N-3 Fatty Acids Cyclo-Oxygenase-1 Pathway

Arachidonic Acid Enhances the Tissue Factor Expression of Mononuclear Cells by the Cyclo-Oxygenase-1 Pathway: Beneficial Effect of n-3 Fatty Acids This information is current as of September 27, 2021. Yves Cadroy, Dominique Dupouy and Bernard Boneu J Immunol 1998; 160:6145-6150; ; http://www.jimmunol.org/content/160/12/6145 Downloaded from

References This article cites 32 articles, 14 of which you can access for free at: http://www.jimmunol.org/content/160/12/6145.full#ref-list-1

Why The JI? Submit online. http://www.jimmunol.org/

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

*average by guest on September 27, 2021

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 1998 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. Arachidonic Acid Enhances the Tissue Factor Expression of Mononuclear Cells by the Cyclo-Oxygenase-1 Pathway: Beneﬁcial Effect of n-3 Fatty Acids1

Yves Cadroy,2 Dominique Dupouy, and Bernard Boneu

Monocytes express tissue factor (TF) upon stimulation by inflammatory agents. Dietary administration of fish oil rich in eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) results in an impairment of TF expression by monocytes. EPA and DHA are metabolized differently from arachidonic acid (AA), the major fatty acid present in cell membranes. We examined the effects of AA on the TF expression of isolated human PBMC, and we determined whether EPA and DHA modulated this phenomenon differently. Nonstimulated PBMC had a low TF-dependent procoagulant activity. When PBMC were incubated with increasing concentrations of AA, the TF-dependent procoagulant activity increased in a dose-dependent manner to 190% at 7.5 ␮M. Indo- Downloaded from methacin, a cyclo-oxygenase inhibitor, totally abolished the stimulating effect of AA, whereas specific pharmacologic inhibitors of cyclo-oxygenase-2 or of 5-lipoxygenase had no inhibitory effect. A thromboxane (TX)A2/endoperoxides receptor antagonist and a

TX synthase inhibitor blocked the potentiating effect of AA. Puriﬁed PGG2 and carbocyclic TXA2, a TXA2 agonist, enhanced the procoagulant activity of PBMC in a dose-dependent manner whereas, in contrast, PGE2 inhibited it. Finally, contrary to AA, EPA or DHA did not increase TXB2 production or TF expression by PBMC. The TF-dependent procoagulant activity of isolated PBMC http://www.jimmunol.org/ was increased by AA through the production of cyclo-oxygenase-1 metabolites; the combined action of PGG2 and TXA2, which potentiated it, was greater than that of PGE2, which inhibited it. Dietary n-3 fatty acids exert part of their beneﬁcial effect by modulating this procoagulant activity differently from AA. The Journal of Immunology, 1998, 160: 6145–6150.

rachidonic acid (AA;3 20:4, n-6) is a polyunsaturated atherosclerosis. The molecular mechanisms by which dietary n-3 fatty acid present in cell membranes that modulates di- fatty acids exert their beneficial effects are complex and have not A verse physiologic and pathologic responses. These ef- been totally clarified. Some of these effects have been attributed to fects are primarily mediated by its metabolic products, eico- the activity of n-3 fatty acids on monocytes (4, 8, 11–13). sanoids. Eicosanoids include notably the cyclo-oxygenase Monocytes are centrally involved in numerous pathophysiologic by guest on September 27, 2021 products, i.e., PG and TX, and the lipoxygenase products, i.e., processes, such as thrombosis, atherosclerosis, wound repair, and leukotrienes and hydroxyeicosatetraenoic acids (HETEs). Their ef- inflammation. These properties are partly related to their ability to fects, which depend on the nature of compounds, principally con- express various procoagulant activities (14–17). The procoagulant cern inflammation and thromboregulation (1). Eicosapentaenoic activities of monocytes are mediated to a large extent by cell sur- acid (EPA; 20:5, n-3) and docosahexaenoic acid (DHA; 22:6, n-3) face-associated tissue factor (TF). TF is the cellular receptor and are other polyunsaturated fatty acids contained in fish oil. EPA is cofactor for plasma factor VII(a), which initiates the coagulation metabolized into eicosanoids that are different from the corre- protease cascade leading ultimately to the generation of thrombin sponding compounds derived from AA and are considered to de- and fibrin (18). Dietary administration of fish oil results in an im- velop reduced biologic activities (2–5). DHA is not metabolized by pairment of TF expression by monocytes (8, 12, 13). The molec- cyclo-oxygenase or lipoxygenase enzymes (6). Experimental (7, ular mechanisms underlying TF activity reduction after n-3 fatty 8), epidemiologic, and interventional (9, 10) studies suggest that acids intake are unknown. However, since monocytes are impor- consumption of fish oil, which causes a progressive decrease in the tant generators of eicosanoids and possess both cyclo-oxygenase AA content of tissue phospholipids and an increase in the tissue and lipoxygenase enzymes, changes in the monocyte capacity to level of n-3 fatty acids, exerts beneficial effects on thrombosis and generate eicosanoids by these pathways have been suspected to influence monocyte TF activity. Thus, several recent studies have indicated that eicosanoids may modulate the procoagulant proper- Laboratoire de Recherche sur l’He´mostase et la Thrombose, Centre Hospitalo-Uni- ties of monocytes. For example, prostacyclin analogues and PGE1 versitaire Purpan, Toulouse, France cause a decrease in cytokine-induced TF activity by monocytes Received for publication October 3, 1997. Accepted for publication February (19, 20), whereas, conversely, platelet 12-HETE enhances PBMC 12, 1998. procoagulant activity (21). The costs of publication of this article were defrayed in part by the payment of page To clarify the mechanisms of the beneficial effect of dietary fish charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. oil on the procoagulant activity of monocytes, we investigated in 1 This work was supported in part by a grant from Pierre Fabre, Castres, France. vitro the effects of AA, EPA, and DHA on TF expression by iso- 2 Address correspondence and reprint requests to Dr. Yves Cadroy, Laboratoire lated PBMC. We show for the first time that AA, but not EPA or d’He´matologie, Pavillon Lefe`bvre, Centre Hospitalo-Universitaire Purpan, 31059 DHA, enhanced the TF-dependent procoagulant activity of PBMC. Toulouse Cedex, France. The roles of the cyclo-oxygenase and lipoxygenase pathways were 3 Abbreviations used in this paper: AA, arachidonic acid; TX, thromboxane; HETE, also examined. We describe a novel role for the cyclo-oxygenase-1 hydroxyeicosatetraenoic acid; EPA, eicosapentaenoic acid; DHA, docosahexaenoic pathway in modulating the TF expression of AA-stimulated acid; TF, tissue factor; CTXA2, carbocyclic thromboxane A2; AU, procoagulant activity units. PBMC that appeared to be balanced by the opposite effects of

