Effect of eicosapentaenoic and docosahexaenoic acid on natural killer cell activity in human peripheral blood lymphocytes

Cl.INlCAL

IMMUNOLOGY

AND

IMMUNOPATHOLOGY

59, 335-345 (1991)

Effect of Eicosapentaenoic and Docosahexaenoic Acid on Natural Killer Cell Activity in Human Peripheral Blood Lymphocytes NAOHIROYAMASHITA.~MIJNEHARU MARUYAMA,KATSUYA TOMOHITOHAMAZAKLAND SABUROYANO First Department of Internal Medicine. Toyamu Medical and Pharmaceutical Medicine, Sugitani. Toyama 930-01, Japan

YAMAZAKI, University. School of

The effects of eicosapentaenoic acid (EPA)and docosahexaenoic acid (DHA) on natural killer (NK) cell activity in human peripheral blood lymphocytes were studied. The direct addition of trieicosapentaenoyl-glycerol (EPA-TG) or tridocosahexaenoylglycerol (DHA-TG)emulsion to a cytotoxicity assay system significantly suppressed NK cell activity. The addition of lipoxygenase inhibitor AA861 also inhibited NK cell activity. The inhibition was proportional to the concentration of EPA-TG emulsion, DHA-TG emulsion, or AA861. The presence of both EPA-TG emulsion or DHA-TG emulsion and AA861 at the same time led to a greater inhibitory effect on NK cell activity than when these emulsions were used separately. The inhibitory effect caused by these lipids or lipoxygenase blockade could not be reversed by adding back exogenous leukotrienes to the assay system. Preincubation of effector cells with EPA-TG or DHA-TG emulsion resulted in a significant inhibition of their NK cell activity. NK cell activity of human lymphocytes was markedly decreased after the infusion of EPA-TG emulsion into healthy volunteers. Thus, in vivo use of EPA-TG or DHA-TG emulsion may influence immune reactivity of the host. although the mechanism has not yet been elucidated. ‘!‘I 1991 Academic Pre\s. Inc.

INTRODUCTION

Arachidonic acid released from membrane phospholipids during cell activation is metabolized by the cyclooxygenase pathway to prostaglandins and thromboxanes and by the Slipoxygenase pathway to leukotrienes. These derivatives are known to act as mediators in the regulation of many physiological systems, including the immune system. Prostaglandins of the E series are mediators of acute inflammatory responses and influence the proximal limb of immunological function (l-3). Excess or imbalanced production of leukotrienes may exacerbate rheumatoid arthritis, bronchial asthma, and psoriasis (4-6). Eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA), termed n-3 fatty acids to indicate the position of the double bond furthest from the carboxylic acid, are each prominent in fish oil-enriched diets. Both these marine fatty acids can act as inhibitors of arachidonic acid metabolism (7-10). EPA, a 20-carbon analog of arachidonic acid, is known to be metabolized to a series of prostaglandins, thromboxanes, and leukotrienes structurally differing only in one additional double bond from those of arachidonic acid. The physiological effects of these ’ To whom correspondence should be addressed. 335 0090-1229/91 $1.50 Copyright All rights

Q 1991 by Academic Press. Inc. of reproduction in any form reserved.

336

b'AMASHITA

L1I Al

EPA products differ from those produced by arachidonic acid metabolites (IO. Ii). DHA is probably essential for the functional development of the nervous system, including the retina. DHA has some other biomedical effects such as depression of platelet aggregation and augmentation of efficacy of anticancer treatment (12, 13). Recently, many investigators have examined the effect of fish oils enriched in EPA and/or DHA in the alleviation of rheumatoid arthritis, chronic renal diseases, psoriasis, and in modulating the immune function ( 14-17). To receive more information about the effects of EPA and DHA in human organisms, we made infusible emulsions of trieicosapentaenoyl-glycerol (EPATG) and tridocosahexaenoyl-glycerol (DHA-TG). Because the natural killer (NK) cells appear to be involved in multiple effector, regulatory, and developmental steps of the immune system (18-20). we examined the effects of EPA and DHA on NK cell activity as one of the immunological parameters, in t,itro and in \~ir~o. MATERIALS

