ARCHIVES OF BIOCHEMISTRY AND BIOPHYSICS Vol. 269, No. 2, March, pp. 603-611,1989
Effects of Eicosapentaenoic and Docosahexaenoic Acid Supplements on Phospholipid Composition and Plasmalogen Biosynthesis in P388D, Ceils MERLE L. BLANK,
ZIGRIDA
L. SMITH,
Y. JOSEPH LEE, AND FRED SNYDER’
Medical and Health Sciences Division, Oak Ridge Associated Oak Ridge, Tennessee 37831
Universities,
Received July 12,1988, and in revised form November
P.O. Box 117,
15,1988
This investigation describes the influence of n-3 fatty acid supplements on the phospholipid composition and the metabolism of plasmalogens in P388Dr cells. The cellular content of phospholipid classes and subclasses was unchanged in P388Dr cells (a macrophage-like cell) grown for 24 h in media supplemented with 10 yM sodium eicosapentaenoate or sodium docosahexaenoate. However, phospholipids from these cells were highly enriched in acyl groups of the corresponding fatty acid supplement, with the largest increases occurring in the ethanolamine plasmalogens (e.g., 46% of the ethanolamine plasmalogens from cells supplemented with docosahexaenoate contained this acyl group at the an-2 position). Eicosapentaenoate supplements lowered the levels of oleate in phosphatidylinositol/serine, diacyl-sn-glycero-3-phosphoethanolamine (GroPEtn), and alk-1-enylacyl-GroPEtn in the P388Dr cells but had little or no effect on the amounts of arachidonate in the cellular phospholipids. In contrast, supplementation of the cells with docosahexaenoic acid not only reduced the level of oleate but also decreased the amount of arachidonate by one-third in the alk-1-enylacyl-GroPEtn. When P388Dr cells were incubated for 1 h with [3H]alkyllyso-GroPEtn both [3H]alkylacyl-GroPEtn and [3H]alk-l-enylacyl-GroPEtn were formed. The sn-2 acyl composition of these two ethercontaining GroPEtn lipids reflected the fatty acid supplement that the cells had received (e.g., 68% of the [3H]alk-l-enylacyl-GroPEtn from cells supplemented with docosahexaenoate contained this acyl group at the m-2 position). Cells from both the controls and supplemented groups contained greater amounts of docosahexaenoate in the [3H]alk-1-enylacyl-GroPEtn (plasmalogen) than in the [3H]alkylacyl-GroPEtn subclass. Analysis of molecular species from pulse-chase experiments with intact cells and examination of the molecular species of [3H]alk-l-enylacyl-GroPEtn produced by the A’-desaturase system in cell-free membrane fractions suggest that the docosahexaenoate-containing species of [3H]alk-l-enylacyl-GroPEtn have a higher turnover rate than other molecular species. Possible biological implications of our findings are also discussed. 0 1989 Academic
Press. Inc
Plasmalogens are a significant component of the ethanolamine-containing phospholipids in many mammalian tissues (1). The 1-alk-1’-enyl-2-acyl-sn-glycero-3phosphoethanolamine (alk - 1 - enylacyl-
1 To whom correspondence
GroPEtn)2 subclass appears to act as a storage site for arachidonic acid in the testes of rats fed an essential fatty acid defi’ Abbreviations used: Ptd, phosphatidyl; Ser, serine; Ins, inositol; GroPCho, sn-glycero-3-phosphocholine; GroPEtn, sn-glycero-3-phosphoethanolamine; BSA, bovine serum albumin; TLC, thin-layer chroma-
should be addressed.
603
0003-9861/89 $3.00 Copyright All rights
0 1989 by Academic Press, Inc. of reproduction in any form reserved.
604
BLANK
cient diet (23). It has also been suggested that in some cells ethanolamine plasmalogens may be involved in the release of arachidonic acid (4-11) for subsequent synthesis of eicosanoids. However, a definitive functional role(s) for plasmalogens has yet to be firmly established. One interesting observation is that the ethanolamine plasmalogens contain relatively high amounts of polyunsaturated acyl groups, particularly arachidonic acid (1). We have recently shown that there is a preference for polyunsaturated fatty acids in the transacylation of 1-[3H]alkyl-2lyso - sn- glycero - 3 - phosphoethanolamine ([3H]alkyllyso-GroPEtn) by a membrane preparation from Madin-Darby canine kidney cells (12). The 1-alkyl-2-acyl-snglycero-3-phosphoethanolamine (alkylacyl-GroPEtn) desaturase (EC 1.14.99.19) that produces alk - 1 - enylacyl - GroPEtn demonstrated a slight selectivity for the alkyl - acyl - GroPEtn molecular species that contained docosahexaenoate at the sn-2 position. This molecular species of alk-1-enylacyl-GroPEtn also showed the highest turnover rate in pulse-chase experiments with intact cells (12). Others have also reported a rapid incorporation and/or turnover of the docosahexaenoate species in the ether-containing subclasses of the ethanolamine phospholipids of rat brain (13-15) and Ehrlich ascites cells (16-18). Because of the dramatic influence of dietary supplements of marine oil fatty acids on lipid components, we have examined the effect of supplementing cultured P388D1 cells with two prominent n-3 fatty acids present in marine fish oils (eicosapentaenoic and docosahexaenoic acids) on the content of phospholipid classes and subclasses as well as their acyl composition. We also determined what effect the supplementation of cells with these two n-3 fatty acids had on the acyl specificity of enzymes involved in the acylation of vH]alkyllyso-GroPEtn tography; GLC, gas-liquid chromatography; Hepes, 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid; PAF, platelet activating factor; HPLC, high-performance liquid chromatography.
