Contributing factors in the trafficking of [3H]arachidonate between phospholipids

Contributing factors in the trafficking of [3H]arachidonate between phospholipids

Riochimicu et Biophysics Ac,la, 1124 (19YZ)X2-272 0 1992 Elsevier Science Publishers B.V. All rights reserved OOOS-2760/92/$OS.O(1 BBALIP 53X.55 Con...

1MB Sizes 0 Downloads 185 Views

Riochimicu et Biophysics Ac,la, 1124 (19YZ)X2-272 0 1992 Elsevier Science Publishers B.V. All rights reserved OOOS-2760/92/$OS.O(1

BBALIP 53X.55

Contributing factors in the trafficking of [ 3H]arachidonate between phospho~ipids Merle L. Blank, Zigrida L. Smith and Fred Snyder Oak Ridge Associated Unil,ersiries,Medical Scicwce.~Dirisiort, Oak Ridge, TN (USA) (Received 6 August 1991)

Key words: Arachidonic acid; Arachidonate; Arachidonate movement; HL-60 cell; Phospholipid, plasmalogen and ether-linked

Cultured human promyelocytic leukemia cells (HL-601, depleted of arachidonic acid by continued growth in serum-free media, were used as a model system to examine various factors that control the incorporation and distribution of r3Hlarachidonic acid into classes and subclasses of cellular lipids. Increasing the culture media concentration of [3HIarachidonic acid from 1 - 10 -8 M to 1 * 10 -s M caused a greater percentage of the cellular tritium to be distributed into triacylglycerols (from < 1% at 1 - 10 -* M to 38% at 1 * 10 -5 M) with a corresponding decrease in cellular [3H1diradylglycerophosphoethanolarnine (from 53% at 1 - 10 -s M to 12% at 1 . 10 -5 M) during 2 h incubations. A greater proportion of the tritium present in diradylglycerophosphoethanolamine and diradyiglycerophosphocholine, at the higher media concentration of [3Hlarachidonic acid (1 * 10 -’ Ml, was found in the diacyi subclasses of these two lipids than was observed at the lower concentrations ( < 1 - 10 -6 M) of i3Hlarachidonic acid. Signiticant amounts of diarachidonoyl molecular species were found in the phosphatidylethanolamine (10%) and phosphatidylcholine (15%) of HL-60 cells that were labeled for 2 h with 1 * LO-5 M L3Hlarachidonic acid. This was the only molecular species of phosphatidylcholine to completely disappear when prelabeled cells were placed in arachidonate-free media for 22 h. Prelabeling-chase experiments with 1 * 10 -’ M r3Hlarachidonic acid were consistent with movement of [3Hlarachidonate from t~acylglycero~s into diradylglycerophosphatides and from diacyIphospholipids into ether-linked phospholipids. increasing the concentration of HL-60 cells in the incubations influenced the distribution of L3Hlarachidonic acid in cellular lipid classes in a manner analogous to decreasing the concentration of 13Hlarachidonic acid in the media. Increasing the endogenous level of cellular arachidonate in phospholipid classes with supplements of unlabeled arachidonic acid changed the subsequent lipid class distribution of a low concentration (1 * 10 -’ M) of [3Hlarachidonic acid to resemble results obtained with a much higher mass level of [3Hlarachidonate in arachidonate depleted cells. HL-60 cells differentiated into granulocytes by treatment with dimethyl sulfoxide incorporated less t3Hlarachidonic acid but had a greater proportion associated with.alkylacylglycerophosphocholine and alk-l-enylacylglycerophosphoethanolamine than undifferentiated HL-60 cells. Introduction Cellular arachidonic acid is not only important as a precursor of bioactive eicosanoid metabolites but also occupies a central position in the metabolism and synthesis of platelet-activating factor (l-alkyl-2-acetyl~~-glycero-3-phosphocholine; al~lacety~-GPC; PAF) [l-61. Radioactive arachidonic acid is commonly used to study the cellular metabolism of this important eicosanoid, but results are sometimes inconsistent and difficult to interpret. For example, significant differences were observed between the lipid class distribution of radioactivity and the distribution of arachi-

Correspondence: F. Snyder, Oak Ridge Associated Universities, Medical Sciences Division, P.O. Box 117, Oak Ridge, TN, U.S.A.

donate mass in human neutrophils prelabeled for 5 min with [~~larachidonic acid and then stimulated with ionophore A23187 [71. Differences between the mass distribution of arachidonic acid and the distribution of radiolabeled arachidonic acid in glycerolipids has also been pointed out in other cell systems, such as platelets [8-l l] and vascular smooth muscle cells 1121. Therefore, although the use of radiolabeled arachidonic acid can be a vaiuabie too1 for the study of its metabolism by intact cells, care must be taken to insure that its use does not confound the resulting conclusions. Many laboratories have described the effects of time and cell stimulation on the incorporation, movement and release of radiolabeled arachidonic acid in various celi systems. Therefore, the main objective of our work was to examine the role of other factors that

263

influence the distribution and movement of radiolabeled arachidonic acid among classes and subclasses of glycerolipids in human leukemia cells (HL-60). Advantages of the HL-60 cell system are: (1) the cells can be grown in a serum-free, chemically defined media [13], thus eliminating serum-derived variables; (2) HL-60 cells that are passaged in the serum-free growth media can still be differentiated into granulocytes by treatment with dimethylsulfoxide [13]; and (3) arachidonic acid can induce arachidonate-depleted (via serum-free growth media) differentiated HL-60 cells (granulocytes) to produce greater amounts of PAF upon ionophore A23187 stimulation [6], thereby providing an excellent model to study the interrelationships of arachidonate and PAF metabolism. To our knowledge, this is the first detailed investigation of factors that influence arachidonic acid metabolism in an arachidonate-free, human cell model. Materials

and Methods

Materials [5,6,8,9,11,12,14,15-“HlArachidonic acid (76 Ci/ mmol) and [9,10-“Hloleic acid (7.4 Ci/mmol) were purchased from DuPont-NEN. Unlabeled arachidonic acid, oleic acid and triolein were from NU-CHEKPREP. RPM1 1640 culture media, L-glutamine and penicillin-streptomycin were supplied by GIBCO. The media supplements of fatty acid-free bovine serum albumin (BSA) and the insulin-transferrin-sodium selenite mixture were from Sigma Chemical, who also supplied the dimethylsulfoxide (DMSO), phospholipid standards and materials for the determination of nitroblue tetrazolium reduction by differentiated cells. Cell cultures

