Inhibition of leukotriene biosynthesis by the leukocyte product 15-hydroxy-5,8,11,13-eicosatetraenoic acid.

Rabbit peritoneal polymorphonuclear leukocytes, elicited with glycogen, metabolized added [1-14C]arachidonic acid to the 5-lipoxygenase products 5-hydroxy-6,8,11,14-eicosatetraenoic acid and 5,12-dihydroxy-6,8,10,14-eicosatetraenoic acid (leukotriene B) and the 15-lipoxygenase product 15-hydroxy-5,8,11,13-eicosatetraenoic acid. These metabolites were isolated by high pressure liquid chromatography and converted to the trimethylsilyl-ether methyl ester derivatives, and the structures were confirmed by gas chromatography-mass spectrometry. When polymorphonuclear leukocytes were preincubated with 15-HETE (16 microM), the formation of 5-hydroxy-6,8,11,14-eicosatetraenoic acid and 5,12-dihydroxy-6,8,10,14-eicostatetraenoic acid from [1-14C]arachidonic acid was strongly suppressed. The concentration required for 50% inhibition of the 5-lipoxygenase pathway in these cells was approximately 6 microM, which is comparable to the concentrations (0.20 to 1.8 microM) of 15-hydroxy-5,8,11,13-eicosatetraenoic acid produced in incubations of polymorphonuclear leukocytes with arachidonic acid alone. Recent reports indicate that slow-reacting substance of anaphylaxis (leukotriene C/D) and chemotactic substance leukotriene B are arachidonic acid metabolites formed via the 5-lipoxygenase pathway. Our observations thus suggest that 15-hydroxy-5,8,11,13-eicosatetraenoic acid can regulate the formation of these vasoactive and inflammatory mediators intracellularly.

and hydroperoxyeicosatetraenoic acids (1, 2). The fist three groups of compounds are all derived via the cyclooxygenase pathway, whereas the latter two classes are lipoxygenase metabolites. Mammalian lipoxygenases can be classified into three groups according to the product specificity observed when arachidonic acid is used as a substrate. The platelet lipoxygenase metabolizes arachidonic acid to yield 12-hydroperoxy-5,8,10,14-eicosatetraenoic acid (12-HPETE), and human and rabbit polymorphonuclear * This work was supported by Grants 5R01 HL-22527-03 and lROl HL-24830-01 from the National Institutes of Health and by Grant 78.767 from the American Heart Association. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. leukocytes contain the 5-and 15-lipoxygenases which produce 5-HPETE' and 15-HPETE, respectively (3)(4)(5). 5-HPETE was recently shown to be the precursor of leukotrienes including slow-reacting substance of anaphylaxis (2,6). All these hydroperoxy fatty acids are readily reduced enzymatically to the corresponding hydroxy derivatives. One approach to modulating immediate hypersensitivity reactions would be to inhibit the formation of slow-reacting substance of anaphylaxis.
Recently, two drugs, benoxaprofen and the pyrazoline derivative BW755C were shown to inhibit the 5-lipoxygenase pathway (7,8 ) . We now report that 15-HETE, previously shown to be the stable 15-lipoxygenase end product in human and rabbit polymorphonuclear leukocytes, when added exogenously, inhibits the formation of the 5-lipoxygenase arachidonic acid metabolites 5-HETE and 5,12-diHETE in these leukocyte preparations.
Arachidonic acid was converted into 15-HETE via soybean lipoxygenase (9) and a reference sample of 5-HETE was prepared via single oxygen oxidation of arachidonic acid followed by purification by HPLC according to the method of Porter and co-workers (IO). GC-MS analysis of the methyl ester trimethylsilyl (Me.,Si) ether derivatives of the hydroxy fatty acids was carried out using a GC column (2 mm inside diameter X 91 cm) packed with 3% SP2100 (Supelco, Inc,, Bellefonte, PA) operated at 220'C (11). The methyl ester Me:Si ether derivative of 5-HETE (C21.1) gave prominent high mass ions at m / e (relative abundance and probable mode of origin in parenthe- (4), 255 ( l l ) , 216 (8), 215 (2), 203 (9), and 190 (9). This spectrum was very similar to that reported by Borgeat and co-workers (4). The methyl esters of 15-HETE and 5-HETE were hydrogenated to the respective methyl hydroxyeicosanoates as described by Borgeat and Samuelsson (12) 313 (25), and 203 (100). These data are in agreement with published spectra (4, 13).
