Composition and incorporation of [3H]arachidonic acid into molecular species of phospholipid classes by cultured human endothelial cells

211

Biochimica et Biophysics Acta 877 (1986) 211-215 Elsevier

BBA Report

BBA 50148

Composition and incorporation of [3H]arachidonic acid into molecular species of phospholipid classes by cultured human endothelial cells

Merle L. Blank a, Arthur A. Spector b, Terry L. Kaduce b and Fred Snyder a3* aMedical and Health Sciences Division, Oak Ridge Associated Universities, Oak Ridge, TN 37831, and b Departments of Biochemistry and Internal Medicine, University of Iowa, Iowa City, IA (U.S.A.) (Received

Key words:

Arachidonic

acid; Phospholipid

February

2nd, 1986)

composition;

Phospholipid

class; (Human

endothelial

cell)

Based on quantitative high-performance liquid chromatographic analyses of molecular species in selected phospholipid subclasses from cuhured human umbilical vein endothelial cells, the relative degree of unsaturation was ethanolamine plasmalogens > phosphatidylethanolamine > phosphatidylcholine. A total of 36 different molecular species were identified in the phosphatidylcholine fraction. Interestingly, the phosphatidylcholine contained a significant amount (11.7%) of the dipahnitoyl species, a lipid normally associated with lung surfactant. The arachidonoyl-containing molecular species of phosphatidylserine/ inositol were labeled to the highest extent and the ethanolamine plasmalogens contained the lowest specific radioactivity after incubating [3H]arachidonic acid with human endothelial ceils for 4 h. Within each phospholipid subclass the arachidonoyl species where both acyl groups of the phospholipid are unsaturated (20 : 4-20 : 4, 18 : 2-20 : 4 + 16 : l-20 : 4, and 18 : l-20 : 4) had higher specific radioactivities, after labeling with [3H]arachidonic acid, than those that contained saturated aiiphatic chains (16: O-20: 4 and 18 : O-20: 4). This indicates that the unsaturated species have higher turnover rates.

Endothelial cells synthesize a number of cyclooxygenase and lipoxygenase products from arachidonic acid [l-3]. These eicosanoid products, especially prostaglandin I, (prostacyclin), mediate several important physiologic actions, including the inhibition of platelet aggregation and the prevention of arterial constriction [4]. When grown in a medium containing serum, cultured endothelial cells do not convert appreciable amounts of linoleic acid, the usual 20 : 4 precursor, to arachidonic acid [5-71. Instead, the endothelial cells obtain arachidonic acid derived from extracellular sources, either from the plasma-free fatty acids [8,9] or plasma lipoproteins [lO,ll]. Nothing is presently known about the distribution of arachidonic acid in specific molecular species of phospholipids and the factors that regulate * To whom correspondence

should be addressed.

0005-2760/86/$03.50 0 1986 Elsevier Science Publishers

arachidonic acid release from phospholipids in endothelial lipids. Recent studies have shown that when arachidonic acid is initially taken up by rat testes, a substantial amount is incorporated into phosphoglyceride molecules that contain two polyenoic acyl groups, primarily the diarachidonoyl and linoleoylarachidonoyl species [12]. These diacylpolyenoic phosphoglycerides have a high turnover rate, suggesting that they may be involved in arachidonic acid processing into cellular phospholipid storage pools. Because of the importance of arachidonic acid in endothelial functions, it was of interest to determine whether diacyl-polyenoic phosphoglyceride molecular species also are formed when cultured endothelial cells take up arachidonic acid. Primary cultures of human umbilical vein endothelial cells were prepared according to a slight

B.V. (Biomedical

Division)

