results showed that plasma membranes and azurophilic granules were enriched with ethanolamine-(PE) rela- tive to choline-(PC) containing phosphoglycerides ...
THEJOURNAI. OF BIOLOGICAL CHEMISTRY 1989 by The American Society for Biochemistry and Molecular Biology, Inc
Vol. 264, No. 30,Issue of October 25, pp. 17718-17726, 1989 Printed in U.S.A.
Distribution of Arachidonic Acid in Choline- and Ethanolaminecontaining Phosphoglycerides in SubfractionatedHuman Neutrophils* (Received for publication, June 23, 1989)
James I. S . MacDonald and Howard Sprecher$ From the Department
of
Physiological Chemistry, The Ohio State University, Columbus, Ohio 43210
Human neutrophils were fractionated on Percoll gra- human neutrophils are PC’ and PE which contain, respecdients and the various subcellular fractions were anatively, 19 and 68% of the total endogenous arachidonate (1). lyzed for phospholipid and fattyacid composition. The Moreover, this arachidonate is localized to specific subclasses, results showed that plasma membranes and azurophilic 66% of the PC-associated arachidonatebeing found at the sngranules were enriched with ethanolamine-(PE) rela- 2 positionof the 1-0-alkyl-2-acyl (ether linked) subclass while tive to choline-(PC) containing phosphoglycerides. A 71% of the PE arachidonate is at the sn-2 positionof the 1remarkable degree of uniformity existed throughout O-alk-l’-enyl-2-acyl (plasmalogenic-linked) subclass (1). the gradient with respect to the subclass composition Tightly coupled to arachidonate release and metabolism is of the subcellular PC and PE components. In each the productionof PAF, a potent mediator of aggregation and fraction 50-60% of the PC was diacyl, 40-45% was 1- degranulation in platelets and neutrophils (2-4), the precursor 0-alkyl-2-acyl (ether linked), and2-5% was 1-0-alk- of which is l-O-alkyl-2-arachidonoyl-sn-glycero-3-phosphol’-enyl-2-acyl (plasmalogenic). For PE, 20-25% was choline (2, 4-6), and a number of PC-derived metabolites diacyl, 7-12% ether linked, and 64-76% plasmalo- which may serve as second messengers during cellular actigenic. vation (7, 8). In addition a plasmalogenic-linked ethanolaWhen neutrophils were incubated for 15 min with mine-containing PAF analog hasalso been detected in stim[ l-’4C]arachidonic acid and subfractionated most of ulated cells (9). the PC-associated label was intracellularly localized. Resting cells maintain low intracellular concentrations of A similar result was observed in PE, however, when free arachidonate (10) and neutrophils rapidly acylate [“CC] the cells were allowed to stand for 2 h in fatty acid- arachidonic acid into cellular phospholipids, primarily PI and free buffer following the 15 min of labeling and then PC (11-14). Recent studies have indicated that exogenous subfractionated there wasa sizable migration of [‘“C] arachidonate is preferentially acylated into 1-acyl-linked choarachidonate into plasma membrane PE. In all cases line and ethanolamine containingglycerophospholipids via a the diacyl subclass waslabeled most heavily after 15 CoA-dependent acyl transferase and is then remodeled by a min but after an additional 2 h of incubation in fatty CoA-independent mechanism to respective the ether and plasacid-free buffer there was a direct transfer of label to malogenic subclasses (12, 13). the ether- and plasmalogenic-linked P C and PE subIn light of the important function of ether- and plasmaloclasses. It wasalso found that arachidonoyl-coenzyme genic-linked PC and PE, both as precursors and reservoirs, it A 1-acyl-lysophosphatide acyltransferase activity was would be of interest to determine whether any compartmeninherent in all three major membrane types but was tation of these lipidsoccurs withinneutrophils.Previous enriched in theendoplasmic reticulum/secondary gran- studies have localized the acetyl-coA transferase activity to ule fraction. Arachidonate consistently accounted for the endoplasmic reticulum (15) and exogenous 3H-PAF has roughly 5%of the PC and 17%of the PE fatty chain been shown to bind preferentially to the plasma membrane composition in each subcellular fraction. (16). Inaddition,protein-mediatedtransfer of PAFfrom These findings demonstrate that, despite the uniform cytosol to artificial membranes has been documented (17). arachidonateand P C andPEsubclass composition Smith and Waite (18) have presented evidence suggesting a within the various neutrophil subcellular fractions, preferential the loss of [3H]arachidonate from azurophilic granbulk of the PC- and PE-associated arachidonate is ules when neutrophils were stimulated with ionophore localized in intracellular membranes. A23187, however, Chiltonand Connell (1) have recently shown that release of radioactivity alone was not an accurate measure of the mass loss of arachidonate upon stimulation. Moreover, Smith and Waite(18)did not analyze the PC and Wheninflammatory cells suchasneutrophilsarechallenged with ionophore A23187 they release arachidonic acid P E subclass composition of the subcellular fractions nor did composition of the respective from intracellular stores andmetabolize this to a number of they determine the arachidonate cholineand ethanolamine-containing phosphoglycerides. compounds which play an importantrole in the inflammatory of this study was to determine whether therewas The aim response. Two major sources of arachidonate within resting any specific compartmentation within neutrophilsof arachi* This work was supported by Grants DK-20387 and DK-18844 from the National Institutes of Health. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “aduertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate thisfact. $ To whom correspondence should be addressed Dept. of Physiological Chemistry, The Ohio State University, 5148 Graves Hall, 333 W. 10th Ave., Columbus, OH 43210.