AND

METHODS

Cell preparation. Peripheral blood mononuclear cells (PBMC) were separated from heparinized normal human blood by Ficoll-Hypaque density gradient centrifugation, washed, and resuspended in RPM1 1640 medium supplemented with 5% heat-inactivated fetal bovine serum (GIBCO, Grand Island, NY), 2 mM Lglutamine, 60 kg/ml gentamicin (complete medium). PBMC were then depleted of adherent cells by plastic adherence. Nonadherent cells were passed over a nylon wool column to remove B lymphocytes and residual monocytes. The remaining cells, referred to hereafter as peripheral blood lymphocytes (PBL), were washed and resuspended in complete medium. Monocyte contamination was less than 2% as judged by morphology and nonspecific esterase staining. Cytotoxicity assays. NK cell activity was assessed through the standard 5’Cr release assay (21). Briefly, 3 x 10” effector cells and I x IO4 “Cr-labeled K562 target cells, an erythroleukemia cell line. were incubated in U-bottomed microtitration plates for 4 hr at 37°C in 5% CO,. In some experiments, various compounds and/or lipids were added to effector-target cell cultures during the assay. Percentage cytotoxicity was calculated as % Cytotoxicity

=

Experimental cpm - spontaneous cpm x 100. Maximum cpm - spontaneous cpm

All experiments were performed in triplicate. Percentage inhibition of NK cell activity was calculated as Percentage inhibition = Percentage control cytotoxicity - percentage experimental cytotoxicity / percentage control cytotoxicity X 100. In experiments in which a given manipulation enhanced cytotoxicity, this value is denoted as negative ( - ). Preparation of infusible emulsion. EPA ethyl ester (EPA-EE) was obtained as described previously (22). 1,2,3-Trieicosapentaenoyl-glycerol was synthesized by chemical condensation of glycerol and free EPA, which was obtained by hydrolysis of EPA-EE. 2-Eicosapentaenoil-phosphatidylcholine (EPA-PC) was synthe-

N-3

FATTY

ACIDS

AND

NK

CELL

337

ACTIVITY

sized by chemical condensation of lysophosphatidylcholine, which was enzymatically prepared from soybean oil phosphatidylcholine with phospholipase A2 and free EPA. EPA-TG was emulsified with EPA-PC according to the method of Geyer et al. (23). One hundred milliliters of the EPA-TG emulsion contained 10 g of EPA-TG, I.2 g of EPA-PC, and 2.5 g of glycerol. DHA ethyl ester (DHA-EE) was obtained as described previously (24). 1,2.3Tridocosahexaenoyl-glycerol was synthesized by chemical condensation of glycerol and free DHA. 2-Docosahexaenoic-phosphatidylcholine (DHA-PC) was synthesized by chemical condensation of lysophosphatidylcholine. DHA-TG was emulsified with DHA-PC. One hundred milliliters of DHA-TG emulsion contained 10 g of DHA-TG, 1.2 g of DHA-PC, and 2.5 g of glycerol. Incomplete reactive products of EPA-TG and DHA-TG were separated by a silica gel column. Each lipid contained a-tocopherol(O.295, w/w) as an antioxidant and 2.5% glycerol solution. This concentration of cr-tocopherol and glycerol had no effect on NK cell activity (data not shown). The fatty acid compositions of these lipids mentioned above are shown in Table 1. Reagent. Leukotriene (LT) B,, LTC,, and LTD, were purchased from Sigma Chemical Co. (St. Louis, MO). LTB, was stored in ethanol at -8O”C, and LTC, and LTD, were stored in H,O at -80°C. Indomethacin (Sigma) and AA861 (Takeda Pharmaceutical, Osaka, Japan) were dissolved in ethanol and diluted in medium to the appropriate concentration immediately before use. Experimental design. EPA-TG emulsion, DHA-TG emulsion, and/or AA861 was added directly to mixtures of effector and target cells in 4-hr 5’Cr release assays. PBL were also preincubated with EPA-TG or DHA-TG emulsion for 16

