Neurochemical Research, VoL 17, No. 2, 1992, pp. 167-172

Dietary Manipulation with High Marine Fish Oil Intake of Fatty Acid Composition and Arachidonic Acid Metabolism in Rat Cerebral Microvessels

Jfinos Kdlmdn ~, Arpfid Gecse 2, Tibor Farkas 3, Ferenc J064, Gyula Telegdy 2, and Abel Lajtha 5

(Accepted June 25, 1991)

Male weanling Wistar rats were maintained on one of two semisynthetic diets, differing only in the type of oil used: (i) 10% by weight marine fish oil (MFO group) containing 20% eicosapen- taenoic acid (EPA) and 17% docosahexaenoic acid (DHA), or (ii) 10% by weight sunflower oil (SFO group). The control group was kept on standard diet for 4 weeks. Blood-free microvessels were isolated from brain cortex by a rapid micromethod, and their fatty acid composition was determined by gas chromatography. It was found that the proportion of n-3 fatty acids (including EPA and DHA) increased significantly in the microvessels of the MFO group, accompanied by a decrease of the n-6 fatty acid series. The changes in fatty acid composition of endothelial cells were not significant in the SFO group in comparison to the control. The amounts of lipoxygenase and cyclooxygenase metabolites were determined. Dietary fish oil decreased the percentage of total products of arachidonate by 50%, while the SFO diet had no effect on it. The amount of lipoxy- genase products in the MFO group decreased significantly from 16931 --- 3131 dpm to 6399 +_ 357 dpm/300 mg wet weight of brain. Significantly less PGF-I,, PGF-2,~ and 12-hydroxyhepfa- decatrienoic acid (HHT) were found in the capillaries of MFO treated animals, in comparison to the SFO group. The ratios of vasoconstrictor and vasodilator metabolites of arachidonate cascade were not modified by the diets. Our results suggest that fish oil diet reduces the arachidonate cascade in cerebral microvessels. This effect may explain for the efficiency of n-3 fatty acids in vascular diseases.

KEY WORDS: Fish oil; diet; arachidonic acid cascade; cerebral endothelial cells; polyunsaturated fatty acids.

INTRODUCTION

The role of metabolites of the arachidonate cascade in the pathogenesis of cerebral ischemia has been em-

1 Department of Neurology and Psychiatry, 2 Department of Pathophysiology Albert Szent-Gy6rgyi Medical Uni-

versity, Szeged, Hungary, 3 Institutes of Biochemistry and 4 Biophysics, Biological Research Center, Hungarian Academy of Sci-

ences, Szeged, Hungary 6701, P.O.B. 521. Center for Neurochemistry, Division of the N.S. Kline Institute for Psychiatric Research, Orangeburg, New York 10952

167

phasized earlier (17, 30). After the microcirculatory stasis, during the time of reperfusion vasoactive products of cyclooxygenase (Co) enzymes, such as PGI-2, TXA-2 and different hydroperoxy and hydroxy derivatives of the lipoxygenase (Lo) pathway are formed which then initiate the opening of the blood-brain barrier (8).

Eicosapentaenoic acid (EPA), the major fatty acid component of marine fish oils (MFO), is known to be a potent inhibitor of diene prostaglandin formation. In the endothelial cells and platelets, EPA competitively inhib- its the production of PGI-2, TXA-2, and leukotrienes from arachidonic acid, while it is converted to trien me-

0364-3190/92/0200-0167506.50/0 �9 I992 Plenum Publishing Corporation

168 K~ilm~in, Gecse, Farkas, Jo6, Telegdy, and Lajtha

tabolites (PGI-3, TXA-3, LTA-5, LTB-5, etc.) in the same enzymatic pathway.

The alterations of the arachidonate cascade after high dietary intake of MFO have been studied in the vascular cells of different organs; however, only a little attention has been focused on the effect of EPA-rich diet on the cerebral microvasculature.

