Suicidal inactivation of the rabbit 15-lipoxygenase by 15S-HpETE is paralleled by covalent modification of active site peptides

Free Radical Biology & Medicine, Vol. 34, No. 3, pp. 304 –315, 2003 Copyright © 2003 Elsevier Science Inc. Printed in the USA. All rights reserved 0891-5849/03/$–see front matter

PII S0891-5849(02)01244-3

Original Contribution SUICIDAL INACTIVATION OF THE RABBIT 15-LIPOXYGENASE BY 15S-HpETE IS PARALLELED BY COVALENT MODIFICATION OF ACTIVE SITE PEPTIDES RAINER WIESNER,* HIROSHI SUZUKI,† MATTHIAS WALTHER,* SHOZO YAMAMOTO,‡

and

HARTMUT KUHN*

*Institute of Biochemistry, University Clinics Charite´, Humboldt University, Berlin, Germany; †Department of Biochemistry, Graduate School of Medicine, University of Tokushima, Tokushima, Japan; ‡Department of Food and Nutrition, Faculty of Home Economics, Kyoto Womens’s University, Kyoto, Japan (Received 12 September 2002; Revised 10 October 2002; Accepted 10 October 2002)

Abstract—Lipoxygenases (LOXs) are multifunctional enzymes that catalyze the oxygenation of polyunsaturated fatty acids to hydroperoxy derivatives; they also convert hydroperoxy fatty acids to epoxy leukotrienes and other secondary products. LOXs undergo suicidal inactivation but the mechanism of this process is still unclear. We investigated the mechanism of suicidal inactivation of the rabbit 15-lipoxygenase by [1-14C]-(15S,5Z,8Z,11Z,13E)-15-hydroperoxyeicosa-5,8,11,13-tetraenoic acid (15-HpETE) and observed covalent modification of the enzyme protein. In contrast, nonlipoxygenase proteins (bovine serum albumin and human ␥-globulin) were not significantly modified. Under the conditions of complete enzyme inactivation we found that 1.3 ⫾ 0.2 moles (n ⫽ 10) of inactivator were bound per mole lipoxygenase, and this value did depend neither on the enzyme/inactivator ratio nor on the duration of the inactivation period. Covalent modification required active enzyme protein and proceeded to a similar extent under aerobic and anaerobic conditions. In contrast, [1-14C]-(15S,5Z,8Z,11Z,13E)-15-hydroxyeicosa-5,8,11,13-tetraenoic acid (15HETE), which is no substrate for epoxy-leukotriene formation, did not inactivate the enzyme and protein labeling was minimal. Separation of proteolytic cleavage peptides (Lys-C endoproteinase digestion) by tricine SDS-PAGE and isoelectric focusing in connection with N-terminal amino acid sequencing revealed covalent modification of several active site peptides. These data suggest that 15-lipoxygenase-catalyzed conversion of (15S,5Z,8Z,11Z,13E)-15-hydroperoxyeicosa-5,8,11,13-tetraenoic acid to 14,15-epoxy-leukotriene leads to the formation of reactive intermediate(s), which are covalently linked to the active site. Therefore, this protein modification contributes to suicidal inactivation. © 2003 Elsevier Science Inc. Keywords—Lipid peroxidation, Eicosanoids, Oxygenases, Affinity labeling, Free radicals

INTRODUCTION

as in cardiovascular disorders [10,11]. LOXs are multifunctional enzymes, since they are capable of converting their primary oxygenation products the hydroperoxy fatty acids to an array of secondary metabolites, and there are three major types of secondary reactions (Scheme 1): 1) Double or triple oxygenation. Since several hydroperoxy fatty acids (5-, 12-, and 15-HpETE but not 13-HpODE) contain bisallylic methylenes, they constitute potential substrates for LOX-catalyzed dioxygenation. In fact, the formation of various dihydro(pero)xy- and trihydro(pero)xy [DiH(p)ETE, TriH(p)ETE] isomers has been reported for various LOXs [12,13]. 2) Hydroperoxidase reaction. At reduced oxygen concentration or under the conditions of

Lipoxygenases (LOXs) constitute a heterogeneous family of nonheme iron-containing fatty acid dioxygenases that spread widely in plants and animals [1,2]. They catalyze the oxygenation of polyunsaturated fatty acids to their corresponding hydroperoxy derivatives and the reaction mechanism involves the formation of enzymebound radicals [3,4]. Mammalian LOXs have been implicated in the biosynthesis of anaphylactic mediators [5,6] in cell differentiation and maturation [7–9] as well Address correspondence to: Dr. Hartmut Kuhn, Humboldt University, University Clinics Charite´, Institute of Biochemistry, Monbijoustrasse 2, 10117 Berlin, Germany; Tel: ⫹49 30 450 528 040; Fax: ⫹49 30 450 528 905; E-Mail: [email protected]. 304

