Calcium/Calmodulin-dependent Conversion of 5-Oxoeicosanoids to 6,7-Dihydro Metabolites by a Cytosolic Olefin Reductase in Human Neutrophils*

We previously showed that 6-trans isomers of leukotriene B4 but not leukotriene B4itself are converted to dihydro metabolites by human neutrophils. The first step in the formation of these metabolites is oxidation of the 5-hydroxyl group by 5-hydroxyeicosanoid dehydrogenase. The objective of the present investigation was to characterize the second step in the formation of the dihydro metabolites, reduction of an olefinic double bond. We found that the olefin reductase reduces the 6,7-double bond of 5-oxoeicosanoids, is localized in the cytosolic fraction of neutrophils, and requires NADPH as a cofactor. Neutrophil cytosol converts a variety of both 5-oxo- and 15-oxoeicosanoids to dihydro products. However, conversion of 5-oxoeicosanoids to their 6,7-dihydro metabolites is inhibited by EGTA and a calmodulin antagonist and stimulated by the addition of calcium and calmodulin, whereas the reduction of 15-oxoeicosanoids to their 13,14-dihydro metabolites is slightly inhibited by calcium. Furthermore, eicosanoid Δ6- and Δ13-reductases could be separated by chromatography on DEAE-Sepharose. 5-Oxo-6,8,11,14-eicosatetraenoic acid (5-oxo-ETE) is converted by the Δ6-reductase to 6,7-dihydro-5-oxo-ETE, which is 1000 times less potent than 5-oxo-ETE in mobilizing calcium in neutrophils. We conclude that neutrophils contain both 5-oxoeicosanoid Δ6-reductase and prostaglandin Δ13-reductase. Metabolism of 5-oxo-ETE by the Δ6-reductase results in loss of its biological activity.

A major pathway in the metabolism of many eicosanoids is initiated by oxidation of one of the hydroxyl groups by an NAD ϩ -or NADP ϩ -dependent dehydrogenase. This is usually followed by reduction of an adjacent double bond by an olefin reductase in the presence of NADH or NADPH. A number of distinct cytosolic 15-hydroxyprostaglandin dehydrogenases oxodize various prostaglandins (PGs) 1 to their biologically inactive 15-oxo metabolites (1)(2)(3). These products can then be reduced by cytosolic PG ⌬ 13 -reductases to biologically inactive 13,14dihydro-15-oxo-PGs (1;4), which in turn can be further reduced to dihydro-PGs by ketoreductases (5).
We previously showed that neutrophils convert 6-trans isomers of LTB 4 , which are formed nonenzymatically from LTA 4 , to dihydro metabolites (15,16). This reaction proceeds by a sequence analogous to that described above for LTB 4 , the initial step being oxidation of the 5-hydroxyl group, followed by reduction of one of the double bonds and the oxo group (16). We initially speculated that the dihydro products of these reactions might have been 6,11-dihydro metabolites, due to migration of the two remaining double bonds. However, mass spectral evidence subsequently suggested that the products were 6,7-dihydro metabolites (17). The initial step in the formation of these substances is oxidation of the 5-hydroxyl group by a microsomal NADP ϩ -dependent dehydrogenase that is highly specific for eicosanoids containing a (5S)-hydroxyl group followed by a 6-trans double bond (18). LTB 4 , which has a 6-cis double bond, is not metabolized by this pathway. The best substrate for 5-hydroxyeicosanoid dehydrogenase is (5S)-hydroxy-6,8,11,14eicosatetraenoic acid (5-HETE), which is converted to 5-oxo-ETE (18), a potent activator of neutrophils (19,20) and eosinophils (21)(22)(23).
Relatively little is known about the olefin reductase that converts 5-oxoeicosanoids to their dihydro metabolites. The objectives of this study were to investigate the regulation of this enzyme, its substrate specificity, and its subcellular localization. We also wanted to determine whether 5-oxo-ETE could be converted to a dihydro metabolite by this pathway and, if so, how this would affect its biological activity.