FATTY

ACID

COMPOSITION

TABLE 1 OF LIPIDS USED IN THE EXPERIMENT

Lipids (mol % I 2-EPA-PC Fatty acids 16:O 18:O 18:l (n-9) 18:2 (n-6) 18:3 (n-3) 18:4 (n-3) 20:4 (n-6) 20:4 (n-3) 205 (n-3) 22:4 (n-3) 225 (n-3) 22:6 (n-3)

EPA-TG

Position I

Position 2

2-DHA-PC .-____ .~ Position Position DHA-TG

2

26 I 20 45 2

26 I 20 45 2 2 5 2 89

I

2 5 2 89

1

I

1 2 93

I 2 93

Ndte. One hundred milliliters of EPA-TG emulsion contained 10 g of EPA-TG and 1.2 g of EPA-PC. One hundred milliliters of DHA-TG emulsion contained 10 g of DHA-TG and 1.2 g of DHA-PC.

338

LAMASHI-FA

1.1 ?I

hr, washed, and tested for NK cell activity. Data were compared for statistical significance by means of Student’s t test and considered significant for P K 0.05. H~mun studies. Three normal healthy volunteers participated in a study of intravenous infusion of EPA-TG emulsion. Thirty milliliters of EPA-TG emulsion was infused to an antecubital vein over IO min. Blood samples were taken just before and 24 hr after the infusion. PBL were prepared and assayed NK cell activity. RESULTS Effect of EPA-TG emulsion, DHA-TG emulsion, and/or AA861 on NK cell uctivity. As shown in Table 2. the direct addition of 3 x 10 -4 to 3 x 1Om6 M (EPA equivalent) EPA-TG emulsion to effector-target cell cultures significantly inhibited NK cell activity in a dose-dependent manner at an E:T ratio of 3O:l. The direct addition of 3 X 10e4 M (DHA equivalent) DHA-TG emulsion also significantly inhibited NK cell activity. More dilute concentrations of DHA-TG emulsion did not significantly alter cytotoxicity, although some suppression of target lysis was observed. Since several observations indicate that products of arachidonic acids, particularly products of lipoxygenation, are required for NK cell activity (25, 26), lipoxygenase inhibitor AA861 was also tested. As expected, AA861 inhibited NK cell activity (Table 2). The inhibitory effect of these lipids or AA861 occurred at various effector to target cell ratios (data not shown). The presence of both EPA-TG or DHA-TG emulsion and AA861 at the same time during incubation led to greater inhibitory effects on NK cell activity than when these emulsions were used separately (Table 3). The viability of PBL was not affected by these lipids or AA861 and was comparable with that of the control culture (data not shown). Effect of leukotrienes on inhibitory activity qf‘EPA-TG emulsion or AA&l. One likely explanation for the marked inhibition of NK cell activity by EPA-TG emulsion or a lipoxygenase inhibitor is that these are interfering with the production of TABLE 2 EFFECTS OF EPA-TG EMULSION. DHA-TG EMULSION. OR AA861 ON NK CELI. ACTIWTY Concentration (Ml 3x10 4

Agent tested None EPA

DHA

Inhibition (%)

54.5 L I.?

-

3x 10 5 3 x IO h 3 x 10 ’

16.2 35.1 46.6 49.8

3x

20.2 + 5.5***

70.3 35.6 14.5 X.6 62.9 26. I 8.1 -0.6 46.8 17.4

IO 4

3x10 5 3 x 10 h 3x IO ’

AA861

Cytotoxicity (%Tr)

1 X 10 ( I x IO -h

40.3 50.1 54.8 ‘9.0 45.0

k 2 -t 2 2 2 + * ?I

3.9* ?.I** 1.5** 4.2 3.3 I.0 1.3 3.t** 4.5

Nore. EPA-TG emulsion, DHA-TG emulsion, or AA861 was added directly to mixtures of effector and target cells in 4-hr “Cr release assays. The results are the means 2 SEM of three experiments. * P < 0.01;

**p

-c 0.025;

***p

< 0.05.