Lipoxygenase affinity labeling

305

Scheme 1. LOX-catalyzed secondary reactions of hydroperoxy fatty acids; 15-HpETE was selected as an example. Radical intermediates and the principle reaction products are indicated.

fatty acid substrate limitations, LOXs are capable of catalyzing the decomposition of hydroperoxy fatty acids [14,15]. This reaction, which requires the presence of hydrogen donors, has been implicated in xenobiotic metabolism [16,17]. It is initiated by a homolytic cleavage of the hydroperoxy bond and leads to a complex mixture of secondary products [3]. 3) Epoxy-leukotriene synthesis [18 –20]. This reaction involves a stereoselective hydrogen abstraction from a doubly allylic methylene and a homolytic cleavage of the peroxide bond forming a bisradicalic intermediate that is stabilized by epoxide formation. LOXs undergo suicidal inactivation during the oxygenation of polyenoic fatty acids [21–27] and mechanistic aspects of the inactivation process have been reviewed recently [28]. It has been suggested that hydroperoxy fatty acids, the primary products of the LOX reaction, may oxidize catalytically relevant amino acids at the active site and, thus, may contribute to suicidal inactivation. In fact, treatment of the rabbit 15LOX with 13S-HpODE was accompanied by a selective oxidation of a methionine residue [29]. Although later experiments confirmed methionine oxidation, this process was shown to be unrelated to suicidal inactivation since site-directed mutagenesis of this methionine to an oxidation-resistant alanine did not reduce suicidal inactivation [30]. More recent studies on the inactivation of other LOX isoforms by 15S-HpETE suggested a covalent linkage of reactive metabolites as the mechanistic reason for suicidal inactivation [31,32]; but, the specificity of the labeling process and its functional consequences have not been addressed. When we reinvestigated the suicidal inactivation of the rabbit reticulocyte 15-LOX, we observed that the

inactivation rate with arachidonic acid was higher than the one with linoleic acid, suggesting mechanistic differences between these two reactions. This finding and the literature data on other LOX isoforms prompted us to explore the inactivation mechanism of the rabbit enzyme in more detail to find out whether or not the inactivation is associated with a specific modification of active site peptides. MATERIALS AND METHODS

Chemicals The chemicals used were from the following sources: [1-14C]-arachidonic acid (specified activity of 2.15 GBq/mmole) from Amersham (Braunschweig, Germany); arachidonic acid and linoleic acid from Serva (Heidelberg, Germany); sodium borohydride from Merck (Darmstadt, Germany); Lys-C endoproteinase from Roche Diagnostics GmbH (Mannheim, Germany); Tricine (N-tris-(hydroxymethyl)-methylglycine from ICN (Eschwege, Germany); urea, iodoacetic acid, and dithioerithritol from Sigma Chemical Co. (St. Louis, MO, USA). Preparation of the rabbit reticulocyte 15-LOX and enzyme activity assay The native reticulocyte-type 15-LOX was prepared from a stroma-free hemolysis supernatant of a reticulocyte-rich blood cell suspension by sequential ammonium sulfate precipitation, hydrophobic interaction chromatography, and anion exchange chromatography on a preparative Mono-Q column (Amersham Pharmacia, Biotech Europe, Freiburg, Germany). The final enzyme prepara-

306

R. WIESNER et al.

tion was electrophoretically pure (⬎ 90%) and exhibited a linoleic acid oxygenase turnover rate of 30 s⫺1. The oxygenation of polyenoic fatty acids was assayed spectrophotometrically monitoring the increase in absorbance at 235 nm. For this purpose, the enzyme was incubated at room temperature in 1 ml of 0.1 M phosphate buffer, pH 7.4, containing 0.2% sodium cholate and 0.25 mM fatty acid substrate (arachidonic acid or linoleic acid). Protein concentration of the 15-LOX preparations was determined with the Roti-Quant kit (Carl Roth GmbH, Karlsruhe, Germany).

tip-signifier (Braun, Messenger, Germany), and kept overnight at 37°C under argon atmosphere. Samples were diluted 5-fold with water and aliquots were applied either to SDS-PAGE or liquid scintillation counting. After electrophoresis, the gels were stained (Compassion-staining), dried, and exposed for 24 h or longer to either a radiographic film or an autoradiography imaging plate. Radioactivity on the plates was quantified with a BAS2000 imaging analyzer (Fuji Photo Film Co., Tokyo, Japan). Protein derivatization and Lys-C-endoproteinase digest

