et Biophysics &ta ELSEVIER Biochimica et Biophysics Acta 1212 (1994) 193-202

Vitamin E suppresses diacylglycerol (DAG) level in thrombin-stimulated endothelial cells through an increase of DAG kinase activity

Khai Tran, Pierre R. Proulx, Alvin C. Chan *

Department of Biochemistry, Faculty of Medicine, University of Ottawa, 451 Smyth Rd., Ottawa, Ontario, KIH 8M5, Canada

(Received 2 September 1993)

Abstract

The present study has examined the role of vitamin E, a natural lipid antioxidant, in the production of diacylglycerol (DAG) and phosphatidic acid (PA) in thrombin-stimulated human endothelial cells. Cells were labelled with [“Hlmyristate and the incorporation and distribution of [3H]myristate into cellular lipids was not affected by vitamin E. However, in response to thrombin stimulation, considerably more PA and less DAG were formed in cells enriched with vitamin E. The time-course of thrombin stimulation indicated that vitamin E attenuated the accumulation of sustained DAG levels with a concomitant increase in PA. Direct determination of DAG mass further confirmed that vitamin E suppresses the accumulation of DAG induced by thrombin. In the presence of ethanol, the formation of [3H]phosphatidylethanol (PEt) in [3H]myristate-labelled cells stimulated by thrombin was unaffected by vitamin E enrichment. oL-Propranolol, a PA phosphohydrolase inhibitor, caused an accumulation of PA, without affecting DAG formation in either vitamin E-treated and untreated cells. This indicated that the increase in PA and decrease in DAG in vitamin E-treated cells was not due to a stimulation of phospholipase D or an inhibition of PA phosphohydrolase. Determination of inositol phosphates formation in response to thrombin showed that the change of DAG levels elicited by vitamin E was independent of phospholipase C-induced hydrolysis of inositol phospholipids. In contrast, analysis of DAG kinase activity revealed that vitamin E enrichment enhanced the activity of the enzyme in both basal and thrombin-stimulated cells. Taken together, these data indicated that vitamin E caused an increased conversion of DAG to PA by activating DAG kinase activity without causing any change in the activities of phospholipase D, PA phosphohydrolase or phospholipase C.

Key words: Vitamin E; Diacylglycerol; Endothelial cell; Thrombin; Diacylglycerol kinase; Phosphatidic acid

1. Introduction

The binding of hormones or neurotransmitters to

specific receptors of many cell types and tissues trig-

* Corresponding author. Fax: + 1 (613) 7876732. Abbreviations: ATP, adenosine triphosphate; DAG, diacylglycerol; DME, Dulbecco’s modified Eagle’s medium; DMSO, dimethylsulf- oxide; HBSS, Hank’s balanced salt solution; HPLC, high pressure liquid chromatography; IP,, inositol 1,4-bisphosphate; IP,, inositol 1,4,5_trisphosphate; PA, phosphatidic acid; PAF, platelet-activating factor; PBS, phosphate-buffered saline; PC, phosphatidylcholine; PE, phosphatidylethanolamine; PEt, phosphatidylethanol; PGI,, prosta- cyclin; PI, phosphatidylinositol; PIP, phosphatidylinositol 4-phos- phate; PIP,, phosphatidylinositol 4,5-bisphosphate; PKC, protein kinase C; PLA,, phospholipase A,; PLC, phospholipase C; PLD, phospholipase D; PMA, phorbol 12-myristate 13-acetate; TG, triacyl- glycerol; TLC, thin-layer chromatography; vitamin E, in this paper it refers to RRR-cY-tocopherol.

gers multiple signal transduction processes which oper- ate in a tightly regulated manner to release various second messengers for many cellular responses (for review see Ref. 1). For instance, the binding of throm- bin to its receptors in endothelial cells leads to the activation of phospholipase C (PLO, phospholipase D (PLD) and phospholipase A, (PLA,) [2-41. Thrombin activation of endothelial phosphatidylinositol-specific PLC elicited the hydrolysis of phosphatidylinositol 4,5- bisphosphate (PIP,) resulting in the accumulation of

inositol 1,4,Strisphosphate UP,) and sn-l,Zdiacyl- glycerol (DAG) [2]. However, there is now strong evi- dence that DAG is not solely generated from PIP,, but is also derived from phosphatidylcholine (PC> in many cell types [5,6] including endothelial cells [71. Both PC-specific PLC and PLD activities were found to be present in endothelial cells and were responsible for the hydrolysis of PC generating DAG and PA, respec-

0005-2760/94/$07.00 0 1994 Elsevier Science B.V. All rights reserved SSDI 0005-2760(93)E0264-K

194 K. Tran et al. / Biochitnica et Biophysics Acta 1212 (1994) 193-202

tively [8,9]. Moreover, DAG and PA are readily inter- converted by DAG kinase and PA phosphohydrolase via phosphorylation and dephosphorylation reactions. DAG is responsible for activating a large family of protein kinase C (PKC) isoenzymes [lO,ll] while IPx is a calcium secretagogue 1121. The increase in intra- cellular calcium and PKC activation are intimately related to PLA, activation which is the rate-limiting step in the synthesis of platelet activating factor and prostacyclin [4,13,14].

Vitamin E, the membrane bound, lipid-soluble, chain breaking antioxidant, has recently been reported to inhibit PKC in smooth muscle cells leading to the inhibition of cell proliferation [15]. In addition to its unique role in protection of biological membranes from injury induced by free radicals, we have previously demonstrated that vitamin E can influence the turnover of cellular phospholipids and the release of arachi- donic acid. For example, vitamin E enrichment was shown to inhibit phosphoiipase A, activity in platelets and myocardium [l&17] and it exerted a differential effect on thromboxane and prostacyclin synthesis [18- 211. Furthermore, we have recently shown that vitamin E enrichment caused an increase in a CoA-independ- ent transacyiase in endothelial cells [22]. The present investigation was undertaked in order to further delin- eate the involvement of vitamin E in the turnover of phospholipids and the generation of significant cellular lipid mediators.