Preparation of Subcellular Fractions from Neutrophils-Human neutrophils were purified by treatment of blood with Dextran T-500, centrifugation over Ficoll-Paque, and hypotonic lysis of the remaining red cells (30). The cells (25 ϫ 10 6 /ml) were suspended in 20 mM phosphate buffer, pH 7.4, containing 0.3 M sucrose, phenylmethylsulfonyl fluoride (1 mM), leupeptin (2 g/ml), and aprotinin (2 g/ml). The neutrophils were then disrupted by sonication (model 4710 Ultrasonic Homogenizer; Sonics & Materials, Danbury, CT) in an ice bath for 2 ϫ 5 s at a power setting of 1 and for a further 5 s at a power setting of 2. The sonicate was centrifuged successively at 1500 ϫ g for 10 min, 10,000 ϫ g for 10 min, and 200,000 ϫ g for 60 min. The 10,000 ϫ g and 200,000 ϫ g pellets were suspended in phosphate-buffered saline (1.25 times the original volume) containing calcium and magnesium.
Analysis of Metabolites by Precolumn Extraction/Reversed-phase High Pressure Liquid Chromatography (RP-HPLC)-Fractions obtained as described above were incubated with various substrates, and the reactions were terminated by the addition of methanol (0.6 ml) and stored at Ϫ80°C until analysis by RP-HPLC. After the samples were thawed, water was added to give a final volume of 4 ml (i.e. 15% methanol). Eicosanoids were analyzed by automated precolumn extraction/RP-HPLC as described previously (31). Products were detected and UV spectra recorded using a Waters model M991 diode array detector. Dihydro products were quantitated on the basis of UV absorbance or (when 5-oxo- were the substrates) measurement of radioactivity. Different conditions were used for the analysis of metabolites of oxoeicosanoids (i.e. 5-oxo-ETE and 8-trans-5-oxo-ETE), hydroxyoxoeicosanoids, and 15-oxo-PGF 2␣a . All mobile phases contained 0.02% acetic acid. Conditions for oxoeicosanoids were as follows: Spherisorb ODS-2 column (3.2 ϫ 250 mm; 5-m particle size; Phenomenex); 60% acetonitrile in water, isocratic for 40 min at 0.5 ml/min. Conditions for hydroxyoxoeicosanoids were as follows: Novapak C 18 column (3.9 ϫ 150 mm; Waters); linear gradient between 37 and 45% acetonitrile over 30 min at 1 ml/min. Conditions for 15-Oxo-PGF 2␣a were as follows: Novapak C 18 column; 31% acetonitrile, isocratic at 1 ml/min. PGB 2 (250 ng/sample) was used as an internal standard.
Localization of the Positions of the Double Bonds in Dihydro-5-oxo-12-epi-6-trans-LTB 4 -12-epi-6-trans-LTB 4 (2 M) was incubated with the 1500 ϫ g supernatant fraction from human neutrophils for 90 min at 37°C in the presence of NADP ϩ (1 mM). The reaction was terminated by the addition of methanol (0.5 volumes). Water was added to give a final concentration of methanol of 15%, and the mixture was centrifuged at 1000 ϫ g for 10 min. The supernatant was extracted without acidification on a C 18 Sep-Pak (Waters-Millipore) as described previously (32). The methyl formate fraction was evaporated to dryness under a stream of nitrogen, and the residue (containing 5-oxo-12-epi-6trans-LTB 4 ) was incubated with the 200,000 ϫ g supernatant fraction from human neutrophils for 90 min at 37°C in the presence of calcium (1 mM) and NADPH (1 mM). The products were extracted using octadecylsilyl silica as described above. RP-HPLC analysis of an aliquot of the methyl formate fraction after the first extraction (material from the incubation with the 1500 ϫ g supernatant) confirmed that the major product was 5-oxo-12-epi-6-trans-LTB 4 (18). Dihydro-5-oxo-12-epi-6trans-LTB 4 , the major product of the incubation with the 200,000 ϫ g supernatant fraction, was purified by RP-HPLC as described above and incubated at a concentration of 2 M with the 200,000 ϫ g pellet obtained from porcine neutrophils, prepared as described previously (6), in the presence of NAD ϩ (1 mM). The products of the reaction were analyzed by precolumn extraction/RP-HPLC as described above.
Protein Determination-Protein concentrations were determined as described by Bradford (33).
Measurement of Cytosolic Calcium Levels-Calcium levels were measured in indo-1-loaded neutrophils as described previously (34), using a Photon Technology International (PTI) Deltascan 4000 spectrofluorometer with a temperature-controlled cuvette holder equipped with a magnetic stirrer.