Human leukemic cells (HL-60), from the American Type Culture Collection, were grown in RPM1 1640 media supplemented with insulin (10 pg/ml), transferrin (10 pg/ml), sodium selenite (10 ng/ml>, BSA (O.l%), penicillin (100 units/ml), streptomycin (100 pg/ml) and r_-glutamine (4 mM) as previously described [6,13]. The cells were passaged at least 30 times in serum-free media before being used in the experiments. Differentiation of HL-60 cells into granulocytes with dimethylsulfoxide and the degree of differentiation as measured by nitroblue tetrazolium reduction have been described [6]. Unless otherwise specified in the tables or figures, the specific radioactivities of [“Hlarachidonic and [3H]oleic acids (1 PCi) were adjusted by addition of appropriate amounts of unlabeled fatty acid in 15 ~1 of ethanol; the latter aliquot was added to 5 ml of serum-free media containing 1 . lo6 HL-60 cells per ml. Cells were incubated for the indicated times at 37°C; in a humidified air atmosphere containing 5% CO,. At the end of the incubation, cells

were pelleted by centrifugation (1200 X g for 5 min), washed once with 5 ml of serum-free media and the cellular lipids were either extracted or the cells were resuspended in fresh media for subsequent incubation. Lipid analysis

Cellular lipids were extracted by the method of Bligh and Dyer [14], except 2% glacial acetic acid was included in the methanol. Neutral lipids were separated by thin-layer chromatography (TLC) on plates coated with Silica Gel G using a solvent system of hexane/diethyl ether/glacial acetic acid (85 : 15 : 1, v/v>. Phospholipids were separated by TLC on Silica Gel H layers developed in chloroform/methanol/ glacial acetic acid/water (50 : 25 : 8 : 2, v/v>. Subclasses and in some instances molecular species, of the cholineand ethanolamine-containing glycerophosphatides were analyzed as benzoate derivatives of their diradylglycerols as previously described [15]. To determine whether HL-60 cells had extensively converted the [“Hlarachidonic acid into acyl groups other than arachidonate, the distribution of tritium in various acyl chains was assessed by first reducing selected total lipid extracts with Vitride to convert acyl groups to their corresponding fatty alcohols [16]. Total products from the Vitride reduction were passed through a short silicic acid column and, after evaporation of solvent, the samples were benzoylated [17]. Alcohol benzoates were isolated from Silica Gel G coated TLC plates developed in benzene/hexane (50 : 45, v/v) where the alcohol benzoates migrate just below (R, = 0.41) the cholesterol benzoate derivative (RF = 0.50). The alcohol benzoates were then separated by reverse-phase high performance liquid chromatography (HPLC) 1153 using isocratic elution with 15% isopropanol (v/v) in acetonitrile (1 ml/min). Radioactivity was measured from aliquots of the effluent (0.2 to 0.4 ml> collected in a fraction collector. The only significant amount (> 2%) of radioactivity other than [“HI20 : 4 that eluted from the HPLC analysis of the alcohol benzoates from cells labeled with [“Hlarachidonic acid (2 h with 1 . 10m5 M) was associated with a peak having the same retention volume as 22 : 4(n - 6) (19.0 _t 0.3% of the total tritium). Because benzoate derivatives of 14 : 0, 16 : 1, 18 : 2, and 20 : 3tn - 6) alcohols also elute at or very near the same retention volume as 22 : 4, this peak was collected, hydrogenated [171 and analyzed again by HPLC with an eluting solvent of 30% isopropanol (v/v> in acetonitrile at 1 ml/min. Greater than 85% of the tritium in the hydrogenated [“H]22: 4 peak eluted with the benzoate derivative of 22 : 0. Therefore, even though there was some chain elongation of [“Hlarachidonic acid to adrenic acid by HL-60 cells, the major amount of cellular tritium (80%) remained associated with 20:4 acyl groups.

264 For mass measurements of triacylglycerols, TLC plates developed in the neutral lipid solvent system were sprayed with concentrated sulfuric acid and charred by heating in an oven at 185°C for 1 h. The amount of carbon in the charred triacylglycerol spots was quantitated by photodensitometry [ 181 using a LKB Ultrascan XL laser densitometer. Different amounts of triolein on the same TLC plates with the samples were used to generate a standard response curve for quantitation of the triacylglycerols.

6oI

Results Effect of /-‘H/fatty acid concentration on the distribution of tritium in cellular lipids The percentage of [ 3H]arachidonic acid incorporated by HL-60 cells during a 2 h incubation decreased as the concentration of arachidonic acid was increased; in contrast, the percent cellular uptake of [“Hloleic acid did not change with increasing concentrations of oleic acid (Table I). The mole fraction of oleic acid that was taken up by the cells was lower than that of arachidonic acid at all concentrations examined. Nearly all of the incorporated radioactivity from both fatty acids was found in various glycerolipid classes. In fact, the highest percentage of tritium found in free fatty acids was only 1.3 + 0.2% of the total cellular radioactivity from cells supplemented with 10 PM [“Hlarachidonic acid. The masses of [‘Hlarachidonic acid and [3H]oleic acid that were incorporated by the HL-60 cells are also shown in Table I and demonstrate that increasing the media concentrations of either fatty acid resulted in higher absolute amounts of the fatty acid being incorporated into cellular lipids. If needed, percentages given in subsequent tables and figures can be easily converted into mass units by utilizing the mass data in Table I. Diradyl-GPE was the major radiolabeled lipid (> 50% of the cellular tritium) in HL-60 cells supplemented for 2 h with the lower concentrations (1 . lo-’ M and 1 . 10~’ M) of [‘Hlarachidonic acid and triacylglycerols were barely labeled (< 1% of the cellular TABLE

10 I; l

0

~~~

105

,_-. Tz!BcL 10'

~~~__~ 105

106

Molarity of rH]Arachidonic

Acid

Fig. 1. Effect of [‘H]arachidonic acid concentration on the distrihution of tritium in lipid classes of undifferentiated HL-60 cells (5 ml of 1.10” cells/ml; 2 h at 37°C). Data points are meanskS.E. from two separate experiments, each with duplicate samples tn = 4). Symbols represent: 0, diradyl-GPC; 0, diradyl-GPE: 0, triacylglycerols; and X, diradyl-GPS/GPI.

tritium) at these lower concentrations of [3H]arachidonic acid (Fig. 1). However, at higher concentrations of [“Hlarachidonic acid (1 . 10ph M and 1 . lo-’ M) the lipid class distribution of cellular tritium was significantly different after 2 h of incubation than the distribution pattern found at lower arachidonic acid concentrations for this time period. This difference was especially noticeable with the 1 . lo-’ M concentration of [‘Hlarachidonic acid, where there was a dramatic decrease in the fraction of cellular radioactivity associated with diradyl-GPE and a near commensurate increase in the percent of cellular tritium associated with triacylglycerols. The mass of triacylglycerols was also higher in cells that were supplemented for 2 h with 1 . IO-” M arachidonic acid (1.96 &-0.06 wg/lOh cells) than in cells incubated for 2 h with 1 . lo-’ M arachidonic acid (1.56 + 0.10 pg/lOh cells; P < 0.02). This increase of triacylglycerol mass (0.40 pg/lOh cells) in

I

Eff&r of’firtty acid concentration on the cellular uptake

of [.‘H]uruchidonic and (H]oleic

acids by undijjerentiated HL-60 cells

Fatty acids were incubated with 5 ml of 1.10” HL-60 cells/ml for 2 h at 37°C; cells were washed once in fresh media, and the cellular lipids then extracted as described in Materials and Methods. Values are means k S.E. from two separate experiments, each with duplicate samples (n = 4). I’ Significantly Fatty acid concentration

1, IO-’ M 1 IO-” M 1~lO~‘M 1~10~” M 2.7, ltll’ M

higher (P < 0.05) than the value obtained Percent

at the 1 lo-’

51.3*3.2 57.2 + 0.6 62.6 i 2.7 “ 63.4 + 1.Y ‘I _

of arachidonic

acid, based on Student’s

nmol uptake’lO_’

uptake

[‘Hlarachidonic acid

M concentration

[‘Hloleic acid

[‘Hlarachidonic acid

[ ‘Hloleic

30.5 k 0.6 30.8 + 1.4 30.5 * 1.0 _

2560 *IhO 286 t 3 31.3 5 1.4 3.17& 0.10

1520 +30 154 +7 15.2 + 0.5 _

27.4 f 3.3

acid

3.70f

0.45

t-test.

