important role that diacylglycerol kinase plays in the attenuation of diacylglycerol signals in agonist-stimulated cells (14)(15)(16)(17) makes an understanding of its regulation essential. Our in uitro studies on pig brain diacylglycerol kinase have defined further its requirements for phospholipid cofactors and its substrate specificity. In addition we have begun to screen for potentially interesting and biologically useful inhibitors of this enzyme.
Co-sonicated mixtures of diacylglycerol and phospholipid should provide a substrate closely resembling the physiological one. Using this approach, a variety of anionic phospholipids were found to be activators of the kinase, whereas zwitterionic phospholipids did not support activity. These results suggest that diacylglycerol kinase may require anionic phospholipids as essential cofactors. Such a surface would be present on the inner surface of the plasma membrane which is rich in PS. Alternatively, surfaces containing anionic phos-pholipids may affect activity by physically modulating diacylglycerol accessibility.
Substrate Specificity of Diacylglycerol Kinase-Our data indicate that pig brain diacylglycerol kinase has high specificity for sn-1,2-diacylglycerol. DiCs-2-amide was the only analog tested which was phosphorylated at a significant rate. 2-Monooleoylglycerol and diC8-1-amide were also substrates for the kinase, but low maximal velocities were observed. Phosphorylation of 2-monoacy1glycero1, but not l-monoacylglycerol, has also been demonstrated by Kanoh's Testing of diCs analogs as substrates of diacylglycerol kinase may aid in the identification of cell-permeable activators of protein kinase C which will not be attenuated. For example, diCs-butanetriol has been found to activate protein kinase C (22) but is not a substrate for diacylglycerol kinase, and therefore should be longer lived in uiuo than diC8. Such compounds will be useful for in uiuo studies on protein kinase C.
In Vitro Inhibition of Diacylglycerol Kinase-Inhibitors of diacylglycerol kinase will be valuable tools in studying the role of this enzyme in uiuo. Although most of the analogs tested were inhibitory using PE/PG/CL or 0.1 mM PS as activators, inhibition by a number of these was overcome by raising the PS concentration. This suggests that inhibition was due to alterations in the physical state of the diacylglycerol/phospholipid mixtures. The concentration of diCs employed (30 p~) represents 2.3 mol % in the PE/PG/CL assay and 30 mol % in the 0.1 mM PS assay. At 0.5 mM, the mole % of diacylglycerol analog becomes very high and may induce such perturbations. The most potent inhibitors, however, exhibited the same inhibition at 0.1 or 0.5 mM PS, suggesting that inhibition was due to interaction with the kinase; these were competitive inhibitors with respect to diacylglycerol (Fig.  4).
Effects of Analogs on Human Platelet Diacylglycerol Metabolisrn-DiC8-ethylene glycol also inhibited diacylglycerol kinase in human platelets, as indicated by the decreased basal and stimulated levels of [32P]PA. This effect was exerted without any suppression of diacylglycerol formation. DiCsethylene glycol may prove a useful tool for investigation of diacylglycerol kinase in other cells.
I-MonooIeoylglycerol inhibition of the pig brain kinase was not reflected in platelets when [32P]PA production was measured. Pretreatment with this compound, however, had profound effects on diacylglycerol levels in both unstimulated and stimulated platelets. The elevation of diacylglycerol levels induced by 1-monooleoylglycerol alone resulted in only slight activation of protein kinase C. This observation indicates that distinct intracellular pools of diacylglycerol exist. Diacylglycerol generated by acylation of monoacylglycerol may enter an internal membrane pool where it can be used in phospholipid biosynthesis, and where it cannot activate protein kinase C.
The levels of diacylglycerol attained upon thrombin stimulation of 1-monooleoylglycerol-treated platelets were extremely high. These elevated diacylglycerol concentrations probably masked inhibition of the kinase when assessed by [32P]PA formation.
The possibility that diacylglycerol generated by phospholipase C action is metabolized by alternate pathways has not been excluded. Human platelets are known to contain a diacylglycerol lipase specific for the sn-1 position (36). The 2monoacylglycerol generated is metabolized further by a lipase activity capable of using 1-or 2-monoacylglycerol (36). Evidence that this pathway functions in stimulated platelets (37) H. Kanoh, T. Iwata, T. Ono, and T. Suzuki, personal communication. and 3T3 cells (38) has been presented. Other workers, however, have suggested that this pathway does not play an important role in metabolism of diacylglycerol in stimulated platelets (39, 40).
Our data are consistent with a role for diacylglycerol lipase in stimulated platelets. This could account for (i) the ability of platelets to metabolize thrombin-induced diacylglycerol in spite of apparent inhibition of the kinase by diC8-ethylene glycol, and (ii) the effects of 1-monooleoylglycerol on diacylglycerol levels in stimulated platelets. This compound, at high concentrations, could compete for the monoacylglycerol lipase, preventing flux of endogenously generated diacylglycerol through this pathway. The results seen with l-monooleoylglycerol treatment are similar to those obtained using platelets treated with indomethacin, a compound reported to inhibit lipase activity in vitro (27).
Further analysis of diacylglycerol metabolism in platelets would profit from potent and specific inhibitors of both the diacylglycerol kinase and diacylglycerol lipase. These, in conjunction with mass quantitation of diacylglycerol levels as used in the current work, could provide further insight into the regulation of diacylglycerol second messengers.