Subcellular Localization of Olefin
Reductase Activity in Human Neutrophils-We had previously shown that 12-epi-6trans-LTB 4 is converted to dihydro and dihydro-5-oxo metabolites by a 1500 ϫ g supernatant fraction from human neutrophils. To investigate the subcellular localization of the olefin reductase required for the formation of these products, 5-oxo-12-epi-6-trans-LTB 4 was incubated with subcellular fractions from neutrophils in the presence of different cofactors. When 5-oxo-12-epi-6-trans-LTB 4 was incubated with a microsomal fraction from neutrophils in the presence of NADPH, the major metabolite was the ketoreductase product, 12-epi-6trans-LTB 4 (12e-6t-B 4 ) (Fig. 1A). Only a small amount of a dihydro product (dh-12e-B 4 ) was formed under these conditions. In contrast, the major product formed when 5-oxo-12-epi-6-trans-LTB 4 was incubated with the cytosolic fraction from neutrophils in the presence of NADPH was its dihydro metabolite (dh-5o-12e-B 4 ) (Fig. 1B).
The amounts of the above metabolites of 5-oxo-12-epi-6trans-LTB 4 formed by different subcellular fractions from neutrophils are shown in Table I. The major product formed by both the 10,000 and 200,000 ϫ g pellets was the ketoreductase product 12-epi-6-trans-LTB 4 . Formation of this product by particulate fractions was dependent upon the presence of NADPH, NADH being much less effective. It was formed to a lesser extent by the 200,000 ϫ g supernatant, but in this case its formation was not affected by the addition of cofactors, perhaps because the level of endogenous cofactor was sufficient. In contrast, the two particulate fractions tested displayed relatively little or no olefin reductase activity, whereas the 200,000 ϫ g supernatant was quite active in the formation of dihydro metabolites of 5-oxo-12-epi-6-trans-LTB 4 . The cytosolic olefin reductase activity was highly dependent on the presence of NADPH, with much smaller amounts of dihydro metabolites being formed in the absence of exogenous cofactors or in the presence of NADH.
Position of Double Bond Reduced by the Olefin Reductase-We had previously suggested that 12-epi-6-trans-LTB 4 was converted to a 6,11-dihydro metabolite by intact neutro-phils (16). However, identification of the positions of the double bonds in this product was not very conclusive, because the diagnostic fragment ions in its mass spectrum were not very intense and could possibly have arisen as a result of rearrangements. A recent study employing mass spectral analysis of fragments formed by oxidative ozonolysis provided evidence that it is the 6,7-double bond of 6-trans isomers of LTB 4 that is reduced by these cells (17). We used a different approach to investigate the position of the reduced double bond of 12-epi-6-trans-LTB 4 . As shown in Fig. 2, 5-oxo-12-epi-6-trans-LTB 4 could potentially be reduced to three products by an olefin reductase, resulting in 6,7-dihydro, 6,11-dihydro, or 10,11-dihydro metabolites. Reduction of the 9,10-double bond in the middle of the triene chromophore is unlikely, whereas reduction of the 14,15-double bond is theoretically possible but would not result in a change in the UV spectrum of the product. Reduction of the 10,11-double bond of 12-epi-6-trans-LTB 4 ( Fig.  2) can also be excluded, since the resulting product would have a max around 280 nm, which is not observed.
To determine whether 5-oxo-12-epi-6-trans-LTB 4 is converted to a 6,7-dihydro or a 6,11-dihydro metabolite, this substance was incubated with microsomal 12-hydroxyeicosanoid dehydrogenase from porcine neutrophils in the presence of NAD ϩ . Oxidation of the 12-hydroxyl group of 6,7-dihydro-5oxo-12-epi-LTB 4 by 12-hydroxyeicosanoid dehydrogenase would give a 5,12-dioxo product absorbing at 280 nm, whereas oxidation of the 12-hydroxyl group of 6,11-dihydro-5-oxo-12epi-LTB 4 would not result in any change in the max of the substrate (Fig. 2). RP-HPLC analysis of the metabolites of dihydro-5-oxo-12-epi-LTB 4 formed by porcine microsomes indicated that two fewer polar products were formed, presumably due to oxidation of the 12-hydroxyl group (Fig. 3A). The major product had a max at 280 nm and was therefore a 6,7-dihydro product (structure I in Fig. 2), whereas a minor product had a max at 231 nm and was probably identical to the 6,11-dihydro compound (structure II in Fig. 2) (Fig. 3B). This demonstrates that the cytosolic fraction from human neutrophils reduces 5-oxo-12-epi-6-trans-LTB 4 principally by 1,2-addition to the triene chromophore and suggests that 1,6-addition may also occur to some extent. For the purpose of clarity, this activity will be referred to below as ⌬ 6 -reductase activity. Time Course for the Formation of Dihydro Metabolites of 5-oxo-15-HETE-Other 5-oxoeicosanoids were also metabolized by cytosolic fractions from neutrophils in a manner analogous to that shown for 12-epi-6-trans-LTB 4 in Fig. 1. 5-Oxo-15-HETE was converted to 5,15-diHETE and two dihydro products, presumably 6,7-dihydro-5-oxo-15-HETE and 6,7-dihydro-5,15-diHETE. The time course for the formation of these three metabolites is shown in Fig. 4. The initial ⌬ 6 -reductase and ketoreductase products (dihydro-5-oxo-15-HETE and 5,15-diHETE) were formed fairly rapidly and reached maximal levels by about 90 min, after which time the amounts declined. The product formed by a combination of the two pathways (dihydro-5,15-diHETE) was formed much more slowly and did not appear to have reached maximal levels by 120 min.