It has been shown previously that the phospholipids of rat brain capillary endothelial cells contain relatively high amounts of n-6 and n-3 fatty acids (27, 28). Ho- mayoun et al. (13) found a compensatory increase in the amount of n-6 fatty acids in rat brain capillary endothe- lial ceils after a diet, deficient in n-3 fatty acids. Brown et al. (6) reported replacement of n-6 by n-3 fatty acids, and decreased synthesis of PGI-2 and PGF-2~ after di- etary manipulation with linseed oil (rich in 18:3 n-3) in rats. Recently, Yerram et al. (31) provided evidence for reduced PGI-2 and PGF-2,~ formation of cultured and isolated mouse brain capillary endothelial cells after preincubation with EPA and docosahexaenoic acid (DHA), respectively. Knapp et al. (20) suggest that there are differences in the formation of PGI-2 and PGI-3 by en- dothelial cells, depending on the mode of administration (in vivo or in vitro).

However, it remained to be seen if the fatty acid and the arachidonic acid composition of the cerebral mi- crovessels could be influenced by the dietary intake of MFO. The aim of our present study was to examine the possible changes in the cerebral microvasculature after in vivo administration of MFO. We report here on the replacement of n-6 by n-3 fatty acids, and, the ~ltered production of cycloxygenase and lipoxygenase mktabo- lites of the arachidonate cascade in the capillary endo- thelial cells of rats after high dietary intake of MFO.

EXPERIMENTAL PROCEDURE

EPA rich fish oil was generously provided by J.C. Marstens G.C.; Bergen, Norway. [1-14CJarachidonic acid (spec. act. 2035 MBq/mM) was obtained from Amersham (England). TC Medium 199 was pur- chased from DIFCO Laboratories, Detroit, MI (USA). Silica gel G thin-layer plates (0.25 mm, art. no. 5721) were obtained from Merck AG, Darmstadt (FRG). PGE-2, PGD-2, TXB-2, PGF-2,~, and 6-oxo- PGF-L, (the stable metabolite of PGI-2) were generously provided by Dr. J. E. Pike, Upjohn Co. Kalamazoo, MI (USA). Hydroxyeicosa- tetraenoic acid (HETE) standards were purchased from Calbiochem, La Jolla, CA (USA). Chemicals were of analytical grade and obtained commercially unless otherwise stated.

Animals and Diets. Male weanling Wistar rats (5 weeks old) were divided into three groups (30 rats in each). Two groups were main- tained on one of two semisynthetic diets (composition: dry weight 86%, starch 69%, protein 20.8%, fat 4.7% and fibers 4.1%) differing

only in the type of oil used: 10% (by weight) fish oil (MFO), or 10% (by weight) sunflower oil (SFO) for four weeks. The third (control) group (CO) was kept on the standard rat chow (Lati, Budapest, Hun- gary).

The fatty acid compositions of the diets are shown in Table I. The experimental diets were prepared fresh daily, using oils stored at -20~C. Diets and oils were periodically checked for deterioration by determination of peroxide content (18). The Vitamin E content of the diet was 1.0 I.U./g lipid. The food and water were given ad libitum, and the diurnal light cycle of 12 hours was maintained.

Isolation of Microvessels from Brain Cortex. The brains of the rats were perfused with 0.9% NaCI (containing EDTA at 5.8 mM) under light ether anesthesia, to eliminate blood from the central ner- vous system. Blood-free brain microvessels were isolated by a micro- method of Hwang et al. (14) with modifications. Briefly the cerebral cortex was freed of myelin and pial membranes. The chopped tissue (900 mg wet weight) was homogenized with a teflon-glass homoge- nizer in 9 volumes of 3 mM HEPES buffer (containing 0.32 M sucrose, pH 7.3) with 4 strokes (1000 revolution/rain), then centrifuged at 1000 g for 10 rain at 2 ~ The pellet was resuspended in 5 volumes of 3 mM buffer homogenized with 2 strokes (1000 revolutions/rain), then centrifuged again at 1000 g for 10 min at 2 ~ The pellet was resus- pended in the above mentioned HEPES-sucrose buffer, centrifuged at 100 g for 15 sec, then diluted with 1 volume of 3 mM HEPES buffer and followed by centrifugation at 100 g for 15 sec. The supernatants of the last two centrifugations were collected and centrifuged at 200 g for I rain. The pellet was resuspended in 2.5 volumes of TC Medium 199 (pH 7.4) and centrifuged at 200 g for 1 rain. The pellet, which contained the microvessels, was resuspended in TC Medium 199. An aliquot was examined for purity under light and phase-contrast micro- scopes and for viability using the trypan blue exclusion test. The con- tamination was less than 6% and the microvesszls were free of blood cells. More than 90% of the endothelial cells were viable.