N-3 FATTY

COMBINED

EFFECTS

OF

339

ACIDS AND NK CELL ACTIVITY

TABLE 3 EPA-TG OR DHA-TG EMULSION __.

--~__

AND AA861 ON NK CELL ACTIVITY ~~ ~~~- _.-.---~-.~~~

Agent tested EPA (M) 3 x lo-” 3 x lOFh 3x IO 6 -

DHA (M) 3 x lo-6 3 x lo-b 3 x lo-6 -

AA861 Ml

Cytotoxicity (5%) 59.3 48.8 74.7 32.2 46.4 23.7 30.6 31.6 48.4

lo- 5 lomh 10 -r 10 -h IO -5 10 h

2 + + 2 + 2 t k i

Inhibition (5%) 17.7 58.3 45.1 21.8 60.0 48.4 46.7 18.4

1.9 3.3 2.9 2.1 2.7 1.8 4.2 4.3 2.7

__. Note. EPA-TG emulsion, DHA-TG emulsion. and/or AA861 was added directly to mixtures of effector and target cells in 4-hr “Cr release assays. The results are the means 2 SEM of three experiments.

some lipoxygenase metabolite of arachidonic acid that normally plays an important role in NK cytotoxicity. We attempted to address this possibility by asking if exogenous leukotrienes added to the assay could reverse the inhibitory effect of EPA-TG emulsion or a lipoxygenase inhibitor. We added back LTB,, LTC4, and LTD, at various concentrations (lop6 to lo-‘* M). In no case did we find any reversal of the inhibitory effects of EPA-TG emulsion or AA861 by the addition of these lipoxygenase metabolites. One representative experiment is shown in Table 4. TABLE EFFECT

OF LTB, _-___

ON THE INHIBITORY

Agent tested EPA CM) 3x10 .l 3 x 10 ~1 3 x 10~’ 3 x 10~r 3 x lo-’ 3 x lo-’ 3x10 5 -

AA861 CM) ~.. - ~_.____._~ IO-” 10.” 1O-5 -

4

ACTIVITY

OF EPA-TG

EMULSION

OR AA861

.-. _.-.

LTB, CM) 10m9 lo- I0

Cytotoxicity (%) 53.8 15,s 13.9 13.6 32.7 32.0 30.9 34.6 31.1 29.9 30.6 56.5 56.6 55.1

-+ 3.1 + 4.7 -+ 4.8 -’ 4.9 -+ 2.1 t 3.6 ” 5.5 I 2.4 t 2.1 k 1.4 2 1.1 +- 1.3 t 1.4 2 0.8

Inhibition (5%) 71.2 74.2 14.7 39.2 40.5 42.6 35.7 42.2 44.4 43.1 -5.0 -5.2 -2.4

10 ~8 10-9 lo- ‘O 10 9 lo- ‘O 10 8 10~ 9 lomio .-. __~. __.Note. EPA-TG emulsion, AA861, and/or LTB, was added directly to mixtures of effector and target cells in 4-hr “Cr release assays. The rest&s are the means + SEM of three experiments.

Pwir~crrhliotr (!I‘ l.vtr~pizoc~ytt~s riitll EPA-I’G c~ttl~rlsion or l)NA-I’(; c~ttltrisior~ with EPA-TG or DHA-‘IX rrtttl NK CPII rrc*fil’ily. The effect of preincubation

