Eggs of the sea urchin, Strongylocentrotus purpuratus, contain a prominent (11R) and (12R) lipoxygenase activity.

Recent work has shown that oocytes of the starfish synthesize (8R)-hydroxyeicosatetraenoic acid and that this eicosanoid has a potent and highly specific action in induction of oocyte maturation. These striking results prompted us to examine the lipoxygenase activity of eggs of the sea urchin Strongylocentrotus purpuratus. Four hydroxyeicosanoids were formed in homogenates of sea urchin eggs; their structures and stereochemistry were characterized by high pressure liquid chromatography, UV spectroscopy, and gas chromatography-mass spectrometry. The compounds were identified as (11R)-hydroxy-5,8,12,14-ZZEZ-eicosatetraenoic acid and (12R)-hydroxy-5,8,10,14-ZZEZ-eicosatetraenoic acid (from arachidonic acid) and the corresponding (11R)- and (12R)-hydroxy analogs of eicosapentaenoic acid. The formation of these egg products was not blocked by a cyclooxygenase inhibitor, indomethacin (10 microM), and their precise structures are consistent with their formation by a lipoxygenase reaction. Eicosapentaenoic acids with a prochiral tritium label in the 10-D or 10-L position were used to investigate the mechanism of biosynthesis. The formation of (12R)-hydroxyeicosapentaenoic acid proceeded with the stereoselective abstraction of the 10-D hydrogen from the substrate. This reaction was shown to be opposite to the (12S) oxygenation catalyzed by porcine leukocyte 12-lipoxygenase. These results with S. purpuratus eggs constitute the first demonstration of (11R)- or (12R)-lipoxygenase activity in any cell type or tissue.

dented in that a single HETE of one chiral configuration expresses an activity which is not surpassed by any other prostaglandin, leukotriene, or hydroxyeicosanoid.
Our interest in these findings drew our attention to a report that addition of Ca2+ to egg homogenates and the fertilization of eggs of the sea urchin, Strongylocentrotus pupuratus, induces the formation of arachidonic acid oxidation products (3, 4). The eggs were reported to synthesize a "HETE-like" product. We have now examined the nature of this oxygenase activity. Although we discovered that lipoxygenase activity is abundant in the sea urchin egg homogenates, no 8-HETE synthesis could be detected. Herein we characterize the unique (11R)-and (12R)-lipoxygenase activity of the eggs of the sea urchin, S. purpuratus.

EXPERIMENTAL PROCEDURES
Chemicals and Reagents-The sources of isotopes, reagents, and chemicals were as follows: purity,>97%) and [1-"C]eicosapentaenoic acid were from Amersham International; unlabeled arachidonic acid (purity, 799%) was from NuChek Prep (Elysian, MN); the unlabeled eicosapentaenoic acid was 90% grade from Sigma and it was purified by RP-HPLC prior to use. Nitrosomethylurea was from ICN Pharmaceuticals. Bis (trimethylsily1)-trifluoroacetamide was obtained from Supelco (Bellefonte, PA). (-)-Menthylchloroformate was obtained from Aldrich. Organic solvents were distilled-in-glass grade obtained from Burdick and Jackson (Muskegon, MI). Trizma (Tris base) and indomethacin were obtained from Sigma.
Isolation of Sea Urchin Eggs-Spawning of the sea urchins, S. purpuratus (Marinus, Inc., Long Beach, CA), was induced by the intracoelomic injection of -5 mi of 0.5 M KCl. After injection with KCl, the sea urchins were inverted in plastic dishes and the orange eggs were harvested in ice-cold artificial seawater containing 454 mM NaCl, 9.7 mM KC1,24.9 mM MgCI,, 27.1 mM MgSO,, 4.4 mM NaHC03, 10 mM Tris (pH 7.9) with either 9.6 mM CaC12 or 2 mM EGTA with only 2.7 mM MgSO, and 2.5 mM MgC1,.