Lipid Extraction and Fatty Acid Analysis. Lipids were extracted from cerebral capillaries according to the method of Folch et al. (9). Fatty acids were transmethylated, and fatty acid methyl ester analysis was performed with a Hitachi 263-80 gas chromatograph equipped with a hydrogen flame detector connected to a Hitachi M263 data

Table I. Fatty Acid Composition of Dietary Lipids (% of the total)

Fatty acid Standard Marine Fish Oil Sunflower Oil

14:0 1.3 6.6 - - 16:0 22.4 7.1 6.0 16:1 3.1 9.9 0.7 18:0 9.1 3.3 4.9 i8:1 32.6 17.3 16,3 18:2 n-6 25.7 1.2 63.2 20:1 n-9 - - 4.9 - - 20:4 n-6 0.4 0.7 - - 20:5 n-3 - - 19.7 - - 22:4 n-6 - - 0.8 - - 22:5 n-3 - - 1.9 - - 22:6 n-3 - - 17.3 - - others (incl. 20:3 n-6) 5.4 9.3 8.9 Total n-3 - - 41.6 - - Total n-6 26.1 2.7 63.2 n-3/n-6 - - 14.4 - -

Dietary Manipulations in Cerebral Microvessels 169

processor. The stainless steel columns (6 feet long, 3 mm ID) were filled with 15% carbowax 20 M on Supelcoport, 100/120 mesh (Su- pelco, Bellefonte, PA). The temperature was programmed from 145~ to 200~ at 1~ The fatty acid methyl esters were identified by comparison of their relative retention times with those of known methyl esters. Peak areas were corrected for detector response obtained from pure 16:0, 18:1, 20:4, 20:5 and 22:6.

Analysis of Eiocosanoids. The microvessels were preincubated in TC Medium 199 at 37~ for I0 min. The enzyme reaction was started by the addition of [14C]arachidonic acid (3.7 kBq; in 10 Izl ethanol) as tracer substrate to the incubation mixture (1.0 ml). Ten minutes later the enzyme reaction was quenched by bringing the pH to 3.0 with formic acid. The samples were extracted twice with 3 volumes of ethyl acetate. The organic phases were pooled and evaporated to dryness under nitrogen. The residues were reconstituted in 150 ~1 ethyl acetate and quantitatively applied to silica gel thin-layer plates. The plates were developed to a distance of 15 cm in an organic phase of ethyl acetate : acetic acid : 2,2,4 trimethylpentane : water (110:20:30:100) using an over pressure thin-layer chromatograph (Chrompres, Newman Howells, Winchester UK.). Each 3 mm band of the chromatograms was scraped off and the radioactivity was de- termined in a Beckmann LS 1800 Liquid Scintillation Counter, using 5 ml of toluene containing 0.4% W/V PPO, 0.02% W/V POPOP, and 10% V/V ethanol. The radiolabelled products of arachidonic acid were identified with unlabeled authentic standards, which were detected with anisaldehyde reagent (18).

High Performance Liquid Chromatography. Reversed-phase high- performance liquid chromatography (HPLC, Isco Mod. 2350) was per- formed on a column (4.6 x 250 mm) packed with Liehrosorb C18 (7- urn particles). For the identification of HETEs both UV detection and liquid scintillation were used. The solvent used for the separation of HETEs was a mixture of 255 mM phosphoric acid in 70% acetonitrile. Monitoring was at 235 nm with a 308 UV detector (Labor MIM Hun- gary), integration by Hewlett-Packard integrator (HP 3396 A).

Statistical analysis was performed by Student's t-test.

RESULTS

Fatty Acid Composition of the Rat Brain Microves- sels. The fatty acid profile of the capillary endothelial ceils of rats fed on different diets is shown in Table II. The relative amount of n-3 fatty acids (20:5 n-3, 22:5 n-3, and 22:6 n-3, respectively) increased slightly but significantly in the brain microvessels of the marine fish oil diet group. The increase of n-3 polyenes was accom- panied by a decrease of n-6 fatty acids, 22:4 n-6 and 22:5 n-6, while the level of 20:4 n-6 remained unaf- fected. The n-3/n-6 ratio significantly increased in the MFO-treated animals, but the sum of the n-3 and n-6 polyunsaturated fatty acids remained unchanged in all groups. We found no significant differences in the fatty acid composition after sunflower oil treatment of the animals.