In the present study, we have examined the effect of vitamin E on the formation of DAG and PA induced by thrombin in cultured endothelial cells isolated from veins of human umbilical cords. This model has a distinct advantage because after the primary passage, the cells contain an undetectable amount of vitamin E and therefore, cellular vitamin E levels can readily be controlled [23]. By using [“Hlmyristate and myo- [iH]inositol labelling as well as direct quantitation of DAG mass. we report here that vitamin E suppresses the sustained level of DAG but increases PA formation when cells were stimulated by thrombin. Indirect stud- ies indicate that vitamin E has no effect on the activity of PLC, PLD or PA phosphohydrolase but activates DAG kinase activity which is responsible for the in- creased conversion of DAG to PA.

2. Experimental procedures

Materials ~~H]Myristic acid (60 Ci/mmoI) was from American

Radiolabelled Chemicals (St. Louis, MO, USA). rye-

[“H]Inositol (50.2 Ci/mmol) and [Y-“~PIATP (30 Ci/mmol) were from New England Nuclear (Lachine, PQ, Canada). The ~~-1,2-diacylgiycerol (DAG) assay kit was from Amersham Canada (Qakville, ON,

Canada). Human thrombin 3000 NIH units,/mg pro- tein, heparin, gelatin, collagenase type IV, gentamycin sulfate, Hepes, DL-propranolol, dithiothreitol and octylglucoside were from Sigma (St. Louis, MO, USA). Medium 199, inositol free-Dulbecco’s modified Eagle’s medium, Iyophilized antibiotic-antimycotic, trypsin- EDTA, heat-inactivated fetal bovine serum and all tissue culture plasticware were from Gibco (Burling- ton, ON, Canada). Endotheiial cell growth supplement was from Collaborative Research (Bedford, MA, USA). Phosphatidylethanol was from Avanti Polar Lipids (Alabaster, AL, USA) and all other standard lipids were from Serdary Research Laboratories (London, ON, Canada). Dowex (AG-1 x 8 resin, formate form) and plastic chromatography columns were from Bio- Rad Laboratories (Mississauga, ON, Canada). RRR- L-r-tocopherol was donated by the vitamin E Research Information Services (LaGrange; IL, USA). Thin layer plates (silica gel 60A) were from Canlab (Mississauga, ON, Canada). All solvents used were HPLC grade and were purchased from BDH Chemicals (Toronto, ON, Canada). Glassware were silanized before use.

Culture of endothe~ia~ cells Human endothelial cells were isolated from fresh

umbilical veins after incubation with 0.2% collagenase and seeded on 100 mm Petri dishes precoated with 0.2% gelatin as previously described 121,241. The cells were grown to confluence in medium-199 (pH 7.4) supplemented with heat-inactivated fetal calf serum (lo%), endothelial cell growth supplement (30 pg/ml), heparin (90 ,ug/ml), Hepes (25 mM), gentamycin sul- fate (40 pg/ml), antibiotic-antimycotic (100 units/ml penicillin G sodium, 100 pg/ml streptomycin sulfate and 0.25 pug/ml amphotericin B). The endothelial monolayer was detached using 0.05% t~psin-EDTA and centrifuged at 800 X g for 10 min. The cells were subcultured in a 1 : 3 ratio on 35 mm or 100 mm Petri dishes. Cells of the first passage were routinely used in all experiments. Confluence was reached in 5 to 7 days. The cells were identified as of endothelial origin by their cobblestone appearance under phase constrast microscopy and by the presence of factor VIII-reiated antigen as seen by immunofluorescent microscopy [23].

[ 3H]Myristate labelling and thrombin stimulation Cell monolayers in 35 mm dishes were rinsed twice

with 37°C phosphate-buffered saline (PBS) and radio- labelled with 1 pCi/ml of [ 3H]myristate in medium-199 containing 10% fetal calf serum. After 20 h of incuba- tion, radiolabelled medium was removed, cells were rinsed twice with 37°C PBS and further incubated for 4 h with vitamin E-enriched medium.

For thrombin stimulation, vitamin E-enriched medium was removed and the cells were rinsed twice with warm PBS. The stimulation was started by adding

IC Tran et al. /Biochimica et Biophysics Acta 1212 (1994) 193-202 195

1.5 units/ml of thrombin dissolved in Hanks’ balanced salt solution (HBSS) (pH 7.41, containing 10 mM Hepes, 0.8 mM MgSO, and 5 mM CaCl,. The same incuba- tion buffer, but without thrombin, was added in the unstimulated control dishes. In some experiments, cells were incubated with or-propranolol (inhibitor of phos- phatidic acid phosphohydrolase) for 5 min before the addition of thrombin. In the experiments designed to determine the presence of phospholipase D activation in human endothelial cells, ethanol (O.l-2%) was in- cluded during thrombin stimulation. At appropriate times, stimulation was stopped by adding 1 ml of ice-cold methanol/cone. HCl (100: 1, v/v). The cells were scraped with a rubber policeman and collected into test tubes containing 40 pg of lipid standards (lysoPC, PI, PC, PE, PA, DAG, TG or PEt). Total lipids were extracted by the method of Bligh and Dyer [25]. The lower chloroform phase of the extraction was collected and stored at -20°C for lipid analysis.

myo-13H]Inositol labelling Cells grown at about 80% confluence in 35 mm

dishes were radiolabelled with 5 &i/ml of myo- [ 3H]inositol in free Dulbecco’s modified Eagle’s medium (DME) containing 10% of dialyzed fetal calf serum. After 20 h of incubation, cells were rinsed three times with serum-free medium-199, reincubated with vitamin E-enriched medium-199 for 4 h and stimulated with thrombin in the presence of 10 mM LiCl. The stimulation was stopped with 1 ml of ice-cold methanol; cells were scraped and cell lipids were extracted by chloroform/methanol by the method of Bligh and Dyer 1251. The upper aqueous phase of the extraction was collected and saved at -20°C for the determination of inositol phosphates.