The abbreviationsused are: PC, choline-containingphosphoglycerides; PE, ethanolamine-containingphosphoglycerides; PI, inositolcontaining phosphoglycerides; PS, serine-containing phosphoglycerides; SM, sphingomyelin; PAF, platelet-activating factor; PBS, phosphate-buffered saline; HPLC, high pressure liquid chromatography; ER, endoplasmicreticulum; PIPES, 1,4-piperazinediethanesulfonic acid.
17718
Subcellular Distributionof Arachidonic Acid donate containing ether- and plasmalogenic-linked PC and P E subclasses. Concomitant with such a localization would be specific transacylation of [14C]arachidonate from diacyl to ether and plasmalogenic subclasses. The results indicate that the bulk of thearachidonic acid inPC was localized in intracellularmembranes,althoughwithineach individual subcellular fraction the percent composition of arachidonate and the various PC and PE subclasses remained relatively constant. It wasalso found that arachidonoyl-CoA l-acyllysophosphatide acyltransferase activity was not confined to any one membrane type but rather was distributed nonuniformly throughout thecell. Finally, despite the preponderance of [14C]arachidonate incorporation into PC relative to PE, most of the arachidonate associated with these phospholipids was confined to the ethanolamine-containing fraction, a result which further emphasizes the difficulties in using radiolabels in measuring arachidonic acid pools as sources of prostanoid synthesis. EXPERIMENTALPROCEDURES
Materials-[l-'4CC]Arachidonic acid (54.5 pCi/pmol) was purchased from Du Pont-New England Nuclear. LK5 Linear K silica gel thin layer chromatographyplates were from Whatman,andthe lipid standards were from Avanti Polar Lipids (Birmingham, AL). Essentially fatty acid-freebovine serumalbumin wasfromSigma, and Percoll was purchased from Pharmacia LKB Biotechnology, Inc. All solvents were of HPLC grade fromFisherorreagent gradefrom Mallinckrodt. Preparation of Neutrophils-Human neutrophils were obtained fromfreshly prepared buffy coats supplied by the AmericanRed Cross. The cells were mixed with a 1:5 volume of 6% (w/v) dextran T-400 (Pharmacia LKB13iotechnology Inc.) in 0.15 M NaCl containing 7.7 mM EDTA, made up to thedesired volume with 0.15 M NaCl and allowed to stand at room temperature for 30 min. This facilitated the sedimentation of residual red blood cells. The supernatant was then siphoned off and Centrifuged a t 800 X g for 15 min. Each cell pellet was then resuspended in 20 ml of PBS (0.14 M NaCI, 2.7 mM KC], 8.0 mM Na,HPO,, 1.5 mM KHzHPO,, 0.9 mM CaC12, and 0.5 mM MgCl,), layered over an equal volumeof Ficoll-Paque (Pharmacia LKB Biotechnology Inc.'t and centrifuged a t 800 X g for 20 min. To get rid of any contaminatingred blood cells, the neutrophil pelletwas resuspended in 20 ml of ahypotonic solution containing 0.155 M NH4C1,0.01 M KHCO?, and 0.1 mM EDTA. Thecells were centrifuged as before and washed twice with PBS and resuspended in the same. Incorporation of p'C]Arachidonic Acid into Neutrophils-Fifty p1 of an aqueous solution of [l-'4C]arachidonic acid complexedto bovine serum albumin (5.0 mg/nnl) was added to each of 20 flasks containing 1.95 mlof the neutrophil suspension(2.0 X lo7 cells/ml) in PBS (final arachidonate concentration of 10 & I ) . T h eincubation flasks were shaken at 37 "C for 15 min afterwhich an equal volume of cells was removed to each of two 50-ml centrifuge tubes andwashed three times with 40 ml of ice-cold PBS containing 0.5 mg/ml bovine serum albumin. The contentsof one tube was then resuspended in 6.0 ml of ice-cold lysis buffer (0.1 M KCI, 5.0 mMMgC12, 0.1 mM leupeptin, and 