TABLE I
Metabolism of 5-oxo-12-epi-6-trans-LTB 4 by subcellular fractions from human neutrophils 5-Oxo-12-epi-6-trans-LTB 4 (2 M) was incubated for 30 min at 37°C with different subcellular fractions from human neutrophils (equivalent to 2 ϫ 10 7 cells/ml) in the presence or absence of NADH (1 mM) or NADPH (1 mM). The products were quantitated by RP-HPLC as described in the legend to Fig. 1 metabolized more slowly, but this may have been due at least in part to the conversion of these substances by a competitive pathway to 15-hydroxy products due to the presence of 15lipoxygenase in the cytosol. The ⌬ 6 -reductase appears to prefer substrates with an 8-trans double bond, since both 8-trans-5-oxo-ETE and 8-trans-5-oxo-15-HETE were metabolized more rapidly than the corresponding 8-cis compounds. However, of all of the products tested, 15-oxo-5-HETE was by far the best substrate, being metabolized at a rate at least 3 times that of the 5-oxoeicosanoids tested. This raised the possibility that the cytosolic reductase was actually a ⌬ 13 -reductase that was also capable of reducing the 6,7-double bond of 5-oxoeicosanoids. Effects of Calcium on the Formation of Dihydroeicosanoids by Neutrophil Cytosol-All of the experiments described above were performed in the presence of calcium (1 mM). To determine whether the conversion of oxoeicosanoids to dihydro products was affected by calcium, neutrophil cytosol was incubated with various substrates in calcium-free medium in the presence of EGTA (1 mM). Removal of calcium inhibited the conversion to dihydro metabolites of the three 5-oxoeicosanoids tested (5-oxo-6-trans-LTB 4 , 5-oxo-15-HETE, and 8-trans-5-oxo-15-HETE) by between 70 and 80% (p Ͻ 0.01) (Fig. 5). In contrast, conversion of 15-oxo-5-HETE to its dihydro metabolite was stimulated by about 25% (p Ͻ 0.05) in the presence of EGTA. This experiment thus provides strong evidence that neutrophil cytosol contains at least two distinct olefin reductases and that the activity of one of these is enhanced by calcium.
Separation of Eicosanoid ⌬ 6 -and ⌬ 13 -Reductases by Ion Exchange Chromatography-The experiments described above suggested that neutrophil cytosol contains both ⌬ 6 -and ⌬ 13reductases. To attempt to separate these two activities, the cytosol was applied to a column of DEAE-Sepharose, which was washed with 20 mM phosphate buffer, pH 7.4, and eluted with increasing concentrations of NaCl in the same buffer (Fig. 7). The ⌬ 6 -reductase activity of each of the column fractions was estimated by incubation with 5-oxo-6-trans-LTB 4 in the presence of Ca 2ϩ (1 mM) and NADPH (1 mM). ⌬ 13 -Reductase activity was determined by incubating column fractions with 15-oxo-5-HETE in the presence of EGTA and NADPH. The ⌬ 13 -reductase activity was not retained by the DEAE-Sepharose and appeared in the flow-through fraction (Fig. 7). On the other hand, the ⌬ 6 -reductase was strongly retained by the column and was eluted with 250 mM NaCl. This fraction did not contain significant 15-lipoxygenase activity, which resulted in the metabolism of 5-oxo-ETE and 8-trans-5-oxo-ETE to 15-hydroxy products when they were incubated with neutrophil cytosolic fractions (data not shown).