Rabbit peritoneal polymorphonuclear leukocytes were prepared according to a modification of the method of Borgeat and Sarnuelsson (12). In short, glycogen-induced (110 ml, 0.2%, 4 h) rabbit peritoneal polymorphonuclear leukocytes were collected from the opened peritoneal cavity after killing the rabbit, filtered, centrifuged, and treated with buffered ammonium chloride solution to lyse contaminating red blood cells. After centrifugation and washing, the cells were resuspended in Dulbecco's phosphate-buffered saline (0.9% NaCl solution) containing 11 mM D-glucose. These cells, identified as polymorphonuclear leukocytes by Wright's stain, exhibited greater than 90% viability, as measured by the trypan blue exclusion test.
The cells (10 to 20 X 1O6/ml), were preincubated either with ethanol (2 pl, control) or with different concentrations of 15-HETE in ethanol for 10 min a t 37°C. [l-'4C]Arachidonic acid (5.4 pg, 1.0 pCi) was then added and the incubations were terminated after 5 min by the addition of 4 volumes of methanol. The mixture was extracted with 8 volumes of chloroform and the chloroform phase was evaporated under N,.
The residue was dissolved in a small amount of chloroform, applied to preactivated Silica Gel G TLC plates, and developed in petroleum ether/diethyl ether/acetic acid (50501, v/v) along with authent.ic standards of 5-HETE and 15-HETE. Radioactive zones were detected by autoradiography, scraped from the TLC plates, and counted by a liquid scintillation counter. Control experiments indicated that the level of platelet contamination of the polymorphonuclear leukocyte preparation did not contribute to the formation of any of the observed arachidonic acid metabolites.
The chemical characterization of the radioactive arachidonic acid metabolites from polymorphonuclear leukocytes was camed out on a chloroform extract from an incubation mixture of polymorphonuclear leukocytes (3 ml, 35 X 10" cells/ml) and arachidonic acid diluted with [I-'4C]arachidonic acid (0.39 pmol, 2.2 pCi/pmol). After evaporation of the chloroform, the arachidonic acid metabolites in the residue were separated by HPLC using a Lichrosorb Si-100 column (1.0 X 25 cm) supplied by Alltech Associates, Arlington Heights, IL. The column was eluted with the solvent hexane/isopropyl alcohol/ methanol/acetic acid (97522:261, v/v) as the mobile phase with a flow rate of 4 ml/min. Ten minutes after injection of the sample, a linear solvent gradient was established to reach the composition of 891:99:91 by 20 min after injection. The effluent was monitored simultaneously at 254 nm (fLued wavelength detector) and 235 nm (variable wavelength detector), and 4-ml fractions were collected. After 20 min, the variable wavelength detector was changed to 275 nm. Aliquots of each fraction were assayed for radioactivity. The ultraviolet absorbance and radioactivity profiles indicated that [I4C115-HETE, ['"C]5-HE:TE, and ["C15,12-diHETE (see below) eluted a t 6.3, 16.5, and 25.2 min. respectively. In separate assays. reference samples of 12-HETE. 15-HETE, and 5-HETE eluted a t 6.16, 6.50, and 16.56 min, respectively. The HPLC-purified, polymorphonuclear leukocyte-derived samples of 15-HETE and 5-HETE and their hydrogenated derivatives were converted to the methyl ester Me:Si ether eicosatetraenoates and methyl ester Me& ether eicosanoates, respectively. The GC retention times (on 3% SI'2100) and mass spectra of these derivatives were in agreement with the data (see above) obtained from authentic 15-HETE and 5-HETE. The 5,12-diHETE fraction ( R P 0.05 in the TLC solvent system described above), showed three ultraviolet absorptions bands a t A?: : ? = 282 * 1, 271 & 1, and 261 f 1 nm, indicative of a triene chromophore (12). This fraction was then methylated, silylated, and subjected to CC-MS analysis using a column packed with 1% SE-30 on Cas-Chrom Q (Supelco). Two major components, identified as methyl ester bis-Me23 derivatives of 5.12-diHETE, eluted with equivalent chain lengths C23.7 and C25.0. Their relative proportions were approximately 2 to 1, respectively. Rabbit platelets were obtained from citrated blood, drawn from the ear vein of New Zealand white rabbits. After centrifugation a t 900 X g for 2 min a t room temperature to sediment the red blood cells, the platelet-rich plasma was removed and recentrifuged a t 900 X g for 20 min. The resultant platelet pellet was washed and resuspended in Dulbecco's phosphate-buffered saline containing glucose. Platelets (2 X 10"/ml) were preincubated for 10 min a t 37°C with 15-HETE. [ I -"C1Arachidonic acid (2.7 to 5.4 pcg, 0.5 to 1.0 pCi) was then added. The identification and quantitation of the platelet arachidonic acid products were carried out as previously described (9).