PGE2, which inhibited it, and endoperoxides/TXA2, which Determination of TF, TXB2, and PGE2 potentiated it. TF Ag was measured on cell lysates by commercially available ELISA (Imubind Tissue Factor, American Diagnostica). Cell lysates were prepared by lysing PBMC in PBS containing 1% Triton X-100, 1 mM EDTA Materials and Methods (Merck, Chelles, France), 16 mM octyl PD glucopyranoside (Boehringer Cell isolation and cell cultures Mannheim, Meylan, France), 10 ␮M pepstatin A, 10 ␮M leupeptin, 0.1 mM PMSF (Sigma), and 100 kallkrein inhibitor units/ml aprotinin (Sanofi- PBMC were isolated using a modification of the method described by Choay, Paris, France). The cell lysates were then frozen and thawed three Ϫ Balter et al. to efficiently deplete platelet from the cell preparation (22). times. They were stored at 80°C until assayed. TXB2 and PGE2 produced Whole blood was obtained with a 19-gauge needle from healthy volunteers by cultured PBMC were measured in the supernatant by enzyme immu- who had not taken aspirin or other nonsteroidal anti-inflammatory drugs in noassays (Cayman). the 7 days preceding the donation. Blood was anticoagulated with triso- dium citrate (0.129 M; Becton Dickinson, Meylan, France) and centrifuged Statistical analysis at 280 ϫ g for 15 min at 4°C. Platelet-rich plasma was removed. The All results represent the mean Ϯ 1 SEM of 4 to 10 separate experiments. sedimented cells were diluted to twice the original blood volume with PBS Depending on the data, statistical analysis was performed using Student’s (pH 7.4; Seromed, Biochrom, Berlin, Germany), layered onto Ficoll- test for paired variables or ANOVA followed by a Neuman-Keuls test Hypaque PLUS (Pharmacia Biotech, Uppsala, Sweden), and centrifuged at when p Յ 0.05. Differences were considered significant at p Յ 0.05. 400 ϫ g for 35 min at 4°C. The resulting PBMC were washed in 5 mM EDTA-PBS four to six times to remove remaining platelets The resulting mononuclear fraction contained less than two platelets per leukocyte. Non- Results specific ␣-naphtyl-acetate esterase staining indicated that the mononuclear AA enhanced the procoagulant activity and TF synthesis of fraction contained 28.3 Ϯ 6.4% (n ϭ 4) monocytes. Cells (500 ␮l) were mononuclear cells Downloaded from incubated at 2.5 ϫ 106 PBMC/ml in medium 199 (ATGC Biotechnologie, Noisy-le-Grand, France) containing 2 mM glutamine, 100 U/ml penicillin, Nonstimulated PBMC had a very low procoagulant activity and 100 ␮g/ml streptomycin in stoppered, sterile, pyrogen-free tubes. (138 Ϯ 17 AU/106 cells; n ϭ 7). Considering that the procoagulant Sodium arachidonate, eicosapentaenoate, and docosahexaenoate (Sig- activity of the cell suspension was solely due to monocytes, which Ϫ ma, Saint Quentin Fallavier, France), stored under argon at 80°C, or represented 28% of the PBMC suspension, the procoagulant ac- carbocyclic TXA2 (CTXA2; Sigma), TXB2 (Cayman, Ann Arbor, MI), 6 ␣ tivity expressed by monocytes (493 Ϯ 61 AU/10 monocytes) was PGG2,E2,D2, and F2 (Calbiochem, La Jolla, CA) were dissolved in http://www.jimmunol.org/ medium 199 just before use and added to the cell suspension at the con- comparable to that expressed by cultured resting endothelial cells centrations indicated below for 20 h. Indomethacin (10 ␮M; Sigma), (498 Ϯ 110 AU/106 endothelial cells; n ϭ 9), indicating that NS398 (1 ␮M; Calbiochem), furegrelate (10 ␮M; Cayman), baicalein (10 monocytes were not significantly stimulated by the procedure of ␮M; Calbiochem), or L655,238 (1 ␮M; Calbiochem), selective inhibitors of cyclo-oxygenase, cyclo-oxygenase-2, TX synthase, 12-lipoxygenase, isolation. The procoagulant activity was TF dependent, since it and 5-lipoxygenase, respectively, and SQ29,548 (10 ␮M; Cayman), a was inhibited Ͼ90% by preincubating the cells with anti-TF ␮ TXA2/endoperoxides receptor antagonist, were added before adding so- mAbs. When PBMC cells were incubated with AA (5 M) for 4 dium arachidonate to the cells. The concentration used for each of these and 20 h, respectively, their procoagulant activity increased, but inhibitors was chosen based on the IC reported by the respective man- 50 this increase was less marked with 4-h incubation (i.e., by 132% at ufacturers. Depending on the reagents, they were dissolved in water, eth- ϭ anol, or DMSO and added (1 ␮l) to the cells. Control experiments per- 4 h vs 335% at 20 h; n 1). Therefore, all additional experiments by guest on September 27, 2021 formed with the respective solvents indicated that they did not affect the were performed with 20-h incubation. In these conditions, AA in- procoagulant activity of PBMC (data not shown). LPS was obtained from creased the procoagulant activity in a dose-dependent manner (Ta- Escherichia coli 0111:B4 (Sigma) and incubated with PBMC for 20 h in ble I) to a mean of 190% at 7.5 ␮M(n ϭ 10), which corresponds some experiments (see below). Human endothelial cells were isolated from umbilical veins and cul- to a mean reduction in clotting times of 40 s (from 240 to 200 s). tured according to the method of Jaffe et al. (23). The culture medium was Higher concentrations resulted in cell lysis. Similarly, AA signif- composed of RPMI 1640 and medium 199 (ATGC Biotechnologie, Noisy- icantly enhanced the cell content of TF Ag (Table II). By com- Le-Grand, France) supplemented with 20% human pooled serum (Institut parison, LPS (100 ng/ml) increased the procoagulant activity of Jacques Boy, Reims, France). The cells were identified by their typical PBMC by 294% (406 Ϯ 50 AU/106 cells; n ϭ 7). morphology. Cells were incubated for4hinmedium 199 and carefully scraped with a rubber policeman just before measurement of their proco- The effect of AA was mediated by the cyclo-oxygenase-1 agulant activity. Cell viability, as assessed by the measurement of lactate dehydrogenase pathways release in the supernatant of cultured cells and by trypan blue exclusion, Preincubation of PBMC with indomethacin (10 ␮M), a cyclo-ox- was Ͼ90%. All reagents used for cell isolation and culture were prepared ygenase-1 and -2 inhibitor, which blocked TXA2 formation by with endotoxin-free water. The level of endotoxin contamination in the Ϯ different reagents incubated with PBMC, as assessed by a sensitive chro- 96 1%, totally abolished the stimulating effect of AA, whereas mogenic Limulus assay (Chromogenix, Molndal, Sweden), was very low specific pharmacologic inhibitors of cyclo-oxygenase-2 (NS398) (Ͻ0.001 ng/ml, final concentration). This level of endotoxin did not en- or 5-lipoxygenase (L655,238) had no inhibitory effect (Table I). hance the procoagulant activity of PBMC (data not shown). Importantly, the stimulating effect of AA was not due to a low contamination by endotoxins, since the procoagulant activity in- Measurement of procoagulant activity duced by LPS remained unchanged in the presence of indometh- Ϯ Ϯ 6 Procoagulant activity was measured on intact cells using a one-step plasma acin (563 123 vs 570 88 AU/10 cells with and without recalcification time assay performed on a coagulometer (KC4, Amelung, indomethacin, respectively; n ϭ 8). Therefore, these results indi- Lemgo, Germany). The PBMC were placed on ice for 10 to 20 min to cate that AA enhanced the procoagulant activity of PBMC by the remove any adherent monocytes. A 20-␮l sample of PBMC or scrape- cyclo-oxygenase-1 pathway. harvested endothelial cells was added to 90 ␮l of citrated normal human platelet-poor plasma. One hundred microliters of 25 mmol/l CaCl was 2 PGG2 and CTXA2 enhanced the procoagulant activity and TF added to initiate the reaction. Coagulation times were converted into pro- synthesis of PBMC coagulant activity units (AU) using reference curves determined with a standard human brain TF preparation containing 106 AU/ml (Thromborel The major metabolites of AA produced by the cyclo-oxygenase-1 S, Behring, Marburg, Germany); the logarithm of the procoagulant activity pathway are represented by the endoperoxides PGG2/PGH2 and was related to the logarithm of the coagulation time. The procoagulant ␣ activity of mononuclear cells was characterized by incubating the cells further metabolized into TXA2, PGE2, PGD2, and PGF2 .Tode- with a mixture of two mouse anti-human TF mAbs (10 ␮g/ml; American termine those that are responsible for the enhancing effect of AA Diagnostica, Greenwich, CT) for 30 min at 37°C. on the procoagulant activity of PBMC, we tested the effect of The Journal of Immunology 6147