Arachidonic Acid Metabolism. There was a marked reduction (50%) in the amount of total [14C]arachidonate

Table II. Fatty Acid Composition of Cerebral Capillaries of Rats Fed on Different Diets

Marine Fatty acid Control Fish Oil Sunflower Oil

14:0 6.85• 6.14• 6.52• 14:1 0.83_-2-0.15 0.74---0.09 0.70-+0.09 16:0 23.16-+1.13 22.16• 22.96-+1.79 16:1 n-9 0.83-+0.09 0.95• 0.85-+0.09 16:1 n-7 1.33---0.12 1.22-+0.10 1.39-+0.14 18:0 13.31-+0.72 13.87-+0.83 12.73-+0.79 18:1 n-9 17.85-+1.39 16.34• 18.11• 18:2 n-6 3.46 • 0.23 3.25 • 0.27 3.93 • 0.22 20:0 1.26--_0.28 1.01-+0.27 1.35-+0.12 20:1 n-9 1.05-+0.07 0.93-+0.12 1.12• 20:4 n-6 14.79• 13.65• 15.17• 20:5 n-3 n.d. 0.80• n.d. 22:4 n-6 1.50-+0.11 1.01-+0.13" 1.59-+0.08 22:5 n-6 0.76• 0.46-+0.05* 1.13-+0.14 22:5 n-3 0.51• 0.89-+0.07* 0.44-+0.09 22:6 n-3 6.87-+0.52 9.28• 5.73•

n-3 (total) 7.19• 10.97• 7.31• n-6 (total) 20.31• 18.37-+2.15 21.72-+2.59 n-3/n-6 0.34-+0.02 0.52• 0.32• n-3+n-6 27.05-+3.07 29.13-+3.25 28.89-+1.61

Data are mean • SEM, and expressed as a percentage of the identified fatty acids. Each experiment was repeated 3 times. One sample is a pooled fraction of 7 animals, n.d. = not detectable. The statistical comparison was made between the marine fish oil and sunflower oil groups. *p<0.05; **p<0.02;

metabolites in the microvessels of the high dietary EPA group (Table III). The total amount of cycloxygenase metabolites decreased significantly from 6334 + 689 dpm/10 min/300 mg wet weight of brain cortex, to 4622 + 407 dpm/10 min/300 mg wet weight of brain cortex. Marked reduction of lipoxygenase metabolites was found in the microvessels prepared from the MFO-treated group (6399 _ 1358 dpm/10 min/300 mg wet weight of brain cortex) compared to the control value (16931 _ 3131 dpm/10 min/300 mg wet weight of brain cortex). Sig- nificantly less 6-keto PGF-I,~, PGF-2,~, and 12-hydrox- yheptadecatrienoic acid (HHT) was found in the capillaries of MFO-treated animals than that comparing to those of found in the sunflower oil group. Non-converted [14C]- 20:4, n-6 was measured together with the metabolites in each experimental group and recovered arachidonic acid was found to be increased correspondingly in the MFO- treated group.

The ratios of vasoconstrictor and vasodilator me- tabolites of the arachidonate cascade were not modified by the diets. The animals fed on sunflower oit showed no significant differences in the production of different cycloxygenase and lipoxygenase metabolites compared to the control group.

�9 170 Kfilmfin, Gecse, Farkas, Jo6, Telegdy, a n d L a j t h a

Table III. The Effect of Different Diets on the Arachidonic Acid Metabolites in Rat Cerebral Microvessels

Marine Control % Fish Oil % Sunflower Oil %

Total prod. 22907 • 3873 (100) 11021 __ 1747" (100) 20104 • 2448 (100) 6-keto PGF-I~ 803 -- 56 3.5 566 - 92* 5.1 909 --- 84 4.5 PGF-2,, 839 • 80 3.6 583 • 70* 5.3 770 • 55 3.8 TXB-2 1011 • 129 4.4 921 • 68 8.3 1176 • 147 5.8 PGD-2 966 • 152 4.2 939 • 101 8.5 1271 • 54 6.3 PGE-2 703 • 84 3.0 542 • 83 5.0 858 • 91 4.2 HHT 1846 • 410 8.0 1072 • 247* 9.7 1554 • 334 7.7 HPETE; HETE 16573 • 3131 72.2 6399 • 1358'* 58.0 3526 • 2280 67.2 Co. 6334 • 689 27.6 4622 • 407* 41.9 6578 • 351 32.7 Lo./Co. 2.50 • 0.30 1.33 • 0.16"* 2.03 • 0.30 C/D 0.76 -4- 0.04 0.78 • 0.10 0.64 • 0.05