0.025 M Tris-C1, pH 9.6) (19) fornitrogen cavitation.The remaining cells were resuspended in 20 ml of PBS and reincubated a t 37 "C for an additional 2 h after which they were centrifuged and resuspended in lysis buffer as described above. For mixing experiments, neutrophilswere incubated asabove with [5,6,8,9,11,12,14,15-3H]arachidonic acid (Du Pont-New England Nuclear, 100 Ci/mM) (10 flasks with 2 pCi/flask), washed, lysed, and fractionated asdescribed below. The azurophilic granule portion was then mixed with a nonhbeled cavitate and this was centrifuged on Percoll as described. Subfractionation of Neutrophils-All procedures were carried out at 4 "C on ice. The disruption and fractionation procedure was an adaptation of the methods of Record et al. (19) and Borregaardet al. (20). Neutrophils suspended inlysis buffer were subjected to 500 p.s.i of nitrogen in a Parr bomb (Parr Instruments,Moline, IL) for 20-25 min. The cavitate was collected dropwise into a chilled flask on ice, and 2000 units of DNase I (Sigma)and 20 units of micrococcal nuclease (Sigma) were added to the lysate. The homogenate was then centrifuged for 5 min a t 3500 r.p.m. in a Beckman Acu-Spin benchtop centrifuge to remove unbroken cells and nuclei.
17719
The supernatant was made up to 5 ml, and this was then applied
to a discontinuous Percoll gradient consisting of 15 ml of Percoll, density 1.05 g/ml in 0.106 M KCI, 5.3 mM MgCL, 1.0 mM Na*ATP, 0.1 mM phenylmethylsulfonyl fluoride, and 0.106 M Tris-C1, pH 9.4, layered over 11 ml of Percoll, density 1.1 g/ml made up in the same buffer, except the pH was 9.0. The gradients were centrifuged for 30 min a t 22,000 r.p.m. in a Beckman Ti-60 rotor. Fractions of 2 ml were collected by aspiration from the topof the gradient, mixed with Percoll, density 1.12 g/ml in 0.06 M KC1 and 0.06 M PIPES, pH 6.0, and centrifuged for 2 h a t 39,000 r.p.m. in a Beckman Type 40 rotor. The resulting membrane pellets, which were found resting overa solid core of Percoll, were removed and either extracted for lipid analysis or resuspended in a small volume of 0.025 M Tris-CI, pH 7.0, containing 0.1 mg/ml leupeptin for assay of enzyme markers. Lipid Analysis-Lipids were extracted by the method of Bligh and Dyer (21). Phospholipidswere separated from neutral lipids by silicic acid column chromatography. The total lipid extract was applied to the column in CHCI, and the neutral lipids were eluted in the same solvent. Phospholipids were removed from the column by elution with CHBOH.Individual phospholipids were resolved by thin layer chromatography using CHC13/CH30H/40% CH3NH2 (60:20:5), (22) as the mobile phase. The phospholipids were visualized by spraying the plate with 0.1% 2',7'-dichlorofluorescein in ethanol and the PC and P E bands were scrapedintotesttubesandeluted with CHC13/ CH,OH/H,O (50:50:10). Extraction of each phospholipid was determined to be better than 90% based on the amount of radioactivity left in the silica gel. Individual PC and PE fractions were converted to theirrespective diradylglycerols by treatment with phospholipase C (23), extracted, and converted to benzoates according to theprocedure of Blank etal. (24). The subclasses were resolved by thin layer chromatography using benzene/hexane/diethyl ether (50:45:5) as the solvent. In order to quantitate the subclasses, it was first necessary to remove thesmallamount of dimethylaminopyridine which coextracted with the benzoates. This wasaccomplished by silicic acid column chromatography using CHCla as the eluting solvent. Separation of the subclasses was achieved by normal phase HPLC using a pm