Table I. Effect of AA and of selective inhibitors of the cyclooxygenase and lipoxygenase pathways on the procoagulant activity of PBMC a

AA (␮M)

0 2 5 7.5

Control (AU/106 cells, n ϭ 10) 138 Ϯ 17 193 Ϯ 19** 236 Ϯ 29** 261 Ϯ 26** Indomethacin (AU/106 cells, n ϭ 9) 124 Ϯ 22 128 Ϯ 20 123 Ϯ 22 145 Ϯ 40 NS398 (AU/106 cells, n ϭ 8) 128 Ϯ 15 236 Ϯ 31** 262 Ϯ 44** 238 Ϯ 55** L655,238 (AU/106 cells, n ϭ 9) 144 Ϯ 24 177 Ϯ 40 184 Ϯ 27 288 Ϯ 28** Baicalein (AU/106 cells, n ϭ 10) 116 Ϯ 12 ND ND 241 Ϯ 33** Furegrelate (AU/106 cells, n ϭ 10) 171 Ϯ 21 191 Ϯ 35 210 Ϯ 34 194 Ϯ 36 SQ29,548 (AU/106 cells, n ϭ 10) 148 Ϯ 22 158 Ϯ 24 182 Ϯ 22 188 Ϯ 29

a PBMC (2.5 ϫ 106 cells/mL) were incubated for 20 h with increasing concentrations of AA. Indomethacin (10 ␮M), NS398 (1 ␮M), L655,238 (1 ␮M), baicalein (10 ␮M), or furegrelate (10 ␮M), selective inhibitors of cyclooxygenase, cyclooxygenase-2, ␮ 5-lipoxygenase, 12-lipoxygenase, and thromboxane synthase, respectively, and SQ29,548 (10 M), a TXA2/endoperoxides receptor antagonist, were added to the cells before incubating PBMC with AA. For each inhibitor, statistical comparisons of the procoagulant activity of PBMC incubated in presence of AA vs that without AA are represented by asterisks, where * ϭ p Ͻ 0.05, and ** ϭ p Ͻ 0.01. Data are the mean Ϯ 1 SEM of n experiments. Downloaded from ␮ purified PG (Table III). PGG2 and CTXA2, a TXA2 agonist, en- CTXA2 (0 to 1 M). As found previously (Table III), the addition hanced, in a dose-dependent manner, the procoagulant activity of of PGE2 to PBMC decreased their baseline procoagulant activity Ϯ Ϯ 6 PBMC and TF Ag (from 0.32 0.12 to 1.15 0.33 ng/10 cells by 54% (Fig. 2). When the molar ratio CTXA2/PGE2 was in- ␮ Ͻ ϭ at 2 M PGG2; p 0.05; n 8). In contrast, PGE2 significantly creased from 0 to 2, the procoagulant activity of PBMC increased decreased the procoagulant activity of the cells. TXB2, PGD2, and in a ratio-dependent manner. However, although the increase in ␣ PGF2 had no significant effect. procoagulant activity produced by the addition of CTXA2 to Ͻ http://www.jimmunol.org/ To further test the hypothesis that the formation of endoperox- PGE2-treated cells was statistically significant ( p 0.01), the pro- ides and TXA2 was involved in the potentiating effect of AA on the coagulant activity of PGE2-treated cells remained significantly Ͻ procoagulant activity of PBMC, we studied the effect of furegre- lower ( p 0.05) than that of cells incubated with CTXA2 but Ͻ late, a TX synthase inhibitor, which blocked TXA2 formation by without PGE2 ( p 0.05). Ϯ 94 1%, and that of SQ29,548, a TXA2/endoperoxides receptor antagonist. Both molecules inhibited the potentiating effect of AA EPA and DHA did not enhance the procoagulant activity and on the procoagulant activity of PBMC (Table I). As expected, the TF synthesis of mononuclear cells procoagulant effect of CTXA2 was not inhibited by furegrelate The n-3 fatty acids EPA and DHA are metabolized differently from 6

Ϯ Ϯ by guest on September 27, 2021 (266 51 vs 142 30 AU/10 cells with and without CTXA2;1 AA by cyclo-oxygenase and lipoxygenase enzymes. For example, ␮ Ͻ ϭ M; p 0.01; n 4). when these fatty acids were incubated with PBMC, EPA, unlike The potentiating effect of AA on the procoagulant activity of AA, did not increase the basal production of TXB2, and DHA PBMC and the respective amounts of TXB2, the stable metabolite inhibited it (Table II). In these conditions, EPA and DHA did not of TXA2, and PGE2 released by PBMC varied among the different enhance the procoagulant activity (Fig. 3) and TF Ag level (Table blood donors from whom the cells were isolated. The effect of AA II) of PBMC. At 5 and 7.5 ␮M, the procoagulant activity was on the procoagulant activity of cells was not directly related to the significantly lower with EPA and DHA ( p ϭ 0.01) than with AA. Ͼ amount of TXB2 or PGE2 released ( p 0.20), but was signifi- Therefore, these results confirm that the effect of AA was mostly cantly and positively related to the ratio of TXB2/PGE2 formed by mediated by the production of endoperoxides and TXA2. cells in presence of AA (r ϭ 0.61; p Ͻ 0.001; Fig. 1). This cor- relation remained significant when the outlying data with the high- est procoagulant activity (937% of the baseline procoagulant ac- Discussion tivity) were deleted (r ϭ 0.44; p Ͻ 0.01). Therefore, the variability We have shown in the present study that the procoagulant activity of response of blood donors to AA appeared to be related to the of isolated PBMC was enhanced by AA. This increase in proco-

respective amount of endoperoxides/TXA2 and PGE2 produced by agulant activity was due to an increased generation of TF and not PBMC; there was a balance between these eicosanoids, which to an increased surface expression of procoagulant activity (Table showed opposite effects on the procoagulant activity of PBMC. II). The effect of AA was mediated through the production of cy-