Incubation and Extraction of Egg Homogenates-Eggs, freshly har-

7629
(11 R )and (lBR)-Lipoxygenase Activity in Sea Urchin Eggs vested in artificial seawater, were routinely concentrated by centrifugation for 10 min at 1000 rpm in a Sorvall RC-3 refrigerated centrifuge. The sedimented eggs (1 ml) were homogenized on ice with a Kontes ground glass homogenizer in 10 volumes of 20 mM Tris, pH 7.9; added substrates and/or inhibitors were included in the homogenization buffer. Microscopic examination revealed that eggs begin to spontaneously lyse within minutes after suspension in this hypotonic buffer and that the integrity of the eggs is disrupted completely by a few passes (510) with the ground glass homogenizer pestle. Egg lysate incubations (10 ml) were terminated by the addition of 0.5 ml of a sodium formate buffer (10% formic acid adjusted to pH 3.0 with NaOH) and immediate extraction with methanol/chloroform by the Bligh and Dyer procedure (7). After phase separation, the organic phase contained 95% of added radiolabel; the aqueous phase (pH 3.5-4) was discarded. Incubations with Porcine Leukocyte (12S)-Lipxygenase-The 100,000 X g supernatant fraction was used as the source of enzyme (8). (12S)-HETE and (12s)-HEPE were prepared by incubation of 2 ml of supernatant with 100 p~ arachidonic or eicosapentaenoic acids, respectively. After 5 min at room temperature, the incubations were terminated by acidification with sodium formate buffer and the mixtures applied to C-18 Bond Elut cartridges (Analytichem International, Harbor City, CA) preconditioned with 10 ml of methanol washed with 10 ml of water and the fatty acids eluted with 5 ml of followed by 10 ml of water. After application of sample, columns were methanol. The extracts were treated with a 1.5 molar excess of triphenylphosphine in methanol prior to isolation by HPLC.
Purification of Products by High Pressure Liquid Chromatography-An Altex ODS 5 S Ultrasphere RP-HPLC column (250 X 4 mm) equipped with a 3-cm guard column was eluted at 1 ml/min with a solvent system of MeOH/H20/HAc (75:250.01, by volume). Lipid extracts were vortexed in 100 p1 of methanol, centrifuged briefly, and then the MeOH mixed with an equal volume of 0.05% aqueous acetic acid and injected on the reversed-phase column. The effluent was monitored with a Hewlett-Packard 1040A diode-array detector recording at 205, 235, and 270 nm. Fractions of 0.5 ml were collected. The egg-derived products were further purified by straight-phase HPLC using an Alltech 5-pm silica column eluted at 2 ml/min with solvent systems of n-hexane:2-propanol:acetic acid (1001.6:0.1, by volume) for free acids and n-hexane:2-propanol (1000.5, v/v) for methyl ester derivatives. The effluent was monitored at 235 nm with a Waters model 450 UV detector, and 0.25-1-ml fractions were collected.
Chromatography on a silver-loaded cation exchange column was carried out with a Sepralyte SCX 5-pm cation exchange HPLC column (250 X 4.6 mm, Analytichem International, Harbor City, CA) with a solvent system of hexane:2-propanol:water (4:6:0.2 by volume for HETE methyl esters and 460.5 for HEPE methyl esters).
Derivitizatwn Procedures for Gas Chromatography and Mass Spectrometry Analyses-Methyl esters were prepared by treatment of the compound in methanol with an excess of ethereal diazomethane for 5 min at room temperature. For catalytic hydrogenation, about 20 pg of the methyl ester derivative was dissolved in 0.1 ml of absolute ethanol, platinum oxide (1 mg) was added, and hydrogen was bubbled through the suspension for 1 min. Water was then added and the product recovered by immediate extraction with ethyl acetate. Silylation was performed by dissolving the fatty acid methyl esters in 10 p l of pyridine and 15 pl of bis(trimethylsily1)-trifluoroacetamide. After at least 30 min at room temperature, the pyridine and bis-(trimethylsily1)-trifluoroacetamide were evaporated under nitrogen and the trimethylsilyl ether methyl ester derivatives dissolved in dodecane for GC-MS analysis.