265 60

cells supplemented with the higher arachidonic acid concentration corresponds reasonably well with the amount that can be calculated (0.58 pg/lOh cells) from the cellular uptake (Table I) and percentage (Fig. 1) of [‘Hlarachidonic acid in triacylglycerols. Not as pronounced was an increase in the percent of radioactivity in diradyl-GPC and a decrease in tritium found in the diradylglycerophosphoinositol/serine fraction when at the lower compared to the percentages [“Hlarachidonic acid concentrations. It was not our intention to carry out a detailed kinetic investigation of [“Hlarachidonic acid movement with respect to time, but because of the large differences between cells labeled with high versus low concentrations of [“Hlarachidonic acid in their relative content of cellular [ ‘Hldiradyl-GPC, [ “Hldiradyl-GPE and [“Hltriacylglycerols (Fig. l), we did analyze lipids from cells incubated for shorter times (5, 10, 15 and 30 min.) with M) of the lowest concentration (1 . lo-’ [“Hlarachidonic acid. Based on duplicate samples, the ratio of radioactivity in diradyl-GPC : diradyl-GPE was greater than unity in lipids from cells incubated for only 5 (ratio = 1.52) and 10 min (ratio = 1.35). This ratio continued to decrease after both 15 min (0.92) and 30 min (0.89). Moreover, there was no evidence of significant levels of tritium (< 1%) in the triacylglycerols at any of these times. To summarize, with incubation times > 30 min the cellular diradyl-GPE fraction accumulated the highest proportion of tritium when low media concentrations of [ ‘Hlarachidonic acid were used. However, at the highest concentration of [“Hlarachidonic acid the greatest levels of tritium were found in cellular triacylglycerols and diradyl-GPC. In contrast to the extensive labeling of diradyl-GPE observed after 2 h with low concentrations of [“Hlarachidonic acid, the preponderance of tritium (50% to 58% of the cellular radioactivity) was found in diradyl-GPC of HL-60 cells at all concentrations of

TABLE

_-I

1---------i-,

; 50

10

0 10.0

10.'

IO.5

104

Molarity of [%]Oleic Acid Fig. 2. Effect of [‘Hloleic acid concentration on the distribution of tritium in lipid classes of undifferentiated HL-60 cells. Symbols and conditions are the same as Fig. I.

[3H]oleic acid that were tested (Fig. 2). Although a greater amount of radioactivity was found in triacylglycerols at the highest concentration of [“Hloleic acid (1 . lo-” M) used, the increase in the labeling of triacylglycerols between the low and high concentrations of [“Hloleic acid (6% to 20% of the cellular tritium) was less than the increase seen for [“Hlarachidonic acid (1% to 38% of the cellular tritium). Furthermore, the cellular diradyl-GPE fraction did not exhibit the same drastic decrease in [“Hloleic acid labeling at the higher concentration of oleic acid supplement as observed with the [“Hlarachidonic acid additions. Effect of [-‘HIfatty acid concentration on the distribution of tritium in subclasses of diradyl-GPC and diradyl-GPE

The diacyl subclass of the diradyl-GPC fraction was labeled to the greatest extent in HL-60 cells supple-

II

Effect of [.‘H]arachidonic cells

acid concentration on the distribution of tritium in subclasses of diradyl-GPC and diradyl-GPE in undifferentiated HL-60

[3HlArachidonic acid was incubated with 5 ml of 1. lo6 then extracted, and the tritium distribution determined means+S.E. from two separate experiments, each with concentrations of arachidonic acid, based on Student’s Subclass

HL-60 cells/ml for 2 h at 37°C; cells were washed once in fresh media, the cellular lipids in subclasses of diradyl-GPC/-GPE as described in Materials and Methods. Values are duplicate samples (n = 4). Significantly different than the values at either of the lowest t-test; a P < 0.001; ‘P < 0.01; ’ P < 0.02; ’ P < 0.05.

% of total 3H in diradyl

class

l.lO-‘M

l.lO-’

M

l.lOWh M

1.10-5

Diacyl Alkylacyl-GPC Alk-1-enylacyl-GPC

54.3 + 1.4 40.3 f 1.7 5.4 + 0.4

56.3 + 1.5 38.9 f 1.8 4.8 k 0.3

67.3& 1.1 ’ 28.6* 1.3 ’ 4.0 + 0.2 d

83.7 + 0.7 a 14.3 +0.7 ‘I 2.0+0.1 a

Diacyl-GPE Alkylacyl-GPE Alk-1-enylacyl-GPE

36.9* 1.8 11.2+0.9 51.9k2.4

37.1* 1.5 12.5 + 0.2 50.4kl.7

36.9 + 0.5 15.5 * 0.7 c 47.6 + 0.3

47.7+ 1.0 h 17.8k 1.2 h 34.5 * 2.0 h

M

266 mcnted with all levels of [‘Hlarachidonic acid (Table II). Except at the highest concentration (1 . lo-’ M), a relatively high percentage of radioactivity from [jH]arachidonic acid was also found in the alkylacylGPC subclass (29-40%X As the media concentration of [jH]arachidonic acid was increased to 1 . 10mh M and 1 . 1OF” M, a greater percentage of the tritium associated with the cellular diradyl-GPC fraction was found in the diacyl subclass with a corresponding decrease in the alkylacyl subclass. Only small amounts of radiolabel were associated with the alk-I-enylacyl-GPC subclass at all media concentrations of [ “Hlarachidonic acid. However, the alk-1-enylacyl subclass contained most of the total radioactivity present in the diradylGPE fraction from HL-60 cells incubated with from 1 . 10~’ M to 1 . lo-’ M [“Hlarachidonic acid. At the highest concentration of [ ‘Hlarachidonic acid tested (1 10p5 M) there was a decrease of radioactivity in the alk-I-enylacyl subclass (from 50% to 34%) relative to both the diacyl and alkylacyl subclasses of cellular diradyl-GPE. In general, as the media concentration of [3H]arachidonic acid was increased to 1 . IO-’ M there was a tendency for greater proportions of the tritium in the diradyl-GPE/GPC classes to locate in the diacyl subclasses at the expense of etheracyl subclasses. The diacyl subclasses of both diradyl-GPC and diradyl-GPE contained the highest percentages, 85-88% and 59-61%, respectively, of tritium after labeling HL60 cells with [ZH]oleic acid (Table III>. However, significant amounts of tritium were also associated with alkylacyl-GPC (IO-13%), alkylacyl-GPE (17-19%) and alk-1-enylacyl-GPE (20-23%) in HL-60 cells supplemented with [3H]oleic acid. In contrast to the results obtained with [‘Hlarachidonic acid, increasing the media concentration of [3H]oleic acid from 2.7. lo-” M to 1 . lo-’ M had virtually no effect on the cellular distribution of radioactivity in the subclasses of either diradyl-GPC or diradyl-GPE.