emulsion on NK cell activity was examined. PBL were preincubated with I-IPATG or DHA-TG emulsions at different concentrations for- I6 hr. washed. and tested for cytotoxicity. Figure 1 demonstrates the effect of preincubating effector cells with 3 X 10 5 M EPA-TG or DHA-TG emulsion. These lipids significantly inhibited NK cell activity when compared to control effecters preincubated with media alone. Lower concentrations of these lipids tested produced similar but lesx pronounced effects at concentrations as low as 3 x 10 * M (data not shown). At the concentrations tested, the viability of effector cells preincubation with EPATG or DHA-TG emulsion was greater than 90% as measured by the trypan blue dye exclusion test. On the other hand, when we added back LTB, at concentnttions of IO ‘to IO ’ M to effector-target cell cultures. LTB, did not significantly alter cytotoxicity (Fig. I). ,Cf~w of itljirsion of EPA-E rmrrlsion into normal rdrtnterrs. The effect of NK cell activity after the infusion of EPA-TG emulsion into healthy volunteers was investigated. After the infusion of EPA-TG emulsion. NK cell activity of PBL was markedly depressed (Fig. 2). DISCUSSION We previously

demonstrated

EPA

that the addition of an emulsion of EPA-TG

EPA i-

EPA t

LTB4 (lo+

LTB4 Mi

!1O-g MJ

DHA

DHA +

DHA c

LTB4 ElII*

emul-

LTB4 M!

ilObgMl

FIG. I. Effect of pretreatment of effector cells with EPA-TG or DHA-TG emulsion and effect of LTB, of pretreated effector cells on NK cell activity. (A) PBL were preincubated for 16 hr at 37°C with 3 x IO-’ M EPA-TG emulsion. washed, and tested for NK cell activity. LTB, was added directly to mixtures of pretreated effector and target cells in a 4-hr 5’Cr release assay. (B) PBL were preincubated for 16 hr at 37°C with 3 x IO-’ M DHA-TG emulsion, washed, and tested for NK cell activity. LTB4 was added directly to mixtures of pretreated effector and target cells in a 4-hr “Cr release assay. The results are expressed as mean percentage inhibition _f SEM of three experiments.

N-3

FATTY

ACIDS

AND

NK

CELL

ACTIVITY

341

A. Y.

T. H.

0

24 hr

FIG. 2. Inhibition of NK cell activity of PBL by intravenous infusion of EPA-TG emulsion. EPA-TG emulsion inhibited NK cell activity when 30 ml was infused to an antecubital vein 24 hr prior to assay.

sitied with phosphatidylcholine from krill to a cytotoxicity assay system resulted in a marked depression of NK cell activity, whereas the addition of soybean oil emulsion lacking EPA resulted in no inhibition (21). In this study, we further demonstrated that emulsions of EPA-TG emulsified with EPA-PC and DHA-TG emulsified with DHA-PC significantly inhibited NK cell activity of human lymphocytes in a dose-dependent manner. On the other hand, the addition of egg yolk-PC emulsion without any other lipids resulted in no inhibition of NK cell activity (data not shown). The decrease in cytotoxicity was not due to direct toxicity of effector cells, because PBL treated with EPA-TG or DHA-TG emulsion showed viability comparable to that of untreated control cultures as measured by the trypan blue dye exclusion test. The lipoxygenase inhibitor AA861 suppressed NK cell activity, whereas the specific cyclooxygenase inhibitor indomethacin had no effect (data not shown). Other reseachers have also reported that lipoxygenase inhibitors suppressed natural killing of PBL depleted of monocytes (25. 26). Both EPA-TG and DHA-TG emulsions inhibit natural killing by PBL that have been depleted of monocytes, indicating that their effect is on the lymphocyte subpopulations of PBMC, perhaps directly on NK cells. The major sources of lipoxygenase products of arachidonic acid in the peripheral blood cell compartment are monocytes, neutrophils, and eosinophils. To date, there is no convincing evidence that NK cells produce any arachidonic acid metabolites. NK cells share features with T cells, but they also have certain features of monocytes

