In vitro incorporation and metabolism of some icosaenoic acids in platelets. Effect on arachidonic acid oxygenation

52

Biochimica et Biophysrca Acta 833 (1985) 52-58 Elsevier

BBA 51825

In vitro incorporation and metabolism of some icosaenoic acids in platelets. Effect on arachidonic acid oxygenation Michel Lagarde, INSERM

Brigitte Drouot,

Michel Guichardant

and Marc Dechavanne

U.63, Institut Pasteur, Laboratoire d’H&nobiologie, Facultd Alexis Carrel. 69372 Lyon Cedex 2 (France) (Received

Key words:

Arachidonate

oxygenation;

July IOth, 1984)

Phospholipid;

Icosaenoic

metabolism;

(Platelet)

Three icosaenoic acids (20: 3(n - 6), 20: 5(n - 3) and 20: 3(n - 9)) which may arise in platelet phospholipids under certain dietary conditions and which may affect platelet functions have been taken up by human platelets. Each acid was pre-coated onto delipidated albumin and then incubated with platelets isolated from their plasma. The distribution study of each acid in cellular lipids revealed that around 80% of the acid taken up was located in phospholipids, of which the bulk was in phosphatidylcholine. The percentage incorporation of each acid into the different glycerophospholipids was similar to their endogenous percentage profiles, therefore simulating the in vivo situation. The icosaenoic acids then incorporated were liberated from phospholipids when platelets were incubated with thrombin or calcium ionophore A23187 and subsequently oxygenated through the cyclooxygenase and/or lipoxygenase pathway. Whereas 20: 3(n - 6) was readily converted into cyclooxygenase products, 20: 5(n - 3) was more specifically converted into lipoxygenase products, and this latter conversion was comparable to that of 20: 3(n - 9) which is not a prostanoid precursor. Finally, only 20:3(n - 6)- or 20:5(n - 3)-rich platelets exhibited a reduced availability of endogenous arachidonic acid from phospholipids when induced by thrombin. It is concluded that inhibitory polyunsaturated fatty acids (20: 3(n - 6) and 20: S(n - 3)) could act both by reducing prostaglandin H,/thromboxane A, production from endogenous arachidonic acid and in generating platelet inhibitory substances (cyclooxygenase and/or lipoxygenase products of 20: 3(n - 6) and 20: 5(n - 3)). On the other hand, 20:3(n - 9), a fatty acid which potentiates platelet aggregation through its lipoxygenase end product, could produce sufficient amounts of this compound to enhance the aggregation when platelets are triggered with inducers of phospholipase activity such as thrombin or calcium ionophore.

Introduction Great interest arose during the last decade for certain polyunsaturated fatty acids of nutritional Abbreviations: 20: 5( n - 3), 5,8,11,14,17-icosapentaenoic acid; 20: 3(n -6), 8,11,14-icosatrienoic acid; 20: 3(n -9), 5,8,11icosatrienoic acid; 20 : 4(n -6), arachidonic acid; HHT, 12-hyacid; 12-HETE, 12-hydroxydroxyheptadecatrienoic icosatetraenoic acid; PC, phosphatidylcholine; PE, phosphatidylethanolamine; PI, phosphatidylinositol; PS, phosphatidylserine.

0005-2760/85/$03.30

0 1985 Elsevier Science Publishers

B.V.

value which might interact with platelet arachidonic acid (20 : 4( n - 6)) metabolism and then inhibit platelet functions (for recent reviews see Refs. 1 and 2). Among these fatty acids, 5,8,11,14,17icosapentaenoic acid (20 : 5( n - 3)) the trienoic series prostaglandin precursor, was widely studied as a component of fish oil [3]. 20 : 5(n - 3) has been described as a competitive inhibitor of 20 : 4(n - 6) oxygenation by platelet cyclooxygenase [4] and that its own oxygenation is markedly dependent on the presence of peroxides, particularly from arachidonic acid [5,6]. The pre-