To conﬁrm that the procoagulant activity of PBMC was deter- clo-oxygenase-1 metabolites, especially endoperoxides and TXA2. mined in part by a balance between TXA2 and PGE2, PBMC were These ﬁndings differ from those of a previous study in which AA ␮ incubated with PGE2 (0.5 M) and increasing concentrations of enhanced the procoagulant activity of PBMC when they were

a Table II. Effect of AA, EPA, and DHA on TF Ag synthesis and TXB2 production by human mononuclear cells

AA EPA DHA

TFag TXB2 TFag TXB2 TFag TXB2 (ng/106 cells, (pmol/106 cells, (ng/106 cells, (pmol/106 cells, (ng/106 cells, (pmol/106 cells, n ϭ 7) n ϭ 7) n ϭ 5) n ϭ 7) n ϭ 5) n ϭ 7)

0 ␮M 0.4 Ϯ 0.1 1.1 Ϯ 0.1 0.4 Ϯ 0.1 1.1 Ϯ 0.1 0.4 Ϯ 0.1 1.1 Ϯ 0.1 5 ␮M 1.2 Ϯ 0.4** 11.5 Ϯ 1.2** 0.3 Ϯ 0.1 1.1 Ϯ 0.4 0.3 Ϯ 0.1 0.2 Ϯ 0.1**

a PBMC (2.5 ϫ 106 cells/mL) were incubated for 20 h with increasing concentrations of AA, EPA, or DHA. Statistical comparisons of PBMC incubated in presence of fatty acids vs PBMC without fatty acids are represented by asterisks, where ** ϭ p Ͻ 0.01. TFag, tissue factor Ag. Data are the mean Ϯ 1 SEM of n experiments. 6148 ARACHIDONIC ACID AND MONONUCLEAR CELL TISSUE FACTOR

Table III. Effect of puriﬁed prostaglandins on the procoagulant activity of mononuclear cells a

PG (␮M)

0 0.1 0.5 1 2

6 ϭ Ϯ Ϯ Ϯ Ϯ PGG2 (AU/10 cells, n 8) 108 10 ND 114 32 146 33** 192 40** 6 ϭ Ϯ Ϯ Ϯ Ϯ CTXA2 (AU/10 cells, n 4) 130 33 174 62 211 28 272 28* ND 6 ϭ Ϯ Ϯ Ϯ Ϯ TXB2 (AU/10 cells, n 5) 122 30 ND 120 30 116 18 114 16 6 ϭ Ϯ Ϯ Ϯ Ϯ Ϯ PGE2 (AU/10 cells, n 9) 108 10 80 9754** 99 8783 6 ϭ Ϯ Ϯ Ϯ Ϯ Ϯ PGD2 (AU/10 cells, n 8) 154 14 163 30 160 24 158 20 182 28 6 ϭ Ϯ Ϯ Ϯ Ϯ Ϯ PGF2␣ (AU/10 cells, n 5) 149 31 164 40 139 26 160 46 108 22 a ϫ 6 PBMC (2.5 10 cells/mL) were incubated for 20 h with increasing concentrations of PGG2, CTXA2, TXB2, PGE2, PGD2, and PGF2␣. Statistical comparisons of the procoagulant activity of mononuclear cells incubated in presence of PG vs that without PG are represented by asterisks, where * ϭ p Ͻ 0.05 and ** ϭ p Ͻ 0.01. Data are the mean Ϯ 1 SEM of n experiments.

stimulated by endotoxins and in presence of platelets. This result The enhancing effect of AA observed in our study was mediated was related to the platelet lipoxygenase product, 12-HETE (21). by cyclo-oxygenase-1. It was inhibited by indomethacin, a cyclo- The role of this platelet lipoxygenase product was negligible in our oxygenase-1 and -2 inhibitor (29), but it remained unchanged with

study, since the platelet/PBMC ratio was low (i.e., less than two a selective inhibitor of cyclo-oxygenase-2 (NS398; Table I). The Downloaded from platelets per leukocyte) and since baicalein, a pharmacologic in- latter result is not surprising, since the cyclo-oxygenase-2 pathway hibitor of platelet 12-lipoxygenase, did not inhibit the effect of AA is only functional when monocytes are exposed to inﬂammatory (Table I). Our study was performed with nonstimulated PBMC. stimuli, which were absent in our study (24, 25). Interestingly, AA and other reagents were not contaminated with LPS; no de- previous works and our own preliminary data indicate that path- tectable amount of endotoxin was found in the various molecules ways other than cyclo-oxygenase-1 are involved in the enhancing

incubated with the cells, and the effect of AA was inhibited with effect of AA on the procoagulant activity of PBMC when they are http://www.jimmunol.org/ indomethacin, contrary to that induced by LPS. stimulated with LPS (21, 30). AA is metabolized by PBMC through different enzymatic path- Our experiments were performed with exogenous AA. How- ways, notably cyclo-oxygenase-1, cyclo-oxygenase-2, and 5- and ever, we do consider the role of cyclo-oxygenase-1 pathway in 12-lipoxygenases (22, 24–26). Leukotrienes, which are important AA-induced TF expression to be pathophysiologically relevant. modulators of the inﬂammatory response (27), were not involved Indeed, this polyunsaturated fatty acid was effective at concentra- in the enhancing effect of AA on the procoagulant activity of non- tions as low as 5 ␮M. The plasma concentrations of polyunsatu- stimulated PBMC, since it was not affected by 5- and 12-lipoxy- rated fatty acids have been reported to vary between 0.2 and 2 mM genase inhibitors (L655,238 and baicalein, respectively; Table I). (31), and the percentage of AA in plasma free fatty acid can reach by guest on September 27, 2021 In addition, previous studies have shown that leukotriene B4, pro- 9% (32). Interestingly, and in contrast to AA, endotoxin-stimulated duced by the 5-lipoxygenase pathway, has no effect on the proco- TF expression of PBMC was not inhibited by inhibitors of the agulant activity of PBMC (21, 28). 12-HETE is a potent cofactor cyclo-oxygenase or lipoxygenase pathway. Thus, endogenous AA, for TF generation by PBMC (21). However, in our working con- present in the cell membrane, and/or the cyclo-oxygenase or li- ditions, the amount of 12-HETE released may have been too low poxygenase pathways did not appear to be involved in endotoxin- to have a role, since there were very few platelets, an important induced TF expression.