Cas Chromatography-Mass Spectrometry-Analyses were performed using a Nermag R-10-10-C quadrupole instrument equipped with a 6-m DB-1 capillary column (0.25-mm inner diameter; coating thickness, 0.25 pm) and temperature programmed from 190 to 250 "C at 10 "C/min. Mass spectra were recorded in the electron impact mode with an electron energy of 70 eV.
Steric Analysis by HPLC of Menthoxycarbonyl Derivatives-This method was used for 11-hydroxyeicosanoid. The methyl ester (-20 p g ) was converted to the (-)-menthoxycarbonyl derivative by treatment with 60 pl of (-)-menthylchloroformate (5% v/v solution in d v toluene) plus 12 pl of dry pyridine. After 20 min of reaction at room temperature, the solvents were evaporated under nitrogen. The sample was dissolved in dichloromethane, and pyridinium hydrochloride was removed by washing with water. The menthoxycarbonyl methyl ester derivative was isolated by RP-HPLC (Altex Ultrasphere ODS 5 S, 250 X 4 mm; solvent, methanol/water, 1003 (v/v), elution volume, -15 ml). The menthoxycarbonyl diastereomers were resolved by SP-HPLC using an Alltech 5-pm silica column (250 X 4.6 mm) with a solvent system of n-hexane:2-propanol (100:0.1, v/v), with elution at 1 ml/min and UV detection at 235 nm. The absolute configuration of the two peaks corresponding to the R and S enantiomers of racemic HETE standards was established using the method of Hamberg (9) as previously described (10). Briefly, the two peaks from a racemic standard were collected separately and each subjected to oxidative ozonolysis, followed by re-esterification and gas chromatographic comparison of the resulting derivative to menthoxycarbonyl methyl ester derivatives prepared from authentic R and S malic acids.
Steric Analysis Uszng a Chiral Phuse HPLC Column-The methyl esters of 12-HETE and 12-HEPE enantiomers were resolved without further derivatization by analysis on a Baker dinitrobenzoylpbenyl glycine chiral phase HPLC column (250 X 4.6 mm) using a solvent system of n-hexane:2-propanol (100:0.5, v/v) and a flow rate of 0.5 ml/min. The R and S assignments were established using a HETE standard of known chirality, e.g. use of platelet or leukocyte-derived (12s)-HETE.

RESULTS
Identification of Products-Four products of polyunsaturated fatty acid oxygenation were formed from endogenous substrates during the incubation of homogenized eggs at 22-24 "C. On RP-HPLC analysis these appeared as two pairs of peaks with strong absorbance at 235 nm, and they were designated as Compounds I, 11,111, and I V in order of elution.
When incubations were carried out in the presence of 100 PM ['4C]eicosapentaenoic acid, Compounds I and I1 were radiolabeled and the amounts were increased 3-5-fold. Incubation with 100 PM [14C]arachidonic acid resulted in incorporation of radioactivity in Compounds I11 and I V and an increase in their relative abundance (Fig. 1). The preponderance of the two products within each pair was within a 4:l to a 1:2 ratio of peak heights in 10 separate experiments. Typically, about 50% of the radiolabel was recovered in these products, the remainder being unmetabolized substrate and polar lipids. No products were synthesized in the presence of calcium-free artificial sea water containing 2 mM EGTA. The enzymatic activity was calcium dependent, and it was not stimulated by addition of 1 mM ATP. The cyclooxygenase inhibitor indo- The column (ODs 5 S 5 pm, 4 X 250 mm) was eluted with 75% methanol, 25% water, and 0.01% acetic acid at 1 ml/min, and 0.5-ml fractions were collected and radioactivity measured by liquid scintillation counting. UV-absorbing products eluting at 18.3, 19.5, 27.2, and 28.8 ml were designated Compounds I, 11,111, and IV, respectively. methacin (10 WM) did not affect the production of Compounds I-IV.
