The Rat Leukocyte-type 12-Lipoxygenase Exhibits an Intrinsic Hepoxilin A 3 Synthase Activity*

Hepoxilins are biologically relevant eicosanoids formed via the 12-lipoxygenase pathway of the arachidonic acid cascade. Although these eicosanoids exhibit a myriad of biological activities, their biosynthetic mechanism has not been investigated in detail. We examined the arachidonic acid metabolism of RINm5F rat insulinoma cells and found that they constitutively express a leukocyte-type 12 S -lipoxygenase. Moreover, we observed that RINm5F cells exhibit an active hepoxilin A 3 synthase that converts exogenous 12 S -HpETE (12 S -5Z,8-Z,10 E ,14Z-12-hydro(pero)xy-eicosa-5,8,10,14-tetraenoic acid) or arachidonic acid predominantly to hepoxilin A 3 . 12 S -lipoxygenase and hepoxilin A 3 synthase activities were co-localized in the cytosol; immunoprecipitation with an anti-12 S -lipoxygenase antibody co-precipitated the two catalytic activities. A synthase of the lipoxygenase. To test this we cloned the leu-kocyte-type 12 S -LOX from RINm5F cells, Pichia pastoris found that the recombinant enzyme exhibited both 12 S -lipoxygenase and hepoxilin A 3 syn- thase activities. The recombinant human platelet-type S -lipoxygenase and the porcine leukocyte-type 12 S hepoxilin 3 ac- In the native rabbit reticulocyte-type This was by site-di-rected mutagenesis studies that altered the positional specificity of the rat 12 S - and the rabbit reticulocyte-type 15-lipoxygenase. it may be that 12 S -lipoxygen-ases an intrinsic hepoxilin A 3 synthase activity that is minimal in lipoxygenase isoforms with different positional specificity. of hepoxilin isomers quantified by gas chromatography-mass spec-trometry (GC-MS) and/or high performance liquid chromatography (HPLC). The stopped acidification 3.0, mix-tures incubate acidic con- ditions for min (epoxide hydrolysis), lipids twice extracted three of ethyl acetate. evaporated, lipids reconstituted 0.1 methanol, aliquots HPLC after de- rivatives, GC-MS. lipids after upper evaporated, the reconstituted ethyl acetate, aliquots methylsilyl for GC-MS. HPLC the the 100 g one hour HPLC (cid:1) a mobile eluate 254 Shimadzu chromatograms measurements in the range between for in the range between this range we observed deviation from linearity in Lin-eweaver-Burk may

Hepoxilins form a family of eicosanoids that are biosynthesized via the 12S-lipoxygenase (12S-LOX) 1 pathway of the arachidonic acid (AA) cascade. Chemically they constitute epoxy-hydroxy eicosanoids that can be classified as HxA 3 (8hydroxy-11,12-epoxyeicosatetraenoic acid) and HxB 3 (10-hydroxy-11,12-epoxyeicosatrienoic acid) (1). Hepoxilins exhibit a myriad of biological activities. They stimulate glucose-induced secretion of insulin and increase the intracellular calcium levels in rat pancreatic islets cells (2). In human neutrophils a dose-dependent increase in intracellular calcium concentration (3,4) and an augmented cellular diacylglycerol content (5) were observed. The latter effect was blocked by pertussis toxin, suggesting a receptor-mediated mechanism involving GTPbinding proteins. In neurons, HxA 3 induces hyperpolarization of the membrane potential and increases the amplitude and duration of the inhibitory postsynaptic potential (6,7). In platelets and aplysia neurons, HxA 3 antagonized the expansion of cell volume and the opening of potassium channels (8,9).
Although hepoxilin biosynthesis has been studied for a long time, some mechanistic details remain unclear. The first step of hepoxilin formation is conversion of AA to 12S-HpETE by a 12S-LOX (10). In mammalian cells this primary metabolite may be further metabolized via two alternative routes: (i) reduction to 12S-HETE (reductive pathway), and (ii) isomerization to hepoxilins (hepoxilin synthase pathway). When the reductive pathway, which involves glutathione peroxidase isoforms, is up-regulated hepoxilin biosynthesis is minimal because of substrate exhaustion (11). In contrast, cells with diminished reductive capacity are capable of biosynthesizing large amounts of hepoxilins provided that they express a hepoxilin synthase (11). A critical point in the mechanism of cellular hepoxilin formation has been the question of whether or not intracellular conversion of 12S-HpETE to hepoxilin isomers is an enzymatic process. Recent investigations on the stereospecificity of hepoxilin formation in rat pineal glands suggested an enzymatic pathway. This conclusion was based on the stereoselective conversion of 12S-HpETE. Under strictly comparable conditions, 12R-HPETE remained unmetabolized (10). Moreover, it became evident during the past several years that an active hepoxilin biosynthesis was always associated with a high expression level of the 12S-LOXs. This observation was usually explained by the fact that a 12S-LOX was required to convert AA to 12S-HpETE, the immediate substrate of hepoxilin biosynthesis (12). On the other hand, these data prompted the possibility that 12S-LOX itself might be involved in the isomerization of 12S-HpETE to hepoxilins (11).