Preparation of vitamin E-enriched medium Vitamin E-enriched medium-199 containing 10% fe-

tal calf serum was prepared as previously described [20]. Appropriate amounts of RRR-cY-tocopherol dis- solved in dimethyl sulfoxide (DMSO) are added to fetal calf serum. The amount of DMSO added did not exceed 0.2% of the total volume. The mixture was vortexed vigorously and incubated at 37°C in the dark for 15 min. Medium-199 and antibiotics were added last and the mixture was incubated for another 15 min before adding to the monolayers. Control medium con- tained the same amount of DMSO as the vitamin E-enriched medium.

Analysis of lipids by thin-layer chromatography The lower chloroform phase of the Bligh and Dyer

extraction was evaporated under N, gas. The lipid was resuspended in chloroform/methanol (9 : 1, v/v) and spotted on precoated glass silica gel G plates. The plates were developed with chloroform/ methanol/

acetic acid/water (85 : 15 : 10 : 3, by volume) as solvent system I. In this system, major phospholipids such as 1ysoPC (RF = 0.051, PI (RF = 0.14), PC (RF = 0.20), PE (RF = 0.40) and PA CR, = 0.63) are separated from each other and from neutral lipids. To separate neutral lipids, the plates were developed with hexane/diethyl- ether/acetic acid (70 : 30 : 1, by volume) as solvent sys- tem II. In this system, DAG (RF = 0.13), free fatty acid (RF = 0.36) and triacylglycerol (RF = 0.82) were sepa- rated from major phospholipids (RF = 0). To separate PEt from other phospholipids, the plates were devel- oped with the upper organic phase of ethylacetate/ iso- octane/ acetic acid/water (110 : 50 : 20 : 100, by volume) (system III) by which PA (RF = 0.07) and PEt (RF = 0.14) were separated from PC, PE, PI and LysoPC, all displaying an R, = 0). In this system, the neutral lipids migrate close to the solvent front. When it was neces- sary to simultaneously separate PA, PEt and DAG, the plates were first developed half-way with solvent sys- tem III, dried and redeveloped to the top of the plates with solvent system II [26]. The individual spots comi- grated with lipid standards were visualized by iodine vapor, scraped and quantified by liquid scintillation spectrometry.

Analysis of inositol phosphates Inositol phosphates were separated by anion-ex-

change chromatography [27]. About two thirds of the aqueous phase of the Bligh and Dyer extract was applied to a commercial plastic column containing 1 ml of Dowex (AG l-x8; formate form). Free inositol was first eluted with 20 ml of water. Glycerophospho- inositol, inositolmonophosphate, inositolbisphosphate and inositoltrisphosphate were eluted with 10 ml of 5 mM disodiumtetraborate/60 mM sodium formate, 10 ml of 0.2 M ammonium formate/O.l M formic acid, 10 ml of 0.4 M ammonium formate/O.l M formic acid and 5 ml of 1.0 M ammonium formate/O.l M formic acid, respectively. A 1.0 ml portion of each fraction was used for liquid scintillation counting.

Determination of diacyglycerol kinase activity Cells were harvested from 100 mm diameter dishes

by scraping in ice-cold phosphate-buffered saline (pH 7.41, containing 5 mM EDTA. The suspension was sedimented at 800 X g for 5 min and the pellet was resuspended in an homogenizing buffer (50 mM Tris- HCl (pH 7.4), 10% glycerol, 0.5 mM EDTA and 2 mM dithiothreitol). The cell suspension was then sonicated twice for 15 s periods at a power setting of 50% output using an ultrasonic cell disruptor. After the removal of nuclei and unbroken cells by a 800 x g centrifugation for 5 min, the postnuclear fraction was subjected to ultracentrifugation (100 000 x g, 30 min) to obtain the membrane and cytosolic fractions. The membrane pel- let was resuspended in homogenizing buffer and dis-

196 K. Tran et al./Biochimica et Biophysics Acta 1212 (1994) 193-202

rupted with the aid of 1 ml syringe fitted with a 25 gauge needle.

Diacylglycerol kinase activity was determined by measuring the incorporation of ‘*P from [y- 32P]ATP into DAG to form [“2P]PA. The standard assay mix- ture was adopted with slight modifications from the published methods [28,29]. To determine diacylglycerol kinase activity using exogenous DAG, the membrane or cytosolic fraction (lo-30 pg protein) was added in a total reaction mixture of 100 ~1 containing 50 mM Tris-HCl (pH 7.41, 50 mM NaCl, 15 mM MgCl,, 1 mM EGTA, 2 mM dithiothreitol, 20 mM NaF, 6.8 mol% (2 mM) diolein, 3 mol% (0.9 mM) PS and 51 mM octyl- glucoside. The reaction was started by adding [y- 32P]ATP (4000 cpm/lOO pmol Na, ATP) and the mixture was incubated at 30°C. After 5 min of incuba- tion, the reaction was terminated by adding 0.4 ml of methanol/chloroform (1 : 1, by volume) and 0.1 ml of 200 mM CaCl, [30]. The lower chloroform phase was washed three times with 1% of perchloric acid and spotted on silica gel G plates which were developed in the solvent system I. Radiolabelled phosphatidic acid spots were visualized by autoradiogram and quantified by Iiquid scintillation spectrometry.

Determination of DAG mass DAG mass was enzymatically quantified by using

the assay kit purchased from ~ersham. This assay was based on the conversion of cellular DAG to [ j2 P]PA by Escherichia coli diacylglycerol kinase in the presence of [ y-32P]ATP [31,32]. Briefly, one half of the lipid extracts obtained from 35 mm diameter dishes was first solubilized with detergent solution (n-octyl- glucoside/cardiolipinl using a waterbath sonicator. The soIuble lipid solution was treated with excess E. cob diacylglycerol kinase in the presence of [ y- j2P]ATP (specific activity: 50 000-100 000 cpm/nmoll. The [ 32P]PA formed was separated by thin-layer chro- matography using solvent system I, identified by auto- radiogram, and quantitated by liquid scintillation counting. The mass of [ .‘*P]PA was calculated from the known specific activity of [ y- ” PIATP.