RESULTS A N D DISCUSSION
Several lipoxygenase products were formed when rabbit polymorphonuclear leukocytes were incubated with [ I4C]arachidonic acid. These include 5-HETE, 5,12-diHETE, and 15-HETE (Fig. 1, Lane 1 ) whose structures were confirmed by GC-MS. The first two metabolites were formed via the 5lipoxygenase pathway and we now confirm that the 15-lipoxygenase, analogous to that observed with human pol.ymorphonuclear leukocytes, also functions in rabbits (4, 5). When rabbit leukocytes were incubated with arachidonic acid in the presence of exogenously added 15-HETE (8 to 16 PM), the formation of 5-HETE and 5,12-diHETE (leukotriene B (2)) was inhibited (Fig. 1, Lanes 2 and 3, 68 to 81% inhibition.' This inhibition was not time-dependent over a 15-min preincubation period and was also observed in the presence of the calcium ionophore A23187. The concentration of exogenouslv added 15-HETE required for 50% inhibition of the 5-lipoxygenase pathway in these leukocytes was approximately 6 PM, which is comparable to the concentrations (0.20 to 1.8 PM) of 15-HETE produced in incubations of polymorphonuclear leukocytes with arachidonic acid alone. The production of [I4C]15-HETE is not appreciably affected by concentrations of exogenous 15-HETE which inhibits the synthesis of the 5lipoxygenase products. However, a t higher concentrations, a feedback type of inhibition of [I4C]15-HETE formation was observed. 15-HETE also blocked the 5-lipoxygenase pathway in human polymorphonuclear leukocytes with approximately the same inhibitory potency (data not shown).
15-HETE is also a very selective inhibitor of rabbit platelet lipoxygenase, as was shown earlier in humans (9). At levels of 15-HETE that inhibited formation of the stable platelet lipox-"HPLC analysis based upon ultraviolet absorbance profiles (cfi "Materials and Methods") confirmed that exogenously added 15-HETE inhibited the formation of 5-HETE and 5.12-diHETE. ygenase end product, 12-HETE, by 90 to 95%, synthesis of the cyclooxygenase products thromboxane B2 and 12-hydroxyheptadecatrienoic acid was not significantly affected (data not shown). Table I summarizes the inhibitory effects of 15-HETE on the formation of arachidonic acid metabolites from three different rabbit lipoxygenases. The rabbit platelet 12-lipoxygenase was 17 times more sensitive to 15-HETE than the polymorphonuclear leukocyte 5-lipoxygenase. Since the concentration of 15-HETE required to inhibit production of 5-HETE and 5,12-diHETE by 50% is approximately the same (Table I), it suggests that 15-HETE affects the lipoxygenase enzyme directly instead of subsequent enzymes such as the peroxidase component of the pathway. The effect of 15-HETE on polymorphonuclear leukocyte lipoxygenases contrasts with that of eicosatetraynoic acid, a well-known inhibitor of platelet 12-lipoxygenase (3) and soybean 15-lipoxygenase (15). Eicosatetraynoic acid inhibits the leukocyte 15-lipoxygenase without blocking the 5-lipoxygenase pathway ( 4 ) , whereas 15-HETE is a selective inhibitor of the latter enzyme (Table 11). This inhibition does not appear to be a general property of hydroxy fatty acids since both the 5-and 15-lipoxygenase activities were stimulated by ricinoleic acid (Table 11).

Inhibition of Leukotriene Biosynthesis by 15-HETE
It has been reported recently that 5,12-diHETE (leukotriene B) is a major inducer of chemotactic responses by polymorphonuclear leukocytes (16, 17). The present results suggest that stimulation of 15-lipoxygenase activity could inhibit the formation of 5,12-diHETE and could thus decrease chemotaxis induced by this compound. Since various arachidonic acid metabolites appear to be involved in the inflam-matory response (l), other aspects of the potential regulatory role of 15-HETE on arachidonic acid metabolism are under investigation.
From the data presented, it can be seen that the stable 15lipoxygenase end product 15-HETE could regulate the formation of different lipoxygenase arachidonate metabolites in platelets and polymorphonuclear leukocytes. A proposed scheme for this regulation is shown in Fig. 2. Preliminary observations indicate that 15-HETE decreases ionophore (A23187)-induced production of slow-reacting substance in the rat peritoneal ~a v i t y .~ Since slow-reacting substance of anaphylaxis (leukotriene C/D) is a 5-lipoxygenase metabolite (2,6), these findings provide additional support that 15-HETE can regulate the formation of slow-reacting substance of anaphylaxis intracellularly. Consequently, 15-HETE (or a derivative) might be of potential therapeutic value in the treatment of inflammatory processes and allergic responses such as anaphylaxis and asthma.