212

modification of the method of Jaffe et al. [13] as previously described [5,8]. Briefly, the cells were suspended in a modified Medium 199 containing 20% heat-inactivated fetal bovine serum, counted with a hemocytometer, and seeded into plastic flasks (75 cm*) at a concentration of 9. lo4 cells per cm*. After incubation for 24 h at 37°C in an atmosphere containing 5% CO,, the medium was replaced with fresh Medium 199 containing 20% fetal bovine serum and 25 mM 4-(2-hydroxyethyl)-1-piperazineethansulfonic acid. Cultures were then maintained for three days at 37°C in a 5% CO, atmosphere. After the maintenance medium was removed, cell monolayers were washed three times with 2 ml of Dulbecco’s phosphate-buffered saline and then incubated in a 5% CO, atmosphere with 15 ml of Medium 199 containing 2% fetal bovine serum enriched with 19.2 PCi of [5,6,8,9,11,12,14,153Hlarachidonic acid. The labeled arachidonic acid was added to the serum as a warm solution of the sodium salt [8]. Reactions were terminated after 4 h by removal of the incubation medium and washing the cells twice with 2 ml of ice-cold Dulbecco’s phosphate-buffered saline, followed by the addition of 1.5 ml of ice-cold methanol into each dish. Cells were harvested by scraping, and the methanol-cell extract was quantitatively transferred to separatory funnels with an additional methanol rinse. The lipids were extracted with chloroform/methanol (2 : 1, v/v) by the method of Folch et al. [14]. After combining the upper phase wash with the original lower phase, the solvent was evaporated under N,, and the lipids were resuspended in 1 ml of chloroform/methanol. A portion of this extract was used for measurement of total radioactivity contained in cellular lipids [8]. Two separate experiments were done, using several pooled endothelial cell cultures for each experiment (7 mg of cellular protein [15] in the first experiment and 11.4 mg in the second), to obtain enough unlabeled phospholipid to determine the composition of HPLC peaks that contained more than one molecular species. Aliquots of the total lipid extracts of the cells were separated by TLC on 250micron layers of either silica gel G layers developed in hexane/diethyl ether/glacial acetic acid (80 : 20 : 1, v/v) for neutral lipids or silica gel HR layers developed in

chloroform/methanol/glacial acetic acid/water (50 : 25 : 8 : 3, v/v) for phospholipids. Lipid standards were cochromatographed with aliquots of the labeled samples and, after staining with iodine vapor to locate the separated lipid classes; each area representing a specific lipid class was scraped into vials for assay of radioactivity by liquid scintillation spectrometry [16]. With HPLC analyses, phospholipid classes were first isolated on silica gel HR layers (2 X 8 inch TLC plates) developed in the solvent system described above for phospholipids. The developed TLC plates were exposed to ammonia vapor, sprayed with a solution of 0.05% 2,7-dichlorofluorescein in 50% ethanol, and viewed under ultraviolet light to locate the separated phospholipid classes for subsequent extraction. Each of the major glycerol-containing phospholipid classes was then treated with phospholipase C [17] to obtain the diradylglycerol fractions present in each class. After extraction of these fractions with hexane, the diradylglycerols (diacyl, alkylacyl, and alk-1-enylacyl subclasses) were converted to diradylglycerobenzoates for resolution of each by TLC as previously described [18]. The subclasses that contained sufficient mass and radioactivity were then analyzed by reversephase HPLC to determine the distribution of molecular species and their specific radioactivities [12,18]. For detailed determination of the molecular species in unlabeled cells from the pooled cultures, it was necessary to collect the HPLC peaks that contained a mixture of species, methylate the mixture, and analyze the methyl esters by GLC. The relative amount of diacyl, alkylacyl, and alk-1-enylacyl subclasses present in the choline glycerophospholipids was 88.6, 4.6, and 6.8 mol%, whereas the ethanolamine glycerophospholipids consisted of 52.6, 4.9, and 42.5 mol%, respectively. Table I presents a detailed analysis of the molecular species in phospholipid subclasses of human endothelial cells for those fractions with sufficient mass to do such an analysis. The major molecular species in the phospholipid subclasses were 16 : O18 : 1, 18 : O-20: 4, and 16 : O-20 : 4 for diacyl-snglycero-3-phosphocholine (diacylglycerophosphocholine), diacyl-sn-glycero-3-phosphoethanolamine (diacylglycerophosphoethanolamine), and alk-1-enylacyl-sn-glycero-3-phosphoethanolamine