Poracil column (Waters) eluted isocratically with cyclohexane/ diethyl ether/glacial acetic acid (97:3:0.7), (25) a t a flow rate of 1.0 ml/min. Integrated peaks were obtained using a Varian model 4290 integrator. Under these conditions the retention times for the three subclasses were roughly 5 min for the plasmalogenic species, 8-10 min for the ether and about 14 min for the diacyl. Alterations in the retention times of the three subclasses could be made by changing the concentration of diethyl ether relative to cyclohexane. Inclusion of a radiolabeled internal standard (about 40,000 cpm of l-acyl-2linoleoyl-PC, Du Pont-New England Nuclear, 58.3 mCi/mmol) revealed a recovery rate of about 80%. For fatty acid analysis, thebenzoates or intact phospholipids were dissolved in a small volume of 5% anhydrous HCI in methanol and heated a t 70 "C for 1 h. About 7 pg of trieicosoin (Nu Chek Prep, Elysian, MN) and40,000 cpm of 1-acyl-2-linoleoyl-PC were added as internal standards. Methylated fatty acids were extracted with hexane and analyzed on a Varian model 6000 gas chromatograph containing a 10-ft by 2-mm internal diameterglass column packed with10%SP2330 on a 100:120 mesh Supelcoport (Supelco, Bellefonte, PA). Components were quantitated by integration with a Varian model 4290 integrator. The oven was maintained a t 180 "C with a 5-min ramp to 190 "C after 17 min. Assays-phospholipid phosphorous was measured by the procedure of Rouser et al. (26). Protein was measured using the BCA protein assayreagent supplied by PierceChemical Co. Lysozyme and gglucuronidase were assayed as described by Voetman et al. (27), and myeloperoxidase was determined according to theprocedure of Migler and DeChatelet (28) with o-dianisidine (final concentration 1%)as the substrate. Alkaline phosphatase wasassayedaccording to the method of Nauseef and Clark (29) and glucose-6-phosphatase was determined asdescribed by Klempner etal. (30). Arachidonoyl-CoA:lacyl-lysophosphatide acyltransferase was assayedaccording to the procedure of Fuse et al. (31) with minor modifications. Briefly, the assay system contained 10 nmol of arachidonoyl-CoA, 12 nmol of Ll-stearoy~-2-lyso-phosphatidylcholine or ~-l-o1eoyl-2-lyso-phosphatidylethanolamine,membraneproteinand 35,000 CPM of [1-"C] arachidonoyl-CoA (53.4 mCi/mM) in a final volume of 0.8 ml of 50 mM PIPES, pH6.6. The reaction was incubated at 37 "C for 2-5 min and was terminated by the addition of 3.5 ml of CHC13/CHsOH/HCI (50:lOO:l). Lipids were extracted and separated by thin layer chromatography on Whatman LK5 Silica Gel TLC plates using CHCln/
Distributio
17720
Subcellular
CH,,OH/H,O/CH,COOH (75:25:3:1) as the mobile phase. PC and PE were visualized by exposure of the plate to I, vapor and the radioactivity was determined by liquid scintillation counting. RESULTS
Fig. 1 shows the distribution of enzyme markers following centrifugation of a neutrophil cavitate on Percoll gradients. Alkaline phosphatase, the plasma membrane marker(20, 30, 32), was localized in fractions 5 and 6 while the endoplasmic reticulum (ER)marker, glucose-6-phosphatase (30), was found in fractions 10 and 11, and the azurophilic granule enzymes myeloperoxidase and /3-glucuronidase (16, 20, 32)
I
I i
0.5
3
IO
I3
FRACTION
h
U
FRACTION
FIG. 1. Distribution of neutrophil subcellular marker enzymes onPercoll gradients. Neutrophils were ruptured by nitrogen cavitation and centrifuged through Percoll as described under “Experimental Procedures” and collected in 2-ml fractions. A , alkaline phosphatase and lysozyme; B,myeloperoxidase and glucose-6-phosphatase; C, B-glucuronidase and protein.