53

cursor of the monoenoic series prostaglandins, 8,11,14-icosatrienoic acid (dihomogammalinolenic acid, 20 : 3(n - 6)) is also known as a platelet inhibitor [7], presumably through the formation of of platelet prostaglandin E,, a potent inhibitor aggregation [8]. 20 : 3(n - 6) is a moderate substrate for platelet oxygenases and its oxygenation is slightly increased by 20 : 4( n - 6) [5]. In contrast to these inhibitory fatty acids, 5,8,11-icosatrienoic acid (20 : 3( n - 9)), which does not possess the structural requirement to be a prostaglandin precursor [9] appears to potentiate platelet aggregation through its lipoxygenase end product [lo]. In order to investigate the behaviour of these three icosaenoic acids in platelets, they were incorporated into membrane phospholipids and the cells were aggregated by thrombin or the calcium ionophore A23187. The metabolism of the acids was then studied. The effect of the fatty acid incorporation on 20 : 4(n - 6)oxygenation was also studied. Material and Methods Reagents. 14C-labelled 20 : 4( n - 6) 20 : 3( n - 6) and 20 : 5( n - 3) were purchased from New England Nuclear, Boston. The unlabelled acids, human albumin and thrombin were provided by Sigma, St Louis. The calcium ionophore A23187 was furnished by Boehringer, Mannheim. Thinlayer chromatography was performed on silicagel G plates obtained from Merck, Darmstadt. Organic solvents and other reagents were provided by Prolabo, Paris. Platelet preparation. Human platelets were obtained from normal volunteers who had not taken any drug for at least ten days. They were isolated from their plasma as described previously [ll] and resuspended in a Tyrode/Hepes buffer (pH 7.4) containing 3.5 g/l of fatty acid-free albumin precoated or not (controls) with one of the three fatty acids studied. For this purpose, each acid was dried and incubated overnight at 37°C with Tyrode/ Hepes buffer-albumin under nitrogen. The molecular ratio fatty acid/albumin was 2. After 2 h incubation at 37°C under gentle shaking, platelet suspensions (300000/mm’) were acidified to pH 6.4 with citric acid and centrifuged for 10 min at 700 X g. Pellets were then resus-

pended studies.

into

Tyrode/Hepes

buffer

for

further

Extraction and separation of lipids. Total lipids of platelets which incorporated radiolabelled fatty acids (1.6 Ci/mol) were obtained with a double extraction with 9 volumes of the mixture chloroform/ethanol 2: 1 (v/v) containing 5. lo-’ M butylated hydroxytoluene as an antioxidant [5]. Lipids were separated on thin-layer chromatography (TLC) with the mixture hexane/diethyl ether/acetic acid 80 : 20: :l (v/v) into sterol esters (R, 0.87), triacylglycerols (RF 0.50), fatty acids (RF 0.25), monohydroxy fatty acids (RF 0.10) and phospholipids (R, 0). A second migration with the mixture diethyl ether/ methanol/ acetic acid 90 : 1 : 2 (v/v) allowed the separation of prostanoids into prostaglandins D (RF 0.46), thromboxanes B (RF 0.27), prostaglandins E (RF 0.18) and prostaglandins F, (RF 0.09). Finally, a third elution with the mixture chloroform/ methanol/ acetic acid/water 85 : 15 : 14 : 4 (v/v) separated phospholipids into PE (RF 0.5) PC (RF 0.26) PI + PS (R, 0.15) and sphingomyelin (R, 0.02). After each run a quantitative radiochromatogram was developed. Measurements arachidonic

of

oxygenated

acid. Oxygenation

derivatives

of

of 20 : 4( n - 6) was took up unlabelled

studied on platelets with icosaenoic acids. Exogenous [14C]20 : 4( n - 6) was incubated with platelets for 4 min at 37’C and its three main oxygenated products, thromboxane B, and HHT through the cyclooxygenase pathway, and 12HETE through the lipoxygenase, were measured after TLC separation as described earlier [5]. After thrombin stimulation, thromboxane B, formation from endogenous 20 : 4( n - 6) was determined by capillary gas-liquid chromatography (GLC) as described elsewhere 1121. 1ZHETE was measured by reverse-phase HPLC using 15-hydroxy-8,11,13-icosatrienoic acid as an internal standard [13]. Analysis of fatty acids. After treatment of glycerophospholipids with BF,/ methanol, fatty acid methyl esters were analysed by capillary GLC as described previously [14]. The possibility for an icosaenoic acid to be modified (saturation/ desaturation, elongation/ retroconversion) during its incorporation was also