source of 12-HETE, and since monocytes produce only very low In the presence of AA, monocytes release endoperoxides (PGG2 amounts of 12-HETE, especially in the absence of LPS (26). and PGH2) through the cyclo-oxygenase-1 pathway that are further metabolized into two major metabolites, TXA2 and PGE2 (22, 24, 25). Puriﬁed PGG2 and CTXA2, a TXA2 agonist, enhanced the TF

FIGURE 2. Effects of increasing concentrations of CTXA2 on the pro- ϭ FIGURE 1. Relationship between the procoagulant activity and the mo- coagulant activity of PBMC, treated or not with PGE2 (n 5). PBMC ϫ 6 lar ratio of TXB2/PGE2 produced by PBMC isolated from different blood (2.5 10 cells/ml) were incubated for 20 h with PGE2 and CTXA2. donors and incubated with AA. PBMC (2.5 ϫ 106 cells/ml) isolated from Results are expressed as a percentage of the procoagulant activity of resting ␮ different blood donors were incubated without or with AA (5 M) for 20 h. PBMC incubated for 20 h without PGE2 or CTXA2. Statistical compari- ␮ The procoagulant activity with 5 M AA was expressed as a percentage of sons of the procoagulant activity of PBMC incubated with puriﬁed PGE2 the baseline procoagulant activity obtained in the absence of AA. TXB2 and CTXA2 vs that obtained without PGs are represented by asterisks ϭ Ͻ Ͻ and PGE2 were measured in the supernatant of PBMC (n 35). (* indicates p 0.05; ** indicates p 0.01). The Journal of Immunology 6149

inducing a rise in the concentration of ionized calcium in the cell cytoplasm (1). Our study adds one more mechanism by

which TXA2 may be prothrombotic. Other mechanisms include its capacity to cause platelet aggregation and vasoconstriction, to augment monocyte and neutrophil adhesiveness (1, 35–38), and to promote diapedesis (3). Finally, our results also suggest that part of the antithrombotic effect of aspirin, which irrevers- ibly inactivates cyclo-oxygenase and inhibits the production of

endoperoxides/TXA2, may be mediated by a diminution of the procoagulant activity of PBMC.