,nof Arachidonic Acid TABLEI Distribution of myeloperoridase activity and PHlarachidonate in choline- and inositol-containingphosphoglyceridesin neutrophil subcellular fractions Data are the mean and standard deviation from the mean of four separate experiments. Plasma membrane comprises fractions 5 and 6, ER/secondary granule comprises fractions 10 and 11 and fractions 13-15 make up theazurophilic granule. % Radioactivity % Subcellular fraction Myeloperoxidase PC
PI
activitv
1.2 k 0.5 11.2 & 5.0
1.4 t 0.5 9.8 & 4.6
2.9 t 1.6 4.6 k 1.0
76.9 & 8.8
75.6 k 7.7
80.9 k 2.7
total
Plasma membrane ER/secondary granule Azurophilic granule
were confined to the bottom three fractions of the gradient. Fractions 1-4 comprised the cytosolic portion while fractions 7-9 contained very little enzyme activity or membrane material and were routinely discarded. Lysozyme was distributed throughout the lower half of the gradient and may indicate secondary granule localization with ER. Lysozyme appears to be a nonspecific enzyme found on both azurophilic (primary) and secondary granular membranes (16, 20, 32). Whenthe azurophilic granularfractionfrom cells prelabeled with [3H]arachidonic acid was mixed with a nonradioactivelysate andfractionatedon Percoll as before, roughly 80% of the radioactivity was recovered in the same fraction (Table I). In addition, the distribution of the azurophilic granule markerenzyme myeloperoxidase paralleledthat of the radiolabel (Table I). This indicates that, although some membrane mixing may occur during the fractionation procedure, it is probably not sufficient to significantly influence the results. The phospholipid composition of the various fractions is shown in Table 11. Plasma membranes contained very little PI relative to PS and SM. In addition, there was roughly twice theamount of PE relative toPC.Thisresult was reversed in the ER/secondary granular membranes where PC predominated. About 55% of the homogenate PC was diacyl, 43% ether, and 4% plasmalogenic whereas 21% of total PE was diacyl, 12% ether, and 69% plasmalogenic (Tables 111and IV). Analysis of the subclass composition of the respective choline- and ethanolamine-containing phospholipids from each subcellular fraction revealed a remarkable degree of uniformity throughoutthe cell (Tables 111 and IV). The onlydeviationwas observed intheplasmamembranefraction,andthis was slight. Human neutrophils rapidly incorporated [ 14C]arachidonate into cellular lipids (33). After 15 min about 25-3096 of the cell-associatedradioactivity was incorporatedinto cellular phospholipids (predominantly PIandPC),theremainder residing mainly in triacylglycerol (33).Tables V and VI show theincorporation of labeled arachidonate, expressed as pmoles of [14C]arachidonate incorporated, into the PC and PE components of the individual subcellular fractions after an initial 15 min exposure ( t = 0) and following a 2-h reincubation in fatty acid-free buffer. As can be seen, most of the PC- andPE-associated radioactivitywas localized within intracellular membranes.Acylation into PC exceeded that observed into PE after both timeperiods (Tables V and VI). Moreover, a large time-dependent increase in the amount of [14C]arachidonate was observed in each subcellular fraction for bothphospholipids. With respect to PC this increasewas fairly uniform and did not result in any overt changes in the
Subcellular Distribution
of
Arachidonic Acid