ET AL.
and the subsequent desaturation of rH]alkylacyl - GroPEtn to [3H]alk - 1 - enylacylGroPEtn. The P388D1 cell line used in our experiments has been characterized as a macrophage-like cell (19,20) and, therefore, the results described in this report form a fundamental basis for subsequent studies of the effects of fatty acid modifications on selected macrophage functions. Although the P388D1 cells have been used in many biological experiments since the cell line was established in 1956 (21), as far as we were able to ascertain, only limited data are available on their lipid composition (22,23). Our detailed analyses of these cells provide some needed, basic information about the phospholipid composition of an important macrophage-like cell model that can be cultured in quantities large enough for biochemical studies. MATERIALS
AND
METHODS
Cell cultures. P388Di cells (from the American Type Culture Collection, Rockville, MD) were grown in RPM1 1640 culture media containing 9% fetal calf serum, 258 mg of L-glutamine/liter, 89 units of penicillin/ml, and 89 fig of streptomycin/ml (all from Grand Island Biological Co., Grand Island, NY). Cultures were maintained at 37°C in an atmosphere of 5% CO, in air. For supplementation of cells with eicosapentaenoic (98% purity, Biomol Research Laboratories Inc., Philadelphia, PA) and docosahexaenoic acids (>95% purity, Nu Chek Prep Inc., Elysian, MN), O.l-ml solutions of filtration-sterilized sodium salts of the fatty acids (24) were added (final concentration = 10 pM) to 20 ml of culture media containing 5 X lo7 P388Di cells. After a 24-h incubation, the supplemented cells and unsupplemented controls were collected by centrifugation (15Og for 6 min) and either extracted (25) immediately for the lipid compositional studies or washed once with 0.1% bovine serum albumin (BSA, essentially fatty acid free; Sigma Chemical Co., St. Louis, MO) in saline and resuspended in 4.75 ml of RPM1 1640 medium without serum at a concentration of 0.8 to 1.0 X lo7 cells/ml. With the washed cells, 1.8 to 2.2 &i of [3H]alkyllyso-GroPEtn dissolved in 0.25 ml of 0.2% BSA was added to each 4.75 ml of suspended cells. After incubation for 1 h at 37OC the cells were pelleted by centrifugation and cellular lipids were extracted by the method of Bligh and Dyer (25). In two experiments, unsupplemented cells that had been prelabeled for 1 h with [3H]alkyllyso-GroPEtn were transferred to unlabeled, unsupplemented serum-containing media and incubated for an additional
EFFECT
OF n-3 FATTY
ACIDS
ON PHOSPHOLIPIDS
23 h before extraction of the cellular lipids. Cell viability was determined by exclusion of trypan blue dye. Enzymatic assay with cell membranes. To examine the selectivity of the transacylase and A’-desaturase enzymes in a cell-free system, lo9 cells in 10 ml of 0.25 sucrose and 0.1 M phosphate buffer (pH 7.1) were broken using 600 psi of nitrogen pressure and two passages through a Mini-Bomb cell disruption chamber (Kontes Glass Co., Vineland, NJ). This resulted in >98% breakage of the cells as judged by light microscopy. The cell homogenates were centrifuged at 1lOOg for 10 min and the resulting supernatants were centrifuged at 100,OOOgfor 60 min. These membrane pellets were resuspended in 0.25 M sucrose and after centrifugation again at 100,OOOg for 60 min the final membrane preparations were suspended in 2 ml of water for use as the enzyme source (average protein = 11.0 + 0.3 mg, n = 6). All operations in preparing the membranes were conducted at 0 to 4°C and only freshly prepared membrane fractions were used for the enzyme assays. Enzymatic reactions were conducted essentially as previously described (12) except that 0.4 nmol of [aH]alkyllyso-GroPEtn was used per tube and the incubations were for 1 h. Protein was determined by the method of Bradford (26) using the Bio-Rad assay kit with BSA as a standard. of [‘H]alkyllyso-GroPEtn. [“H]AlkylPreparation lyso-GroPEtn was prepared by slight modifications of a previously described method (27). Briefly, [9,103H]hexadecanol (prepared by Vitride reduction of [9,10-“Hlpalmitic acid, 28.5 mCi/pmol, from DuPont NEN, Wilmington, DE) was added (0.5 mCi, 17.5 nmol) in 10 ~1 of ethanol to each of six tubes containing 2 X lo7 Ehrlich ascites cells suspended in 2 ml of Dulbecco’s modified Eagle