54

5000

TABLE

22 6n-3

t

I

AMOUNT OF ICOSAENOIC ACID INCORPORATED INTO PLATELETS AFTER 2 h INCUBATION WITH ALBUMIN PRE-COATED WITH EACH ACID Results in nmol/lO’ platelets. Oxygenated products represent the sum of cyclooxygenase products and/or lipoxygenase end product. 20:3(n

0

5

IO

15 x,

25 30 35

40 45

50-65

70

19.9 k 2.8

PC

15.1 +2.5

20:5(n-3)

20:3(n

13.1 + 3.7

16.6f

-9) 1.8

75 80

Fig. 1. HPLC separation of standard polyunsaturated fatty acids. The column (20 cm X 3.2 mm) was packed with Rosil Cl8 HL 3 pm and eluted with 0.017 M acetonitrile/H,PO, (74:26, v/v) at a flow rate of 0.6 ml/min. Fatty acids were detected by their radioactivity measured every 30 s.

investigated. For that purpose, total phospholipids were hydrolysed with porcine pancreas phospholipase A2 and free fatty acids (mainly polyunsaturated fatty acids) were analysed by reversephase HPLC [15] and detected by their radioactivity. A separation profile of standard fatty acids is presented in Fig. 1. Statistics. Results represent the mean f SE. of five determinations obtained on different platelet preparations. The paired t-test was used for comparing the data. Results Incorporation

-6)

Glycerophospholipids

of icosaenoic acids into platelet lipids

The incorporation of 20 : 3(n - 6) 20 : 5(n - 3) or 20 : 3(n - 9) into platelet lipids was monitored by their radioactivity. No significant labelling appeared in either sterol esters or triacylglycerol. The bulk of the radioactivity was located in phospholipids, although a significant amount of each fatty acid remained in its free from. Based upon the specific radioactivity of the acid we found 19.9 5 2.8, 13.1 k 3.7 and 16.6 k 1.8 nmol/lO’ platelets (X * S.E., n = 5) of 20 : 3(n - 6) 20 : 5(n - 3) and 20 : 3(n - 9) respectively, incorporated into phospholipids. In comparison, 3.5 k 1.4, 2.7 + 1.2 and 3.0 k 1.3 nmol of each fatty acid, respectively, remained in its free form (Table I). We also measured the phospholipid enrichment by capillary GLC analysis. The data (not shown)

9.0*

. ..5

11.5i1.2

PE

2.7 * 0.6

1.9f0.6

1.5*0.3

PI+Ps

1.8+0.6

2.0 f 0.7

3.4 k 0.6

Free fatty acid pool

3.5 * 1.4

2.7i1.2

3.0+ 1.3

Oxygenated

1.6f0.5

0.8 k 0.4

0.2 * 0.1

16.6 f 4.4

19.8 * 3.0

Total

products

25.0f

3.8

agreed with those obtained by measuring radioactivity, although the percentage enrichments were slightly lower. In these measurements, the total amounts of fatty acids in control and modified platelets were not different, indicating that no significant synthesis of phospholipids occurred during the enrichment. Taking into account the amount of icosaenoic acid incubated with platelets in the presence of albumin, we can calculate that around 5% of this amount was taken up by the cells. After TLC separation of phospholipids, the radioactivity attached to each of them was calculated. These measurements have shown that sphingomyelin did not incorporate any of the icosaenoic acid. Only glycerophospholipids (mainly PC) were labelled. Table II shows the percentage distribution of each icosaenoic acid in separated glycophospholipids in comparison to that of control platelets, as measured by capillary GLC. Patterns are quite similar with about 70% of the phospholipid 20 : 3( n - 6) 20 : 5( n - 3) or 20 : 3( n - 9) in PC. In contrast, the percentage distribution of endogenous 20 : 4( n - 6) (control platelets) is completely different, with 60% in PE and only 30% in PC. In addition, the possible transformation of each fatty acid into a more or less saturated, longer or