3. Results

In the present study, cultured human endotheliai cells were used to investigate the role of vitamin E in the regulation of thrombin-stimulated DAG and PA formation. The generation of DAG in stimulated cells is known to be biphasic and originates from muItipie phospholipid sources. The earlier rise of DAG is de- rived from the hydrolysis of inositol phospholipids but the sustained level of DAG is formed through the combined actions of PC hydrolysis by PLD and the subsequent dephosphorylation of PA to DAG by PA

Table 1 Distribution of radioactivity in lipids of human endothelial cells

incubated with [3H]myristate

Cellular lipids Incorporation (% of total radioactivity)

-E +E

LysoPC 1.43 + 0.33 1.39 * 0.48

PI 5.05&0.17 5.16t: 1.45

PC 69.80 f 0.70 69.60 It 1.40

PE 3.33 $- 0.12 3.62 +0.21

PA 0.46 f 0.09 0.50 f 0.02

DAG 0.73+0.12 0.85 t 0.05 FA 1.26f0.26 0.95*0.14

TG 10.96 + 0.91 10.67 f 0.50

Others 6.98 * 0.20 7.26 F 0.30

Cell monolayers were labelled with [3H]myristate (1 pCi/ml) for 20

h and then incubated in the absence f-E) or presence (+ E) of

vitamin E (46 FM) for 4 h. Total lipids were extracted and quanti-

tated by TLC as described under Section 2. Values are means+S.D.

(n = 3 dishes) and they represent the percentage of radioactivity

among the bands on the TLC.

phosphohydrolase [33]. Since myristate is known to be preferentially incorporated into PC [34], cells were therefore labelled with this fatty acid.

Data in Table 1 shows that the inco~oration of [ “Hlmyristate was highest in PC (70%), followed by TG (ll%), PI (5%), and PE (3%). Tritium was incorpo- rated the least into DAG and PA, indicating the low steady state levels of these two lipid products. How- ever, vitamin E enrichment had no effect on the incor- poration pattern of myristate into various cellular lipids.

Effects of v~tarn~ E on t~rorn~~n-inda~ed changes in cellular lipid levels

In order to evaluate the role of vitamin E in the regulation of lipid metabolism, we sought to determine whether vitamin E enrichment has any influence on the cellular levels of lipids after they were stimulated by thrombin. Cell monolyers were labelled with [3H]myri- state for 20 h followed by vitamin E enrichment for 4 h. The time course of cellular lipid changes induced by thrombin was monitored. Fig. 1 shows that vitamin E enrichment caused a marked 40% decrease in DAG level after 5 min of thrombin stimulation. This was accompanied by a concomitant 2-fold increase in Ia- belled PA and a marked increase in PI. At the early time of stimulation (O-60 s), thrombin elicited a rapid increase in DAG and a sharp decrease in PI. In con- trast, the generation of PA was slow and reached maximum at 5 min after addition of thrombin. The effects of vitamin E on the change of DAG and PA were clearly observed between 2.5-10 min of stimula- tion. Despite its initial drop, the PI level in cells that were not enriched with vitamin E returned to its steady level between 1 and 5 min of thrombin stimulation. In contrast, the level of PI in the vitamin E-enriched cells was 2-fold higher than that of the control cells. In

K. Tran et al. /Biochimica et Biophysics Acta 1212 (1994) 193-202 197

addition, thrombin induced a gradual decrease of [3H]PC, and by 5 min there was a 5% drop in PC level in both vitamin E-treated and unteated cells. The amount of labelled PE in both vitamin E-treated and -untreated cells remained unaffected by thrombin in all time points tested (data not shown).

Effect of vitamin E on the change cellular DAG mass in thrombin-stimulated cells

Although vitamin E was found to affect the changes in radiolabelled DAG and PA observed in Fig. 1, the total mass pool of these lipid mediators remained unknown. We therefore sought to determine whether the treatment of vitamin E would lower the mass of DAG in thrombin-stimulated endothelial cells. Direct measurement of cellular DAG mass was performed in thrombin-stimulated cells pretreated with various con- centrations of vitamin E. Fig. 2 shows that vitamin E dose-dependently suppressed the elevation of DAG mass caused by thrombin-stimulation. The change of DAG level after 5 min of thrombin stimulation was detectable at 11.5 PM of vitamin E, and the inhibition was maximum (40-45%) when vitamin E concentra-

2.6

c? 0

s 2.1

E

t 1.4

2 0.7

1.2

C CI

5 0.9

5

5 0.6

2

0.3

= ; 70

5

f 65

L M 5 6

z

25

60 1, , , , , ,j 4

0 2 4 6 6 10

Time (min)

Fig. 1. Time course of radiolabelled lipids change by thrombin-

stimulated human endothelial cells enriched with or without vitamin

E. Cell monolayers were labelled with L3HImyristate (1 pCi/ml) for

20 h and then incubated in the absence (0) or presence (0) of

vitamin E (46 PM) for 4 h before being stimulated with thrombin

(1.5 U/ml) for the indicated times. Labelled DAG, PA, PI and PC

were quantified by TLC as described under Experimental proce-

dures. Values are means of two dishes from one representative

experiment out of three.

0.0 1 ’ I I I I I 0 23 46 69 92

Vitamin E (FM)

Fig. 2. Effect of vitamin E on the mass of DAG in thrombin-stimu-

lated human endothelial cells. Cell monolayers were incubated with

indicated concentrations of vitamin E for 4 h. Cells were then

treated with either 1.5 U/ml of thrombin (0) or buffer alone (0) for

5 min. The extracted lipids were subjected to DAG mass determina-

tion as described under Experimental procedures. Values are means

f S.D. (n = 5 dishes).

tions approached 46-69 PM. However, vitamin E had no effect on the basal level of DAG mass (Fig. 2).