213 TABLE I DISTRIBUTION OF MOLECULAR SPECIES IN SELECTED PHOSPHOLIPID SUBCLASSES OF HUMAN ENDOTHELIAL CELLS Values represent the mol% of each species. Values for molecular species that elute as single components from HPLC are given as the means f SE. (n = 3). HPLC peaks that contained a mixture of molecular species required more lipid for subsequent GLC analyses and are reported as the averages from two separate pooled samples. The first number in the molecular species represents the length of the carbon chain and the second number the number of double bonds in the chain (the vinyl ether bond is excluded). N.D., not detected. GPC, ~ycerophosph~holine; GPE, gIycerophosph~thanolamine. Molecular species 20:4-20:4 16:1-20:4 18 : Z-20: 4 > 18:1-22:6 16:0-22:6 18:1-22:5(n-3) 16:0-22:5(n-3) 18:1-20:4 16:0-20:4 18:1-22:5(n-6) 16:0-22:5(n -6) 18:1-18:2 18:0-22:6 18:1-20:3(n -6) 18:1-22:4 16:1-18:l 16:0-18:2 14:0-18:l 16:0-16:l 16:0-20:3(n -6) 16:0-22:4 18:0-22:Sfn -3) 18:0-20:4 16:0-20:3(n-9) 18:0-22:5(n-6) 16:0-20:2 18:1-18:l 18:1-20:2 16:0-18:l 18:0-22:4 18:0-20:3(n -6) 18:0-18:2 18:0-16:l 16:0-16:O 17:0-18:l 18:1-2O:l 18:0-18:l 16:0-18:O

DiacylGPC

DiacylGPE

Alk-l-enylacylGPE

0.3*0.1

0.4kO.l

N.D.

0.9fO.l

0.5kO.l

N.D.

0.7*0.1 1.5*0.0 1.2+0.1 0.7 3.2 5.0 0.1 0.5 * 0.0 2.5 0.7 1.2 0.8 2.0 3.0 1.3 3.9 1.4 1.4 3.oio.7 5.2 0.1 1.5 *0.2 1.3 5.4 1.8 20.8 0.4 1.2 1.8 0.6 11.7io.3 1.3*0.1 0.910.1 5.t3*0.3 2.3kO.l

2.5 k 0.3 2.2Ito.2 1.6fO.l 1.0 5.6 2.5 0.7 0.7 * 0.1 1.4 5.1 1.2 2.8 1.0 0.8 N.D. 0.3 0.6 1.6 3.6kO.O 21.6 N.D. 1.6+0.1 6.6 < 0.1 4.8 7.4 4.0 2.2 0.7 0.5 * 0.2 0.8 f 0.0 0.7 *0.1 11.21f10.7 0.6kO.l

3.0+0.2 7.4 f 0.4 1.7kO.2 4.0 9.0 26.7 -=z0.1 1.2i0.5 < 0.1 4.0 0.2 2.1 1.8 0.3 N.D. < 0.1 0.8 7.4 2.4*0.3 13.4 N.D. 1.1*0.2 1.3 *0.3 ( 1.7 2.8 0.4 0.2 1.2 0.3+0.1 -=z0.1 < 0.1 0.7 * 0.1 < 0.1

(ok-1-enylacylglycerophosphoethanolamine), respectively. The relative unsaturation of the phospholipid subclasses, as determined by the total amount of species that contained at least four double bonds/mol, was alk-l-enylacylglycerophosphoethanolamine > diacylglycerophosphoethanolamine > diacylglycerophosphocholine. It is also interesting that the diacylglycerophosphocholine contained a significant amount of the 16: O16 : 0 molecular species (11.7%), whereas the other phospholipid subclasses had less than 1% of this species. Because considerable variation was found in the mass distribution of molecular species in the diacyl-sn-glycero-3-phosphoserine (diacylglycerophosphoserine)/diacyl-sn-glycero-3-phosphoinositol (diacylglycerophosphoinositol) fractions, these results are not included in Table I. However, the two major molecular species present in all samples of diacylglycerophosphoserine/diacylglycerophosphoinositol analyzed were 18 : O20 : 4 (13.5-28.4%) and 18 : O-18 : 1 (16.2-24.3%). In a preli~nary experiment on the time course of arachidonic acid uptake by cultures of human endothelial cells (data not shown), we found that a 4 h incubation time, although not reaching the maximum uptake of arachidonic acid, would provide enough radioactivity in the lipids to carry out our experiments. Cellular lipids from duplicate culture flasks that were incubated 4 h with [3H]arachidonic acid contained 48.1% and 51.1% of the added tritium. Based on the average of duplicate analyses of cultures (which agreed within less than i_ 10% relative error), over 90% of the cellular tritium in lipids was found in four fractions: triacylglycerols (13.2%), diradylglycerophosphoethanolamine (12.2%), diradylglycerophosphoserine/diradylglycerophosphoinositol(23.1%), and diradylglycerophosphocholine (43.3%). Only 2.0% of the cellular tritium remained as the free fatty acid. TLC analyses of the phospholipid subclasses (as diradylglycerobenzoates) revealed that diacylglycerophosphoethanolamine, diacylglycerophosphocholine, diacyIglycerophosphoserine/ diacylglycerophosphoinositoI, and alk-l-enylacylglycerophosphoethanoIa~ne were labeled to the greatest extent with [ 3Hfarachidonic acid (Table II). In contrast, very little (3H]arachidonic acid was incorporated into alkylacylglycerophospho-