medium containing 20 mM Hepes buffer (pH ‘7.3), 50 units of penicillin/ml, and 50 pg of streptomycin/ml (all from Grand Island Biological Co., Grand Island, NY). After incubation with gentle agitation for 2 h at 37°C the cells were pelleted by centrifugation (5009 for 10 min) and the cellular lipids were extracted by the method of Bligh and Dyer (25). The radiolabeled diradyl-GroPEtn fraction was isolated from the Ehrlich ascites cell lipids by preparative thin-layer chromatography (TLC) on layers of silica gel H developed in a solvent system of chloroform:methanol:glacial acetic acid:water (50:25:8:2, v/ v; solvent system I). This diradyl-GroPEtn fraction was first exposed to HCl gas to cleave vinyl ether linkages (plasmalogens) and then treated with a monomethylamine reagent (28) to hydrolyze acyl groups, [“HlAlkyllyso-GroPEtn was finally isolated by TLC using a developing solvent of chloroform:methanol: glacial acetic acid:water (50:25:8:4, v/v). Based on zonal scans of TLC plates (29), the radiopurity of the [3H]alkyllyso-GroPEtn was >96% Analysis of benzoate derivatives by high-performance liquid chromatography (HPLC) (12) showed that >91% of the tri-
OF P388Di CELLS
605
tium in the alkyl chain was associated with the hexadecyl species and that the specific radioactivity as [‘H]alkyllyso-GroPEtn was 1.5 &i/nmol. Lipid analyses. After separation of cellular phospholipids by TLC (solvent system I) the distribution of mass or radioactivity in each class was determined either by assay for phosphorus content (30) or by zonal scraping for tritium-labeled lipids (29), respectively. For determination of subclasses, diradylGroPCho and/or diradyl-GroPEtn were isolated from the cell lipids by preparative TLC (solvent system I) and subjected to hydrolysis with phospholipase C followed by benzoylation of the resulting diradyl-Gro (12, 31). Diradyl-Gro benzoate derivatives were separated by TLC (31) and the areas of silica gel containing the individual subclasses were either directly methylated (2% HaSO, in methanol for 45 min at 100°C) in the presence of an internal standard (pentadecanoic acid) for gas-liquid chromatography (GLC) analysis of unlabeled lipids or extracted from the gel for HPLC analysis (31) of tritium-labeled subclasses. When applicable, statistically significant differences were calculated using Student’s t test; all data are expressed as means f SE. RESULTS
Based on three separate experiments, we found no significant difference (P > 0.2 vs controls) in the average number of cells per flask at the end of the 24-h supplementation period with the n-3 acids. Cell viability was >95% in all three groups. Therefore, supplementation of the P388Di cells at the 10 PM level with the sodium salts of eicosapentaenoic and docosahexaenoic acids did not exhibit any cytotoxicity. Eflect of fatty acid supplementation on the phospholipid composition of P3880, cells. The data in Table I indicate that supplementation of P388Di cells with eicosapentaenoic and docosahexaenoic acids did not influence the cellular levels of phospholipid classes. Although present in small amounts, the content of alkylacylGroPEtn was too low (estimated at about 5% of the total diradyl-GroPEtn fraction) for accurate measurement and, therefore, was not included in Table I. It is clear from data in Table II that P388D1 cells grown in the presence of eicosapentaenoic and docosahexaenoic acids readily incorporated these acids into acyl groups of cellular phospholipids. Although
606
BLANK
ET AL.
TABLE
I
PHOSPHOLIPID COMPOSITION OF P388D1 CELLS GROWN IN SUPPLEMENTED (20~5 and 22:6) AND UNSUPPLEMENTED MEDIA Supplement Lipid class
None
+20:5”
+22:6”
(nmol per 10s cells) Sphingomyelin Diacyl-GroPCho Alkylacyl-GroPCho Diacyl-GroPEtn Alk-l-enylacyl-GroPEtn Ptd-Ins/Ser
15.7 f 514.0 f 108.7 * 147.8 + 89.2 f 129.0 f
3.9 21.4 11.3 6.2 4.3 5.5
13.9 f 503.0 + 119.0 * 150.1+ 102.2 2 131.1 f
2.6 8.4 3.5 5.1 9.8 4.5
18.5 f 519.0 f 99.2 2 152.0 + 104.3 f 132.2 f
3.2 7.5 7.5 5.7 10.5 2.9
a Cells were supplemented with 10 FM sodium eicosapentaenoate (20:5) or docosahexaenoate (22:6) for 24 h, extracted, and analyzed for phospholipid composition as described under Materials and Methods. Values represent the average nanomoles per lo8 cells f SE from two separate experiments each done in duplicate (n = 4).