The UV spectrum of each product exhibited the characteristic chromophore of a conjugated diene, Amax = 237 nm for Compounds I, 11, and IV and 235 nm for Compound 111. Comparison of the UV spectra of authentic standards' and further analyses on RP-HPLC and SP-HPLC indicated the probable structures of the four products as 11-HEPE, 12-HEPE, 11-HETE, and 12-HETE. Each product and corresponding standard were analyzed by GC-MS as the methyl ester (Me3Si ether derivative), before and after catalytic hydrogenation of the double bonds. Positive ion electron impact spectra of the MeaSi derivative of Compounds I and I1 are shown in Fig. 2. The spectra were indistinguishable from the corresponding spectra of ll-hydroxy-5,8,12,14,17-eicosapentaenoate (11-HEPE) and 12-hydroxy-5,8,10,14,17-eicosapentaenoate (12-HEPE), respectively. In each case the high mass ions are found with the expected shift of 2 atomic mass units relative to the well characterized spectra of their eicosatetraenoate analogs (11). The position of the hydroxyl group was confirmed by analysis of the Me3Si derivatives of the hydrogenated compounds; the mass spectra of the saturated products were essentially identical to the published spectra of 11hydroxy-and 12-hydroxyeicosanoates (11) The position and configuration of the double bonds in Compounds I-IV appear to correspond to a cisltrans-conjugated diene unit allylic to the hydroxyl group, with the other bonds in their original positions and retaining the cis configuration. This is the most likely result on biochemical grounds, and it is supported by the UV and HPLC data. (i) The UV spectra are identical to the authentic standards prepared by controlled autoxidation (5) in agreement with the cisltrans nature of the conjugated system; (ii) products and standards cochromatograph on RP-HPLC, SP-HPLC, gas-liquid chromatography, dinitrobenzoylphenyl glycine-chiral phase HPLC (vide infra), and, significantly, on a silver-loaded cation exchange HPLC column used with normal phase solvent systems (12). Silver-loaded normal phase systems are especially sensitive to the cis or trans configuration of double bonds (13, 14), and, therefore, the co-chromatography supports the result one might anticipate, namely that the bonds remote from the site of oxygenation remain unaffected by the metabolic transformation.
Steric Analyses-The chirality of the 11-hydroxy products from the eggs was analyzed by SP-HPLC of the methyl ester (-)-menthoxycarbonyl derivatives (10). The racemic ll-hydroxy standards resolved as two menthoxycarbonyl diastereomers with elution in the order (11R) and ( l l S ) , as previously established by oxidative ozonolysis of the individual compounds and assignment of stereochemistry by comparison with derivatives of authentic R and S malic acid (10). The methyl ester menthoxycarbonyl derivative of Compound In 12-HETE, the "nonconjugated" 5-cis double bond contributes about a 2-nm bathochromic shift to the conjugated diene, and the Xmax is near 237 nm. This extension to the chromophore is absent in 11-HETE (and 9-HETE), and the X , , is near 235 nm. In the 20.5~3 series, the corresponding pairs at 235 nm are 9-and 14-HEPEs; both 11-HEPE and 12-HEPE have a nonconjugated double bond extending the Xmax to near 237 nm.
A dinitrobenzoylphenyl glycine chiral phase HPLC column was used for the assignment of chirality of the 12-hydroxy products. Elution volumes were compared with those of racemic standards and with enantiomers of the (12s) configuration prepared using the well characterized (12s)-lipoxygen-(11 R )and (12R)-L&oxygenaseActivity in Sea Urchin Eggs ase of porcine leukocytes (15). The results established that in each case the egg products were of the (12R) configuration (Fig. 4).