Properties of the Partially Purified ⌬ 6 -Reductase-The requirements of the DEAE-Sepharose-purified ⌬ 6 -reductase for cofactors and calcium were investigated using 5-oxo-6-trans-LTB 4 as a substrate. Removal of calcium by chelation with EGTA inhibited ⌬ 6 -reductase activity in the 250 mM NaCl column fraction by about 40%, whereas the addition of calmodulin resulted in an increase in enzyme activity of about 73% (Table III). The effect of calmodulin was nearly completely inhibited by the addition of EGTA. The reductase reaction was dependent on NADPH. No products could be detected in the absence of cofactors, whereas activity was substantially reduced when NADH was substituted for NADPH.
The substrate specificity of the ⌬ 6 -reductase was investigated (Table II). These experiments were performed in the presence of calcium and NADPH. However, calmodulin was not included for reasons of cost, and for this reason, the rates of metabolism of most of the substrates were lower in the DEAE-Sepharose fraction than in the cytosol. A variety of 5-oxoeicosanoids are substrates for the ⌬ 6 -reductase. As observed for the cytosolic fraction, substrates containing 8-trans double bonds  (8-trans-5-oxo-15-HETE and 8-trans-5-oxo-ETE) are better substrates than their 8-cis counterparts (5-oxo-15-HETE and 5-oxo-ETE, respectively). Of the substrates tested, 8-trans-5oxo-ETE is the best. Unlike neutrophil cytosol, the 250 mM NaCl fraction was unable to reduce either 15-oxo-5-HETE or 15-oxo-PGF 2␣ to dihydro metabolites. Table II, both 5-oxo-ETE and 8-trans-5-oxo-ETE are metabolized to dihydro products by the neutrophil ⌬ 6 -reductase. To confirm the identities of these products, 5-oxo- [11,12,14, H]ETE and 8-trans-5-oxo- [11,12,14, H]ETE, both in the absence of unlabeled substrates, were incubated with the 250 mM NaCl DEAE-Sepharose fraction in the presence of NADPH and Ca 2ϩ . After termination of the reactions, authentic chemically synthesized 6,7-dihydro-5-oxo-ETE was added, and the products were analyzed by RP-HPLC. As shown in Fig. 8A, the major metabolite of 5-oxo-[ 3 H]ETE cochromatographed with 6,7-dihydro-5-oxo-ETE, which was detected at 200 nm, whereas a smaller amount of 5-HETE was formed. 8-trans-5-Oxo- [11,12,14, H]ETE was also converted principally to a dihydro metabolite that had a longer retention time than 6,7-dihydro-5-oxo-ETE, presumably because of the different configuration of the 8,9-double bond (Fig. 8B). Although only a small amount of 8-trans-5-HETE was detected, a product with a slightly longer retention time, presumably identical to 6,7-dihydro-8-trans-5-HETE, was present.

Formation of 6,7-Dihydro Metabolites of 5-Oxo-ETE and 8-trans-5-Oxo-ETE-As shown in
To determine the K m for the ⌬ 6 -reductase, various concentrations of 8-trans-5-oxo-[ 3 H]ETE were incubated with neutrophil cytosol, after removal of 5-ketoreductase and 15-lipoxygenase activities by precipitation with ammonium sulfate (40%). This was necessary because the latter enzymes competed with the ⌬ 6 -reductase for the substrate. We did not use the DEAE-Sepharose fraction for this experiment because a substantial degree of enzyme activity was lost, probably due to the removal of calmodulin and other components. Unfortunately, we could use calmodulin only sparingly in these experiments because of the cost, and it was not included in the determination of the K m and V max values. Lineweaver-Burk analysis (Fig. 9) revealed an apparent K m for 8-trans-5-oxo-ETE of 130 Ϯ 20 nM and a V max of 3.0 Ϯ 0.3 pmol/min/mg of protein (n ϭ 3).