TABLE
II
CONTENT OF SELECTED ACYL GROUPS IN PHOSPHOLIPIDS OF P388Di CELLS GROWN IN SUPPLEMENTED (20:5 AND 22:6) AND UNSUPPLEMENTED MEDIA Lipid class
Supplement
20:5
226
20:4
181
(percent in lipid class) Diacyl-GroPCho
None 20:5 22:6
0.1 * 0.0 0.6 + 0.0 0.3 * 0.0
0.1 f 0.0 0.3 + 0.0 1.4 +- 0.1
0.3 f 0.0 0.7 * 0.1 0.6 f 0.0
45.1 f 0.9 45.0 f 0.3 43.5 f 0.7
Alkylacyl-GroPCho
None 20:5 22:6
N.D. 3.4 ?I 0.4 0.2 f 0.1
N.D. 0.3 zk 0.1 4.0 f 0.2
1.7 f 0.0 2.8 * 0.2 2.6 + 0.4
44.9 * 1.2 43.1 + 0.3 39.7 + 0.4
Diacyl-GroPEtn
None 20~5 226
0.9 f 0.1 7.2 f 0.3 1.0 f 0.1
2.5 f 0.1 3.1 f 0.1 11.3 + 0.4
7.2 f 0.2 7.6 f 0.2 6.5 + 0.1
47.4 + 0.2 39.1 * 0.3 38.6 k 0.6
Alk-1-enylacyl-GroPEtn
None 20:5 22:6
3.5 f 0.2 24.9 + 1.2 3.4 + 0.1
13.9 +- 0.3 12.9 f 1.3 46.1 f 2.3
27.9 f 0.6 24.5 k 1.6 18.5 + 1.2
32.5 -+ 1.0 21.5 k 1.5 18.6 221.7
Ptd-Ins/Ser
None 20:5 226
0.3 f 0.0 3.4 f 0.2 0.4 f 0.1
2.8 f 0.2 2.5 +- 0.2 10.3 f 0.5
7.0 f 0.1 8.2 f 0.1 8.1 f 0.1
40.7 f 0.8 33.0 f 0.5 29.5 f 0.2
Note. Phospholipids were isolated from the same samples described in Table I and the acyl compositions determined as described under Materials and Methods. The aeyl composition of the alkylacyl and alk-l-enylacyl subclasses is for the fatty acids at the sn-2 position which equal lOO%, while the acyl composition of the diacyl subclasses represents fatty acids from both the sn-1 and sn-2 positions. All values represent average wt% + SE (n = 4). N.D. signifies not detected.
EFFECT
OF n-3 FATTY
ACIDS
ON PHOSPHOLIPIDS
all cellular phospholipids showed some increase in eicosapentaenoate and docosahexaenoate acyl groups after addition of these two fatty acids to the culture media, the largest increases in these two acyl groups were found in the alk-l-enylacylGroPEtn. Only small increases of acyl groups corresponding to the two fatty acid supplements were observed in the diacylGroPCho subclass. Supplementation of the cells with eicosapentaenoic acid appeared to cause only minor changes in the amount of eicosatetraenoic (n-6) acid in the cellular phospholipids; however, cells supplemented with docosahexaenoic acid contained about one-third less eicosatetraenoate in the alk-1-enylacyl-GroPEtn than was found in unsupplemented cells (Table II; P < 0.001). Growing cells in the presence of eicosapentaenoic acid did not alter the amount of docosahexaenoate (and vice versa) in the cellular phospholipids. Both fatty acid supplements were incorporated into the cellular diacylGroPEtn, alk-1-enylacyl-GroPEtn, and Ptd-Ins/Ser fractions at the expense of oleic acid (P < 0.01). Eflect of fatty acid supplementation on the acylation of [‘H]alkyllyso-GroPEtn and subsequent desaturation of [3H]alkylacylGroPEtn by intact cells. Based on data from three separate experiments, supplemented and unsupplemented P388D1 cells incorporated an average of 35.2 + 3.3% of added [3H]alkyllyso-GroPEtn into the [3H]diradyl-GroPEtn fraction after a l-h incubation. In the controls as well as the supplemented groups, greater than 94% of the tritium in the cellular [3H]diradylGroPEtn fraction was associated with the two ether-containing (alkylacyl and alk-lenylacyl) subclasses, with the major amount of radiolabel (>60%) being present in the [3H]alk-l-enylacyl-GroPEtn subclass. When unsupplemented cells were prelabeled for 1 h with [3H]alkyllysoGroPEtn and then transferred to unlabeled, unsupplemented media for 23 h, tritium in the [3H]alk-l-enylacyl-GroPEtn subclass increased from 65.5 -t 1.5 to 80.7 t 4.7% of the diradyl-GroPEtn lipids and there was a corresponding decrease of tri-
OF P388Di CELLS
607
FIG. 1. Distribution of tritium in selected molecular species of [3H]hexadecylacyl-GroPEtn (shaded bars) and [3H]hexadec-l-enylacyl-GroPEtn (hatched bars) after a l-h incubation of P388Di cells with [3H]hexadecyllysoGroPEtn. Cells in A were unsupplemented controls, those in B were supplemented with eicosapentaenoic acid, and those in C were supplemented with docosahexaenoic acid (see Materials and Methods for details). Data represent the means + SE from three separate experiments.