214 TABLE II PERCENT DISTRIBUTION OF [3H]ARACHIDONIC ACID IN PHOSPHGLIPID SUBCLASSES OF HUMAN ENDOTHELIAL CELLS Values are based on TLC analyses of pooled lipid extracts from two cell culture flasks. [3H]Arachidonic acid was incubated with cultures of human endothelial cells as described in the text. GPC, glycerophosphocholine; GPE, glycerophosphoethanolamine; GPS, glycerophosphoserine; GPI, ~ycerophosphoinositol. Subclass fraction

Alk-I-enylacyl Alkylacyl Diacyl

Phospholipid class (W distribution) DiradylGPC

DiradylGPE

DiradylGPS/GPI

4.6 2.6 92.8

47.0 2.2 50.8

0.6 0.9 98.5

choline, a known precursor [19-21] of a biologically active lipid, platelet activating factor (l-alkyl-2-acetyl~ycerophosphocho~ne). Phospholipid subclasses from cells labeled with [3H]arachidonic acid were analyzed by HPLC to determine the tritium distribution in various molecular species. Although small amounts (none more than 5% of the total) of other tritium labeled species were eluted during the HPLC analyses, we focused our attention only on the arachidonoylcontaining molecular species that we have previously identified [12,18]. In general, those molecular species that occurred in the highest amounts also incorporated more t3HJarac~donic acid than the less abundant species. Specific radioactivities

of the arachidonoyl molecular species in the four major phospholipid subclasses, calculated from the HPLC data, are shown in Table III. When the 20 : 4-20 : 4 peak from ~acyl~ycerophosphocholine was collected from HPLC, hydrogenated, and the hydrogenated sample again analyzed by HPLC [12], we found that more than 70% of the tritium was associated with the 20 : O-20 : 0 species. Using the same meth~ology, we found that the tritium in the original 16 : l-20 : 4 plus 18 : 2-20 : 4 HPLC peak from diacylglycerophosphocholine contained 49% 3H in the 16: l-20:4 and 51% 3H in 18: 220 : 4. If any of the 20 : 4-20 : 4 molecular species of the three diacyl phospho~pid classes contained the [3H]arac~donate at both the sy1- 1 and sn - 2 positions of glycerol, their relative specific activities would be one-half of the values stated in Table III; however, due to the small amounts of sample, positional analyses of fatty acids in these samples were not feasible. Also, under the conditions of our HPLC analyses, the molecular species of 36 : O-22 : 5 (n - 3) and 18 : l-22 : 5( n - 6) would elute with the 18: l-20:4 and the 16:0-20:4 molecular species, respectively. Therefore, except for the diacylglycerophosphose~ne/diacyl~ycer~ phosphoinositol fraction, the specific radioactivities of the 18: l-20: 4 and 16: O-20: 4 molecular species given in Table III were corrected using the analytical data reported in Table I. Within each diacyl phospholipid subclass the specific radioacti~ties, in decreasing order, were 16 : 1 - 20 : 4 + 18 : 2-20 : 4 > 20 : 4-20 : 4 > 18 : l20 : 4 > 16 : O-20 : 4 > 18 : O-20 : 4; these results in-

TABLE III SPECIFIC ~DIOA~IVITIES OF A~CHIDONOYL-CONTAINING MOLECULAR HUMAN ENDOTHELIAL CELLS LABELED WITH [3H]ARACHIDONIC ACID

SPECIES OF PHOSPHOLIPIDS

IN

Values are expressed in dpm 3H*10-5/nmol based on HPLC analyses of pooled lipid extracts from two cell culture flasks. [3H]Arachidonic acid was incubated with cultures of human endothelial cells as described in the text. Molecular species