55

TABLE

II

GLYCEROPHOSPHOLIPID OF EACH ACID

DISTRIBUTION

OF SOME ICOSAENOIC

ACIDS IN PERCENTAGE

OF THE TOTAL

AMOUNT

In control platelets, quantities of 20: 3( n - 6) 20 : 5( n - 3), 20 : 3( n - 9) and 20: 4( n - 6) in total phospholipids were 4.0 k 0.6, 1.6 f 0.4, 0.4+0.2 and 91.8+8.2 nmol/lO’ platelets. respectively. In enriched platelets, the percentage distribution of each icosaenoic acid taken up after 2 h incubation of platelets with albumin precoated with each acid was calculated from the results in Table I. Control

20:3(n 20:5(n 20:3(n 20:4(n

-6) -3) -9) -6)

platelets

Enriched

PC

PE

PI+Ps

PC

PE

PI+Ps

15.9 + 3.3 69.1+ 3.4 74.0 f 9.4 29.8kl.l

21.51 2.6 28.4+ 14.2 23.7* 8.7 61.0* 1.8

2.5 f 2.0 1.8k1.5 2.7 * 2.6 9.2kl.l

75.8 k 14.7 68.7 + 19.0 69.2+ 1.7 _

13.5 k4.0 14.5 + 4.9 9.0 + 2.4 _

9.0*3.2 15.2+5.5 20.4 f 4.4 _

shorter acid was investigated. Total phospholipids of modified platelets were treated by phospholipase A, and resulting free fatty acids were separated by HPLC (see Methods). Only 5% of 20: 5(n - 3) was transformed into a fatty acid, tentatively identified as 22 : 5(n - 3) (according to its retention time). In a similar way, 6% of 20 : 3(n - 9) was converted into an unidentified fatty acid. In contrast, no transformation of 20: 3(n - 6) could be seen. Metabolism of icosaenoic acids under platelet stimulation Platelets pre-enriched with radiolabelled 20:3(n-6),20:5(n-3)or20:3(n-9)weretriggered with thrombin (0.1 U/ml) or the calcium

TABLE

platelets

ionophore A23187 (10e6 M) for 4 min at 37°C. TLC analysis of the total lipid extracts obtained from incubates revealed both a decrease of phospholipid content of the fatty acids and the appearance of the corresponding oxygenated derivatives. However, only production of the lipoxygenase products from each icosaenoic acid induced by thrombin or ionophore was statistically significant. 20 : 5(n - 3) and 20 : 3(n - 9) appeared to be converted quickest by lipoxygenase and 20 : 3(n - 6) by cyclooxygenase (Table III). The appearance of these oxygenated products was accompanied by a decrease of the corresponding fatty acid in phospholipids especially PC and/or PI + PS. The PE fraction seemed very stable in this respect.

III

MOBILISATION WITH 0.1 U/ml

OF PREINCORPORATED 20 : 3( n - 6) 20 : 5( n - 3) or 20: 3( n - 9) WHEN THROMBIN OR 1O-6 M CALCIUM IONOPHORE A23187 FOR 4 min

Results in nmol/lO’

platelets 20:3(n

were compared

to controls.

-6)

PLATELETS

WERE TRIGGERED

n.d., not detectable.

20:5(n-3)

20:3(n

-9)

Control

Thrombin

A23187

Control

Thrombin

A23187

Control

Thrombin

19.9 + 2.8

18.2 k 3.7

17.8+2.5

13.1+3.7

13.2k3.5

10.1 k 2.7

16.6+ 1.8

15.1 k1.4

A23187 12.5 * 1.4 ( p < 0.05)