tium in the [3H]alkylacyl-GroPEtn. However, there was no significant increase in the absolute amount of tritium associated with the [3H]alk-l-enylacyl-GroPEtn subclass after the 23-h chase (0.089 -t 0.006 &i/lo7 cells after 1 h and 0.095 f 0.016 &i/lo7 cells after the 23-h chase). Unsupplemented cells demonstrated an apparent selectivity for different fatty acids in the acylation of [3H]alkyllysoGroPEtn with the order of acyl preference being 20:4 > 22:6 > 18:l > 20:5 (Fig. 1). Cells cultured in media containing eicosapentaenoic or docosahexaenoic acids showed increased acylation of [3H]alkyllysoGroPEtn with the respective fatty acid supplement. In fact, more than 50% of the [3H]alkylacyl-GroPEtn synthesized by cells enriched in docosahexaenoate contained this fatty acid at the sn-2 position (Fig. 1). Very little difference was observed in the amount of radiolabel associated with the eicosapentaenoate species of the [3H]alk-l-enylacyl-GroPEtn compared to the amount of the same molecular species in the [3H]alkylacyl-GroPEtn subclass in
608
BLANK
ET AL.
40 L .TI
T
20:5
22.5
20.4
s-2
Acyl
132
i
181
group
FIG. 2. Distribution of tritium in selected molecular species of [3H]hexadec-l-enylacyl-GroPEtn after a lh pulse labeling (hatched bars) of unsupplemented P388D1 cells with [3H]hexadecyllyso-GroPEtn followed by a 23-h chase (shaded bars) with unlabeled media (see Materials and Methods for details). Data represent the means + SE from two separate experiments. Duplicate samples were pooled before HPLC analysis in one experiment while duplicates were analyzed separately in the second experiment.
lar species to show a decrease in radioactivity (Fig. 2). The decrease of tritium in the hexaenoic species of [3H]alk-l-enylacyl-GroPEtn during the 23-h chase period was about two times greater than the decrease of tritium found in the tetraenoic species. Peaks eluting from HPLC with retention times corresponding to benzoate derivatives of the 16:0-18:l molecular species (and the 16:0-18:2 species, data not shown) had the largest increases in tritium after a 23-h incubation of prelabeled cells (Fig. 2). Efect of fatty acid supplementation on the acylation of [3H]alkyllyso-GroPEtn and the subsequent desaturation of [‘H]aZkyZacyl-GroPEtn by cell membranes. Mem-
brane fractions isolated from P388D1 cells that were supplemented with eicosapentaenoate or docosahexaenoate for 24 h acylated significantly more (P < 0.01) of the [3H]alkyllyso-GroPEtn than was acylated TABLE
the cells from all three experimental groups. Cells grown in the presence or absence of supplements all contained increased amounts of docosahexaenoate and decreased amounts of eicosatetraenoate in the [3H]alk-l-enyl-GroPEtn fraction when compared to the levels of these two acyl moieties found in the [3H]alkylacylGroPEtn precursor (Fig. 1). The increased content of the two n-3 fatty acids in rH]alkylacyl-GroPEtn of cells enriched with eicosapentaenoate or docosahexaenoate was accompanied by decreased acylation of [3H]alkyllyso-GroPEtn with eicosatetraenoic and octadecenoic acids. However, the amount of docosahexaenoate found in cellular [3H]alkylacyl-GroPEtn was not significantly altered (P > 0.1) by growing the cells in the presence of eicosapentaenoate (the reverse was also true). When unsupplemented control cells were prelabeled (1 h) with [3H]alkyllysoGroPEtn and then incubated for another 23 h, the [3H]alk-l-enylacyl-GroPEtn that contained eicosatetraenoic and docosahexaenoic acids were the only two molecu-
III
SYNTHESIS OF [3H]ALKuLACYL-GroPEtn AND ITS SUBSEQUENT DESATURATION TO [3H]A~~-1ENYLACYL-GroPEtn BY A MEMBRANE FRACTION FROM P388D1 CELLS
Supplement None +20:5 * +22:6 *
% [3H]DiradylGroPEtn 33.8 + 2.3 45.7 + 2.8* 46.3 f 1.9*
% [3H]Alk-1enylacylGroPEtn” 9.8 + 1.3 13.1 f 1.6* 18.4 f 1.4*~**
Note. Membrane protein (3.38 f 0.06 mg) was incubated for 1 h at 37°C in the presence of 0.4 GM r3H]alkyllyso-GroPEtn and 2 mM NADH in 1 ml of 0.1 M Tris buffer (pH 7.2). Lipids were extracted and analyzed as described under Materials and Methods. All data are given as means + SE from two separate experiments, each done in duplicate. Based on Student’s t test for paired variables, *P < 0.01 vs controls and **P < 0.05 vs +20:5 group. ’ Percentage of tritium from the [3H]diradyl-GroPEtn fraction that was in the [3H]alk-l-enylacyl-GroPEtn subclass. * Cells were supplemented with 10 PM sodium eicosapentaenoate (20:5) or docosahexaenotate (226) for 24 h and washed, and a membrane fraction was prepared as described under Materials and Methods.