DiacylGPC

DiacylGPS/GPI

DiacylGPE

20:4-20~4

3.11

6.82

4.69

9.61

7.85

2.56 2.42 1.97

1.38 1.26 0.55

18:2-20:4 16:1-20:4 > 18:1-20:4 16:0-20:4 18:0-20:4

2.27 1.72 1.70

Alk-l-enylacylGPE

0.40 0.37 0.33

215

dicate that arachidonoyl species with an unsaturated acyl group at the sn - 1 position are more metabolically active than species with an sn - 1 saturated acyl group. Unlike data obtained with rat testes [12], the 16 : l-20: 4 + 18 : 2-20 : 4 species of human endothelial cells had higher specific radioactivities than the 20 : 4-20 : 4 species in each diacyl subclass. Specific radioactivities of each molecular species were higher in the diacylglycerophosphoserine/diacylglycerophosphoinositol fraction than in the corresponding molecular species of either diacylglycerophosphocholine or diacylglycerophosphoethanolamine, which indicates the arachidonoyl-containing molecular species of the diacylglycerophosphoserine/ diacylglycerophosphoinositol fraction have the highest turnover rate of phospholipids in human endothelial cells. In contrast, the arachidonoylcontaining species in the alk-l-enylacylglycerophosphoethanolamine fraction appears to have a lower metabolic turnover rate in human endothelial cells, since it exhibited the lowest specific radioactivities of the four phospholipid subclasses analyzed. The authors express their appreciation to Edgar A. Cress for his excellent technical assistance during certain parts of this work. The work in Oak Ridge was supported by the Office of Energy Research, US Department of Energy (Contract No. DE-AC05760R00033), the National Heart, Lung, and Blood Institute (Grant HL-27109-05), and the American Cancer Society (Grant BC-70P). The work in Iowa City was supported by an Arteriosclerosis SCOR Grant, HL-14230-14, from NHLBI, NIH.

References 1 Marcus, A.J., Weksler, B.B. and Jaffe, E.A. (1978) J. Biol. Chem. 253, 7138-7141 2 Hopkins, N.K., Oglesby, T.D., Bundy, G.L. and Gorman, R.R. (1984) J. Biol. Chem. 259, 14048-14053 A.R., Revtyak, G. and Campbell, W.B. (1985) 3 Johnson, Fed. Proc. 44, 19-24 4 Moncada, S. (1982) Arteriosclerosis 2, 193-207 5 Spector, A.A., Hoak, J.C., Fry, G.L., Denning, G.M., Stoll, L.L. and Smith, J.B. (1980) J. Clin. Invest. 65, 1003-1012 6 Kaduce, T.L., Spector, A.A. and Bar, R.S. (1982) Arteriosclerosis 2, 380-389 7 Rosenthal, M.D. and Whitehurst, M.C. (1983) B&him. Biophys. Acta 750, 490-496 8 Spector, A.A., Kaduce, T.L., Hoak, J.C. and Fry, G.L. (1981) J. Clin. Invest. 68, 1003-1011 9 Spector, A.A., Kaduce, T.L., Hoak, J.C. and Czervionke, R.L. (1983) Arteriosclerosis 3, 323-331 10 Fleisher, L.N., Tall, A.R., Witte, L.D., Miller, R.W. and Cannon, P. J. (1982) J. Biol. Chem. 257,6653-6655 11 Spector, A.A., Scanu, A.M., Kaduce, T.L., Figard, P.H., Fless, G.M. and Czervionke, R.L. (1985) J. Lipid Res. 26, 288-297 12 Blank, M.L., Cress, E.A., Robinson, M. and Snyder, F. (1985) B&him. Biophys. Acta 833, 366-371 13 Jaffe, E.A., Nachman, R.L., Becker, CC. and Minick, C.R. (1973) J. Clin Invest. 52, 2745-2756 14 Folch, J.M., Lees, M. and Sloane-Stanley, G.H. (1957) J. Biol. Chem. 255,497-509. 15 Lowry, O.H., Rosebrough, N.J., Farr, A.L., and Randall, R.J. (1951) J. Biol. Chem. 193, 265-275 16 Snyder, F. and Smith, D. (1966) Sep. Sci. 1, 709-722 17 Mavis, R.D., Bell, R.M. and Vagelos, P.R. (1972) J. Biol. Chem. 247, 2835-2841 18 Blank, M.L., Robinson, M., Fitzgerald, V. and Snyder, F. (1984) J. Chromatogr. 298, 473-482 19 Albert, D.H. and Snyder, F. (1983) J. Biol. Chem. 258, 97-102 20 Chilton, F.H., Ellis, J.M., Olson, S.C. and Wykle, R.L. (1984) J. Biol. Chem. 259, 12014-12019 21 Albert, D.H. and Snyder, F. (1984) Biochim. 796, 92-101

Biophys.

Acta