Vitamin E-induced increase of PA is not caused by the activation of PLD nor the inhibition of PA phosphohy- drolase

In endothelial cells, evidence for the presence of a PLD activity that generates PA from the hydrolysis of PC has been recently reported 1351. To determine whether the increase in PA caused by the enrichment of vitamin E is due to an enhanced PLD activity, we sought to measure the transphosphatidylation reaction which is unique to PLD. When incubated in the pres- ence of an alcohol, PLD catalyses a direct transfer of the alcohol to the phosphatidyl group of a phospho- lipid substrate to form phosphatidylalcohol [36]. Thus, [ 3H]myristate labelled cells were enriched with or with- out vitamin E and stimulated with thrombin in the presence of ethanol and the formation of [3H]phos- phatidylethanol (PEtI was determined. Fig. 3 shows that there was no difference in the formation of [3H]PEt in both vitamin E treated and untreated cells. When these cells were incubated with increasing con- centration of ethanol, thrombin-induced [ 3H]PEt for- mation was linearly increased with increasing concen- tration of ethanol. However, the amount of PEt formed was independent of vitamin E enrichment. Analysis of the radiolabelled PA showed that vitamin E enrich- ment caused a marked increase in L3H]PA formation at all levels of ethanol tested. If PA was the only product of PLD, the increase of [3HlPEt should be accompa-

198 K. Tran et al. / Biochimica et Biophysics Acta 1212 (1994) 193-202

0.2

0.0

0-U _U--U-,”

ke 0

-P / e

I I I , I I

0.0 0.2 0.4 0.6 0.8 1 .o

EtOH (46)

Fig. 3. Effect of vitamin E on the formation of [3H]PA and [3H]PEt

by thrombin-stimulated human endothelial cells. Cell monolayers

were labelled with [3H]myristate (1 &i/ml) for 20 h and then

incubated in the absence (0, 0) or presence (0, n ) of 46 PM of

vitamin E for 4 h. Cells were stimulated with 1.5 U/ml thrombin (5

min) in the presence of indicated concentrations of ethanol. 13H]PA

(squares) and [3H]PEt (circles) were quantified by TLC as described

under Experimental procedures. Values are means of two dishes

from one representative experiment out of three.

nied by a decrease in [3HlPA due to the competition between the hydrolysis of PC by PLD and its trans-

phosphatidylation. Therefore, judging from the un- changed levels of [3H]PA found in Fig. 3, we could conclude that PA was formed not only from the activity of PLD, but also from other sources, such as DAG kinase that converts DAG to PA. Taken together, these data indicate that PA was not solely derived from the PLD pathway and the enhanced PA formation elicited by vitamin E was not caused by a change in PLD activity.

To determine whether the vitamin E-induced accu- mulation of PA is due to a change in PA phosphohy- drolase activity, the generations [3H]DAG and i3HlPA in thrombin-stimulated endothelial cells enriched with vitamin E was examined in the presence of propra- nolol, a PA phosphohydrolase inhibitor [37,38]. Table 2 shows that in response to 5 min of thrombin, propra- nolo1 elicited an increase in [3H]PA formation in both vitamin E treated and untreated cells, but it had no effect on the levels of [3H]DAG. Irrespective of the presence or absence of propranolol, vitamin E enrich- ment resulted in an increase in PA and a decrease in DAG formation. The fact that propranolol did not mimic the effect of vitamin E in the formation of DAG and PA ruled out the possibility that the increase in PA and the decrease in DAG formation was due to the inhibition of PA phosphohydrolase caused by vitamin E enrichment. However, direct determination of PA phosphohydrolase activity is needed to further prove the lack of vitamin E effect on this enzyme.

Table 2 Effects of propranolol (an inhibitor of PA phosphohydrolase) on the

formation of [“HIDAG and [“HIPA by thrombin-stimulated human

endothelial cells enriched with ( + E) or without t - E) vitamin E

[‘HI radioactivity (% of total)

DAG

-E +E

PA

-E +E

Control 0.76 k 0.02 0.84 k 0.03 0.40 ri_ 0.01 0.45 + 0.04 Thrombin 2.80 + 0.23 1.80+0.10 0.47kO.02 1.20+0.04 Propranolol 2.71 +O.l7 1.7OkO.16 0.78f0.11 1.74 + 0.01 + Thrombin

Cell monolayers were labelled with [“Hlmyristate (1 pCi/ml) for 20

h and then incubated in the absence t-E) or presence (+ E) of

vitamin E (46 FM) for 4 h. Cells were rinsed and preincubated with

propranolol (200 FM) in HBSS for 5 min before adding thrombin

(1.5 U/ml) for an additional 5 min. [‘HIDAG and [“HIPA were

quantitated by TLC as described under Section 2. Values are means

+ S.D. (n = 3 dishes) of one representative experiment out of three.

The vitamin E-induced decrease in DAG is not caused by a change in PI-PLC actiuity

The attenuated DAG level caused by vitamin E enrichment reported in Fig. 1 occurred only after the

4

3

2

1

0

1.5

x .!! 1.2 D

2 0.9 b

r; 0.6

E 0.3 ,”

0.0

0.3

0.2

0.1

0.0

L’ I I I I I_1

I I I I I I

Ip3

, I ! I I I

0 1 2 3 4 5

Time (min)

Fig. 4. Time course of [“Hlinositol phosphates formation by throm-

bin-stimulated human endothelial cells enriched with or without

vitamin E. Cells were labelled with myoj3H]inositol (5 pCi/ml) for

20 h and then incubated in the absence (0) or presence (0) of

vitamin E (46 ,uM) for 4 h before being stimulated with 1.5 U/ml of

thrombin for the indicated times. [‘HlInositol phosphates were quantitated by anion-exchanged chromatography as described in

Experimental procedures. Values are means of two dishes from one

representative experiment out of two.