References 1. Marcus, A. J. 1994. Multicellular eicosanoid and other metabolic interactions of platelets and other cells. In Hemostasis and Thrombosis: Basic Principles and Clinical Practice, 3rd Ed. R. W. Colman, J. Hirsh, V. J. Marder, and FIGURE 3. Effects of AA, EPA, and DHA on the procoagulant activity E. W. Salzman, eds. J. B. Lippincott, Philadelphia, p. 590. 2. Fisher, S., and P. C. Weber. 1983. Thromboxane A3 (TXA3) is formed in human ϭ ϫ 6 of PBMC (n 9). PBMC (2.5 10 cells/ml) were incubated for 20 h platelets after dietary eicosapentaenoic acid (C20:5 omega-3). Biochem. Biophys. with increasing concentrations of AA, EPA, or DHA. Statistical compar- Res. Commun. 116:1091. Downloaded from isons of the procoagulant activity of PBMC incubated in presence of fatty 3. Von Schacky, C., S. Fisher, and P. C. Weber. 1985. Long-term effects of dietary acids vs that obtained without fatty acids are represented by asterisks marine omega-3 fatty acids upon plasma and cellular lipids, platelet function, and Ͻ eicosanoid formation in humans. J. Clin. Invest. 76:1626. (** indicates p 0.01). 4. Lee, T. H., J. M. Mencia-Huerta, C. Shih, E. J. Corey, R. A. Lewis, and K. F. Austen. 1984. Effects of exogenous arachidonic, eicosapentaenoic, and docosachaenoic acids on the generation of 5-lipoxygenase pathway products by ionophore-activated human neutrophils. J. Clin. Invest. 74:1922. expression of the cells in a dose-dependent manner, whereas, in 5. Lee, T. H., R. L. Hoover, J. D. Williams, R. I. Sperling, J. Ravalese, B. W. Spur, contrast and as previously indicated, purified PG of the E series D. R. Robinson, E. J. Corey, R. A. Lewis, and K. F. Austen. 1985. Effect of http://www.jimmunol.org/ dietary enrichment with eicosapentaenoic and docosahexaenoic acids on neutro- decreased the procoagulant activity of PBMC (19, 20) (Table III). phil and monocyte leukotriene generation and neutrophil function. N. Engl. In addition, furegrelate, a TX synthase inhibitor, and SQ29,548, a J. Med. 312:1217. TXA /endoperoxide receptor antagonist, abolished the enhancing 6. Corey, E. J., C. Shih, and J. R. Cashman. 1983. Docosahexaenoic acid is a strong 2 inhibitor of prostaglandin but not leukotriene biosynthesis. Proc. Natl. Acad. Sci. effect of AA on the procoagulant activity of PBMC (Table I). USA 80:3581. Furthermore, our results suggest that there is a balance between the 7. Johnson, B. F., A. S. Daoud, J. Jarnolich, D. Hosmer, and M. H. Johnson. 1986. respective anti- and procoagulant effects of PGE and endoperox- Inhibition of atherosclerosis by cod-liver oil in a hyperlipidemic swine model. 2 N. Engl. J. Med. 315:841. ides/TXA2. The variability of response of PBMC from different 8. Harker, L. A., A. B. Kelly, S. R. Hanson, W. Krupski, A. Bass, B. Osterud, G. A. FitzGerald, S. H. Goodnight, and W. E. Connor. 1993. Interruption of blood donors to AA was positively related to the respective by guest on September 27, 2021 vascular thrombus formation and vascular lesion formation by dietary n-3 fatty amounts of TXB2 and PGE2 released by the cells (Fig. 1). Simi- acids in fish oil in nonhuman primates. Circulation 87:1017. larly, a recent study showed that levels of cytokine production (i.e., 9. Dyerberg, J., H. O. Bang, E. Stoffersen, S. Moncada, and J. R. Vane. 1978. Eicosa- TNF-␣ and IL-1␤) in human monocytes were determined in part pentaenoic acid and prevention of thrombosis and atherosclerosis? Lancet ii:117. 10. Burr, M. L., J. F. Gilbert, R. M. Holliday, P. C. Elwood, A. M. Fehily, S. Rogers, by the balance between TXA2 and PGE2 production (33). P. M. Sweetnam, and N. M. Deadman. 1989. Effects of changes in fat, fish, and We showed that in vitro EPA and DHA, unlike AA, did not fibre intakes on death and myocardial reinfarction: diet and reinfarction trial enhance the TF-dependent procoagulant activity of PBMC (Fig. (DART). Lancet ii:757. 