Each value represents the mean-SEM of 8 determinations. The quantities are expressed: dpm/10 min/300 mg wet weight of brain. HHT = 12-hydroxyheptadecatrienoic acid; HPETE = hydroperoxyeicosatetraenoic acid; HETE = hydroxyeicosatetraenoic acid; Lo. is the sum of the lipoxygenase products (HPETE + HETE). Co. is the sum of the cycloxygenase products; C and D = are the sums of arachidonate metabolites with vasocontrictor (PGF-2 alpha, TXB-2), and vasodilatator (6-keto PGF-1 c~, PGD-2, PGE-2) effects, respectively. *p<0.05; **p<0.01;

DISCUSSION

It has been generally accepted (5, 15) that in most vertebrates the cerebral endothelial cells represent the cellular analogue of the blood-brain barrier. Metabolic coupling of the main functions of cerebral microvessels to the needs of the nerve and glial cells has been de- scribed (25) and certain characteristics of the molecular processes underlying the regulation of vascular tone and permeability have been noted (7). Bourre et al. (3) dem- onstrated that linoleic acid deficiency causes a reduction in content of polyunsaturated fatty acids of the n-3 series in neurons, astrocytes, oligodendrocytes, and brain frac- tions such as myelin and synaptosomes. The manipula- tion of dietary essential fatty acids has been shown to result in changes on the composition of acyl groups of glycerophospholipids in isolated brain microvessels (6, 13). The results of our present study clearly show that dietary ingestion of marine fish oil containing high amounts of EPA and DHA for a relatively short period (4 weeks) can result in a significant increase of 20:5 n- 3; 22:5 n-3, 22:6 n-3, and a decrease in the level of n-6 fatty acids of the C-22 series in the capillary endothelial cells isolated from the rat brain. It can be assumed that the absence of changes in fatty acid com- position after SFO-treatment is due to the short period of time (4 weeks), in contrast to the results of Homayoun et al. (13), who used animals fed through generations with SFO soyabean oil. On the other hand, the arachi- donate production remained unaffected as inferred from fatty acid composition data. In fa~t, the observed changes can be interpreted as a reciprocal replacement of n-6 by

n-3 fatty acids, of the C-22 series similar to that occur- ring in other cells (4, 11, 24).

As we mentioned, the arachidonic acid composition of dietary lipids was similar in control and MFO-treated groups. Therefore, it is not surprising that the fatty acid composition of brain capillaries isolated from rats fed on different diets was also similar. This indicates that the in vivo uptake of arachidonic acid was not modified by dietary manipulation. Consequently, the observed atten- uation in the production of arachidonate cascade can not be interpreted as a reduction of arachidonic acid uptake. In the interpretation of elevated DHA level, it should be remembered that DHA inhibits the cycloxygenase activ- ity (12). In addition, DHA has an important role in main- taining membrane fluidity and permeability, and the function of several membrane bound-enzymes, such as Ca 2+ ATP-ase, Na+-K + ATP-ase, and phosphatidyle- thanolamine methyltransferase (22, 23). So it is reason- able to speculate that the increased level of DHA may alter some transport processes of the blood-brain barrier.

This paper shows, in agreement with Moore et al. (21); Gecse et al. (10); Yerram et al. (31), that the ce- rebral microvessels isolated from the control animals produce much more Lo. than Co. products. The greater ratio of Lo. was found to be characteristic of the isolated microvascular endothelium, while larger vessels produce predominantly PGI-2, a cycloxygenase product (29). This paper also shows that the amount of 6-keto PGF-I,~, PGF-2,~, and HHT decreased significantly in the micro- vessels upon feeding fish oil to the animals. The levels of the other cycloxygenase products detected, PGD-2, PGE-2 and TXB-2, remained unchanged. These are the

Dietary Manipulations in Cerebral Microvessels 171

first data to demonstrate the effect of dietary adminis- tration of fish oil on arachidonate metabolism of brain capillary endothelial cells. Recently, Yerram et al. (31) found decreased formation of PGI-2 and PGE-2 in cul- tured and isolated mouse brain capillary endothelial cells in vitro after incubating them with EPA.