K. Tran et al. / Biochimica et Biophysics Acta 1212 (1994) 193-202 199

first min of thrombin stimulation during which time there was a corresponding drop in the level of PI. However, a change in PI-PLC activity may account for the decrease in the level of DAG observed in the later time points. We therefore sought to determine whether vitamin E had any effect on the hydrolysis of phospho- inositides by phospholipase C (PLC) which might at- tenuate the level of DAG. Cell monolayers were la- belled with myo-[3Hlinositol for 20 h before enrich- ment with or without vitamin E for 4 h. This was followed by the addition of thrombin and the labelled inositol phosphates were monitored over time. Fig. 4 shows that thrombin induced a rapid generation of [3HlIP, and L3H]IP3 that peaked at 15 s, while [3H]IP level increased in a near-linear fashion over a 5 min time course. Vitamin E had no effect on the generation of both [3HlIP and [3HlIP2, but it attenuated the transient elevation of [3H]IP3 at 15 s. At this time

350

300

250

200

150 - 0 23 46

Fig. 5. Effect of vitamin E on the resting activity of DAG kinase in

human endothelial cells. Cell monolayers were incubated with indi-

cated concentrations of vitamin E for 4 h. Cells were scraped,

homogenized and fractionated by ultracentrifugation into a 100000

x g pellet (membrane fraction) and 100000x g supernatant (cyto- solic fraction). DAG kinase activity in the total homogenate (A)

membrane fractions (B) or the cytosolic fractions (Cl was determined

as described under Experimental procedures and expressed as pmol

of [y-32PlATP transferred to exogenous DAG (diolein) per min per mg protein. Values are means of duplicate from one representative experiment out of four.

Total MembraLle Cytosol

Fig. 6. Effect of vitamin E on thrombin-stimulated DAG kinase

activity in human endothelial cells. Cells monolayers were incubated

in the absence (hatched bar) or presence (solid bar) of vitamin E (46

PM) for 4 h followed by 5 min of thrombin (1.5 U/ml) stimulation.

Cells were scraped, homogenized and fractionated by ultracentrifu-

gation into a 100000~ g pellet (membrane fraction) and supernatant

(cytosolic fraction). DAG kinase activity was determined as described

under Experimental procedures. Values were presented as percent

increase of DAG kinase activity caused by thrombin stimulation

when compared with non-stimulated control. The 32P cpm values of

non-stimulated controls in total homogenate, membrane and cytosol

from cells treated without/with vitamin E were 7.0/9.3, 3.5/4.0 and

5.0/7.5 (X lo3 cpm), respectively. Values are means of duplicate

from one representative experiment out of three.

point, [3H]IP3 levels were ten times lower than those of other inositolphosphates. In addition, vitamin E had no effect on the change in [3Hlinositol labelled PI induced by thrombin (data not shown). Taken together, these data show that the attentuated DAG level caused by vitamin E enrichment is not mediated from an inhibi- tion of PI-PLC activity.

Vitamin E enhances DAG kinase activity To further investigate whether vitamin E affected

the pathway that removes DAG, the activity of DAG kinase was determined in the membrane and the cy- tosolic fractions obtained from cells pretreated with and without vitamin E. By using diolein as the sub- strate and the standard assay conditions for DAG kinase described under Experimental procedures, vita- min E enrichment enhanced the resting DAG kinase activity in total cell homogenates, membrane and cy- tosolic fractions (Fig. 5). At the concentration of 46 PM, vitamin E caused a 22% increase of DAG kinase activity in the membrane and 42% increase in the cytosol. Fig. 6 shows that thrombin activates DAG kinase activity in both membrane (40% increase) and cytosol (30% increase). In addition, vitamin E enrich- ment further potentiates DAG kinase activity in all cellular fractions determined (66% increase in cell homogenate, 60% in membrane and 40% in cytosol).

200 K. Tran et al. / Biochimica et Biophysics Acta 1212 (1994) 193-202

Therefore, vitamin E enrichment activates both resting CDP-DAG and its subsequent condensation with my- and stimulated DAG kinase activity in endothelial cells. oinositol [401.

4. Discussion

RRR-cu-tocopherol, a naturally occurring and most potent form of vitamin E, was found for the first time to regulate the level of DAG and PA in human en- dothelial cells in response to thrombin stimulation. Our studies demonstrated that the reduction of cellu- lar DAG level by vitamin E is mediated by activation of DAG kinase and therefore resulting in an up-regu- lation of PA formation. The effect of vitamin E on DAG levels was further supported by direct measure- ment of DAG mass (Fig. 2). Our studies showed that DAG was generated from phosphodiesteri~ cleavage of both phosphoinositides and phosphatidylcholine by ei- ther phospholipase C or phospholipase D and PA phosphohydroIase; but neither of these DAG generat- ing pathways was found to be affected by vitamin E. These findings suggest a new role of vitamin E in signal transduction by down-regulating the levels of impor- tant second messenger such as DAG.

In response to various extracellular signals, there is an increased turnover of phosphoinositides and a tran- sient accumulation of DAG in numerous cell systems [lo]. For example, thrombin stimulates the hydrolysis of PIP, by PI-specific PLC thereby generating IP, and DAG in human endothelial cells [2,14]. Consistently, we observed that thrombin stimulated a transient in- crease of [“HIDAG and caused a rapid decrease in radiolabelled-PI in [ 3H]myristate-labelled cells with a concomitant rise of inositol phosphates in [ “Hlinositol- Iabelled cells at the first IS-30 s of stimulation (Figs. 1 and 4). However, hydrolysis of [~Hlmyristate labelled PI was transient while the level of DAG was main- tained throughout the time course of stimulation (Fig. 1). PC, an important source of arachidonate release and prostacyclin synthesis in human endothelial cells, was found to be a substrate for DAG production in numerous cell types including endothelial cells [7,26,36,39]. We observed that thrombin stimulated a gradual decrease in cellular radiolabelled PC, without any change in the level of PE. When comparing be- tween the changes in DAG, PI and PC in response to thrombin, our results confirmed that PI hydrolysis con- tributed to the initial rise of transient DAG level whereas PC hydrolysis contributed to the maintenance of the sustained accumulation of DAG. The increase in radiolabelled PI in vitamin E enriched cells could be explained by its rapid resynthesis from PA, the level of which was increased by vitamin E. In most cell systems, DAG undergoes rapid phosphorylation by DAG kinase to PA and the resynthesis of inositol phospholipids occurs through the reaction of PA with CTP to form