11. Endres, S., R. Ghorbani, V. E. Kelley, K. Georgilis, G. Lonnemann, 3). Similarly, administration of diets rich in n-3 fatty acids to an- J. W. M. van der Meer, J. G. Cannon, T. S. Rogers, M. S. Klempner, P. C. Weber, imals or humans results in a reduction of TF activity by monocytes E. J. Schaeffer, S. M. Wolff, and C. A. Dinarello. 1989. The effect of dietary sup- (8, 12, 13). Our results indicate that this effect is related to changes plementation with n-3 polyunsaturated fatty acids on the synthesis of interleukin-1 and tumor necrosis facto by mononuclear cells. N. Engl. J. Med. 320:265. in the cyclo-oxygenase metabolites produced in the presence of 12. Hansen, J. B., J. O. Olsen, L. Wilsgard, and B. Osterud. 1989. Effects of dietary fatty acids contained in fish oil compared with those produced by supplementation with cod liver oil on monocyte thromboplastin synthesis, coag- AA. When fish oils are fed to humans, a substantial amount of the ulation and fibrinolysis. J. Intern. Med. 225(Suppl.):133. 13. Tremoli, E., S. Eligini, S. Colli, P. Maderna, P. Rise, F. Pazzucconi, n-3 polyunsaturated fatty acids, EPA and DHA, accumulates in F. Marangoni, C. R. Sirtori, and C. Galli. 1994. n-3 fatty acid ethyl ester admin- cell membrane phospholipids, and the AA content decreases (3). istration to healthy subjects and to hypertriglyceridemic patients reduces tissue Thus, the reduction of TF activity by n-3 fatty acids is not imme- factor activity in adherent monocytes. Arterioscler. Thromb. 14:1600. 14. Gregory, S. A., R. S. Kornblath, H. Helin, H. G. Renold, and T. S. Edgington. diate and appears well correlated to changes in the fatty acid com- 1986. Monocyte procoagulant. J. Immunol. 137:3231. position of cell membranes (13). With n-3 fatty acids, lower levels 15. Altieri, D. C., J. H. Morrissey, and T. S. Edgington. 1988. Adhesive receptor of TXA , a procoagulant mediator, are produced (3). EPA is con- Mac-1 coordinates the activation of factor X on stimulated cells of monocytic and 2 myeloid differentiation: an alternative initiation of the coagulant protease cas- verted by cyclo-oxygenase to the 3-series TX and PG (2, 3). PGH3 cade. Proc. Natl. Acad. Sci. USA 85:7462. 16. McGee, M. P., and L. C. Li. 1991. Functional difference between intrinsic and and TXA3 are relatively inactive with regard to platelet aggrega- extrinsic coagulation pathways: kinetics of factor X activation on human mono- tion and are synthesized in small amounts, since EPA is a rela- cytes and alveolar macrophages. J. Biol. Chem. 266:8079. tively poor substrate for cyclo-oxygenase. DHA is not metabolized 17. Robinson, R. A., L. Worfolk, and P. B. Tracy. 1992. Endotoxin enhances the by cyclo-oxygenase and directly inhibits this enzyme (6), as sug- expression of monocyte prothrombinase activity. Blood 79:406. 18. Rapaport, S. I., and L. V. M. Rao. 1992. Initiation and regulation of tissue factor- gested by the results presented in Table II. dependent blood coagulation. Arterioscler. Thromb. 12:1111. This study is the first to show that endoperoxides and TXA2 19. Crutchley, D. J., and M. J. Hirsh. 1991. The stable prostacyclin analog, Iloprost, and enhance the TF-dependent procoagulant activity of PBMC. Pre- prostaglandin E1 inhibit monocyte procoagulant activity in vitro. Blood 78:3852. 20. Archipoff, G., A. Beretz, K. Bartha, C. Brisson, C. de la Salle, C. Froget-León, vious works have indicated that PGE2 decreases endotoxin-in- C. Klein-Soyer, and J. P. Cazenave. 1993. Role of cyclic AMP in promoting the duced TF synthesis in human monocytes by elevating cAMP thromboresistance of human endothelial cells by enhancing thrombomodulin and through stimulation of adenylate cyclase (34). Therefore, we decreasing tissue factor activities. Br. J. Pharmacol. 109:18. 21. Lorenzet, R., J., Niemetz, A. J. Marcus, and M. J. Broekman. 1986. Enhancement suggest that endoperoxides/TXA2 enhanced TF expression of of mononuclear procoagulant activity by platelet 12-hydroxyeicosatetraenoic PBMC by inhibiting adenylate cyclase, lowering cAMP, and acid. J. Clin. Invest. 78:418. 6150 ARACHIDONIC ACID AND MONONUCLEAR CELL TISSUE FACTOR