We should like to emphasize the marked, 42% re- duction in the amount of HHT in the microvessels of marine fish oil treated animals. HHT was found to be an important regulator of PGI-2 synthesis (26), so its reduction would contribute to the decreased prostacyclin formation.

The lipoxygenase products of arachidonic acid are, in general, known to increase vascular permeability (1, 16). Microvessels isolated from fish oil treated animals produce 60% less of the products of Lo., compared to those of the control and sunflower treated groups.

Our observation, namely the decreased production of the Lo. compounds could provide an explanation for the results of Black et al. (2), who had found that the same diet prevented the development of post-ischemic cerebral edema.

The present results indicate that a relatively short treatment with dietary marine fish oil causes a reciprocal replacement of n-6 by n-3 fatty acids, and a decreased formation of Co. and Lo. products of arachidonate cas- cade in the brain capillary endothelial cells of rats. It is conceivable that these alterations affect the functions of the blood-brain barrier system under normal and patho- logical circumstances, but further investigations are needed to clarify this important point.

ACKNOWLEDGMENTS

This work was partly supported by OTKA Committee (Proj. Number: 903/90) and by the Scientific Research Council, Ministry of Public Welfare, Hungary (Proj. Number: T-04, T-130). The authors are grateful to Mrs. Zsuzsanna Gonda for the excellent secretarial help. Our thanks are due to J.C. Martens, A.S. (Bergen, Norway) for sup- plying fish oil (Activepa-30) for the experiments.

REFERENCES

1. Asano, T., Gotoh, O., Koide, T., and Takarura, K. 1985. Isch- emic brain edema following occlusion of the middle cerebral ar- tery in the rat. II: Alterations of the eicosanoid synthesis profile of brain microvessels. Stroke 16:110-113.

2. Black, K. L., Hoff, J. T., Radin, N. S., and Deshnach, G. D. 1984. Eicosapentaenoic acid: Effect on brain prostaglandins, ce- rebral blood flow and edema in ischemic gerbils. Stroke 15:65- 69.

3. Bourre, J. M, Pascal, G., Durand, G., Masson, M., Dumont, O., and Piciotti, M. 1984. Alteration in fatty acid composition of rat

brain cells (neurons, astrocytes, oligodendrocytes), and of sub- cellular fractions (myelin and synaptosomes) induced by a diet devoid of n-3 fatty acids. J. Neurochem. 43:342-348.

4. Bourre, J. M., Bonneil, M., Dumont, O., Piciotti, M., Naltone, G., and Lafont, H. 1988. High dietary fish oil alters the brain polyunsaturated fatty acid composition. Biochim. Biophys. Acta 960:458--461.

5. Bradbury, M. W. B. 1979. The Concept of a Blood-Brain Barrier. Wiley, Chichester, England.

6. Brown, M. L., Marshal, L. A., and Johnston, P. A. 1984. Al- teration in cerebral and microvascular prostaglandin synthesis by manipulation of dietary essential fatty acids. J. Neurochem. 43:1392-1400.

7. Eisenberg, H. M., and Suddith, R. L. 1980. The cerebral micro- vasculature. Adv. Exp. Med. Biol. 131, Plenum Press, New York.

8. Feuerstein, G., and Hallenbeck, J. M. 1987. Leukotrienes in health and disease. FASEB J. 1:186-192.

9. Folch, I., Lees, M. and Sloane-Stanley, G. W. 1957. Simple method for the isolation and purification of total lipids from animal tissues. J. Biol. Chem. 226:497-503.

10. Gecse, ,/~., Ottlecz, A., Mezei, Zs., Telegdy, G., Jo6, F., Dux, E., and Karnushina I. 1982. Prostacyclin and prostaglandin syn- thesis in isolated brain capillaries. Prostaglandins 23:287-297.

11. Gibson, R. A., McMurchie, E. J., Chamocu, J. S., and Knee- bone, G. M. 1984. Homeostatic control of membrane fatty acid composition in the rat after dietary lipid treatment. Lipids 19:942- 951.