Preparation of radio-labeled hydroperoxy fatty acids Radioactively labeled 15S-HpETE was prepared by incubating [1-14C]-arachidonic acid (0.1 mM final concentration) with soybean LOX-1 (grade 1, Serva; 5 ␮g pure enzyme) in 5 ml 0.1 M borate buffer, pH 9.0. The increase in absorbance at 235 nm was monitored. When the absorbance did not increase any more, the reaction was terminated by the addition of an equal volume of ice-cold methanol and 15S-HpETE was purified by RPHPLC using the solvent system methanol/water/acetic acid (80:20:0.1, by volume). 15S-HETE was obtained from the peroxide by reduction with sodium borohydride and subsequent RP-HPLC purification. 13S-HpODE was prepared in a similar way using linoleic acid as substrate. For production of radio-labeled 12S-HpETE, the I418A mutant of the rabbit 15-LOX [33] was incubated with [1-14C]-arachidonic acid (0.1 mM final concentration) in 0.1 M phosphate buffer, pH 7.4, containing 0.2% sodium cholate. Product workup and HPLC purification was carried out as described for 15S-HpETE. The chemical structures of the hydroperoxy fatty acids were confirmed by coinjections with authentic standards in RP- and SP-HPLC. 15S-HpETE-induced inactivation of the rabbit 15-LOX and radioactive labeling of the enzyme For routine experiments, 1 nmole of pure 15-LOX was incubated in 1 ml of 0.1 M phosphate buffer with 10 nmoles of [1-14C]-15S-HpETE for 60 min at room temperature. The residual linolenic acid oxygenase activity was assayed to prove enzyme inactivation (typically we observed an inactivation of ⬎ 90%). Then, 1 ml of ice-cold acetone was added and the sample was kept at ⫺20°C for 2 h. The protein precipitate was spun down and washed three times to remove excess of radioactivity (first wash: 5 ml of a 1:2 mixture of methanol/chloroform; second wash: 5 ml of a 8:2 mixture of methanol/ water; third wash: 5 ml acetone). The protein pellet was dried under vacuum, reconstituted in 5-fold concentrated SDS-PAGE loading buffer (100 mM Tris-HCl, pH 6.8, 2% SDS, 20% glycerol, 1 M ␤-mercaptoethanol, and 0.1% Serva-blue), solicited for 5 min with a Lab sonic U

The 15S-HpETE-inactivated rabbit 15-LOX was precipitated with ice-cold acetone (see above); the precipitate was spun down and dried in a vacuum centrifuge. The pellet was reconstituted in 0.5 ml of 0.4 M Tris-HCl buffer, pH 8.5, containing 2 mM EDTA and 6 M guanidinium hydrochloride. The protein pellet was dispersed by sonication with the tip-sonifier and kept at 20°C overnight. Disulfide bridges were broken reductively by incubation with 0.03 M dithioerithritol for 30 min at 37°C under argon atmosphere. Subsequently, the free SH groups were acetylated with iodoacetate (0.1 M final concentration) and the samples were dialyzed overnight against 50 mM borate buffer, pH 8.0, containing 1.5 M urea. Finally, Lys-C endoproteinase was added to reach a final concentration of 1% to 2% (related to the amount of protein to be digested), and the digestion mixture was incubated at 25°C for 24 h. Mass spectrometric measurements The mass spectra of the native and 15-HpETE-inactivated 15-LOX were recorded with a time of flight MALDI mass spectrometer (Voyager-Elite, Applied Biosystems, Framingham, MA, USA). For analysis, the native rabbit 15-LOX was incubated with a 10-fold molar excess of 15S-HpETE for 30 min at room temperature. This procedure led to a ⬎ 90% inactivation of the enzyme as indicated by spectrophotometric measurements. Excess of 15S-HpETE was removed by gel filtration (Bio-Spin chromatography column, Bio-Rad Laboratories, Munich, Germany) and aliquots of the enzyme solution were mixed with double volume of matrix (sinapinic acid, 10 mg/ml) dissolved in acetonitrile/water/ trifluoracetic acid (3:2:0.1, by volume). Two microliters of this mixture were applied to a gold-plated sample holder and, after drying the holder, were introduced into the mass spectrometer. The spectra were obtained in the linear mode by summing 100 –200 laser slots with an acceleration voltage of 20 kV, a 70% grid voltage, and 300 ns delay. A low mass gate of m/z 500 was adjusted. As reference compound, human serum albumin (theoret-