LOXs are multifunctional enzymes that exhibit a hydroperoxidase (13)(14)(15), a leukotrienes synthase (16,17), and a lipoxin synthase activity (18 -20) in addition to their oxygenase function. We recently observed that RINm5F rat insulinoma cells convert exogenous arachidonic acid or 12S-HpETE to hepoxilins and that HxA 3 , but not HxB 3 , was the major metabolite (11). More detailed mechanistic investigations revealed that these cells are devoid of glutathione peroxidase activities that reduce hydroperoxy fatty acids to the corresponding hydroxy compounds but express a leukocyte-type 12S-LOX at relatively high levels (11).
In the present study, we cloned a HXA 3 synthase from RINm5F rat insulinoma cells and characterized the enzyme as leukocyte-type 12S-LOX. Alterations of the positional specificity of this enzyme by site-directed mutagenesis were paralleled by changes in the intrinsic hepoxilin synthase activity. An inverse mutagenesis strategy converting the rabbit 15S-LOX to a 12-lipoxygenating enzyme species induced significant HXA 3 synthase activity. These data as well as HxA 3 synthase activity assays of other LOX isoforms indicated that mammalian 12S-LOXs exhibit an intrinsic HxA 3 synthase activity.

MATERIALS AND METHODS
Cell Culture-The permanent rat insulinoma cell line RINm5F (a kind gift from Dr. Tiedge, Hannover, Germany) was cultured in RPMI 1640 medium supplemented with 10% fetal calf serum. The cells were grown at 37°C in a humidified atmosphere containing 5% CO 2 .
Activity Assays-HxA 3 synthase activity was assayed by incubating intact cells or enriched enzyme preparations with exogenous AA (100 M) or 12S-HpETE (20 M) for 30 min at 37°C. The formation of hepoxilin isomers was quantified by gas chromatography-mass spectrometry (GC-MS) and/or high performance liquid chromatography (HPLC). The reaction was stopped by acidification to pH 3.0, the mixtures were allowed to incubate at room temperature under acidic conditions for 15 min (epoxide hydrolysis), and the lipids were twice extracted with three volumes of ethyl acetate. The solvent was evaporated, the lipids were reconstituted in 0.1 ml of methanol, and aliquots were directly injected in HPLC or, after making suitable derivatives, in GC-MS.
For measurements of the cellular HxA 3 synthase activity, intact RINm5F cells (about 10 7 cells) were resuspended in 0.1 ml of PBS. 0.1 mM AA was then added, and the reaction was allowed to proceed at 37°C for 30 min. The lipids were extracted by adding a mixture of diethyl ether/methanol/1 M citrate (135:15:1, v/v), and after phase separation the upper organic layer was recovered. The solvents were evaporated, the residues were reconstituted in 0.1 ml of ethyl acetate, and aliquots were converted to methylsilyl derivatives for GC-MS. For convenient detection in HPLC a fluorophore was introduced at the carboxylic group by the reaction with 100 g of 9-anthryldiazomethane for one hour at room temperature (21). HPLC analysis was carried out on a Novapak-C 18 column (250 ϫ 46 mm; 5-m particle size) using acetonitrile/methanol/water (90:6:4; by vol) as a mobile phase. The column eluate was monitored at 254 nm using a Shimadzu diode-array detector. Alternatively, we recorded the chromatograms with a Shimadzu fluorescence detector (excitation 254 nm, emission 400 nm).
To assay the LOX activity, the cellular lysates or enriched enzyme preparations were incubated with exogenous AA, and the LOX products formed were quantified by HPLC. For this purpose 5-10 ϫ 10 6 cells were resuspended in 0.5 ml of PBS, and 0.1 mM of AA (final concentration) was added. Then the cells were lysed on ice by sonication (Labsonic microtip-sonifier; Braun, Melsungen, Germany) two times for 5 s at full power, and the homogenate was incubated for 10 min at 25°C. The hydroperoxy lipids formed were reduced to the corresponding hydroxy compounds by addition of 0.1 ml of saturated sodium borohydride solution (in dry methanol). The incubation mixture was acidified with 50 l of glacial acetic acid, and 0.6 ml of ice-cold methanol was added. The sample was kept on ice for 10 min, and protein precipitate was removed by centrifugation. Aliquots of the clear supernatant were directly injected to HPLC. RP-HPLC was carried out on a Nucleosil C-18 column (Macherey/Nagel, Dü ren, Germany; 250 ϫ 4 mm, 5-m particle size) with a solvent system of methanol/water/acetic acid (85:15:0.1; v/v) at a flow rate of 1 ml/min. The absorbance was monitored at 235 nm. A calibration curve (6-point measurement) for conjugated dienes was established using 13-HODE as standard.
Kinetic measurements were carried out on a Shimadzu spectrophotometer (UV2100), and the linear parts of the kinetic progress curves were evaluated. The basic kinetic characteristics (K m and V max ) were calculated using the Graphpad software package (San Diego, CA).