PC PE PI+Ps Free acids Cyclooxygenase products Lipoxygenase products

15.1 f 2.5 2.7 + 0.6 1.8kO.6 3.5 + 1.4 1.2kO.5 0.4 i 0.3

14.6 f 2.3? 1.4 f 4.9&

3.2 0.6 0.2 1.2

2.6 f 0.7 0.7 * 0.3

12.9 + 2.1 3.0*0.3 1.5kO.4 3.5k1.4

9.0 + 2.5 1.9 * 0.6 2.OkO.7 2.1* 1.2

9.6 + 2.5 2.0 j, 0.6 2.0 f 0.5 1.6kO.6

7.1 f 2.0 1.5kO.5 1.3kO.4 1.3+0.5

11.5k1.2 1.5kO.3 3.4+0.6 3.0+ 1.3

10.7*1.1 1.2kO.6 2.6 k 0.4 2.7 k 0.7

9.2kl.l 1.1+0.5 2.0 * 0.3 1.4kO.3

3.lkO.5 2.1 f 0.5 (P < 0.05)

0.3 rt 0.2 0.5 * 0.4

0.5 +0.3 2.2 + 0.8 (P < 0.005)

0.7 f 0.5 5.0+1.0 (P < 0.02)

n.d. 0.2 * 0.1

nd. 2.1 _t 0.4 (P < 0.01)

n.d. 6.0+1.1 (P < 0.01

56

TABLE

IV

OXYGENATION OF EXOGENOUS ACID-RICH PLATELETS

AND

ENDOGENOUS

ARACHIDONIC

ACID

BY CONTROL

OR

ICOSAENOIC

10e5 M exogenous arachidonate was used. Endogenous arachidonate was liberated from phospholipids after stimulation 0.1 U/ml. Thromboxane B,, HHT and 12-HETE are the main oxygenated products or arachidonic acid obtained cyclooxygenase and lipoxygenase pathways, respectively. Results are expressed as nmol/lO’ platelets. Exogenous

arachidonic

Control

20:3(n

-6)

acid

Endogenous 20:5(n

-3)

20:3(n

6.1 +0.6

0.69kO.05

0.48 f 0.05 (P < 0.05)

0.48 k 0.04 (P < 0.05)

0.62 * 0.05

6.7 + 0.9

7.8* 1.2

_

_

_

_

9.4*0.8

9.6* 1.5

4.0 +0.6

2.6 i0.7 (P -z 0.02)

2.3 +0.6 (P < 0.05)

2.7 *0.3

6.7k0.9

5.OkO.9

5.8 kO.9

HHT

8.1 f 1.2

5.8 * 0.7

12-HETE

8.7+ 1.3

10.6 f 1.6

-9)

-6)

acid

Control

Thromboxane B,

20:3(n

arachidonic

by thrombin through the

20:5(n

-3)

20:3(n

-9)

-

Oxygenation of arachidonic acid The results concerning the formation of oxygenated products from exogenous and endogenous 20 : 4(n - 6) can be seen in Table IV. Cyclooxygenase activity, as judged by thromboxane B, and HHT formation from exogenous 20 : 4( n - 6), was not significantly altered in modified platelets, although it was slightly decreased, particularly in 20 : 3( n - 6)-rich platelets. In a similar way, lipoxygenase activity (12-HETE production) was not affected or reciprocally enhanced. Conversely, both cyclooxygenase (thromboxane B,) and lipoxygenase (12-HETE) products were significantly decreased in 20 : 3(n - 6) as well as 20 : 5( n - 3)-rich platelets when triggered with thrombin, suggesting that the thrombin-induced liberation of 20 : 4( n - 6) from phospholipids was altered in such platelets. Discussion When icosaenoic acids are incubated with blood platelets in the absence of albumin or without precoating onto this protein, they are mainly and quickly oxygenated into lipoxygenase and/or cyclooxygenase products, while only a small part is incorporated into phospholipids [16,17]. Involvement of oxygenases on a large scale also results in their alteration [18]. Under our conditions, incubating precoated icosaenoic acids onto albumin (molecular ratios less than 2) the largest part of