342

YAMASHITA

t:l

.A1

and some cloned NK cells have features of basophils (27, 28). Thus although resting NK cells morphologically belong to lymphocytes, they may share lineage with monocytes and/or granulocytes, cells capable of actively oxygenating arachidonic acid. Because the intravenous infusion of EPA-TG emulsion into rabbits resulted in a marked decrease in LTB, production by granulocytes in 6 hr (unpublished data), we had initially assumed that EPA-TG emulsion inhibited NK cell activity by reducing Slipoxygenase products which normally play an important role in NK cytotoxicity, analogous to the finding of Rola-Pleszczynski rt (11. who reported that lipoxygenase products such as LTB, and LTD, could modulate natural cytotoxic cell activity (29). Unfortunately, two lines of evidence argue against this possibility. First, the addition of Slipoxygenase products, LTB,, LTC,, and LTD, did not reverse the inhibitory effect of EPA-TG emulsion or AA86I. However, this does not rule out the possibility that other lipoxygenase metabolites that we did not test might reverse the effect of EPA-TG emulsion or AA861. Furthermore. the addition of leukotrienes to the cultures may not produce effects comparable to the endogenous production of those metabolites. A second line of evidence against the role played by endogenous S-lipoxygenase products in inhibiting NK cell activity is that DHA-TG emulsion which may not block 5lipoxygenase products (8) inhibited NK cell activity. Recent studies have suggested that specific triggering molecules are present on NK cells. The interaction of these molecules with target cells leads to an increase in phosphatidylinositol metabolism and Ca” influx (30). Contact with target cells elicits an increased breakdown of phosphatidyl-inositol-bisphosphate in NK cells, which is closely correlated to the level of cytolysis (3 I, 32). This suggests that inositol phosphate compounds may act as intracellular messengers as previously reported after stimulation of various cell types, including cultured tumor cells with NK cells. Phosphatidyl-inositol-bisphosphate dissociation leads to the generation of inositol-trisphosphate and the release of Ca”. An inhibitory effect of administration of fish oil enriched with EPA and DHA has recently been reported on inositol-trisphosphate generation by stimulated platelets or cultured vascular muscle cells (33, 34). Thus, a possible mechanism for an inhibitory effect of EPA-TG and DHA-TG emulsions is decreased production of inositol phosphates by NK cells. Preincubation studies indicated that the inhibitory effect of EPA-TG or DHATG emulsion on NK cell activity was directed against effector cells rather than against the target cells. Preincubation of K562 target cells with either EPA-TG or DHA-TG emulsion did not cause a significant inhibition of NK cell activity when the washed target cells were later incubated with effector cells (data not shown). On the other hand, preincubation of PBL with EPA-TG or DHA-TG emulsion resulted in a significant suppression. The addition of LTB, to effector-target cell cultures did not prevent these lipid-mediated inhibitions. We have already reported infusion of fish oil emulsion into rabbits (35) and infusion of pure EPA emulsion into rabbits (36) and humans (37). Recently, we reported an infusion of DHA emulsion into rabbits (38). We also reported that the NK cell activity of Percoll-enriched NK effector cells (LGL) was inhibited after

N-3 FATTY

ACIDS AND NK CELL ACTIVITY

343

an intraperitoneal injection of EPA-TG emulsion into mice (39). In this report we showed that NK cell activity of human lymphocytes was markedly decreased in all healthy volunteers after the EPA-TG emulsion infusion. It is possible that EPA and DHA suppression of NK cell activity of human lymphocytes may be due to the inhibition of a variety of biochemical events leading to the target cell lysis or interference with the release of cytotoxic factors. Further studies will attempt to elucidate the cellular and biochemical mechanisms underlying EPA and DHA suppression of NK cell activity. A number of studies have shown EPA and DHA to act on neutrophils and monocytes-macrophages (10, 40, 41). Endres et al. (42) recently showed that the synthesis of interleukin-1 could be suppressed by dietary supplementation with n-3 fatty acids. The effect of EPA and/or DHA on antigen-presenting cell activity is now being investigated. We confirmed that n-3 fatty acid supplementation potently inhibited the antigen-presenting cell function of murine spleen cells and dendritic cells (unpublished data). Although the clinical importance of our observations is uncertain, we may conclude that the in vivo use of EPA-TG or DHA-TG emulsion appears to influence immune reactivity of the host. ACKNOWLEDGMENTS This work was supported by grants (63770405,01570426J from the Ministry of Education, Science and Culture of Japan. We thank Ms. Akimi Takashima for her secretarial assistance in the preparation of this manuscript.