Immunoblotting-For electrophoresis the cells were reconstituted in lysis buffer (PBS, 1% Nonidet P-40, 0.1% SDS, 0.5% sodium deoxycholate, 1 mM phenylmethylsulfonyl fluoride, and pepstatin and leupeptin, 1 mg/ml each), and the debris was removed by centrifugation at 12,000 rpm at 4°C. Protein concentrations were determined using LowryЈs assay (Bio-Rad). Lysates were electrophoresed on 10% SDS-PAGE, and the protein bands were transferred onto a nitrocellulose membrane (Schleicher & Schü ll, Dassel, Germany) using a semidry blotting procedure. After blocking with 5% skimmed milk in PBS containing 0.05% Tween 20, the blots were probed with an antiporcine leukocyte-type 12-LOX rabbit antibody (kind gift from Prof. S. Yamamoto, Kyoto, Japan) for 1 h and stained using the ECL detection system (Amersham Biosciences). The antibody used cross-reacts with both the rabbit reticulocyte-type 15-LOX and the porcine leukocyte-type 12-LOX but does not recognize platelet-type 12-LOXs.
Immunoprecipitation-RINm5F cell lysates were subjected to immunoprecipitation using the antileukocyte-type 12S-LOX antibody (20 g/ ml) and protein A-Sepharose (1 h of incubation). The total lysate, the resuspended immunoprecipitate, and the depleted precipitation supernatant were incubated with 20 M 12S-HpETE. The production of trioxilin A 3 (TrXA 3 ), the major hydrolysis product of HXA 3 , was determined by GC-MS. Similarly, LOX activity was determined by incubating the different immunoprecipitate fractions with 100 M AA and analyzing the LOX products by HPLC as described above. 13S-Hydroxy-octadecaenoic acid was used as an internal standard for quantification of both TrXA 3 and 12S-HETE.
Gas Chromatography-Mass Spectrometry-Prior to GC-MS the free fatty acid derivatives were converted to their methyl esters by the addition of 300 l of ethereal diazomethane. After 5 min of incubation the solvent was evaporated, 30 l of N-methyl-N-(trimethylsilyl)trifluoroacetamide were added, and the hydroxy groups were silylated for 30 min at 60°C in dry pyridine. GC-MS was performed on a Varian Saturn 4D GC-MS-MS system (Varian, Darmstadt, Germany) equipped with a Supelco DB5-MS column (30 m ϫ 0.25 mm; 0.25 m). The temperature program was started at 150°C and increased to 250°C with a rate of 10°C/min, followed by an isothermic postrun at 250°C for 20 min. The injector and transfer line temperatures were adjusted at 230 and 220°C, respectively.
Cloning of Hepoxilin A 3 Synthase from RINm5F Cells-According to our working hypothesis, the leukocyte-type 12S-LOX of RINm5F cells might exhibit HXA 3 synthase activity. To test this hypothesis, we cloned the leukocyte-type 12S-LOX from these cells. For this purpose we first isolated total RNA from RINm5F cells using an RNeasy kit from Qiagen (Hilden, Germany). Then 3 g of the RNA preparation were reverse-transcribed at 42°C for 60 min with 150 pmol of oligo(dT) primer and 15 units of AMV reverse transcriptase (Roche Applied Science). For amplification of the leukocyte-type 12S-LOX cDNA, 1 l of the reverse transcription reaction was used and the following primers were added to the amplification mixture: 5Ј-cga cat atg tgt cta ccg cat ccg c-3Ј (forward primer), 5Ј-tgg ctc gag tca gat ggc cac gct gtt-3Ј (reverse primer). After initial denaturation for 4 min at 94°C, 30 cycles of PCR were performed. Each cycle consisted of a denaturing period (40 s at 94°C), an annealing phase of 30 s at 60°C, and a synthesis phase (90 s at 64°C). After the last cycle, all samples were incubated for an additional 10 min at 72°C. PCR products were separated by 1% agarose gel electrophoresis, and the DNA bands were stained with ethidium bromide. The purified 12S-LOX PCR fragment was cloned into the NdeI and XhoI sites in PET 15b vector (Novagen, Bad Soden, Germany), and LOX-positive colonies were isolated. Their sequence insert was confirmed by automated fluorescent sequencing.
Expression of the Cloned LOX in P. pastoris GS 115-The 12S-LOX cDNA was cloned into the pPICZ, and the native stop codon prevented formation of a C-terminal His tag fusion protein. The plasmid was linearized and electroporated into competent P. pastoris GS115 cells.
LOX-positive colonies were selected using zeocin as selection marker. The mutant phenotypes of the transformants were tested by plating them on minimal medium containing histidine and methanol or dextrose. All positive clones turned out to be mutants. Some of them were randomly selected and then cultured in minimal medium containing glycerol, biotin, and histidine for 3 h at 37°C with constant stirring to an optical A 600 of 2. Cells were pelleted, resuspended in 1/10 of the initial culture volume (minimal medium containing histidine and 0.5% methanol), and further grown at 30°C. After 24 h the cells were harvested and lysed using glass beads. The lysates were centrifuged, and the stroma-free lysate supernatant was used as enzyme source.