TABLE

70

6o

I

lO%A2h + 22h

Molarity of [3H]Arachidonic Acid, Incubation Time Fig. 3. Redistribution of tritium in lipid classes of undifferentiated HL-60 cells after a 22 h chase of cells that were prelabeled for 2 h with 2 concentrations of 13H]arachidonic acid as described in Fig. I, Data are the means+S.E. from two separate experiments, each with duplicate samples (n = 4). Hatched bars represent diradyl-GPE; open bars represent diradyl-GPC; shaded bars represent diradylGPS/GPI; and solid bars represent triacylglycerols. Data for the 2 h prelabeliny are from Fig. I

Effect of [-‘HI arachidonic acid remorlal on the suhsequent redistribution of tritium in cellular lipids In these experiments HL-60 cells were prelabeled for 2 h with either 1 . lo-’ M or 1 lO_’ M [“Hlarachidonic acid and then transferred to fresh, arachidonic acid-free media for 22 h. There was some decrease in the amount of cellular [3H]arachidonate after the initial 2 h incubation with 1 . lO_’ M [‘Hlarachidonic acid (51.3 f 3.2% of the incubated tritium, Table I> and the amount found after 22 h in fresh, arachidonic acid-free media (42.0 f 2.8%; P > 0.05). In contrast, the amount of cellular radioactivity remained unchanged (61.6 + 1.9% of the incubated tritium) when HL-60 cells were incubated with 1 lo-’

111

&fjtct of [-‘H]olric acid cotzcentrution on the distribution of tritium in sublcasses of diradyl-GPC and diradyl-GPE in undiffrrwtiatrd HL-60 cell.~ [‘H]Oleic acid was incubated with 5 ml of 1. 10h cells/ml for 2 h at 37°C; cells were washed once extracted, and the tritium distribution determined in subclasses of diradyl-GPC/-GPE as described means f S.E. from two separate experiments, each with duplicate samples (n = 4). Subclass

o/c of total ‘H in diradyl 2.7.10-”

M

in fresh media, the cellular in Materials and Methods.

class l.lO-‘M

1.10-h

M

l.lO-’

M

Diacyl-GPC Alkylacyl-GPC Alk-I-enylacyl-GPC

x5.4+ 1.1 13.2+ I.1 1.4iO.l

85.3+ 1.3 13.2* 1.2 1.5iO.l

86.1 f 1.1 12.6k 1.1 1.4*0.1

88.3 * 1.2 10.7+ 1.1 1.0*0.1

Diacyl-GPE Alkylacyl-GPE Alk- I -envlacvl-GPE

60.8 * 1.2 17.3 + 2.4 22.0 * 1.3

60.2 f 2.5 17.3 + 2.7 22.6* 1.0

59.1+ 3.0 17.6k3.1 23.3 + 0.8

60.8 * 3.3 19.1k2.5 20.1* 1.0

lipids then Values are

267 TABLE

IV

Effect of incubating [“Hlarachidonic acid prelabeled HL-60 cells in arachidonic acid-free media for 22 h on the distribution of tritium in subclasses of diradyl-GPC and diradyl-GPE in undifferentiated HL-60 cells [?‘H]Arachidonic acid (1 ‘lo-s and 1~10-’ M) was incubated 2 h with 5 ml of 1.10’ cells/ml as described in Table 11. Prelabeled cells were then washed and transferred to arachidonic acid-free media for 22 h before extraction of cellular lipids. Values are meansi S.E. from two separate experiments, each with duplicate samples (n = 4). Numbers in parentheses are average values at 2 h taken from Table II for easier comparisons and are significantly different from the corresponding values found after the 22 h chase, based on Student’s t-test; a P < 0.001; ‘P < 0.01. Subclass

[‘Hlarachidonic acid concentration f% of total ‘H in diradyl class) 1.10-s

M

l.lO-‘M

Diacyl-GPC Alkylacyl-GPC Alk-1-enylacyl-GPC

47.5 f3.0 (54.3) 42.0 + 1.7 (40.3) 10.6+ 1.3 (5.4) h

47.7+ 3.3 (83.7) ’ 41.6f2.0 (14.3) a 10.7+ 1.3 (2.0) a

Diacyl-GPE Alkylacyl-GPE Alk-1-enylacy-GPE

42.4 k 2.3 (36.9) 5.6kl.l (11.2) h 52.1+ 1.3 (51.9)

39.2 k 1.1 (47.7) h 7.8* 1.5 (17.8) ’ 53.2+ 0.5 (34.5) a

M [‘Hlarachidonic acid and then placed in unlabeled, fresh media for 22 h. A major redistribution of radioactivity occurred in lipid classes (Fig. 3) of HL-60 cells that had been prelabeled with 1 . lo-’ M [3Hlarachidonic acid and then incubated for 22 h in arachidonate-free media. The main changes in these prelabeled cells (1 . 10d5 M [3H]arachidonic acid) after the 22 h chase in arachidonic acid-free media was a decrease from 38.0 f 2.1% to 9.2 + 2.6% in radiolabeled triacylglycerols and an increase from 11.7 + 0.6% to 46.8 + 1.8% in tritium labeled diradyl-GPE. The largest changes in the lipid class distribution of cellular tritium in cells prelabeled with 1 . 10-s M [3H]arachidonic acid was a decrease from 20.4 + 0.9% to 10.6 f 0.7% in labeled phosphatidyl-GPI/GPS and an increase from 53.4 5 0.9% to 64.2 + 0.8% in t3H]diradyl-GPE after the 22 h chase. After 22 h in arachidonic acid-free media the distribution of radioactivity within subclasses of diradyl-GPC was identical regardless of whether HL-60 cells were prelabeled with levels of either 1. lop8 M or 1 . lo-’ M [3H]arachidonic acid (Table IV). This was also true for the distribution of tritium in the subclasses of diradyl-GPE. However, there were much greater changes in [ 3H]arachidonate distribution in subclasses of diradyl-GPC and diradyl-GPE between the end of the 2 h prelabeling and the end of the 22 h chase when the HL-60 cells had been prelabeled with 1. low5 M [3H]arachidonic acid as compared to prelabeling with the lower (1 . 10-s Ml concentration. In fact, the only changes that were statistically different in subclass distribution of tritium between the 2 h prelabeling and