the acids was incorporated into glycerophospholipids, whereas only a tiny amount was oxygenated. These conditions allow, therefore, simulation of the physiological situation where platelets may take up circulating free fatty acids bound to serum albumin [19], especially polyunsaturated fatty acids [20,21]. However, a relatively high amount of fatty acids taken up by platelets remained in the free fatty acid pool as compared to literature data concerning arachidonic acid. This agrees with the higher ratio of free 20 : 3(n - 6)/phospholipid 20 : 3( n - 6) to 20 : 4( n - 6) in resting platelets [22] and with the occurrence of a specific fatty acyl-CoA synthetase for arachidonic acid in human platelets [23]. Besides, the glycerophospholipid repartition of icosaenoic acids taken up was quite similar to . that of the corresponding endogenous acids. Our results also pointed out that this distribution is different from that of 20 : 4(n - 6). It is noteworthy that platelet PE contains the bulk of total 20 : 4( n - 6) [24]. A speculative explanation would be that the plasmalogen form of PE, which is known to have a high content of 20 : 4(n - 6) [25], is devoid of other icosaenoic acids. On the other hand, the very little chain transformation (if any) of the fatty acids studied confirms the weak capacity of blood platelets in fatty acid biosynthesis [2]. When incorporated into phospholipids, icosaenoic acids were available to phospholipase activity induced by thrombin or the calcium ionophore. Oxygenated derivatives 20 : 3( n - 6) 20 : 5( n - 3)

57

or 20 : 3( n - 9) were produced at the expense of both phospholipids and the free fatty acid pool, although this reduction was not significant. However, as far as 20 : 3(n - 6) was concerned, such reduction could only be seen in PC. Besides, PE appeared as a very poor source for each fatty acid studied, confirming previous findings [26]. It was most interesting to find that production of oxygenated derivatives of the icosaenoic acids was quite efficient. This confirms our previous data, showing that arachidonic acid may markedly enhance platelet oxygenase activities towards other icosaenoic acids [5]. Under our conditions, the liberation of 20 : 3( n - 6), 20 : 5(n - 3) or 20 : 3( n - 9) was obviously accompanied with that of endogenous 20 : 4( n - 6). In our previous work [5] we found that 20 : 4( n - 6) stimulated the conversion of 20 : 3( n - 6) into cyclooxygenase products. This fits well with the consistent formation of such products from 20 : 3( n - 6)-rich platelets. Similarly, we reported that 20 : 4( n - 6) markedly increases the oxygenation of 20 : 5(n - 3), particularly through the lipoxygenase pathway. The high production of lipoxygenase products of 20 : 5( n 3) when 20 : 5( n - 3)-rich platelets were treated with thrombin or the calcium ionophore, is in good agreement with that point. Finally, we have recently mentioned that 20 : 3(n - 9) is quite a good substrate of lipoxygenase in intact platelets [27] which also fits very well with the high formation of the lipoxygenase product from this acid. Our data also indicate that platelets enriched with20:3(n-6),20:5(n-3)or20:3(n-9)normally converted exogenous 20 : 4( n - 6) into both cyclooxygenase and lipoxygenase products, suggesting that the oxygenase activities in such modified platelets were not altered. In contrast, the formation of the same oxygenated products from endogenous 20 : 4( n - 6) was significantly reduced in 20 : 3(n - 6) or 20 : 5(n - 3)-rich platelets but not in 20 : 3( n - 9)-rich platelets. Altogether these results strongly suggest that enriching platelets with 20 : 3(n - 6) or 20 : 5(n - 3) decreased the availability of endogenous 20 : 4(n - 6) when induced by thrombin. Similar results have already been published for after platelet enrichment with linoleic acid [28]. Since no difference could be observed in the phospholipid distribution of 20 : 3( n - 6), 20 : 5( n - 3) or 20 : 3(n - 9) in mod-