Expression of Rabbit 15S-LOX Mutants in Escherichia coli-The recombinant wild-type and mutant rabbit 15S-LOX and the various mutants were expressed in E. coli as His-tagged fusion proteins. For this purpose the 15S-LOX cDNA was cloned into the pQE-9 expression plasmid (Qiagen) between the SalI and Hind III restriction sites. Bacteria were transformed with the recombinant plasmids, and 3 liters of Luria Bertani medium containing 100 mg/liter ampicillin and 25 mg/ liter kanamycin were routinely inoculated with a 15-ml overnight preculture. The bacteria were allowed to grow at 37°C for 16 h, and then expression of the recombinant protein was induced by addition of 1 mM isopropyl-1-thio-␤-D-galactopyranoside (final concentration). The cultures were kept for an additional 2 h at 30°C, and then the bacteria were pelleted, washed (with PBS), and resuspended in 30 ml of PBS. Cells were disrupted with an Emulsiflex-C5 high pressure cell homogenizer (Avestin, Ottawa, Canada), and the cell debris was spun down. The clear lysis supernatant was applied to a 0.5-ml Ni-agarose column (Qiagen). The column was washed twice with washing buffer (50 mM NaH 2 PO 4 , 300 mM NaCl, 10 mM imidazole, pH 8.0). The adherent proteins were eluted, rinsing the column with elution buffer (50 mM NaH 2 PO 4 , 300 mM NaCl, 200 mM imidazole, pH 8.0). Five 0.25-ml fractions were collected, and the LOX activity was assayed in each of them. Routinely, more than 90% of the activity was recovered in fractions 2-4. These fractions were pooled, diluted 1:50 with 20 mM Tris-HCl buffer, pH 7.4, and loaded onto a Q-Sepharose column (gel bed volume 500 l; Amersham Biosciences) for further purification by anion exchange chromatography. After loading, the column was washed twice with 2 ml of 20 mM Tris-HCl buffer (pH 7.4), and then the enzyme was eluted with 120 mM KCl dissolved in the same buffer. Fractions of 0.25 ml were collected, activity was assayed, and the active fractions were pooled. For storage the enzyme preparation was supplemented with 10% glycerol and stored at Ϫ80°C.
Site-directed Mutagenesis-Site-directed mutagenesis of leukocytetype 12S-LOX was carried out using the QuikChange mutagenesis kit (Stratagene, Heidelberg, Germany). The following amino acid exchanges were performed: L417Q, I418A, N592S, V593I, and H366L. The mutated plasmids were transformed into E. coli, and five colonies were selected for each mutant for activity assays. The sequence of these colonies was confirmed by automated DNA sequencing.
Kinetic Studies of 12S-HpETE Conversion-Kinetic characterization of 12S-HpETE conversion by the wild-type rabbit 15S-LOX (no HXA 3 formation) and its I418A mutant (5 point measurements) was carried out in the 12S-HpETE concentration range between 15-150 M in 0.1 M phosphate buffer, pH 7.4, at room temperature. The decrease in absorbance at 235 nm was assayed spectrophotometrically, and the catalytic activity was normalized to the LOX protein content of the enzyme preparation. The kinetic constants were extracted from the experimental raw data using the Sigma plot software package (enzyme kinetics module).
Earlier we reported that RINm5F cells exhibit HxA 3 synthase activity (11); these results were confirmed in the present study. To characterize the enzymes involved in HxA 3 biosynthesis, intact RINm5F cells as well as cell lysates were incubated with exogenous 12S-HpETE and the lipid extracts were analyzed by HPLC for the presence of hepoxilin isomers. From Fig. 2 it can be seen that substantial amounts of HxA 3 were formed with intact cells (trace A) or cellular lysate (trace B). In contrast, we did not find any HxA 3 formation when a boiled cell lysate was used as enzyme source (trace C). Preincubation of the cells with the 12S-LOX inhibitor OPP strongly impaired HxA 3 formation; similar results were obtained with arachidonic acid as substrate (data not shown). From these results one may conclude that RINm5F cells are capable of oxygenating AA to 12S-HpETE and further convert this primary LOX product to hepoxilins via a heat-sensitive metabolic pathway that can be inhibited by a 12S-LOX inhibitor. Thus, 12S-LOX appears to be involved in the conversion of 12S-HpETE to HxA 3 .
To obtain additional evidence for the enzymatic character of the hepoxilin A 3 synthase reaction, the enantioselectivity of HxA 3 biosynthesis was tested. For this purpose, lysed RINm5F cells were incubated with 12S-HpETE, the lipids were extracted, and HxA 3 -9-anthryldiazomethane derivatives were prepared and purified by RP-HPLC (Fig. 2). Chiral phase HPLC for the separation of hepoxilin A 3 enantiomers was then Lipids were extracted as described under "Materials and Methods," and LOX products were analyzed by reverse-phase HPLC using methanol: water:acetic acid (85:15:0.1, by vol) as a solvent system. The figure shows a representative chromatogram of three independent analyses. B, cytosolic extracts were prepared from RINm5Fcells, and 30 g of cytosolic protein were applied to 10% SDS-PAGE. The protein bands were transferred onto nitrocellulose, and the blots were probed with a polyclonal anti-12S-LOX antibody raised against the porcine leukocytetype 12S-LOX. carried out. From Fig. 3, trace B, it can be seen that 8(S)-HxA 3 is preferentially formed and only trace amounts of the corresponding 8(R)-enantiomer were detected (S/R ratio 93:7). This high degree of enantioselectivity is not compatible with a nonenzymatic isomerization but strongly suggests its enzymatic character.