the 22 h chase periods, when 1 . lo-' M [“Hlarachidonic acid was used to prelabel the HL-60 cells, were the small increase in percent of the [“Hlalk1-enylacyl-GPC subclass and the small decrease of radiolabel in the alkylacyl-GPE subclass. In contrast, the pattern of tritium distribution was significantly altered in every subclass of diradyl-GPC and diradylGPE during the 22 h chase of HL-60 cells that were prelabeled with 1 . 10e5 M [“Hlarachidonic acid. The largest changes in subclass distribution of [“Hlarachidonate, as a result of the 22 h chase with HL-60 cells prelabeled with 1 . lo-” M [3H]arachidonic acid, were the relative loss of radiolabel from diacyl-GPC (from 84% to 48%) and the increases of tritium in alkylacylGPC (from 14% to 42%) and alk-1-enyl-GPE (from 34% to 53%). In terms of the redistribution of the mass of arachidonic acid (calculated from its specific radioactivity) the greatest gains during the 22 h chase of cells prelabeled with 1 . lo-” M [“Hlarachidonic acid were 5-fold in alk-1-enylarachidonoyl-GPE (from 1.04 nmol to 5.23 nmol), 2.7-fold in diacyl-GPE (from 1.43 nmol to 3.86 nmol) and 2-fold for alkylarachidonoyl-GPC (from 1.12 nmol to 2.26 nmol). As mentioned earlier, these gains in arachidonoyl species were compensated for mainly by transfer of arachidonate from triacylglycerols and diacyl-GPC. Results of the pulse-chase experiments using the higher concentration of [3H]arachidonic acid are reminiscent of other reports on the time-dependent movement of radiolabeled arachidonic acid from diacyl-GPC into ether-linked glycerophosphatides in other resting cell types 119-271. This movement of [“Hlarachidonic acid in HL-60 cells during the 22 h chase was, however, far from obvious when the prelabeling was done with a lower concentration of [“Hlarachidonic acid since very likely the lower concentrations of [“Hlarachidonic acid reached a near equilibrium condition during the 2 h prelabeling step. Effect of preloading HL-60 cells with unlabeled fatty acids on the subsequent distribution of 13H]arachidonic acid

In these experiments HL-60 cells (5 ml, 10’ cells/ml) were first incubated for 2 h with 1 . 10m5 M of either unlabeled arachidonic acid or oleic acid, then washed and transferred to fresh, fatty acid-free media. After 30 min in the fresh media, 1 . lo-’ M [3H]arachidonic acid was added and the cellular lipids were extracted after a 2 h incubation. The distribution of [“Hlaraehidonate in the three major lipid classes of the cells preloaded with arachidonic acid (Fig. 4) more nearly resembled the tritium distribution found when HL-60 cells with no fatty acid preloading were incubated with 1 * lo-* M [3Hlarachidonic acid than it did the distribution pattern observed when 1. lop8 M [3H]arachidonic acid was used with arachidonate-

268 depleted cells with no preloading (Fig. 1). These results clearly indicate that endogenous levels of cellular arachidonate can markedly affect the distribution of radiolabeled arachidonic acid in lipid classes. On the other hand, preloading HL-60 cells with oleic acid had little effect on the subsequent distribution of tritium in cellular lipid classes obtained with the lower level (1 lo-’ M) of [“Hlarachidonic acid (Fig. 4). Effect of cell number on the incorporation and distribution of /-‘Hlarachidonate in cellular lipids Not only do the concentrations of radiolabeled arachidonic acid used in experiments vary from one literature report to another but the concentration of cells also differs. Therefore, experiments were designed to determine if cell concentration also affects the incorporation and distribution of [‘Hlarachidonic acid during a 2 h incubation. At 1 . lO_” M [‘Hlarachidonate, increasing the number of cells from 1 . lob/ml to 2. 10h/ml had little effect on the percentage incorporation of [‘Hlarachidonic acid (48.0 i 0.8 vs. 52.0 f 1.4%, respectively). However, as shown by the results in Fig. 5, cell density does influence the distribution of tritium (1 lo-’ M [‘Hlarachidonate) in lipid classes of HL-60 cells. At the higher cell density (2 . 10h cells/ml) there was a decrease of [‘Hlarachidonate in the triacylglycerols and a corresponding increase in tritium was divided between diradyl-GPC and diradyl-GPE. Even though the ratios of cell density to [3H]arachidonic acid were different by a factor of ten at the two lowest concentrations of [‘Hlarachidonic acid (1 . lo-’ M and 1 . lo-’ M, Fig. 1 and Table II) there was no signifi-

60

(lxlO”M,

(1

x

1”

“MJ

(1

x

1”

“M,

(1

x

I”

-WI,

Cell Condition and Subsequent [3H]20:4 Concentration in ( ) Fig. 4. Effect of preloading arachidonate-deficient cells (5 ml of I 1Oh undifferentiated HL-60 cells/ml) with either unlabeled arachidonic acid or oleic acid (1 10-s M) for 2 h. The cells are then transferred to arachidonic acid-free media for 30 min and the distribution of 1. lo-’ M [‘HI arachidonic acid in lipid classes then determined after a 2 h incubation with HL-60 cells (5 ml of 1’ lOh cells/ml). Data are the meansfS.E. from two separate experiments, each with duplicate samples (n = 4). Hatched bars represent triacylglycerols; open bars represent diradyl-GPE; and shaded bars represent diradyl-GPC. Comparative data for cells not preloaded with arachidonic acid are included from Fig. 1.

TAG

GPE

WC

GPS,GPl

Lipid Class Fig. 5. Effect of concentration (5 ml total volume) of undifferentiated HL-60 cells on the distribution of cellular tritium after incubation for 2 h with 1. IO -’ M [‘Hlarachidonic acid. Data are the meanskS.E. from two separate experiments, each with duplicate samples (12 = 4). TAG represents triacylglycerols.

cant difference in the distribution of tritium in lipid classes or subclass of these HL-60 cells. Therefore, it is unlikely that increasing or decreasing the cell density by a factor of even two or three at the two lower [ “Hlarachidonic acid concentrations would significantly affect the distribution of tritium in cellular lipid classes and subclasses. However, as mentioned earlier, cell density did affect the distribution of tritium in cellular lipids at higher concentrations (1 . lo-’ M) of [ ‘Hlarachidonic acid. Effect of /-7H]arachidonic acid concentration on the synthesis of [“H]diarachidonoyl species of phospholipids by HL-60 cells Significant amounts of glycerophosphatides containing polyunsaturated fatty acids at both the sn-1 and sn-2 positions of glycerol occur in retinal lipids [28, 291, where the major polyunsaturated acid is docosahexaenoic. Dilinolenoyl and didocosahexaenoyl species of phosphatidylcholine were found in cultured rat hepatocytes after incubation with 1 mM concentrations of linolenic and docosahexaenoic acids, respectively [30]. Also, diarachidonoyl species of phospholipids have been observed in rat testes [31], rat erythrocytes [321 and human neutrophils [331. Diacyl-GPC isolated from HL-60 cells (5 ml, 10h cells/ml) incubated for 2 h with 1 . lo-’ M [‘Hlarachidonic acid contained no tritium labeled diarachidonoyl molecular species, whereas at the highest concentration (1 . lo-’ M) of [‘Hlarachidonic acid, 15.6 f 0.7% of the radioactive diacylGPC eluted from HPLC with the retention time of diarachidonoyl-GPC. Based on a pooled sample, this molecular species was also present (10.1%) in phosphatidylethanolamine from cells incubated with 1 10-” M [‘Hlarachidonic acid. Further support for the identity of this peak in phosphatidylcholine was obtained by collecting the peak from HPLC, hydrogenating the