Experiments with [10-3H]Eicosapentuenoic Acids-In the course of (12S)-HETE synthesis in platelets and porcine leukocytes, the (12s)-lipoxygenase catalyzes stereospecific removal of the 10-L-hydrogen from the substrate (15, 16). Because the 12-HETE and 12-HEPE formed by the eggs were of the R configuration, it was of great interest to examine the stereochemistry of hydrogen abstraction in this reaction. Separate incubations of egg homogenates were conducted with [ 10-D-3H]eic~~apentaen~ic acid and [lO-~-~H]eicosapentaenoic acid (one experiment with each enantiomer); [3-14C] eicosapentaenoic acid served as internal standard for measurement of tritium retention. The HEPE products were purified by RP-and SP-HPLC and then the 3H/14C ratio determined by liquid scintillation counting.
In the ~O -D -~H experiment, the 12-HEPE retained 7% of the tritium in the original substrate. In contrast, there was 82% retention of tritium in the 12-HEFE from the incubation with [ 10-~-~H]eicosapentaenoic acid. In both experiments the specific activity of the 11-HEPE product was unchanged (100% and 106% of the original substrate, respectively).
For direct comparison, the [ lO-~-~H]eicosapentaenoic acid was reacted with the (12s)-lipoxygenase of porcine leukocytes. As found before using arachidonic acid (15), the leukocyte lipoxygenase catalyzed (12s)-oxygenation with removal of the 10-L-hydrogen (only 7% retention of tritium in the (12s)-HEPE product). Thus, the chiralities of hydrogen abstraction and stereospecific oxygenation are mirrored in the leukocyte and the sea urchin egg. Both 12-lipoxygenases hold to a characteristic feature of all lipoxygenases studied so far; an antarafacial relationship is observed between hydrogen removal and insertion of oxygen in the substrate.
An additional characteristic of lipoxygenases is that ster- nm. The egg and the leukocyte products migrated as single symmetrical peaks. Steric assignments were established by co-chromatography with the R or S enantiomer of authentic R,S standards; the egg products were of the (12R) configuration, enantiomeric to the (12s) products of porcine leukocytes. PMN, polymorphonuclear leukocytes. eoselective removal of a prochiral tritium atom is associated with a primary isotope effect measured in the unreacted substrate (16)(17)(18)(19). However, in the incubations of sea urchin eggs the measured changes were too slight to be meaningful. Less than 50% of the eicosapentaenoic acids were metabolized in these particular incubations; this incomplete reaction, together with the competing ll-lipoxygenase (which accounted for the larger portion of the total lipoxygenase activity), resulted in relatively trivial changes in the specific activity of the [ 10-D-3H]eicosapentaenoic acid during the incubation (134% of the original value, compared with 120% for the [ 10-L-~H]). Interestingly, the incomplete reaction was associated with a marked secondary isotope effect in the ~O -L -~H experiment; the (12RI-HEPE product showed only 82% (not +go%) tritium retention, in line with similar findings with platelet (12S)-lipoxygenase and the enantiomeric 10-D-labeled substrate (20).

DISCUSSION
In this study we have established that the eggs of the sea urchin, S. purpurutus, contain a prominent (11R)-and (12R)lipoxygenase activity. In the absence of added substrates, products were formed from endogenous arachidonic and eicosapentaenoic acids, implying that these fatty acids may be natural substrates of the enzyme(s). It has been noted before that eggs of S. purpurutus undergo a burst of oxygenase activity with formation of an "HETE-like" product upon fertilization (4). However, this property is not shared by the eggs of all species of sea urchin (3). We confirmed that eggs of the sea urchin Arbuciu punctulatu contained no detectable lipoxygenase activity. Nevertheless, the biosynthesis and biological activity of (8R)-HETE in starfish oocytes sets an important precedent for a specific function for monohydroxyeicosanoids (2). Identification of the lipoxygenase products of S. purpurutw eggs will enable detailed investigation of the role of these oxidized unsaturated fatty acids in fertilization and/or maturation processes. 11-Hydroxyeicosanoids have been reported before both as lipoxygenase (21) and as cyclooxygenase products (22)(23)(24)(25)(26). When 11,14-eicosadienoic acid is used as a cyclooxygenase substrate, formation of an endoperoxide is not possible and the predominant product is the (llR)-hydroxy-12,14-eicosadienoic acid (22). This evidence was used in support of other data which established that prostaglandin biosynthesis is initiated by (11 R) oxygenation of the substrate (23). Interestingly, the chirality of 11-HETE formed as a by-product of prostaglandin synthesis has not been reported. Our colleague Dr. Douglass Taber questioned whether the monohydroxy byproduct was eliminated because it is of the "wrong" (i.e. (11s)) configuration for cyclization to an endoperoxide? To test this hypothesis we determined the chirality of the ll-hydroxy-8,12,14-eicosatrienoic acid which is formed as a by-product of prostaglandin El synthesis in ram seminal vesicle microsomes. Our results (Fig. 5) showed that this by-product was the (11 R)-hydroxyenantiomer, thus disproving this intriguing idea. Our experiments with sea urchin eggs are the first study showing the chirality of 11-hydroxy products of a lipoxygenase enzyme.