17721
mol %phosphorous
5
2.7 f 0.5
14.3 k 3.4
17.7 23.3 f 0.9
10
7.3 f 1.7
6.7 C 0.5
10.7 & 1.9
11
11.3 f 3.1
5.7 f 2.0
11.0 f 4.2
12
11.3 rC_ 1.2
8.7 f 1.2
10.7 f 0.5
13
13.7 f 5.2
9.7 f 2.4
13.725.7 f 0.9
14
7.7 rC_ 0.9
9.0 k 2.4
14.0 F 1.6
15
11.3 f 3.3
10.0 t 1.6
12.0 k 4.2
H
6.9 k 0.7
8.6 f 2.4
16.3 F 1.1
f 2.5 (16.0 f 2.8)" 41.0 f 0.8 (23.0 f 3.7)" 35.3 k 0.9 (11.3 f 2.6)" (9.0 31.0 & 1.4 (11.3 f 1.2)" (10.3 38.0 f 0.9 (10.7 f 1.2)" 25.3 f 1.2 (20.0 f 4.2)" (23.7 24.6 f 0.9 (7.7 f 2.6)" (9.7 32.0 f 2.7
43.4 f 0.5 (21.7 k 2.5)" 34.7 -e 2.0 (16.3 k 2.5) 36.6 rt 1.7 h 2.2)* 38.0 t 2.3 f 1.2)b t 3.3 (10.0 f 1.4)'' 44.0 k 3.7 k 2.5)" 43.0 f 0.8 f 3.2)b 38.4 rt 3.6
Mol % total cellular PC. '' Mol % total cellular PE. a
TABLE 111 Subclass compositionof choline-containing phosphoglyceridesf r o m indiuidual neutrophil subcellular fractions Data are theaverage of three separate experimentsf the standard deviation from the mean. H, cellular homogenate. 1-0-alkyl-2-acyl 1-0-alk-1'-enyl Diacyl Fraction
however, enriched in the ER/secondary granule fraction for both PC and PE, although the total activity in azurophilic the granule was about 60% of that in the ER/secondary granule (Table VII). Moreover, all membranes preferentially acylated 1-acyl-lysophosphatidylcholine,a result which may be indicative of separate enzymesfor each phospholipid. Distinct mol % lysophosphatidylcholine and lysophosphatidylinositol acyl5 58.0 f 2.0 40.0 C 2.0 2.0 f 0.1 transferases have recently been purified from bovine heart 40.1 k 1.1 2.3 f 0.3 10 57.1 f 0.9 muscle (34). Recovery of enzyme activity was only about 30% 11 51.8 f 3.8 45.2 k 2.8 3.0 f 1.0 suggesting that the enzyme might be mainly cytosolic in its 6.5 f 4.5 44.2 k 1.2 49.4 C 3.4 12 localization. This possibility was negated when assays using 4.2 f 1.9 45.4 f 2.4 50.5 k 0.6 13 total membrane and cytosolic preparations showed the en4.0 f 0.1 44.8 f 0.8 14 51.3 f 0.8 zyme to be associated solely withtheparticulatefraction 51.5 15 k 2.5 43.8 f 4.8 4.7 f 2.3 55.8 k 2.8 43.6 f 1.6 3.8 f 1.2 H (data not shown). After the initial 15-min incubation, each of the subcellular of [I4C]arachidonate fractions had an identical percentage TABLE IV associated with the ether-linked PC subclass (Table VIII). Subclass compositionof ~~thanolamine-containing phosphoglycerides from indiuiducd neutrophil subcellular fractions Following a n additional 2 h, the percent radioactivity in the Dataarethe average of threeseparatedeterminations f the ether PC portionof the various fractions hadincreased t.o 38standard deviation from the mean. H, cellular homogenate. 40% (Table VIII). Similarly, the amount of radiolabel in the Diacyl Fraction 1-0-alkyl-2-acyl l-O-Alk-l'-enyl plasmalogenic PE fractionineach subcellular fractioninmol % creased over time with a concomitant decrease in the diacyl76.2 f 1.9 associated radioactivity (Table VIII). No fatty acid remodel16.7 k 0.7 7.2 k 1.2 5 64.8 f 0.8 24.7 C 1.3 10.6 -t 0.6 10 ing occurred when labeled cells were incubated for 2 h under 65.4 f 1.4 9.8 k 0.2 24.9 f 1 2 11 cavitation conditions (i.e. incubated on ice at 4 "C under 450 64.0 f 4.1 24.5 f 4 5 11.7 t 0.4 12 psi Nz) indicating that, under these conditions, the enzymes 69.2 f 2.2 20.0 f 2.1 10.9 f 0.1 13 were inactive (not shown). 