269 compound and then again subjecting the hydrogenated product to reverse-phase HPLC analysis [31,32], where more than 76% of the tritium eluted with the saturated diarachidoyl species. Consistent with the rapid turnover of this molecular species observed in other studies [31,32], [3H]diarachidonoyl-GPC was the only molecular species of diacyl-GPC that disappeared ( < 0.5% of the [ “Hldiacyl-GPC remained as diarachidonoyl species) when HL-60 cells were prelabeled for 2 h with 1 . 10d5 M i3Hlarachidonic acid and then incubated for another 22 h in arachidonic acid-free media. Effect of differentiation on the cellular distribution of / 3H]arachidonic acid HL-60 cells that were differentiated with dimethyl-

sulfoxide (48.3 + 1.7% of the cells reduced nitroblue tetrazolium) incorporated a similar amount of 1 . 10e8 M r3H]arachidonic acid (55.3 + 3.2%; P > 0.05) as the undifferentiated cells (63.4 + 1.9%; Table I). However, differentiated HL-60 cells incubated for 2 h with a higher concentration of [ 3H]arachidonic acid (1 . lop5 M) incorporated a significantly lower amount of the radiolabel (20.8 + 0.6%, P < 0.001) than was incorporated by undifferentiated cells (51.3 + 3.2%; Table I). Compared to undifferentiated cells, the percentage of cellular tritium associated with the diradyl-GPS/-GPI fraction was decreased from 20.4% to 13.0 + 0.8% at 1 . 10-s M and from 7.4% to 4.2 k 0.4% at 1 . 10d5 M [ ‘Hlarachidonic acid concentrations (P < 0.01). The distribution of tritium among other cellular lipid classes was not statistically different from the pattern for undifferentiated cells incubated for 2 h with the same

TABLE

V

Disrribulion of [“Hjarachidonic acid in subclasses of diradyl-GPC and diradyl-GPE of DMSO-differentiated HL-60 cells HL-60 cells were differentiated with DMSO as described in Materials and Methods. [3H]Arachidonic acid (1.10m5 M or l.lO-’ M) was incubated with 5 ml of 1. 10h ceils/ml for 2 h at 37°C; cells were washed once in fresh media, the cellular lipids extracted, and the tritium distribution determined in subclasses of diradyl-GPC/GPE as described in Materials and Methods. Values are means+ S.E. from two separate experiments each with duplicate samples (n = 4). Numbers in parenthesis represent differences between values in this table and values for undifferentiated cells in Table II. Significant differences between differentiated and undifferentiated cells are based on Student’s t-test; a P < 0.001; ’ P < 0.01; ’ P < 0.02. Subclass

% of total ‘H in diradyl

class

l,lO-‘M

1.1O-5

Diacyl-GPC Alkylacyl-GPC Alk-1-enylacyl-GPC

46.9 f 0.7 ( - 7.4) a 47.0 f 0.6 ( + 6.7) a 6.1+ 0.2 (+ 0.7)

74.8 f 0.9 ( - 8.9) a 21.3 + 0.6 (+ 7.0) ’ 3.7f0.4 (+ 1.7) c

Diacyl-GPE Alkylacyl-GPE Alk-1-enylacyl-GPE

39.5 + 4.7 ( + 2.6) 9.2 + 0.4 (- 2.0) 51.3+4.6(-0.6)

46.8+ 1.1 (-0.9) 10.1 kO.5 (-7.7) a 43.1+1.1 (+8.6)h

M

two concentrations of arachidonic acid (data not shown). However, there were several significant differences between differentiated and undifferentiated cells in the relative distribution of tritium from [“Hlarachidonic acid in subclasses of diradyl-GPC and diradyl-GPE (Table V). The alkylacyl subclass of diradyl-GPC in differentiated cells contained a greater proportion of tritium with both high and low concentrations of [‘Hlarachidonic acid than was found in undifferentiated HL-60 cells. An increase of the percent tritium in the alk-1-enylacyl-GPE subclass and a corresponding decrease in the alkylacyl-GPE subclass of differentiated (Table V> compared to undifferentiated (Table II) cells also occurred after 2 h incubation with 1 . lop5 M [‘Hlarachidonic acid. Discussion

An important variable that controls the distribution of [3H]arachidonic acid among lipid classes of arachidonate-depleted, undifferentiated HL-60 cells is the concentration of arachidonic acid present in the media. At higher concentrations (1 . lo-” M) of [ “Hlarachidonic acid, compared to the lower concentrations (1.10-s M and 1 * lo-’ M), a greater proportion of the cellular tritium was channelled into triacylglycerols mainly at the expense of diradyl-GPE. Labeling of cellular triacylglycerols by radioactive arachidonic acid is not unique to HL-60 cells. For example, an increase in the relative amount of radiolabeled arachidonic acid incorporated into cellular triacylglycerols was also found to be associated with higher media concentrations of arachidonic acid in experiments with human neutrophils [33] and bovine endothelial cells [34]. In contrast to our findings, triacylglycerols were labeled to a major extent after incubation of human neutrophils for 5 min with even a low concentration (3. lo-’ M) of [“Hlarachidonic acid [33], probably due to a high level of endogenous arachidonate in the neutrophils. The increase in labeling of triacylglycerols that occurred when we incubated HL-60 cells with higher amounts (1 * 10eh M and 1. lo-’ M) of [“Hlarachidonic acid may be a general phenomena because cultures of rat hepatocytes incubated with several other fatty acids, at much higher concentrations (2. lop4 M and greater), also showed increases in [‘4C]palmitic acid [35] and [“HIglycerol [30,36] labeling of triacylglycerols. None of the other studies described the corresponding major decrease in the percentage of radiolabel in the diradylGPE fraction that we observed when HL-60 cells were incubated with the higher concentrations of L3H]arachidonic acid. In fact, our experiments also differed in showing that low concentrations ( < 1 . lo-” M) of 13Hlarachidonic acid are preferentially incorporated into cellular diradyl-GPE after a 2 h incubation with arachidonate-depleted HL-60 cells, since diradyl-