It has generally been assumed that the stereoselectivity of (12s)-HETE biosynthesis in platelets and leukocytes would prevail in the synthesis of 12-HETE by other cells and tissues.
Recently it was shown that the 12-HETE in human psoriatic scales is of the R configuration; it is not yet established whether a lipoxygenase, P-450, or other enzyme is involved in the biosynthesis (27). Mouse epidermis is known to form 12-HETE (28), and we have examined the chirality of the The 12-HETE was isolated as the major radiolabeled product from incubation of the 10,OOO x g supernatant fraction of mouse (strain C5BL/6) epidermal homogenate with 6.5 p M [1-"CJarachidonic acid in 20 mM Tris, pH 7.3 (37 "C). The identity of the product was confirmed by GC-MS of the Me& derivative. Steric analysis was carried out on the methyl ester derivative exactly as described in the legend to Fig. 4. product; it was (12S)-HETE (Fig. 6). An expected feature of a (12R)-lipoxygenase is that the chirality of both hydrogen abstraction and oxygenation would mirror the well characterized stereochemistry of (12s) oxygenation in platelets and leukocytes (15,16). Invariably it has been found that the lipoxygenase-catalyzed oxygenation of polyunsaturated fatty acid substrates is associated with stereoselective abstraction of hydrogen from the opposite face of the substrate molecule (16)(17)(18)(19). This antarafacial relationship between hydrogen removal and oxygen insertion was found to hold true in the formation of (12R) oxygenation products in the eggs of S. purpurutus. Thus, synthesis of (12R)-H(P)EPE is associated with stereospecific removal of the 10-D hydrogen of eicosapentaenoic acid, a reaction opposite in all respects to the

SCHEME I
(12s)-lipoxygenase (Scheme I). These data provide strong support that the biosynthesis in the sea urchin eggs proceeds via a lipoxygenase mechanism. Although we have no data on the hydrogen abstraction associated with C-11 oxygenation of the substrate, it is very likely that reaction follows the antarafacial rule and that the 13-L hydrogen is removed during (llR)-H(P)EPE biosynthesis. The biosynthesis of (11R)and (12R)-hydroxy products of the sea urchin eggs appears to be closely linked. The Iipoxygenase activity was calcium-dependent, unlike the (12s)lipoxygenases of platelets and porcine leukocytes (29, 30). A requirement for calcium is found with the leukocyte 5-lipoxygenase (31), the 5-and 12-lipoxygenase of rat basophilic leukemia cells (32,33), and the 15-lipoxygenase of rabbit polymorphonuclear leukocytes; the latter is a soluble enzyme which loses calcium sensitivity when purified (34). There are several examples of lipoxygenases which form more than one chiral product. Purified rabbit reticulocyte lipoxygenase will form 5-1076 (12S)-HPETE in addition to the major product, (15S)-HPETE (35). Also, purified (12S)-lipoxygenase of porcine leukocytes expresses a minor (15S)-lipoxygenase activity (16). Based on these and further precedents in the literature (36,37), the possibility exists that a single enzyme is involved in the transformations in the eggs of S. purpurutus.