67.9 f 2.4 14 11.9 k 1.2 20.8 k 1.2 While a considerably greater amount of ['4C]arachidonate 9.3 f 2.7 68.8 f 4.8 21.9 C 2.1 15 68.9 f 0.9 21.2 f 0.9 H 10.0 f 0.0 was incorporated into PC relative to PE, in terms of mass, arachidonic acid accounted for 16-19% of thefatty acid percent distribution of radioactivity among the various sub- content of PE in each subcellular fraction while accounting thePCfatty acid (Tables IX and X). cellular fractions (Table V). The same statement could also foronly 4-795of be made forthe various P E fractions, theonly exception being Moreover, 78% of the totalPC- and PE-associated arachidonthe plasma membrane where the specific activity increased ate was in the ethanolamine-containingphospholipid. Of the PC arachidonate one-third was localized in the ER/secondary 8.7-fold over time as opposed to a general increaseof 3.5-4.5IX). Azurophilic granular PC fold throughout the rest of the cell (Table VI). The probable granularmembranes(Table source of this additional [14C]arachidonateis thehigh specific also contained large amounts of arachidonate. In contrast, a large percentage of total PE arachidonate was found in the activity triacylglycerol pool (33). The rapidityin which the subcellular fractions were labeled plasma membranewith similar amountsbeing localized in the azurophilic granulefractions was somewhat surprising andwhen membranes purified from ER/secondarygranuleand the Percoll gradient were assayed for arachidonoyl-CoA. 1- (Table X). The same results were observed when cells were acyl-lysophosphatide acyltransferase, activity was found in incubated at 37 "C for 2 h prior to extraction. It should be all threemembranetypes(Table VII). The enzyme was, pointed out that the fatty composition at the sn-1 position of
17722
Subcellular Distributionof Arachidonic Acid TABLE V Distribution and specific activity of I'4C/arachidonic acid in choline-containing phosphoglycerides in neutrophil subcellular fractions The data in the first two columns are representative of four separate determinations, the % radioactivity is the average f the standard deviation from the mean of four separate experiments. pmol ["C]arachidonate/ nmol PC
pmol ("CC]arachidonate Fraction
2.5
5 10 264.5 11 12 13 14 15
t=O"
t=2h
63.2 152.6 75.3 76.6 96.1 221.0 55.2
157.5 415.6
1.4 2.5
244.4 326.2 484.1 14.2 133.1
2.4 2.6 4.2 3.8
11.8
t = 2t = h 0" t = O "
8.7
% ' Radioactivity
t=2h
5.4 10.4 10.8 15.4 12.2 11.6
C 3.3 18.9 k 2.8 10.1 -c 2.0 9.2 C 1.5 15.0 C 1.1 26.3 t 5.4 5 4.5
9.5 f 1.8 19.0 t- 0.6 10.2 f 2.0 9.5 f 1.6 14.2 f 2.0 26.7 f 3.0 10.8 4 3.0
" t = 0 represents cells subfractionated and extracted immediately following the 15-min incubation with [I4Cj arachidonate.
TABLEVI Distribution and specific activity of ["C]arachidonic acid in ethanolamine-containing phosphoglyceridesin neutrophil subcellular fractions The data in the first two columns are representative of four separate experiments, the % radioactivity is the mean f the standard errorfrom the mean of four separate experiments. pmol ['T]arachidonate/
('4Clarachidonate pmol
5 IO 11 111.6 12 135.0 13 14 15
% Radioactivity
PE
Fraction t = 0"
tS2h
t = 0"
t=2h
t = 0"
t=2h
31.2 58.7 37.5 34.9 55.1 77.2 32.0
174.2 171.6 11.4 126.6
0.4 1.2 1.4 1.0 1.1 1.1 1.7
3.7 5.0
10.9 f 2.6 20.0 f 3.9 f 2.0 10.3 f 0.8 14.4 f 1.9 22.1 f 4.8 10.7 f 3.5
22.8 f 4.2 17.1 & 1.1 10.1 t 2.1 9.7 1.1 11.9 f 1.7 19.2 f 2.7 8.9 f 3.7
5.7
176.7 66.0
4.6 3.7 3.4 5.1
*
a t = 0 represents cells subfractionated and extracted immediately following the 15-min incubation with [I4Cc] arachidonic acid.