REFERENCES 1. Kunkel. S. L., and Chensue, S. W.. Prostaglandins and the regulation of immune responses. Adv. Inj7ammation Res. 7, 93-109, 1984. 2. Leung, K. H., and Mihich, E., Prostaglandin modulation of development of cell-mediated immunity in culture. Nuture 288, 597-601. 1980. 3. Webb, D. R.. Rogers, T. J., and Nowowiejski, 1.. Endogenous prostaglandin synthesis and the control of lymphocyte function. Proc. Nail. Acad. Sri. USA 332, 262-270, 1980. 4. Davidson, E. M., Rae, S. A., and Smith, M. J. H., Leukotriene B,, a mediator of inflammation present in synovial fluid in rheumatoid arthritis. Ann. Rheum. Dis. 42, 677-679, 1983. 5. Drazen. J. M., and Austen, K. F., Leukotrienes and airway responses. Am. Rev. Resp. Dis. 136, 985-998, 1987. 6. Fogh. K., Temowitz, T.. Kragballe, K., and Herlin, T.. Chemotactic lipoxygenase products in sera from patients with psoriasis. Arch. Dermatol. Res. 278, 173-176. 1986. 7. Yoshino, S., and Ellis, E. F., Effect of a fish oil-supplemented diet on inflammation and immunological processes in rats. Int. Arch. Allergy Appl. Immunol. 84, 233-240. 1987. 8. Corey, E. J., Shih, C., and Cashman, J. R., Docosahexaenoic acid is a strong inhibitor of prostaglandin but not leukotriene biosynthesis. Proc. Nat/. Acad. Sci. USA 80, 3581-3584, 1983. 9. Lokesh, B. R., German, B.. and Kinsella, J. E.. Differential effects of docosahexaenoic acid and eicosapentaenoic acid on suppression of lipoxygenase pathway in peritoneal macrophages. Biochim. Biophys. Acta 958, 99-107, 1988. IO. Lee, T. H.. Mencia-Huerta, J.-M.. Shih, C., Corey, E. J.. Lewis, R. A., and Austen, K. F., Effects of exogenous arachidonic, eicosapentaenoic. and docosahexaenoic acids on the generation of 5-lipoxygenase pathway products by ionophore-activated human neutrophils. J. C/in. Invesr. 74, 1922-1933, 1984. I I. Leitch. A. G., Lee, T. H., Ringel. E. W., Prickett. J. D.. Robinson. D. R., Pyne, S. G.. Corey, E. J.. Drazen, J. M., Austen. K. F.. and Lewis, R. A., Immunologically induced generation of

12. II. 14. IS. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28.

29. 30. 31. 32.

telraene and pentaene leukotriencs m the peritoneal cavittea of menhaden-fed rats. ./. /,r~mrtrl,~/ 132, 2559-256.5. 1984. von Schacky. C. and Weber. P C.. Metabolism and effects on platelet function of the purtfied eicosapentaenoic and docosahexaenoic acids in human\. J. C’lirr. Int,c.vr. 76, 2446-2450. 1985. Guffy, M. M.. North. J. A.. and Burns. C. I’.. Effect ofcellularfatty acid alteration on adriamyctn sensitivity in cultured L1210 murine leukemia cells. C‘trncer RI,.\. 44. 1863-1866. 1984. Kremer, J. M.. Jubiz. W., Michalek, A.. Kynes. K. I.. Bartholomew, I,. E.. Bigaouette. J., Timchalk, M., Beeler. D.. and Lininger. L., Fish-oil fatty acid supplementation in active rheumatoid arthritis: A double-blinded. controlled, crossover study. ilrrr~. Inlerrr. Mrd. 106, 497-503. 1987. Kelley. V. E., Ferretti. A., Izui. S.. and Strom. T. B., A fish oil diet rich in eicosapentaenoic acid reduces cyclooxygenase metabolites, and suppresses lupu\ in MRL-lpr mice. .I. /rtrrnrrm,l. 134, 1914-1919. 1985. Hamazaki, T.. Tateno. S., and Shishido, H.. Eicosapentaenoic acid and IgA nephropathy. L
N-3 FATTY

ACIDS AND NK CELL ACTIVITY

345

0X-34: Possible signaling role for the OX-34 determinant during activation by target cells. Proc. Natl.