Differential Centrifugation Suggests Colocalization of 12S-LOX and HXA 3 Synthase-In the next series of experiments we investigated the subcellular localization of 12S-LOX and HxA 3 synthase activities in RINm5F cells. For this purpose the cells were lysed and several subcellular compartments were prepared by differential centrifugation. Major shares of both 12S-LOX and HXA 3 synthase activities were recovered in the cytosolic fraction (100,000 ϫ g supernatant). In contrast, the washed mitochondrial pellet (20,000 ϫ g pellet) and the microsomes (100,000 ϫ g pellet) only contained small amounts of the two catalytic activities (Table I). These data indicate a similar subcellular localization of 12S-LOX and HxA 3 synthase and exclude cytochrome P-450 isoforms (microsomal localization) and mitochondrial oxygenases as metabolic sources of HxA 3 .
Co-immunoprecipitation of 12S-LOX and HxA 3 Synthase Using an Anti-12S-LOX Antibody-Lipoxygenases are multifunctional enzymes that, in addition to their fatty acid oxygenase activity, exhibit a hydroperoxidase (14,15), a leukotriene synthase (16,17), and a lipoxin synthase activity (18,19). To test whether the leukocyte-type 12S-LOX of RINm5F cells may be involved in HxA 3 synthesis, we immunoprecipitated the enzyme from the cell lysis supernatant using an anti-12S-LOX antibody. Subsequently, the two catalytic activities (12S-LOX and HxA 3 synthase) were assayed in the lysis supernatant, in the precipitate, and in the 12S-LOX-depleted precipitation supernatant. From Fig. 4 it can be seen that immunoprecipitation removed the majority of both 12S-LOX and HxA 3 synthase activities from the lysate supernatant. Control precipitations carried out with a non-immune antiserum did not precipitate any catalytic activity (data not shown). These data suggest that the leukocyte-type 12S-LOX in RINm5F cells is responsible for their HxA 3 synthase activity.
Recombinant Rat Leukocyte-type 12S-LOX Exhibits HxA 3 Synthase Activity-Although the immunoprecipitation experiments suggest that HxA 3 synthase activity may be an intrinsic catalytic property of the rat leukocyte-type 12S-LOX, our data

TABLE I Colocalization of 12-LOX and HXA 3 synthase activity in
RINm5F cells RINm5F cells were disrupted by sonication, and cell debris as well as nuclei were removed by low speed centrifugation (1,000 ϫ g). The 1,000 ϫ g supernatant was then centrifuged at 20,000 ϫ g to obtain the mitochondrial pellet. The 20,000 ϫ g supernatant was then spun at 100,000 ϫ g to obtain the cytosol (100,000 ϫ g supernatant) and the microsomes (100,000 ϫ g pellet). In each fraction we determined the relative 12S-LOX and HxA 3 synthase activities. The activity in the 1,000 ϫ g lysis supernatant was set at 100%.

Subcellular fraction
Relative shares of 12-LOX activity Relative shares of HxA 3 synthase activity % % 1,000 ϫ g supernatant 100 100 20,000 ϫ g pellet (mitochondria) 3 Ͻ1 100,000 ϫ g pellet (microsomes) 7 3 100,000 ϫ g supernatant 93 87 do not prove this working hypothesis. To obtain more convincing evidence we expressed the rat leukocyte 12S-LOX in P. pastoris GS115 and tested the two catalytic activities for the recombinant enzyme. Lysates of yeast cells transformed with the 12S-LOX-containing plasmid converted arachidonic acid to 12S-HpETE (5.3 Ϯ 1.8 g of 12-HETE/ml of lysis supernatant) as indicated by RP-HPLC analysis (Table II). In contrast, wildtype yeast cells did not exhibit measurable LOX activity. Next, we checked the hepoxilin A 3 synthase activity of the cell lysate using 12S-HpETE as substrate. From Fig. 5 it can be seen that large amounts of HxA 3 (measured as the corresponding trioxilin derivative) are formed when a crude enzyme preparation (lysis supernatant of 12S-LOX-transformed P. pastoris) was incubated with 12S-HpETE (trace I). In contrast, much smaller amounts of HxB 3 were detected (trace II), indicating a high degree of product specificity. A heat-denatured enzyme preparation (trace III) and a lysate of wild-type yeast cells (not shown) were catalytically inactive. To obtain additional experimental evidence for the identity of 12S-LOX and HxA 3 synthase, we tested the effect of OPP, which is known as a specific 12S-LOX inhibitor (22, 23), on HxA 3 formation from 12S-HpETE. From Fig. 5, trace D, it can be seen that OPP completely abolished HxA 3 formation. Together with the high degree of enantioselectivity (Fig. 3) and the subcellular localization (Fig. 4), the results shown in Fig. 5 indicate that rat leukocyte-type 12S-LOX is capable of catalyzing isomerization of 12S-HpETE to HxA 3 .