Biological Activity of 6,7-Dihydro-5-oxo-ETE-Because 5-oxo-ETE is a potent stimulator of calcium mobilization in neutrophils, it was important to determine whether reduction to 6,7-dihydro-5-oxo-ETE affected biological activity. Authentic 6,7-dihydro-5-oxo-ETE was capable of inducing calcium mobilization in neutrophils, but only at very high concentrations, and its potency was about 1000 times lower than that of 5-oxo-ETE (Fig. 10). The effect of 6,7-dihydro-5-oxo-ETE on calcium levels would appear to be mediated by a 5-oxo-ETE receptor, FIG. 7. Chromatographic separation of ⌬ 6 -and ⌬ 13 -reductases on DEAE-Sepharose. Neutrophil cytosol (10 ml) was applied to a column of DEAE-Sepharose (1 ml), which was washed with 20 mM phosphate buffer, pH 7.4, and eluted with increasing concentrations (50, 250, and 1000 mM) of NaCl in phosphate buffer (indicated by arrows). ⌬ 6 -(middle) and ⌬ 13 -(top) reductase activities were determined by incubation of column fractions with 5-oxo-6-trans-LTB 4 and 15-oxo-5-HETE, respectively, in the presence of either Ca 2ϩ or EGTA, respectively, as described under "Experimental Procedures." The amounts of dihydro products formed from each substrate were determined by RP-HPLC as described under "Experimental Procedures." The scale for the ordinate showing ⌬ 13 -reductase activity is 5 times that for ⌬ 6 -reductase activity. Absorbance at 280 nm (bottom) was measured as an index of the protein concentration of each of the column fractions.

H]ETE (B).
After termination of the incubations with methanol, authentic 6,7dihydro-5-oxo-ETE was added, and the products were analyzed by RP-HPLC as described under "Experimental Procedures." Dotted lines, absorbance at 200 nm; solid lines, radioactivity. since pretreatment of neutrophils with a high concentration (10 M) of the former compound desensitized these cells to subsequent addition of 5-oxo-ETE (Fig. 10, inset).
Properties of the Partially Purified ⌬ 13 -Reductase-The effects of cofactors and calcium on ⌬ 13 -reductase activity in the DEAE-Sepharose flow-through fraction were examined using 15-oxo-5-HETE as a substrate (Table III). As observed for the ⌬ 6 -reductase, this reaction was dependent upon NADPH, with much lower activity being observed in the presence of NADH. In agreement with the results with cytosol (Fig. 5), calcium inhibited enzyme activity by about 40%.
The specificity of the olefin reductase in the flow-through fraction was also examined by incubating this fraction with various substrates in the presence of NADPH and EGTA (Table  II). None of the 5-oxoeicosanoids tested were metabolized to a detectable extent by this enzyme, whereas both 15-oxo-5-HETE and 15-oxo-PGF 2␣ were excellent substrates. The best substrate for the ⌬ 13 -reductase was 15-oxo-PGF 2␣ .
The K m and V max of the ⌬ 13 -reductase were determined by incubating neutrophil cytosol fractions with 15-oxo-[ 3 H]PGF 2␣ in the presence of NADPH. The unfractionated cytosolic fraction was used in this case because it did not contain any other enzymes that metabolized this substrate to a significant extent. As illustrated by Fig. 9, the K m for conversion of 15-oxo-PGF 2␣a to its 13,14-dihydro derivative was 221 Ϯ 30 nM, whereas the V max was 11.8 Ϯ 0.2 pmol/min/mg of protein (n ϭ 3). DISCUSSION We previously showed that human neutrophils convert 6-trans isomers of LTB 4 to dihydro metabolites (15,16). The enzyme responsible for this reaction does not act directly on these substrates but rather requires prior oxidation of the 5-hydroxyl group (16,18). The position of the double bond that was reduced was not clear from our initial studies (16). However, a recent study based on mass spectral analysis of fragments formed by oxidative ozonolysis of the dihydro metabolite of 6-trans-LTB 4 provided evidence that the reduced double bond is in the 6,7-position (17). The present study indicates that the two remaining double bonds in the dihydro metabolite of 12-epi-6-trans-LTB 4 are in the 8,9-and 10,11-positions, since they are conjugated with the 12-oxo group formed upon oxidation of the 12-hydroxyl group of 6,7-dihydro-5-oxo-12-epi-LTB 4 by 12-hydroxyeicosanoid dehydrogenase (see Fig. 2). We also obtained evidence suggesting that a small amount of a 6,11dihydro product may have been formed, raising the possibility that the ⌬ 6 -reductase may not be completely specific. However, we cannot exclude the possibility that another enzyme is responsible for the formation of the putative 6,11-dihydro product.