12. Hadjiagapiou, C., and Spector, A. A. 1987. Docosahexaenoic acid metabolism and effect on prostacyclin production in endo- thelial cells. Arch. Bioch. Biophys. 253:1-12.

13. Homayoun, P., Durand, G., Pascal, G., and Bourre, J. M. 1988. Alterations in fatty acid composition of adult rat brain capillaries and chloroid plexus induced by a diet deficient in n-3 fatty acids: slow recovery after substitution with a nondeficient diet. J. Neu- rochem. 51:45-48.

14. Hwang, S. M., Weiss, S., and Segal, S. 1980. Uptake of L- psS]cystine by isolated rat brain capillaries. J. Neurochem. 35:41% 424.

15. Jo6, F. 1986. New aspects to the function of the cerebral endo- thelium. Nature 321:197-198.

16. Jo6, F. 1987. A unifying concept on the pathogenesis of brain oedemas. Neumpathol. Appl. Neurobiol. 13:161-176.

17. Jo6, F. and Klatzo, I. 1989. Role of cerebral endothelium in brain oedema. Neurol. Res. 11:67-75.

18. Kaufmann, H. P. 1958. Analyse der Fette und Fettprodukte. Band II. pp. 1293-1298. Springer Verlag.

19. Kiefer, H. C., Johnson, C. R., and Arora, K. L. 1975. Colori- metric identification of prostaglandins in subnanomole amounts. Anal. Biochem. 68:336---340.

20. Knapp, H. R., and Salem, N. 1989. Formation of PGI3 in the rat during dietary fish oil supplementation. Prostaglandins 38:509- 521.

21. Moore, S. A., Figand, P. H., Spector, A. A., and Hant, M. N. 1989. Brain microvessels produce 12-hydroxyeicosatetraenoic acid. J. Neurochem. 53:376--382.

22. Neuringer, M., Anderson, G. I., and Connes, W. E. 1988. The essentiality of n-3 fatty acids for the development and function of the retina and brain. Ann. Rev. in Nutr. 8:517-541.

23. Neuringer, M., and Connor, W. E. 1986. n-3 fatty acids in the brain and retina: evidence for their essentiality. Nutr. Rev. 44:285- 294.

24. Mahadevappa, V. G., Ackman, R. G., and Holub, B. J. 1987. Ingestion of fish oil on a derived n-3 fatty acid concentrate con- taining Eicosapentaenoic acid (EPA) affects fatty acid composi- tions of individual phospholipids of rat brain, sciatic nerve and retina. J. Nutr. 117:1663-1670.

25. Raichle, M. E. 1980. Cerebral blood flow and metabolism in man: past, present and future. TINS 10: VI-X.

26. Sadowitz, P. D., Yamaja Setty, B. N., and Stuart, M. 1987. The

172 Kfilm~in, Geese , F a r k a s , Jo6 , T e l e g d y , a n d L a j t h a

platelet cyclooxygenase metabolite 12-L-Hydroxy-5,8-10-hepta- decatrienoic acid (HHT) may modulate primary haemostasis by stimulating protacyclin production. Prostaglandins 34:749-759.

27. Selivonchick, D. P., and Roots, B. I. 1976. Lipid and fatty acid composition of rat brain capillary endothelia isolated by new tech- nique. Lipids 12:165-169.

28. Selivonehick, D. P. and Roots, B. I. 1978. Isolated brain capillary endothelia: influence of various levels of essential fatty acids on the acyl group composition of glycerophospholipids. Lipids 14:66-- 69.

29. Spector, A. A., Hoak, J. C., Fry, G. L., Denning, G. M., Stoll, L. L., and Smith, J. B. 1980. Effect of fatty acid modification on prostacyclin production by altered human endothelial cells. J. Clin. Invest. 65:1003-1012.

30. Wahl, M., Unterberg, A., Baethmann, A., and Schilling, L. 1988. Mediators of blood-brain barrier dysfunction and formation of vasogenic brain edema. J. Cerebr. Blood Flow Metab. 8:621--634.

31. Yerram, N. R., Moore, S. A., and Spector, A. A. 1989. Eieo- sapentaenoic acid metabolism in brain microvessel endothelium: effect on prostaglandin formation. J. Lipid Res. 30:1747-1757.