270 GPC is normally labeled to a much greater extent than diradyl-GPE during both short < 30 min and long > 2 h incubations of radioactive arachidonic acid with most other cell systems [2,3, 7- 10,12,20-27,30,32,36-451. When HL-60 cells were preloaded with unlabeled arachidonic acid and then incubated with 1 . lo-” M [“Hlarachidonic acid (Fig. 4) the distribution of tritium between diradyl-GPC and diradyl-GPE more nearly resembled the earlier work of others (e.g., [3H]diradylGPC > [“Hldiradyl-GPE). This is a clear indication that depletion of arachidonate from the HL-60 cells by continued growth in a serum-free media, rather than the cell type itself, is the factor responsible for the different distribution pattern of [ ‘Hlarachidonate observed at the lower concentrations of this fatty acid (Fig. 1). At higher media concentrations of [‘Hlarachidonic acid it appears that cell density must be considered when comparing results between different experiments. Expressing the results on the basis of per mg cellular protein or per cell does not solve this problem, but in fact magnifies it (e.g., it can be calculated from the data that at the higher cell density there are 0.536 nmol of [“Hlarachidonate per 10” cells in triacylglycerols and 1.531 nmol per lo6 cells at the lower cell density). Unlike the results obtained with arachidonate preloaded cells, preincubation of HL-60 cells with unlabeled oleic acid had virtually no effect on the subsequent incorporation and distribution of [ ‘Hlarachidonic acid (Fig. 4). Also, the cellular uptake and lipid distribution pattern of [“Hloleic acid, a readily available endogenous fatty acid in arachidonate-depleted HL-60 cells [6], was influenced less than [“Hlarachidonic acid by the media concentrations of oleic acid used (Fig. 2). These observations likely reflect the separate, different pathways utilized in metabolism of the two fatty acids. Free fatty acids must be converted to acyl-CoAs before being esterified to glycerol by acyltransferases during de novo synthesis of diradylglycerols and/or to phospholipids by an acyl-CoA: lysoglycerophospholipid acyltransferase reaction as first described by Lands [46]. Unless HL-60 cells have transacylases that are active only at high concentrations of arachidonic acid in remodeling both the sn-1 and sn-2 positions of intact diacyl-GPC and -GPE, the presence of the [ “Hldiarachidonoyl-GPC and -GPE species argues for their formation via an arachidonoyl-CoA acyltransferase during the de novo synthesis of diacylglycerols as suggested by Chilton and Murphy in experiments with human neutrophils [33] and by Akesson et al. [361 with hepatocytes. In addition, diarachidonoyl species in our experiments could hardly have been produced by remodeling of the sn-1 position of endogenous l-acyl-2arachidonoyl-GPC or -GPE because there is no unlabeled arachidonate to be utilized by transacylases in the depleted cells [6]. Therefore, it appears likely that

both arachidonate-labeled triacylglycerols and diarachidonoyl molecular species of diacyl-GPC and -GPE are produced primarily via the de novo synthesis of diarachidonoylglycero-3-phosphate by acyltransferases in the arachidonate-depleted HL-60 cells. The decrease of tritium labeled triacylglycerols together with a commensurate increase in labeled phospholipids during incubation of prelabeled cells in [‘Hlarachidonate fatty acid-free media (Fig. 3) likely represents a transfer of [‘Hlarachidonate from triacylglycerols to phospholipids during the chase period. Transfer of radiolabeled arachidonic acid from triacylglycerols to phospholipids has also been observed in other types of cells [22,34,37,47-491. The labeling of triacylglycerols at only high concentrations of [ ‘Hlarachidonic acid is consistent with triacylglycerols being a storage site for amounts of arachidonate that are in excess of physiological needs. However, the mechanism responsible for the transfer of arachidonate from triacylglycerols to phospholipids is not known at present. Results from double labeling of triacylglycerols of cultured neuroblastoma cells with [6‘HIglucose and [1-‘“Cllinoleic acid indicate both the ‘H in the glycerol and the 14C in the acyl groups of the triacylglycerols are transferred to phospholipids equally [50] which suggests the transfer of linoleate from triacylglycerols to phospholipids occurs via a lipase catalyzed reaction to produce linoleate-containing 1,2-diacylglycerols for utilization by the choline- or ethanolamine-phosphotransferases in the de novo pathway of phospholipid biosynthesis. Therefore, it is reasonable to expect arachidonate to be transferred from triacylglycerols to phospholipids in the same manner as linoleate. Cessation of cell division occurs when HL-60 cells are differentiated to granulocytes by treatment with DMSO [51]. This lack of cellular proliferation and the decreased need for synthesis of accompanying phospholipids for membrane structure is likely responsible for the lower incorporation of [‘Hlarachidonic acid by the DMSO-differentiated cells when compared to undifferentiated cells in our experiments. The observation that greater proportions of the cellular [ ‘Hlarachidonate was distributed into the alkylacylGPC and alk-l-enylacyl-GPE subclasses of DMSO-differentiated HL-60 cells than in the undifferentiated cells (Table V) suggests these ether-linked phospholipids have a cellular function other than merely structural components of cell membranes. One of these functions in DMSO-differentiated HL-60 cells is to provide the precursor lipid (alkylarachidonoyl-GPC) for synthesis of PAF via the remodeling pathway [61. In differentiated HL-60 cells one of the first steps in the synthesis of PAF (i.e., production of alkyllyso-GPC) appears to be initiated by transfer of the arachidonoyl moiety from alkylarachidonoyl-GPC to alk-1 -enyllyso-

271 GPE via a CoA-independent transacylase [52]. Therefore, production of PAF by differentiated HL-60 cells may require both alkylarachidonoyl-GPC and alk-l-enylarachidonoyl-GPE as participants in this transacylation reaction. Our experiments with arachidonate-depleted HL-60 cells indicate at least four factors in addition to time must be considered in the interpretation of results based on the radiolabeling of cellular lipids with arachidonic acid. These factors are: (1) the concentration of free arachidonic acid; (2) the cell concentration or density (3) the endogenous levels of arachidonic acid in the cellular phospholipids; and (4) the extent of cellular differentiation. The first two variables as well as time, can be easily controlled in suspension cultures of cells but an accurate cell number is more difficult to achieve with monolayer cultures. However, the endogenous levels of arachidonate in the various phospholipid molecules would be the most difficult to regulate, especially with primary cultures of cells taken from animals over which there is little dietary control or cells requiring serum for growth. In fact, it was recently pointed out that variations in the synthesis and release of eicosanoids by cultured porcine aortic endothelial cells appear to be associated with decreases in endogenous levels of arachidonic acid in their cellular phospholipids caused by an increased number of passages in culture [53]. Such conditions can be controlled in the arachidonate-depleted HL-60 cell system we described and emphasizes the usefulness of this system as a model for investigating the metabolism of arachidonic acid and related phospholipids. Acknowledgements

This work was supported by the Office of Energy Research, U.S. Department of Energy (Contract No. DE-AC05-760R000331, The American Cancer Society (Grant BC-,70V) and the National Heart, Lung and Blood Institute (Grant 35495-04Al-02).