ified platelets, it is unlikely that their presence in phospholipids induces only a decreased ‘ phospholipase activity’ in 20 : 3( n - 6) and 20 : 5( n 3)-rich platelets. However, because 20 : 3( n - 6) and 20 : 5( n - 3) but not 20 : 3(n - 9) are prostaglandin precursors, we may speculate that their cyclooxygenase products could be responsible for such a decreased ‘phospholipase activity’. On the other hand, previous findings [13] showed that the platelet lipoxygenase product of 20 : 5( n - 3) may potently inhibit prostaglandin Hz-induced platelet aggregation, whereas 20 : 3( n - 9) could potentiate this aggregation through its lipoxygenase end product [lo]. We may then conclude that 20 : 3( n - 6) and 20 : 5( n - 3) could inhibit platelet functions by both producing inhibitory oxygenated derivatives (cyclooxygenase and/or lipoxygenase products) and reducing the availability of endogenous 20 : 4(n - 6). In contrast, 20 : 3( n - 9) might potentiate platelet aggregation by generating its lipoxygenase product, while prostaglandin Hz/ thromboxane A 2 formation from endogenous 20 : 4( n - 6) would not be altered. Acknowledgements This work was funded by INSERM (CRL 825043). We are indebted to Professor H. Sprecher (Columbus, Ohio) for providing unlabelled and [i4C]20 : 3( n - 9). References 1 Willis, A.L. (1981) Nutr. Rev. 39, 289-301 2 Goodnight, S.H., Harris, W.S., Connor, W.E. and Illingworth, D.R. (1982) Arteriosclerosis 2, 87-113 3 Dyerberg, J., Bang, H.O., Stofferson, E., Moncada, S. and Vane, J.R. (1978) Lancet ii, 117-119 4 Needleman, P., Raz, A., Minkes, M.S., Ferrendelli, J.A. and Sprecher, H. (1979) Proc. Natl. Acad. Sci. USA 76, 944948 5 Boukhchache, D. and Lagarde, M. (1982) B&him. Biophys. Acta, 713, 386-398 6 Morita, I., Takahashi, R., Saito, Y. and Murota, S. (1983) J. Biol. Chem. 258, 10197-10199 7 Silver, M.J., Smith, J.B., Ingerman, C. and Kocsis, J.J. (1973) Prostaglandins 4, 863-875 8 Kloeze, J. (1969) Biochim. Biophys. Acta 187, 285-292 9 Raz, A., Minkes, M. and Needleman, P. (1977) Biochim. Biophys. Acta 488, 305-311 10 Lagarde, M., Burtin, M., Sprecher, H., Dechavanne, M. and Renaud, S. (1983) Lipids 18, 291-294

58

11 Lagarde, M., Bryan, P.A. Guichardant, M. and Dechavanne M. (1980) Thromb. Res. 17, 581-588 12 Lagarde, M., Vericel, E., Guichardant, M. and Dechavanne, M. (1981) B&hem. Biophys. Res. Commun. 99, 1398-1402 13 Croset, M. and Lagarde, M. (1983) Biochem. Biophys. Res. Commun. 112, 878-883 14 Guichardant, M. and Lagarde, M. (1983) J. Chromatogr. 275, 400-406 15 Van Rollins, M., Aveldano, ML, Sprecher, H. and Horrocks, L.A. (1982) Methods Enzymol. 86, 518-530 16 Lagarde, M., Gharib. A. and Dechavanne, M. (1977) Biochimie 59, 935-937 17 Boukhchache, D. and Lagarde, M. (1982) Biochem. Sot. Trans. 10, 237-238 18 Lands, W.E.M. (1984) Prostaglandin Leukotriene Med. 13, 35-46 19 Spector, A.A. (1975) J. Lipid Res. 16, 165-179 20 Bills, T.K., Smith, J.B. and Silver, M.J. (1976) B&him. Biophys. Acta 424, 303-314

21 Schick, B.P., Schick, P.K. and Chase, P.R. (1981) Biochim. Biophys. Acta 663, 239-248 22 Lagarde, M., Guichardant, M. and Dechavanne, M. (1981) Progr. Lipid Res. 20, 439-443 23 Wilson, D.B., Prescott, S.M. and Majerus, P.W. (1982) J. Biol. Chem. 257, 3510-3515 24 Marcus, A.J. (1978) J. Lipid Res. 19, 793-826 25 Mahadavappa, V.G. and Holub, B.J. (1982) Biochim. Biophys. Acta 713, 72-79 26 McKean, M.L., Smith, J.B. and Silver, M.J. (1981) J. Biol. Chem. 256, 1522-1524 27 Lagarde, M., Croset, M., Boukchache, D., Greffe, A., DeM. and Renaud, S. (1984) Prostaglandin chavanne, Leukotriene Med. 13, 61-66 28 Needleman, S.W., Spector, A.A. and Hoak, J.C. (1982) Prostaglandins 24, 607-622