The increase in PA and the decrease in DAG levels caused by vitamin E enrichment could be due to changes in the pathways responsible for either their generation or their interconversion. These include PLD, PA phosphohydrolase, PLC or DAG kinase. Through differential radiolabelling of cellular phospho- lipids and studies with enzyme inhibitor, we have sys- tematically elimated the involvement of PLD, PA phos- phohydrolase and PLC as the cause of the vitamin E elicited change in DAG and PA. In the presence of ethanol (Fig. 31, the formation of PEt in response to thrombin stimulation confirmed the presence of PLD activity. However, vitamin E enrichment did not alter the amount of PEt formed despite of an accumulation in PA (Fig. 3). One possibility to explain for the ele- vated PA level was through an inhibition of PA phos- phohydrolase. To test this possibility, we used bt_-pro- pranolol, a P-adrenergic receptor antagonist [41] and an inhibitor of PA phosphohydrolase [37,38], to deter- mine whether this drug can mimic the vitamin E-in- duced changes in PA and DAG levels. The increase in PA formation in the presence of propran(~lol indicated that there was inhibition of PA phosphohydrolase (Ta- ble 2). However, the lack of effect of propranolol in altering DAG levels in both vitamin E treated and untreated cells indicated that the elevation of PA and the reduction of DAG levels induced by vitamin E were not caused by an inhibition of PA phosphohydro- lase.

We next investigated the aIternative involvement of the PI-specific PLC which directly generates DAG from the hydrolysis of phosphoinositides such as PI, PIP and PIP, [40]. The amount of IP and lPZ formed far exceeded that of IP, (Fig. 41, indicating that the bulk of DAG is mainly derived from the hydrolysis of PI and PIP by a PI-specific PLC. Vitamin E enrich- ment had no effect on the generation of IP and IP,. In contrast, we observed that the early peak of IP, which occurred within 1.5 s following thrombin stimulation was attenuated by vitamin E. However, the attenuation of this early rise in lPS cannot be responsible for the decrease in the sustained accumulation of DAG which occurred after 2 min of thrombin stimulation. There- fore, the reduction of DAG levels induced by thrombin did not result from an inhibition of PLC by vitamin E.

Finally, the effect of vitamin E on DAG and PA levels could be explained entirely on the basis that vitamin E activates DAG kinase. Direct measurement of DAG kinase activity revealed that vitamin E signifi- cantly enhanced the basal activity of this enzyme in both particular and cytosolic fractions. Furthermore, the synergistic activation of DAG kinase caused by a combination of thrombin and vitamin E would explain the lowering of DAG level. The findings that vitamin E

K. Tran et al. /Biochimica et Biophysics Acta 1212 (1994) 193-202 201

dose-dependently suppressed the accumulation of DAG mass induced by thrombin provided further di- rect evidence to substantiate this possibility. However, irrespective to the activation of DAG kinase, vitamin E did not significantly increase PA or decrease DAG levels in resting cells (Figs. 1 and 2). This could be due to the discrimination of DAG kinase towards different DAG pools in the resting and stimulated cells; or alternatively, it is plausible that the changes of DAG and PA levels induced by vitamin E were too little to be detected due to their low steady state levels in the resting cells.

DAG is considered to be an important second mes- senger in many cell types including endothelial cells, and one of its it function is in the activation and translocation of the Cazf-phospholipid dependent pro- tein kinase C [lO,ll]. Activation of PKC is required in many agonist-induced endothelial cell events such as the synthesis of prostaglandins [4,13,42-441, barrier dysfunction [45] and von Willebrand factor secretion [46]. In addition, it was found that PKC participates in the regulation of DAG kinase activation by inducing translocation and protein phosphorylation of this en- zyme [47]. Other mechanisms for DAG kinase regula- tion include the activation of the enzyme by free oleic acid and oleoyl-CoA [48], and by arachidonoyl-DAG [49]. In contrast, DAG kinase activity was found to be inhibited by lipoxygenase metabolites of arachidonic acid such as 15- and 1Zhydroxyeicosatetraenoic acids (15-HETE and 1ZHETE) [50]. Thus, the enhanced DAG kinase activity caused by vitamin E may be a secondary effect of vitamin E in controlling any of the above processes that occur in endothelial cells. We have reported that vitamin E enhanced arachidonic acid release [21] and the acylation of alkylglycerophos- phatidylcholine in endothelial cells [22]. We also ob- served that in endothelial cells, vitamin E increased the PMA-unstimulated membrane PKC activity, but atten- uated PMA-induced PKC translocation (unpublished data).

In summary, vitamin E suppressed the accumulation of thrombin induced DAG levels via an enhanced DAG kinase activity, therefore resulting in an increase of PA in endothelial cells. Enzymes such as PLD, PLC and PA phosphohydrolase were apparently unaffected by vitamin E. The precise mechansim(s) by which vita- min E exert its effect on the changes in these impor- tant lipid mediators and the implication of these changes in endothelial cell metabolism remain to be determined.

Acknowledgements

This study was supported by the Medical Research Council of Canada (MA 7626) to A.C.C. We thank Joanne Barlow for typing the manuscript.

References

[l] Dennis, E.A., Rhee, S.G., Billah, M.M. and Hannum, Y.A.

(1991) FASEB J. 5, 2068-2077.

[2] Brock, T.A. and Capasso, E.A. (1988) J. Cell. Physiol. 136,

54-62.

[3] Garcia, J.G.N.. Fenton II, J.W. and Natarajan, V. (1992) Blood

79, 2056-2067.

[4] Whatley, R.E., Nelson, P., Zimmerman, G.A., Stevens, D.L.,

Parker, C.J., McIntyre, T.M. and Prescott, S.M. (1989) J. Biol.

Chem. 264. 6325-6333.

151 f61 171

fs1

[91

[lOI

[ill

L1-21

[131

[141

[151

1161

1171

fts1 [191

DO1

ml

WI

D31

L’41

[251

P.61

L’71

1281

(291

[301

[311

[321

[331

[341

[351

[361

Billah, M.M. and Anthes, J.C. (1990) Biochem. J. 269, 281-291.

Exton, J.H. (1990) J. Biol. Chem. 265, l-4.