8 Rittenhouse-Simmons, 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33

References Chilton, F.H., Ellis, J.M., Olson, S.C. and Wykle, R.L. (1984) J. Biol. Chem. 259, 12014-12019. Albert, D.H. and Snyder, F. (1984) Biochim. Biophys. Acta 796, 92-101. Purdon, A.D. and Smith, J.B. (1985) J. Biol. Chem. 260, 1270012704. Nakagawa, Y., Kurihara, K., Sugiura, T. and Waku, K. (1986) Biochim. Biophys. Acta 876, 601-610. Ramesha, C.S. and Pickett, W.C. (1986) J. Biol. Chem. 261, 7592-7595. Suga, K., Kawasaki, T., Blank, M.L. and Snyder, F. (1990) J. Biol. Chem. 265, 12363-12371. Chilton, F.H. and Connell, T.R. (1988) J. Biol. Chem. 263, 5260-5265.

34 35 36

37 38 39 40 41

S., Russell, F.A. and Deykin, D. (1977) Biochim. Biophys. Acta 488, 370-380. Comic, M., Breton, M. and Colard, 0. (19861 Pharmacol. Res. Commun. 18, 43-49. Kambayashi, J., Kawasaki, T., Tsujinaka, T., Sakon, M., Oshiro, T. and Mori, T. (1987) Biochem. Int. 14, 241-247. Takamura, H., Narita, H., Park, H.J., Tanaka, K-i., Matsuura, T. and Kito, M. (1987) J. Biol. Chem. 262, 2262-2269. Ford, D.A. and Gross, R.W. (1989) Proc. Natl. Acad. Sci. USA 86, 3479-3483. Breitman, T.R., Collins, S.J. and Keene, B.R. (19801 Exp. Cell Res. 126, 494-498. Bligh, E.G. and Dyer, W.J. (19591 Can. J. Biochem. Physiol. 37, 911-917. Blank, M.L., Robinson, M., Fitzgerald, V. and Snyder, F. (1984) J. Chromatogr. 298, 473-482. Snyder, F., Blank, M.L. and Wykle, R.L. (1971) J. Biol. Chem. 246, 3639-3645. Blank, M.L., Cress, E.A., Lee, T-c., Stephens, N., Piantadosi, C. and Snyder, F. (1983) Anal. Biochem. 133, 430-436. Blank, M.L., Schmit, J.A. and Privett, OS. (1964) J. Am. Oil Chem. Sot. 41, 371-376. Etienne, J., Rybkine, T., Bereziat, G. and Polonovski, J. (1987) Thromb. Res. 47, 573-583. Wey, H.E., Jakubowski, J.A. and Deykin, D. (1986) Biochim. Biophys. Acta 878, 380-386. Takayama, H., Kroll, M.H., Gimbrone, M.A. and Schafer, A.I. (1989) Biochem. J. 258, 427-434. MacDonald, J.I.S. and Sprecher, H. (1989) Biochim. Biophys. Acta 1004, 151-157. Chilton, F.H. and Murphy, R.C. (1986) J. Biol. Chem. 261, 7771-7777. Colard, O., Breton, M. and Bereziat, G. (1984) Biochem. J. 222, 657-662. Yoshioka, S., Nakashima, S., Okano, Y. and Nozawa, Y. (1986) J. Lipid Res. 27, 939-944. Sugiura, T., Katayama, O., Fukui, J., Nakagawa, Y. and Waku, K. (1984) FEBS Lett. 165, 273-276. Emilsson, A. and Sundler, R. (1985) Biochim. Biophys. Acta 816, 265-274. Wiegand, R.D. and Anderson, R.E. (1983) Exp. Eye Res. 37, 159-173. Aveldano, M.I. and Bazan, N.G. (1983) J. Lipid Res. 24, 620-627. Sundler, R., Akesson, B. and Nilsson, A. (1974) J. Biol. Chem. 249, 5102-5107. Blank, M.L., Cress, E.A., Robinson, M. and Snyder, F. (1985) Biochim. Biophys. Acta 833, 366-371. Robinson, M., Blank, M.L. and Snyder, F. (1986) Arch. Biochem. Biophys. 250, 271-279. Chilton, F.H. and Murphy, R.C. (1987) Biochem. Biophys. Res. Commun. 145, 1126-1133. Denning, G.M., Figard, P.H., Kaduce, T.L. and Spector, A.A. (1983) J. Lipid Res. 24, 993-1001. Ontko, J.A. (1972) J. Biol. Chem. 247, 1788-1800. Akesson, B., Sundler, R. and Nilsson, A. (1976) in Lipids, Vol. 1 (Paoletti, R., Porcellati, G. and Jacini, G., eds.), pp. 49-58, Raven Press, New York. Spinedi, A., Pacini, L., Piacentini, M., Melino, G. and Luly, P. (1990) J. Neurochem. 54, 778-782. Colard, O.,Breton, M. and Bereziat, G. (1986) Biochem. J. 233, 691-695. Brown, M.L., Jakubowski, J.A., Leventis, L.L. and Deykin, D. (1987) Biochim. Biophys. Acta 921, 159-166. Rittenhouse-Simmons, S., Russell, F.A. and Deykin, D. (1976) Biochem. Biophys. Res. Commun. 70, 295-301. Trotter, J., Flesch, I., Schmidt, B. and Ferber, E. (1982) J. Biol. Chem. 257, 1816-1823.

272 Y., Kurihara, K., Sugiura, T. and Waku, K. (1985) 42 Nakagawa, Eur. J. Biochem. 153, 263-268. 43 Kroner, E.E., Peskar, B.A., Fischer, H. and Ferber, E. (1981) J. Biol. Chem. 256, 3690-3697. 44 Kramer, R.M. and Deykin, D. (1983) J. Biol. Chem. 258, 1380613811. 45 Billah, M.M., Eckel, S., Myers, R.F. and Siegel, M.I. (1986) J. Biol. Chem. 261, 5824-5831. 46 Lands, W.E.M. (1960) J. Biol. Chem. 235, 2233-2237. 47 Cook, H.W., Clarke, J.T.R. and Spence, M.W. (1982) J. Lipid Res. 23, 1292-1300. 48 Sivarajan, M., Hall, E.R., Wu, K.K., Rafelson, M.E. and Manner. C. (1984) Biochim. Biophys. Acta 795, 271-276.

49 Bachelet, M., Masliah, J., Vargaftig, B.B., Bereziat, G., and Colard, 0. (1986) Biochim. Biophys. Acta 878, 177-183. 50 Cook, H.W. and Spence, M.W. (1985) Can. J. Biochem. Cell Biol. 63, 919-926. 51 Collins, S.J., Ruscetti, F.W., Gallagher, R.E. and Gallo, R.C. (1978) Proc. Nat]. Acad. Sci. USA75, 2458-2462. 52 Uemura, Y., Lee, T-c. and Snyder, F. (1991) J. Biol. Chem. 266, 8268-8272. 53 Brown, M.L. and Deykin, D. (1991) Arterioscler. Thromb. 1 I, 167-173.