EFFECT
OF n-3 FATTY
ACIDS
ON PHOSPHOLIPIDS TABLE
609
OF P388D1 CELLS
IV
DISTRIBUTION OF TRITIUM IN SELECTED MOLECULAR SPECIES OF ALKYLACYL- AND ALK-l-ENYLACYL-GroPEtn PRODUCED BY MEMBRANE FRACTIONS FROM P388Di CELLS Supplement None +20:5
+22:6
GroPEtn
subclass
16:0-20:5
16:0-22:6
16:0-20:4
16:0-18:l
Alkylacyl Alk-l-enylacyl
5.2 f 0.4 4.6
32.5 k 1.0 30.5
48.2 f 1.0 44.4
4.6 f 0.4
Alkylacyl Alk-l-enylacyl
25.2 f 3.4 24.5
27.8 & 2.6
29.7
36.3 f 0.5 33.4
2.8 * 0.2 3.1
Alkylacyl Alk-l-enylacyl
2.7 t 0.5 3.1
63.1 f 3.3 62.6
27.4 + 2.2 26.1
1.8 ~fr0.3
5.9
1.9
Note. The diradyl-GroPEtn lipids were isolated from the same samples described in Table III and the percentage distribution of tritium in the molecular species of each subclass was determined by HPLC as described under Materials and Methods. Data for the alkylacyl-GroPEtn subclass are given as means f SE from two separate experiments, each with duplicate samples (n = 4). In order to have enough radioactivity for accurate HPLC analysis, duplicate samples of alk-l-enylacyl-GroPEtn were pooled within each experiment; these data represent the means from the two separate experiments.
by the membrane fraction from unsupplemented control cells (Table III). Also a greater percentage of the [3H]alkylacylGroPEtn that was synthesized was desaturated to [3H]alk-l-enylacyl-GroPEtn by the membrane fractions from the supplemented cells compared to the control cells (Table III). In contrast, when cells were supplemented with 10 PM oleic acid for 24 h, we found no significant increase in either the percentage of [3H]alkylacylGroPEtn (110 + 6% of unsupplemented controls) or in the amount of [3H]alk-lenylacyl-GroPEtn produced (75 * 9% of unsupplemented controls) by the cellular membrane fractions. In the absence of NADH, the desaturation of [3H]alkylacylGroPEtn by membrane fractions from supplemented or unsupplemented cells was ~1%. The distribution of tritium in the molecular species of [3H]alkylacylGroPEtn, produced by incubation of [3H]alkyllyso-GroPEtn with the cell-free membrane fractions, followed the same pattern (Table IV) that we observed with intact cells (Fig. 1). However, in contrast to the data from experiments with intact cells (Fig. l), there was no apparent selectivity for the docosahexaenoate containing species in the desaturation of [3H]alkylacyl-GroPEtn to [3H]alk-l-enylacyl-
GroPEtn by the membrane fractions from supplemented or unsupplemented P388Di cells (Table IV). DISCUSSION
Supplementation of P388D1 cells for 24 h with 10 pM eicosapentaenoate or docosahexaenoate did not significantly alter the cellular content of phospholipid classes and subclasses. It is noteworthy that significant amounts of alkylacyl-GroPCho (-18% of the GroPCho lipids), a precursor of platelet activating factor (PAF, alkylacetyl-GroPCho), was found in the P388Di cells. Thus, under appropriate conditions of stimulation this cell line could be a useful model for investigating this important autocoid. In fact, it has recently been shown that P388Di cells have specific receptor sites for PAF (32). We found that the acyl composition of the alk-1-enylacyl-GroPEtn subclass was altered more than any other phospholipid subclass by supplements of eicosapentaenoate and docosahexaenoate. With the docosahexaenoic acid supplement, nearly 50% of the alk-1-enylacyl-GroPEtn contained the hexaenoic species at the m-2 position and the arachidonate was greatly decreased. A recent relevant study (33)
610
BLANK
found that the alk-1-enylacyl-GroPEtn subclass of phospholipids was the largest single reservoir of eicosapentaenoic and docosapentaenoic fatty acids in platelets from humans who received dietary supplements of fish oil. If, as has been suggested (4-ll), the alk-1-enylacyl-GroPEtn subclass participates in the release of arachidonic acid for eicosanoid production, it is possible that the reduced formation of eicosanoids observed in some biological systems that have been enriched with eicosapentaenoate/docosahexaenoate (3439) could be explained by the lowering of arachidonate in this particular phospholipid subclass as we observed in the docosahexaenoate-supplemented cells (Table II). Although activities toward phospholipids other than phosphatidylcholine were not examined, a Caa+-dependent phospholipase Aa that could be involved in release of arachidonic acid from cellular phospholipids for eicosanoid production has recently been purified from membranes of P388D1 cells (40). It is also interesting to note that enrichment of cells with eicosapentaenoic acid had little effect on the distribution of docosahexaenoic acid in the cellular phospholipids and that the reverse was also true. Therefore, effects on cell function that occur due to enrichment with one of these n-3 fatty acids are not likely caused by changes in the cellular level of the other n-3 fatty acid. Although the unsupplemented P388D1 cells acylated [3H]alkyllyso-GroPEtn with higher amounts of docosahexaenoate (26%) than we previously found (9%) in our experiments with intact Madin-Darby canine kidney cells (12), the [3H]alkylacylGroPEtn fraction was still not as enriched in this molecular species as was reported for similar experiments (18) with intact Ehrlich ascites cells (64% of the tritium in the hexaenoate species). However, when P388D1 cells were supplemented with docosahexaenoic acid for 24 h before adding the [3H]alkyllyso-GroPEtn to the media, 55% of the tritium in the [3H]alkylacyl-GroPEtn was found in the molecular species containing docosahexaenoate. Whether these variations rep-
ET AL.