Ragab-Thomas, J.M.F., Hullin, F., Chap. H. and Douste-Blazy,

L. (1987) Biochim. Biophys. Acta 917, 388-397.

Martin, T.W., Wysolmerski, R.B. and Lagunoff, D. (1987)

Biochim. Biophys. Acta 917, 296-307.

Martin, T.W. (1988) Biochim. Biophys. Acta 962, 282-296.

Nishisuka, Y. (1986) Science 233, 305-312.

Bell, R.M. and Burns, D.J. (1991) J. Biol. Chem. 266,4661-4664.

Berridge, M.J. and Irvine, R.F. (1984) Nature 308, 693-698.

Garcia. J.G.N., Stasek, J.. Natarajan, V., Patterson, C.E. and

Dominguez, J. (1992) Am. J. Respir. Cell. Mol. Biol. 6, 315-325.

Jaffe, E.A., Grulich, J., Weksler, B.B., Hampel, G. and Watan-

abe, K. (1987) J. Biol. Chem. 262, 8557-8565.

Boscoboinik, D., Szewezyk, A., Hensey, C. and Azzi, A. (1991)

J. Biol. Chem. 266, 6188-6194.

Douglas, C.E., Chan, A.C. and Choy, P.C. (1986) Biochim.

Biophys. Acta 876, 639-645.

Cao, Y.Z., Karmin, O., Choy, P.C. and Chan, A.C. (1987)

Biochem. J. 127, 135-140.

Hamelin, S.-St. and Chan, A.C. (1983) Lipids 18, 267-269.

Chan, A.C. and Leith, M.E. (1981) Am. J. Clin. Nutr. 34,

2341-2347.

Tran, K. and Chart, A.C. (1990) Biochim. Biophys. Acta 1043,

189-197.

Tran, K. and Chan, A.C. (1988) Biochim. Biophys. Acta 963,

468-475.

Tran, K., D’Angelo, A.F., Choy, P.C. and Chan, A.C. (1994)

Biochem. J. (in press).

Chart, A.C. and Tran, K. (1990) Lipids 25, 17-21.

Jaffe, E.A. (1984) in Biology of Endothelial Cells (Jaffe, E.A.,

ed.), pp. 1-13, Martinus Nijhoff, Boston.

Bligh, E.G. and Dyer, W.J. (1959) Can. J. Biochem. Physiol. 37, 911-917.

Billah, M.M., Eckel, S., Mullman, T.J., Egan, R.W. and Siegel,

M. (1989) J. Biol. Chem. 264, 17069-17077.

Berridge, M.J., Dawson, R.M.C., Downes, P., Heslop, J.P. and

Irvine, R.F. (1983) Biochem. J. 212, 473-482.

Walsh, J.P. and Bell, R.M. (1986) J. Biol. Chem. 261, 6239-6247.

MacDonald, M.L., Mack, K.F., Richardson, C.N. and Glomset,

J.A. (1988) J. Biol. Chem. 263, 1575-1583.

Ford, D.A. and Gross, R.W. (1990) J. Biol. Chem. 255, 12280- 12286.

Kennerly, D.A., Parker, C.N. and Sullivan, T.J. (1979) Anal.

Biochem. 98, 123-131.

Preiss, J., Loomis, C.R., Bishop, W.R., Stein, R., Niedel, J.E.

and Bell, R.M. (1986) J. Biol. Chem. 261, 8597-8600.

Pelech, S.L. and Vance, D.E. (1989) Trends Biochem. Sci. 14, 28-30.

Martin, T.W. and Michaelis, KC. (1988) Biochem. Biophys. Res. Commun. 157, 1271-1279.

Martin, T.W. and Michaelis, K. (1989) J. Biol. Chem. 264, 8847-8856.

Bocckino, S.B., Blackmore, P.F., Wilson, P.B. and Extort, J.H. (1987) J. Biol. Chem. 262, 15309-15315.

202 EL Tran et al. / Biochimica et Biophysics Acta 1212 (1994) 193-202

[37] Pappu, AS. and Hauser, G. (1983) Neurochem. Res. 8, 1565- [44] Carter, T.D., Hallam, T.J. and Pearson, J.D. (1989) Biochem. J.

1575. 262, 431-437.

[38] Koul, 0. and Hauser, G. (1987) Arch. Biochem. Biophys. 253,

453-461.

[39] Price, B.D., Morris, J.D.H. and Hall, A. (1989) Biochem. J. 264,

509-515.

[45] Lynch, J.J., Ferro, T.J.. Blumenstock, F.A., Brachenauer, A.B.

and Malik, A.B. (1990) J. Clin. Invest. 85, 1991-1998.

[46] Carew, M.A., Paleolog, E.M. and Pearson, J.D. (1992) Biochem.

J. 286, 631-635.

[40] Michell, R.H. (1986) in Phosphoinositides and Receptor Mecha-

nisms (Putney, J.W., ed.), pp. l-24, Allan R. Liss. [41] Stiles, G.L., Caron, M.G. and Lefkowitz, R.J. (1984) Physiol.

Rev. 64, 661-743.

[47] Kanoh, H., Yamada, K., Sakane, F. and Imaizumi, T. (1989)

Biochem. J. 258, 455-462.

[42] Helter, R., Bussolino, F., Ghigo, D., Garbarino, G., Schriider,

H., Pescarmona, G., Till, U. and Bosia, A. (1991) Biochim.

Biophys. Acta 1093, 55-64.

[48] Kelleher, J.A. and Sun, G.Y. (1989) J. Neurosci. Res. 23, 87-94.

[49] MacDonald, M.L., Mack, K.F., Williams, B.W., King, W.C. and

Glomset, J.A. (1988) J. Biol. Chem. 263, 1584-1592.

[50] Setty, B.N.Y., Graeber, J.E. and Stuart, M.J. (1987) J. Biol.

Chem. 262, 17613-17622.

[43] Zavoico, G.B., Hrbolich, J.K., Gimbrone, M.A. Jr. and Schafer,

A.I. (1990) J. Cell Physiol. 143, 596-605.