TABLEVI1 Distribution of arachidonoyl-CoA:I-acyl-2-lysophosphatide acyltransferase activity in subfractionated human neutrophils Data are representative of four separate determinations. Membrane fractions are asdescribed for Table I. Specific
Total activity
Subcellular membrane activity
PE
Plasma membrane ER/secondary granule Azurophilic granule Homogenate
PC
nmol/rnin/mg protein 9.0 18.1 79.2 36.1 6.9 14.0 10.3 18.6
PE
PC
nmollmin 2.8 27.7 16.9 152.9
5.4 60.8 35.7 271.9
the ether subspecies was not analyzed, therefore, the values in Tables IX and X represent only an approximation of the fatty acid composition of the various subcellular choline and ethanolamine-containing phospholipids. DISCUSSION
The procedure reported herefor neutrophil subcellular fractionation was adapted from the methodsof Record et al. (19) and Borregaard et al. (20). The reasonfor incorporating or the features from these two methods rather than adopt one other was that neither procedure by itself gave satisfactory results in our hands. The cell fractionationtechniquedescribed here gave the same three bands described by Borregaard et al. (20) which were identified as plasma membrane (top band),endoplasmic reticulum/secondary granule (middle
band) and azurophilic (primary) granule (bottom band) with very little or no cross contamination. Analysis of the phospholipid composition of the subceIIular fractions revealedseveral distinct differences. The plasma membrane containedlarge amounts of PS and SM relative to PI and, in addition, PE predominated over PC. In the ER/ secondary granule fraction PC was in excess over P E while in the azurophilic granule P E againwas the predominant phospholipid. PC was found to comprise about 30% of the total cellular phospholipid while P E comprised about 38%. Furthermore, PI and PS were observed to each account for roughly 9% (Table 11). Mueller et al. (35) previously reported the neutrophilphospholipid composition to be 41%(PC), 39% (PE), 4% (PS), 1%(PI), and 14%(SM). Other workers have reported higher values for PS (8%) and PI (5%) and lower values for P E (33%)and SM (5%) (36). In neutrophils, incorporation of radiolabeled arachidonate into PC and PE occurs first into the 1-acyl-linked subclass via a CoA-dependent acyltransferase which is specific for this linkage (12, 13). Subsequently, the label is remodeledover time into the ether andplasmalogenic subspecies by a CoAindependent mechanism (12, 13). In this study we were interested to see if the CoA-independent transfer of arachidonic acidwasconfined to any specific membrane type and, in addition, tolook for intracellular compartmentationof arachidonate and specific PC and PE subclasses. We found that the degree of label incorporation into thevarious subcellular fractions differed (Tables V and VI), although the percent distribution of [''Clarachidonate into the subcellular PC and P E subclasses was uniform. In addition, it was also observed
17723
Subcellular Distributionof Arachidonic Acid TABLE VI11 Distribution of [I -"C]arachidonate i n choline- and ethanolamine-containing phosphoglyceridesubclasses i n individual subcellular fractions Dataaretheaverage of fourseparateexperiments f thestandarddeviationfromthemean. H, cellular homogenate. PE
PC Fraction
t = 0"
t=2h
t = 0"
t=2h
% radioactivity
5 Diacyl 1-0-alkyl-2--acyl 1-0-alk-1'-enyl
72.3 f 3.8 21.6 f 3.1 6.0 f 1.2
53.3 f 4.5 38.8 f 3.5 7.8 f 1.9
46.7 f 6.9 15.5 f 4.4 37.7 f 9.3
20.5 ? 3.4 14.6 +. 3.1 64.9 f 4.7
10 Diacyl 1-0-alkyl-2-acyl 1-0-alk-1'-enyl
66.2 f 9.4 24.8 f 3.5 5.0 f 0.8
53.5 f 4.0 38.8 f 2.7 7.7 f 1.8
36.3 f 2.8 14.8 f 1.3 48.9 f 4.0
22.6 f 4.2 13.9 +. 1.6 63.5 k 2.9
11 Diacyl 1-0-alkyl-2-acyl l-O-alk-l'-enyl
70.6 f 3.9 23.0 k 3.9 6.4 f 1.5
53.1 f 4.6 38.8 f 3.0 8.1 f 2.1
41.1 k 3.1 13.6 f 2.8 45.3 f 2.7
28.3 f 5.8 14.5 k 3.4 57.2 k 3.9
12 Diacyl 1-0-alkyl-2-acyl 1-0-alk-l'-enyl
71.3 f 4.9 22.3 f 4.3 6.4 f 1.1
53.3 f 5.7 38.7 f 3.7 8.1 k 2.2
40.9 f 5.2 14.6 f 5.0 44.3 f 2.1
30.0 f 4.0 15.3 f 4.2 54.7 k 1.3
13 Diacyl 1-0-alkyl-2-acyl 1-0-alk-1'-enyl
72.0 f 4.6 22.4 f 3.8 5.9 f 4.0
52.5 f 3.6 39.6 k 2.5 8.2 f 2.3
41.9 f 9.0 14.8 f 5.1 43.3 f 6.2
28.9 f 4.8 14.8 f 3.6 56.3 f 1.2
14 Diacyl 1-0-alkyl-2-acyl 1-0-alk-1'-enyl 5.1
72.1 f 4.0 22.9 f 3.1 f 0.9
51.4 f 4.9 40.1 f 3.8 8.4 f 2.8
53.4 f 4.8 14.6 f 1.6 32.2 f 3.7
23.4 f 6.2 14.2 f 2.3 62.5 f 4.6
15 Diacyl 1-0-alkyl-2-acyl l-O-alk-l'-