Monocytes have been reported to convert lipoxins to dihydro and dihydro-oxo products. Lipoxin A 4 is converted to 15-oxo, 13,14-dihydro-15-oxo, and 13,14-dihydro metabolites by these cells (35), whereas lipoxin B 4 is converted to dihydro products analogous to those formed from 6-trans isomers of LTB 4 (36). It would seem likely that the enzyme responsible for the formation of 13,14-dihydro metabolites of lipoxin A 4 is PG ⌬ 13 -reductase (35). Although the nature of the enzyme that converts lipoxin B 4 to dihydro products was not investigated further in the above study, it would seem probable that it is identical to the eicosanoid ⌬ 6 -reductase that we have identified in neutrophils.
In the present study, we investigated the specificity of the olefin reductase in neutrophils by synthesizing a series of substrates that could be converted to dihydro products, which could be detected either by UV absorbance or by radioactivity. Initial studies on the metabolism of these substrates by cytosolic fractions from neutrophils suggested that the olefin reductase that converts 5-oxo-12-epi-6-trans-LTB 4 to its dihydro metabolite may not be specific, since 15-oxo-5-HETE was found to be a better substrate than any of the 5-oxoeicosanoids tested. This raised the possibility that the enzyme responsible for this reaction could be a PG ⌬ 13 -reductase. However, the markedly different calcium requirements for reduction of 5-oxo-and 15oxoeicosanoids suggested that this was not the case. This was confirmed when we were able to separate the two activities on a column of DEAE-Sepharose, which retained the ⌬ 6 -reductase but not the ⌬ 13 -reductase. Once the two enzymes were separated, it was apparent that the ⌬ 6 -reductase did not display any ⌬ 13 -reductase activity and vice versa. FIG. 9. Lineweaver-Burk plots for eicosanoid ⌬ 6 -and ⌬ 13 -reductases in neutrophil cytosol fractions. Prior to measurement of ⌬ 6 -reductase activity (q), 5-ketoreductase and 15-lipoxygenase were removed from the cytosol by treatment with ammonium sulfate (final concentration, 40%), followed by centrifugation at 20,000 ϫ g for 10 min. After dialysis against 3 ϫ 2 liters of 20 mM phosphate buffer, pH 7.4, the supernatant was incubated with different concentrations of 8-trans-5-oxo- [11,12,14, 10. Effects of 6,7-dihydro-5-oxo-ETE on calcium mobilization in neutrophils. Various concentrations of 5-oxo-ETE (q) or 6,7dihydro-5-oxo-ETE (E) were added to suspensions of indo-1-loaded neutrophils, and changes in fluorescence were monitored as described under "Experimental Procedures." The data are expressed as percentages of the maximal calcium response to 5-oxo-ETE and are means Ϯ S.E. (n ϭ 3). The inset shows the effects of 6,7-dihydro-5-oxo-ETE (dh-5o; 10 M) on indo-1 fluorescence and on the subsequent response to 5-oxo-ETE (5o; 10 nM). For comparison, the response to 5-oxo-ETE (10 nM) alone is also shown.
The eicosanoid ⌬ 6 -reductase has a fairly low K m (ϳ130 nM) but also a relatively low V max (ϳ3 pmol/min/mg of protein). Thus, this enzyme can efficiently metabolize low, but not high, concentrations of substrate. This is in agreement with our earlier finding that intact neutrophils have only a limited capacity to metabolize 6-trans isomers of LTB 4 to their 6,7-dihydro metabolites (16). The ⌬ 13 -reductase has a somewhat higher K m (ϳ220 nM) than the ⌬ 6 -reductase but also has a higher V max (ϳ12 pmol/min/mg of protein). The difference in the V max values for the two enzymes may be somewhat greater than this, because the ⌬ 6 -reductase was partially purified for this experiment to remove enzymes that competed for the substrate, whereas ⌬ 13 -reductase activity was measured using the unfractionated cytosolic fraction. Thus, the ⌬ 13 -reductase has a considerably higher capacity in neutrophils than the ⌬ 6 -reductase. The K m of the neutrophil ⌬ 13 -reductase is similar to that of a cytosolic NADPH-dependent prostaglandin ⌬ 13 -reductase in rat liver, which was reported to be about 280 nM (4).