HxA 3 Synthase Activity of LOX Isoforms Is Related to Their Positional Specificity of Arachidonic Acid Oxygenation-To
find out whether all LOX isoforms exhibit a HxA 3 synthase activity, we tested several mammalian LOX species for their ability to convert 12S-HpETE to HxA 3 . To directly compare the enzymes with respect to their catalytic activity they were all expressed in E. coli. Among the LOXs tested only the human platelet-type 12S-LOX and the rat leukocyte-type 12S-LOX exhibited measurable HxA 3 synthase activities (Table III). In contrast, the rabbit reticulocyte-type 15S-LOX and the human 5S-LOX were not capable of converting 12S-HpETE to HxA 3 . As negative controls, we carried out activity assays (LOX and HxA 3 activity) in parallel with lysates of wild-type bacteria (not transformed with LOX-containing plasmid) and did not find measurable activities. Because only the 12S-LOX isoforms were capable of catalyzing HxA 3 formation, one may conclude that the HxA 3 synthase activity might be related to the positional specificity of the LOX isoforms.
To obtain additional support for the hypothesis that hepoxilin A 3 synthase is an intrinsic catalytic property of mammalian 12S-LOXs, we performed site-directed mutagenesis studies to alter the positional specificity of the rat leukocyte-type 12S-LOX and the rabbit reticulocyte 15S-LOX. For the rabbit and human 15-LOX1, sequence determinants for the positional specificity have previously been identified, and site-directed mutagenesis of these residues shifted the positional specificity of the enzymes toward arachidonic acid 12-lipoxygenation (24 -27). However, when a similar mutagenesis strategy was carried out with the rat leukocyte-type 12S-LOX, an A418I exchange did not alter positional specificity (28). Construction of chimeric LOX species and additional mutagenesis experiments suggested that other amino acids residues, in particular Leu-353, appear to be more important for the positional specificity of rat and mouse leukocyte-type 12S-LOXs (26). Based on these findings, we first tested the HxA 3 synthase activity of a 12-lipoxygenating mutant (I418A) of the rabbit reticulocyte-type 15S-LOX. In contrast to the wild-type enzyme, we detected significant HXA 3 formation with this mutant (Table II). To identify target amino acids for mutagenesis in the rat leukocyte-type 12S-LOX, we performed an amino acid alignment with the rabbit reticulocyte-type 15S-LOX (Fig. 6). From this alignment it can be seen that the sequence determinants identified for the rabbit enzyme (Phe-353, Gln-417, Ile-418, Ile-593) align with Leu-353, Lys-417, Ala-418, and Val-593 of the rat 12S-LOX. To stress the positional specificity of the rat leukocyte-type 12S-LOX, we constructed several enzyme mutants by interchanging small amino acids with residues carrying more space-filling site chains. The results of these experiments, shown in Table III, can be summarized as follows: (i) all mutants created were enzymatically active, and their specific activities were comparable; (ii) 12-lipoxygenating enzyme mutants exhibit a major HxA 3 synthase activity; and (iii) in contrast, the 15-lipoxygenating mutant (L353F) exhibited a reduced HXA 3 synthase activity. The residual HxA 3 synthase activity of this mutant may be because of its residual 12S-LOX activity. In summary, one may conclude that 12-lipoxygenating enzyme species exhibit a strong HxA 3 synthase activity that is impaired when the positional specificity of arachidonic acid oxygenation is altered in favor of 15-lipoxygenation. In this respect, there appears to be a correlation between AA 12oxygenation and HxA 3 formation.