resent differences in cell types and/or are caused by the availability of docosahexaenoate from a donor phospholipid is not known. The percentage of cellular [3H]diradylGroPEtn associated with the [3H]alk-l-enylacyl-GroPEtn subclass, after a l-h incubation with [3H]alkyllyso-GroPEtn, was a little higher (64%) than we previously found for this subclass (56%) after a 2-h incubation of the same substrate with Madin-Darby canine kidney cells (12). However, both P388D1 cells and MadinDarby canine kidney cells converted much more of the [3H]alkylacyl-GroPEtn to the plasmalogen form than Ehrlich ascites cells (18), which contained ~3% of the tritium in alk-1-enylacyl-GroPEtn. These findings suggest that Ehrlich ascites cells have a lower level of alkylacyl-GroPEtn desaturase activity than the other two cell lines. The higher amounts of the docosahexaenoate-containing molecular species that were found in the [3H]alk-l-enylacylGroPEtn subclass compared to the [3H]alkylacyl-GroPEtn subclass (Fig. 1) suggest either a selectivity for this molecular species as a substrate for the Al-desaturase in the formation of ethanolamine plasmalogens or a higher turnover rate for this molecular species. The former possibility is not supported by the analysis of molecular species of [3H]alk-l-enylacylGroPEtn formed by cell-free membrane preparations (Table IV); however, the latter possibility is supported by the prelabeling-chase experiments done with the unsupplemented control cells (Fig. 2). At the present time it is difficult to attribute a biological significance to the high amounts of docosahexaenoate found in the alk-l-enylacyl-GroPEtn subclass in the cells grown in the presence of docosahexaenoic acid. However, because docosahexaenoic acid can act as an inhibitor of prostanoid formation (41), it is possible that the simultaneous release of docosahexaenoic acid with arachidonic acid from plasmalogens or other cellular phospholipids could attenuate the formation of eicosanoid mediators from arachidonic acid.
EFFECT
OF n-3 FATTY
ACIDS
ON PHOSPHOLIPIDS
ACKNOWLEDGMENTS This work was supported by the Office of Energy Research, U.S. Department of Energy (Contract DEAC05-760R00033), the American Cancer Society (Grant BC-i’OS), the National Cancer Institute (Grant CA-41642-03), and the National Heart, Lung, and Blood Institute (Grant HL-27109-08). REFERENCES 1. HORROCKS, L. A. (1972) in Ether Lipids: Chemistry and Biology (Snyder, F., Ed.), pp. 172-272, Academic Press, New York. 2. BLANK, M. L., WYKLE, R. L., AND SNYDER, F. (1973)Biochim Bio&vs. Acta 316,28-34. 3. WYKLE, R. L., BLANK, M. L., AND SNYDER, F. (1973)Biochim. &o&s. Acta326,26-33. 4. RITTENHOUSE-SIMMONS, S., RUSSELL, F. A., AND DEYKIN, D. (1976) B&hem Biophys. Res. Cowmun. 70,295-301. 5. RITTENHOUSE-SIMMONS, S., RUSSELL, F. A., AND DEYKIN, D. (1977) B&him. Biophys. Acta 488, 370-380. 6. BROEKMAN, M. J., WARD, J. W., AND MARCUS, A. J. (1980) J. Clin Invest. 66,275-283. 7. DANIEL,L. W.,KING,L.,AND WAITE,M.(~~~~)J. BioL Chem. 256,12830-12835. 8. KRAMER, R. M., AND DEYKIN, D. (1983) J. Biol Chem. 258,13806-13811. 9. BROWN,M.L.,JAKUBOWSKI,J.A.,LEVENTIS,L.L., AND DEYKIN, D. (1987) B&him. Biophys. Acta 921,159-166. 10. KAMBAYASHI, J., KAWASAKI, T., TSUJINAKA, T., SAKON,M.,OSHIRO,T.,ANDMORI,T.(~~~~) Bicthem. Int. 14,241-247. 11. CHILTON,F.H, AND CONNELL,T.R.(~~~~)J. Biol. Chem. 263,5260-5265. 12. BLANK, M.L.,LEE,T-C., CRESS,E. A., FITZGERALD,V.,AND SNYDER,F.(~~~~) Arch. Biochem. Biophys. 251,55-60. 13. MASUZAWA, Y., SUGIURA, T., ISHIMA, Y., AND WAKU, K. (1984) J. Neurochem. 42,961-968. 14. ONUMA, Y., MASUZAWA, Y., ISHIMA, Y., AND WAKU, K. (1984) Biochim Biophys. Acta 793, 80-85. 15. OJIMA, A., NAKAGAWA, Y., SUGIURA, T., MASUZAWA,~., AND W~~u,K.(1987)J. Neurochem. 48,1403-1410. 16. WAKU,K.,ANDNAKAZAWA,Y.(~~~~) Eur. J. Biothem. 88,489-494. 17. NAKAGAWA,Y.,ANDWAKU,K.(~~~~) Eur. J. Birr them. 152,569-572. 18. MASUZAWA,Y.,OKANO,S.,NAKAGAWA,Y.,OJIMA, A., AND WAKU, K. (1986) Biochim Biophys. Acta 876,80-90.