The dependence of the ⌬ 6 -reductase on calcium and calmodulin is intriguing and suggests that the activity of this enzyme may be tightly regulated. Other olefin reductases involved in the metabolism of eicosanoids, including PG ⌬ 13 -reductase and the ⌬ 10 -reductase responsible for the formation of 10,11-dihydro metabolites of LTB 4 , are not known to be affected by calmodulin. Similarly, there is little evidence for the regulation of steroid olefin reductases by calmodulin, with the possible exception of a sterol ⌬ 24 -reductase present in hepatoma cells and human skin fibroblasts (37). It is not clear whether calmodulin acts directly on the ⌬ 6 -reductase or whether its actions are mediated by another protein such as a calmodulin-dependent kinase or phosphatase. The relatively low specific activity of the enzyme after chromatography on DEAE-Sepharose suggests that factors other than calmodulin may also be involved. Despite removal of a substantial amount of protein by the chromatographic procedure, the ⌬ 6 -reductase activity in the presence of calmodulin after DEAE-Sepharose (13.5 Ϯ 0.8 pmol/mg/min; Table III) is no higher than that in the cytosol in the presence of calmodulin (15.20 Ϯ 1.48 pmol/mg of protein/ min; Fig. 6, inset).
The 5-hydroxyeicosanoid dehydrogenase/⌬ 6 -reductase pathway was first discovered in studies on the metabolism of 6-trans isomers of LTB 4 , which are formed nonenzymatically from LTA 4 and have little biological activity (15,16). The biological significance of this pathway was unclear until we found that the preferred substrate for the first step, catalyzed by the dehydrogenase, is 5-HETE, which is converted into a biologically active product, 5-oxo-ETE (18). A major objective of the present study was to determine whether 5-oxo-ETE is a substrate for the ⌬ 6 -reductase and, if so, whether the product, 6,7-dihydro-5-oxo-ETE, is more or less potent than its precursor. To accomplish this, we prepared tritium-labeled 5-oxo-ETE (25), which would allow us to monitor the formation of its 6,7-dihydro metabolites, which, unlike other lipoxygenase products, do not absorb significantly in the UV. Furthermore, to enable us to identify the putative 6,7-dihydro metabolite and to test its biological activity, we prepared this compound by chemical synthesis (24).
Our results clearly show that both 5-oxo-ETE and its 8-trans isomer are metabolized by the ⌬ 6 -reductase to dihydro metabolites. Metabolism of 5-oxo-ETE by this enzyme results in a dramatic loss in biological activity, as 6,7-dihydro-5-oxo-ETE is about 1000 times less potent in stimulating calcium mobilization in neutrophils. This further supports the argument that neutrophils possess a highly specific recognition mechanism for 5-oxo-ETE, since a variety of minor structural modifications cause substantial losses in biological activity (19,34,38). It is becoming apparent that metabolism of 5-oxo-ETE by a variety of pathways results in dramatic reductions in biological potency, including metabolism by 20-hydroxylase (100-fold) (34), 12-lipoxygenase (Ͼ10,000), 2 and 5-ketoreductase (100-fold) (19) enzymes. Metabolism of 5-oxo-ETE by the ⌬ 6 -reductase would result in a permanent loss in biological activity, since this reaction is presumably irreversible, in contrast to reduction of 5-oxo-ETE to 5-HETE by a 5-ketoreductase.
It is interesting that the 8-trans isomer of 5-oxo-ETE appears to be a better substrate for the ⌬ 6 -reductase than 5-oxo-ETE itself. Indeed, this enzyme shows a preference for substrates with a 5-oxo group followed by two trans double bonds. Metabolism of 8-trans-5-oxo-ETE could be of some significance, since this substance does possess some biological activity, with a potency about one-fifth that of 5-oxo-ETE (34,38). Moreover, we have detected 8-trans-5-oxo-ETE after stimulation of neutrophils (34), although it is not yet clear whether this compound can be formed enzymatically. However, the results of our specificity studies should be interpreted with some caution, since we do not yet understand completely how this enzyme is regulated, and the conditions employed may not be optimal due to possible requirements for additional factors.
In conclusion, human neutrophils possess two distinct olefin reductases that metabolize eicosanoids, a ⌬ 6 -reductase that converts 5-oxo-ETE and other 5-oxoeicosanoids to 6,7-dihydro metabolites and a ⌬ 13 -reductase that converts 15-oxo-PGs and other 15-oxoeicosanoids to 13,14-dihydro metabolites. The ⌬ 6 -reductase is highly regulated by calmodulin and possibly other factors, whereas the ⌬ 13 -reductase is, if anything, slightly inhibited by calcium. Metabolism of 5-oxo-ETE by the ⌬ 6 -reductase results in a dramatic 1000-fold loss in biological potency.