22. Balter, M. S., G. B. Toews, and M. Peters-Golden. 1989. Different patterns of 31. Miles, J., R. Glasscock, J. Aikens, J. Geich, and M. A. Haymond. 1983. A mi- arachidonate metabolism in autologous human blood monocytes and alveolar crofluorometric method for the determination of free fatty acids in plasma. macrophages. J. Immunol. 142:602. J. Lipid Res. 24:96. 23. Jaffe, E. A., R. L. Nachman, C. G. Becker, and C. R. Minick. 1973. Culture of 32. Heimberg, M., G. D., Dunn, and H. G., Wilcox. 1974. The derivative of plasma- human endothelial cells derived from umbilical veins: identification by morpho- free fatty acids from dietary neutral fat in man. J. Lab. Clin. Med. 83:393. logic and immunologic criteria. J. Clin. Invest. 52:2745. 24. Fu, J. Y., J. L. Masferrer, K. Seibert, A. Raz, and P. Needleman. 1990. The 33. Caughey, G. E., M. Pouliot, L. G. Cleland, and M. J. James. 1997. Regulation of ␣ ␤ induction and suppression of prostaglandin H2 synthase (cyclooxygenase) in hu- tumor necrosis factor- and IL-1 synthesis by thromboxane A2 in nonadherent man monocytes. J. Biol. Chem. 265:16737. human monocytes. J. Immunol. 158:351. 25. Hempel, S. L., M. M. Monick, and G. W. Hunninghake. 1994. LPS induces 34. Ollivier, V., S. Houssaye, C. Ternisien, A. León, H. de Verneuil, C. Elbim, prostaglandin H synthase-2 protein and mRNA in human alveolar macrophages N. Mackman, T. S. Edgington, and D. de Prost. 1993. Endotoxin-induced tissue and blood monocytes. J. Clin. Invest. 93:291. factor messenger RNA in human monocytes is negatively regulated by a cyclic 26. Kaminski, W. E., E. Jendraschak, K. Baumann, R. Kiefl, S. Fisher, A. J. Marcus, AMP-dependent mechanism. Blood 81:973. M. J. Broekman, and C. von Schacky. 1996. Human mononuclear cells express 12-LX: coordinated mRNA regulation with 5-LX and FLAP genes. Blood 87:331. 35. Spagnuolo, P. J., J. J. Ellner, A. Hassid, and M. J. Dunn. 1980. Thromboxane A2 27. Samuelsson, B., S. E. Dahleń, J. A. Lindgren, C. A. Rouzer, and C. N. Serhan. mediates augmented polymorphonuclear leukocyte adhesiveness. J. Clin. Invest. 1987. Leukotrienes and lipoxins: structures, biosynthesis, and biological effects. 66:406. Science 237:1171. 36. Spagnuolo, P. J., J. J. Ellner, A. Hassid, and M. J. Dunn. 1988. Mediation of 28. Crutchley, D. J. 1985. Hydoxyeicosatetraenoic acids and other unsaturated fatty augmented monocyte adhesiveness by thromboxane. Inflammation 12:1. acids inhibit endotoxin-induced thromboplastin activity in human monocytes. 37. Wiles, M. E., R. Welboun, G. Goldman, H. B. Helchtman, and D. Shepro. 1991. Biochem. Biophys. Res. Commun. 132:67. Thomboxane-induced neutrophil adhesion to pulmonary microvascular and aortic 29. Brideau, C., S. Kargman, S. Liu, A. L. Dallob, E. W. Ehrich, I. W. Rodger, and C. C. Chan. 1996. A human whole blood assay for clinical evaluation of bio- endothelium is regulated by CD18. Inflammation 15:181. chemical efficacy of cyclooxygenase inhibitors. Inflamm. Res. 45:68. 38. Goldman, G., R. Welbourn, C. R. Valeri, D. Shepro, and H. B. Hechtman. 1991.

30. Crutchley, D. J. 1984. Effects of inhibitors of arachidonic acid metabolism on throm- Thromboxane A2 induces leukotriene B4 synthesis that in turn mediates neutro- Downloaded from boplastin activity in human monocytes. Biochem. Biophys. Res. Commun. 116:179. phil diapedesis via CD18 activation. Microvasc. Res. 41:367. http://www.jimmunol.org/ by guest on September 27, 2021