Acad.

Sci.

USA 84, 4239-4243.

1987.

33. Medini, L., Colli, S.. Mosconi, C., Tremoli, E., and Galli, C., Diets rich in n-9, n-6. and n-3 fatty acids differentially affect the generation of inositol phosphates and of thromboxane by stimulated platelets, in the rabbit. Biochem. Pharmacol. 39, 129-133, 1990. 34. Lecher, R., Sachinidis, A., Steiner A., Vogt, E., and Vetter, W., Fish oil affects phosphoinositide turnover and thromboxane A metabolism in cultured vascular muscle cells. Biochim. Biophys. Acta 1012, 279-283, 1989. 35. Urakaze. M.. Hamazaki, T., Makuta, M., Ibuki, F., Kobayashi, S.. Yano, S.. and Kumagai, A.. Infusion of fish oil emulsion: Effect on platelet aggregation and fatty acid composition in phospholipids of plasma, platelets, and red blood cell membranes in rabbits. Am. J. Clin. Nutr. 46, 936-940. 1987. 36. Urakaze, M., Hamazaki, T., Soda, Y.. Miyamoto, A., Ibuki, F.. Yano. S.. and Kumagai. A., Infusion of emulsified trieicosapentaenoyl-glycerol into rabbits: The effects on platelet aggregation, polymorphonuclear leukocyte adhesion, and fatty acid composition in plasma and platelet phospholipids. Thromh. Res. 44, 673-682, 1986. 37. Hamazaki. T.. Fischer, S., Schweer, H., Meese, C. O., Urakaze. M.. Yokoyama. A.. and Yano. S., The infusion of trieicosapentaenoyl-glycerol into humans and the in vivo formation of prostaglandin I, and thromboxane A,. Biochem. Biophys. Res. Commun. 151, 1386-1394, 1988. 38. Hamazaki, T.. Urakaze, M.. Yano. S., Soda, Y.. Miyamoto. A., Kubota, K., and Ibuki. F.. Injection of tridocosahexaenoyl-glycerol emulsion and fatty acid composition of blood cells. Lipids 22, 1031-1034, 1987. 39. Yamashita, N., Sugiyama. E., Hamazaki. T., and Yano, S., Inhibition of natural killer cell activity by eicosapentaenoic acid in vivo and in vitro. Biochem. Biophys. Res. Commun. 150, 497-505, lY88. 40. Strasser, T., Fischer, S., and Weber, P. C., Leukotriene B, is formed in human neutrophils after dietary supplementation with icosapentaenoic acid, Proc. Nat/. Acad. Sci. USA 82, 1540-1543. 1985. 41. Lee. T. H.. Hoover, R. L.. Williams. J. D.. Sperling. R. 1.. Ravalese, J.. 111, Spur, R. W., Robinson, D. R., Corey, E. J., Lewis, R. A., and Austen, K. F., Effect of dietary enrichment with eicosapentaenoic and docosahexaenoic acids on in vitro neutrophil and monocyte Jeukotriene generation and neutrophil function. N. Engl. J. Med. 312, 1217-1224, 1985, 42. Endres, S.. Ghorbani, R., Kelley, V. E., Georgilis, K., Lonneman. G., van der Meer, J. W., Cannon, J. G., Rogers, T. S., Klempner, M. S., Weber, P. C.. Schaefer, E. J., Wolff. S. M., and Dinarello, C. A., The effect of dietary supplementation with n-3 polyunsaturated fatty acids on the synthesis of interleukin-1 and tumor necrosis factor by mononuclear cells. N. ,%g/. J. Med. 320, 265-27 I, 1989. Received June 15, 1990: accepted with revision January 28. 1991