OF P388Di CELLS
611
19. KOREN,H.S.,HANDWERGER,B.S.,ANDWUNDERLICH, J. R. (1975) J. ImmunoL 114,894-897. 20. SNYDERMAN,R.,PIKE,M.C.,FISCHER,D.G.,AND KOREN, H. S. (1977) J Immunol. 119, 20602066. 21. DAWE,C.J., AND POTTER, M.(1957) Amer. J Pathol. 33,603. 22. RAMU, A., GLAUBIGER, D., AND WEINTRAUB, H. (1984) Cancer Treat. Rep. 68,637-641. M.,DEGREGORIO,M. W.,GANA23. HOLLERAN,~. PATHI, R., WILBUR,J. R., AND MACHER, B. A. (1986) Cancer Chemother. Pharmacol. 17, ll15. 24. MULLIGAN, J. J., LYNCH, R. D., SCHNEEBERGER, E. E., AND GEYER, R. P.(1977) B&him. Bie phys. Acta 470,92-103. 25. BLIGH, E. G., AND DYER, W. J. (1959) Canad. J. B&hem. Physiol. 37,911-917. Anal. B&hem. 72,24826. BRADFORD,M.M.(~~~~) 254. 27. WYKLE, R. L., BLANK, M. L., AND SNYDER, F. (1972) J. Biol. Chem. 247,5442-5447. Bie 28. CLARKE,N.G.,ANDDAWSON,R.M.C.(~~~~) them. J 195,301-306. 29. SNYDER,F.,AND SMITH,D.(~~~~) Sep. Sci. 1,709722. 30. ROUSER,G.,SIAKOTOS,A.N.,AND FLEISCHER,~. (1966) Lipids 1,85-86. 31. BLANK,M.L.,ROBINSON,M.,FITZGERALD,V.,AND SNYDER,F.(~~~~) J. Chrmatogr. 298,473-482. 32. VALONE, F. H. (1988) J ImmunoL 140,2389-2394. 33. HoLuB,B.J., CELI, B., AND SKEAFF, C. M.(1988) Thromb. Res. 50,135-143. 34. LEE,T.H.,H~ovER,R.L., WILLIAMS,J.D.,SPERLING, R. I., RAVALESE,J., SPUR, B. W., ROBINSON,D.R.,COREY,E.J.,LEWIS,R.A.,ANDAUSTEN, K. F. (1985) N. EngL J. Med. 312, 12171224. 35. CROFT, K. D., CODDE, J. P., BARDEN, A., VANDONGEN,R., AND BEILIN,L.J. (1985) B&him. Biophys. Acta 834,316-323. 36. LoKEsH,B.R.,HsIEH,H.L., ANDKINSELLA,J.E. (1986) J Nutr. 116,2547-2552. 37. CROFT, K. D., BEILIN, L. J., LEGGE, F. M., AND VANDONGEN,R.(~~~~) Lipia?s22,647-650. 38. LoKEsH,B.R., ANDKINSELLA,J.E.(~~~~) Immunobiology 175,406-419. 39. LOKESH,B. R., GERMAN,B., AND KINSELLA,J. E. (1988) Biochim. Biophys. Acta 958,99-107. 40. ULEVITCH,R.J.,WATANABE,Y.,SANO,M.,LISTER, M.D.,DEEMs,R. A., AND DENNIS,E. A.(1988) J. Biol. Chem. 263,3079-3085. 41. COREY,E.J.,SHIH,C.,AND CASHMAN,J.R.(~~~~) Proc. Natl. Acad. Sci. USA 80,3581-3584.