Kinetic Investigation on 12S-HpETE Conversion-Next we determined basic kinetic parameters for 12S-HpETE conversion by a HxA 3 -synthesizing enzyme (I418A mutant of the rabbit 15S-LOX) and its wild-type counterpart, which is not capable of catalyzing HxA 3 formation. We found that the reaction kinetics follow the Michaelis-Menten equation for both LOX isoforms (data not shown), suggesting an enzymatic character of the reaction. The kinetic constants obtained (Table IV) allow the following conclusions: (i) binding affinity of 12S-HpETE at the active site of both enzyme species was lower than that for arachidonic acid. These data are consistent with previous observations indicating that hydroxylated fatty acids exhibit an impaired binding affinity at the active site of LOXs (29); (ii) the reaction rate of arachidonic acid oxygenation was higher for both enzyme species when compared with 12S-HpETE conversion. Thus, the catalytic efficiency (V max /K m ratio) of LOX-catalyzed conversion of 12S-HpETE is lower than that of arachidonic acid oxygenation; and (iii) the wild-type rabbit 15S-LOX, which is not capable of synthesizing HxA 3 (Table III), exhibits a higher affinity for 12S-HpETE than the HxA 3 -synthesizing I418A mutant. In contrast, V max of 12S- HpETE conversion was similar when the two enzyme species were compared. These data in connection with the product analysis indicated that the two enzyme species metabolize 12S-HpETE under V max conditions with comparable rates but exhibit profound differences in their product patterns. The I418A mutant produces HxA 3 , but the wild-type does not (Table IV). Here again, the catalytic efficiency for the mutant enzyme is lower than that of the wild-type species (lower V max /K m ratio). DISCUSSION Hepoxilins are bioactive eicosanoids that have been implicated in regulation of cell physiology (2-9). The mechanism of hepoxilin biosynthesis, in particular the question of whether or not the secondary isomerization of 12S-HpETE to hepoxilins is catalyzed by enzymes or by non-enzymic catalysts, has been a matter of discussion for many years. The apparent stereospecificity of the reaction with the preferential use of 12S-HpETE (12R-HpETE is not a suitable substrate for HxA 3 formation) as well as the heat sensitivity of HxA 3 formation suggested an enzymatic process (10 -12). However, the enzymes involved in hepoxilin biosynthesis in vivo have not been characterized so far. We recently presented experimental evidence suggesting that HxA 3 formation in RINm5F cells from AA requires an oxidative environment and raised the question of whether secondary conversion of 12S-HpETE to hepoxilins may proceed enzyme-controlled (11). However, the proof of this hypothesis was lacking. Here we have presented four lines of experimental evidence implicating the rat leukocyte-type 12S-LOX, which is expressed in RINm5F cells at relatively high level, in cellular HxA 3 synthesis starting from 12S-HpETE: (i) 12S-LOX and HxA 3 synthase are localized in the same subcellular compartment (Table I); (ii) immunoprecipitation with an anti-12S-LOX antibody co-precipitated both the 12S-LOX and the HxA 3 synthase activity (Fig. 4); (iii) the recombinant rat leukocyte-type 12S-LOX-cloned RINm5F cells exhibit an intrinsic HxA 3 synthase activity that is abolished by a specific 12S-LOX inhibitor, OPP (Fig. 5); (iv) conversion of AA or 12S-HpETE by RINm5F cell lysates led to regio- (Fig. 5) and enantioselective synthesis of the 8(S)-HxA 3 isomer (Fig. 3). These data also indicate for the first time that pure 12S-LOXs exhibit an intrinsic hepoxilin A 3 synthase activity. Thus, in vivo hepoxilin formation may involve two consecutive LOX-catalyzed steps (primary step 12S-lipoxygenation of arachidonic acid to 12S-HpETE, followed by secondary step isomerization 12S-HpETE). The chemical mechanism of the secondary step has been discussed for quite a while, but here we provide rigorous experimental evidence to prove its enzymatic nature: (i) pure 12S-LOXs (native and recombinant species) are capable of converting 12S-HpETE to HxA 3 ; (ii) heat denaturation abolished both their LOX and HxA 3 synthase activities; (iii) the reaction is highly stereoselective with respect to substrate (12R-HETE is no substrate) and product (8R-epimer is not formed); (iv) the basic kinetic parameters (Michaelis-Menten kinetics) are consistent with an enzymatic reaction; (v) 12S-LOX inhibitors, such as OPP, abolish hepoxilin A 3 formation from 12S-HpETE. As reported earlier, 5S-LOXs catalyze two consecutive steps during leukotriene biosynthesis (5-lipoxygenation of arachidonic acid to 5S-HpETE and secondary 5S-HpETE dehydration) (16), and multiple LOX reactions are involved in lipoxin formation (19). Here we report that 12S-LOXs can catalyze two consecutive reactions during hepoxilin biosynthesis. Abstracting these findings, one may conclude that LOXs are involved in different (primary and secondary) steps of eicosanoid biosynthesis. This conclu-  3 synthase activities of rabbit 15S-and rat 12S-LOX mutants Wild-type and mutant LOX isoforms were expressed in E. coli as described under "Materials and Methods." Crude lysis supernatants were used as enzyme source for activity measurements. The LOX and HXA 3 synthase activities are expressed as g of HETE or HXA 3 (determined as TrXA 3 s) formation/ml of fermentation culture during a 15-or a 30-min incubation period, respectively. HETE formation was quantified by RP-HPLC after borohydride reduction and TrXA 3 formation by GC-MS after making suitable methylsilyl derivatives. The activity data represent the mean Ϯ error range of duplicate experiments. *, single activity determination.  sion may be important not only for the understanding of hepoxilin formation but may also broaden our view on the catalytic repertoire of LOXs.
To find out whether all LOX isoforms exhibit a major HxA 3 synthase activity we tested different mammalian isoenzymes for this catalytic activity, but only 12-lipoxygenating enzyme species (human platelet-type 12S-LOX, rat leukocyte-type 12S-LOX) turned out to be HxA 3 synthase-positive. In contrast, the human 5S-LOX and the rabbit reticulocyte-type 15S-LOX (Table II) were unable to catalyze HxA 3 formation. Mutation of the sequence determinants for the positional specificity suggested that HxA 3 synthase activity must be considered as intrinsic catalytical property of mammalian 12S-LOXs.
LOXs are multifunctional enzymes that in addition to their oxygenase activity may also exhibit hydroperoxidase (14,15), leukotriene synthase (16,17), and lipoxin synthase activities (18 -20). Our findings that various LOX isoforms are capable of converting 12S-HpETE to HxA 3 increase the functional multiplicity of this enzyme class. Hepoxilin A 3 synthesis from 12S-HpETE resembles the hydroperoxidase reaction (12,14), which, unlike the dioxygenase activity, does not require insertion of atmospheric oxygen. In principle, the hydroperoxidase reaction can be divided into two major steps: (i) homolytic cleavage of the peroxy group forming radical intermediates (14,30), and (ii) radical stabilization via the formation of various secondary products (epoxy-hydroxy isomers, ketodienes, short chain aldehydes alkanes, etc.). During LOX-catalyzed hydroperoxidase reactions the first step appears to be enzymecontrolled (30), whereas the second step may proceed nonenzymatically. This second non-enzymatic share might be responsible for the complexity of secondary reaction products. During LOX-catalyzed HxA 3 synthesis, both reaction steps may proceed enzyme-controlled; thus, a much simpler product pattern is expected. Indeed, we observed that exogenous 12S-HpETE is converted by the recombinant rat leukocyte-type 12S-LOX mainly to HxA 3 , suggesting tight enzymatic control. This could be followed by monitoring the disappearance of absorbance at 235 nm and was completely blocked by OPP. Because no stimulation of 12S-HpETE reduction by OPP was observed, it can be taken for granted that OPP solely inhibits HXA 3 synthase. The mechanistic details for this remarkable specificity remain unclear, but obviously the enzyme prefers to direct the hydroxy group away from the epoxide ring to insert it at C8 of the hydrocarbon backbone in the S position. Taken together, one may conclude that the HxA 3 synthase activity of certain LOX isoforms may be considered a special type of hydroperoxidase reaction in which the two reaction steps appear to proceed tightly enzyme-controlled. HXA 3 has not been detected in all cells and tissues that express either leukocyte-type or platelet-type 12S-LOXs. The mechanistic reason for this observation may be because different cells exhibit different peroxide-reducing capacities. In most mammalian cells peroxide-reducing enzymes, such as seleniumdependent and -independent glutathione peroxidases, are expressed at high levels. These enzymes reduce the cellular peroxides, including 12S-HpETE, and thus remove the substrates for hepoxilin biosynthesis (11). However, when the activity of the cellular glutathione peroxidases, is down-regulated by lowering the intracellular glutathione concentration, e.g. by incubation of the cells with diethyl maleate, HXA 3 formation may be enhanced (11).
Thus, our experimental data indicate that mammalian 12S-LOXs exhibit a hepoxilin A 3 synthase activity in reconstituted in vitro systems as well as in intact mammalian cells. However,  3

synthase activities of various LOX isoforms
Wild-type LOX isoforms were expressed in E. coli as described, and crude lysis supernatants were used as enzyme source for activity measurements. LOX and HxA 3 synthase activities are expressed in g of HETE or g of HxA 3 (determined as TrXA 3 -isomers) formation/ml of fermentation culture during a 15-min incubation period. HETE formation was quantified by RP-HPLC after borohydride reduction and TrXA 3 formation by GC-MS after making suitable methylsilyl derivatives. The activity data represent the mean of triplicate experiments. *, to adjust comparable arachidonic acid oxygenase activities, the bacterial lysis supernatant was diluted with PBS 1:10. n.a., not applicable.  6. Amino acid alignment between the rabbit reticulocyte-type 15-and the rat leukocyte-type 12-LOX. The amino acids that align with the sequence determinants identified for the rabbit 15-LOX are boxed. These residues were targeted by site-directed mutagenesis.

TABLE IV Kinetic constants of arachidonic acid oxygenation and 12S-HpETE conversion by the 12-lipoxygenating I418A mutant of rabbit 15S-LOX
Kinetic measurements were carried out as described under "Materials and Methods," and the kinetic constants were extracted from the Lineweaver-Burk plots (10-point measurements in the concentration range between 5-15 M for arachidonic acid and 5-point measurements in the concentration range between 20 -150 M for 12S-HpETE). Above this concentration range we observed deviation from linearity in Lineweaver-Burk plots, which may be because of substrate inhibition (30). It should be noted that the K m values for arachidonic acid oxygenation are in fair agreement with the values reported before for the native rabbit 15-LOX using linoleic acid as substrate (30 the results do not rule out the possibility that other proteins exhibit a similar catalytic activity. In fact, it appears as if different peroxide-metabolizing enzymes or transition metalcontaining proteins catalyze a similar reaction. Hence, cellular hepoxilin biosynthesis may be considered as the result of several enzymatic and/or pseudoenzymatic processes, and a variety of potential catalysts including 12S-LOXs might be involved. These considerations leave sufficient room for LOXindependent hepoxilin formation. More research is obviously needed to work out the mechanisms involved in hepoxilin biosynthesis in other cells and organs.