Lipoxygenase affinity labeling

307

ical m/z 66,525) was used and an actual m/z of 67,005 was determined. Separation and identification of radioactive cleavage peptides The proteolytic peptides were separated by discontinuous SDS-PAGE using tricine as trailing ion [34]. This method was used because of its high resolution power in the 2–20 kDa range and because omission of glycine and urea prevents problems, which might occur in subsequent N-terminal amino acid sequencing when conventional SDS-PAGE is used. After separation, the gels were fixed for 20 min in 12% trichloroacetic acid, the peptide bands were stained with Coomassie blue (0.1% dye in 10% ethanol in water), and the electropherograms were exposed to radiographic films (Hyperfil-␤max, Amersham) or to an autoradiography plate. The plates were subsequently quantified with a BAS2000 imaging analyzer (Fuji Photo Film Co.). Isoelectric separation of the Lys-C endoproteinase cleavage fragments was carried out on a Biorad “Ready gel” (pH range of 3–10) using a commercial Ready gel focusing unit (Biorad). The Lys-C cleavage mixture was diluted with 2 vol. of 3-fold concentrated Biorad loading buffer and the sample was prefocused for 45 min at 100 V. A focusing current of 6 mA resulted. Then, the sample was focused at 200 V for 4 h and during the run the current went down from 9 mA to 1 mA. The gels were fixed, stained, and exposed to a radiographic film to identify radio-labeled protein bands. For N-terminal sequencing, the Lys-C cleavage peptides were separated by SDS-PAGE and blotted to a PVDF-membrane (Immobilon-P, Millipore, Eschborn, Germany). Bands were excised and subjected to Edman degradation (TOPLAB, Gesellschaft fu¨ r Angewandte Biotechnology, Martinsried, Germany).

Fig. 1. Suicidal inactivation of the rabbit 15-LOX during oxygenation of polyenoic fatty acids. The rabbit 15-LOX (0.24 ␮g/ml) was incubated with 0.25 mM linoleic acid (trace A) or 0.25 mM arachidonic acid (trace B) at room temperature in 0.1 M phosphate buffer, pH 7.4, containing 0.2% sodium cholate, and the increase in absorbance at 235 nm was monitored. The arrows indicate the addition of further substrate.

irreversible enzyme inactivation. Previous studies [21– 30] suggested that enzyme inactivation might be due to LOX/hydroperoxide interaction. To confirm these results, the rabbit 15-LOX was incubated with 15S-HpETE (product of arachidonic acid oxygenation) and 13S-HpODE (product of linoleic acid oxygenation) and the rates of enzyme inactivation were assayed. From Table 1 it can be seen that a 20 min incubation of the enzyme with 10-fold molar excess of 15S-HpETE led to an almost complete inactivation of the rabbit 15-LOX. A similar inactivation was observed when 12S-HpETE was used. In contrast, with 13SHpODE the enzyme was only half inactivated under strictly comparable conditions and the extent was increased to about 80% when the 13S-HpODE concentration was increased. Interestingly, 15S-HETE was only a poor inactivator indicating the importance of the peroxy

RESULTS

Inactivation of rabbit 15-LOX by different substrates LOXs undergo suicidal inactivation during the oxygenation of polyenoic fatty acids. In this study, we compared the rates of suicidal inactivation during linoleic acid and arachidonic acid oxygenation and found that the enzyme was more rapidly inactivated during arachidonic acid oxygenation (Fig. 1). Under our experimental conditions, complete inactivation was observed after about 200 s when arachidonic acid was used as substrate. In contrast, with linoleic acid there was a residual activity of about 15% even after 600 s of incubation. Addition of fresh substrate (arrows in Fig. 1) and further supply of oxygen to the assay sample (not shown) did not lead to an increase in the absorbance at 235 nm, indicating

Table 1. Inactivation of the Rabbit 15-LOX by Various Hydro(pero)xy Fatty Acids Inactivator

Structure None 13-HpODE 15-HpETE 12-HpETE 15-HETE

Concentration (␮M)

Enzyme/inactivator ration

Inactivation (%)

0 20 40 20 20 20

– 1:10 1:20 1:10 1:10 1:10

⬍5 46 80 97 82 13

The pure rabbit 15-LOX (2 nmoles) was preincubated with the inactivators in 1 ml of 0.1 M phosphate buffer, pH 7.4, for 20 min at 25°C. Then the LOX reaction was started by the addition of 0.25 mM linoleic acid and the initial oxygenation rates assayed spectrophotometrically.

308

R. WIESNER et al.

Fig. 2. Time course of rabbit 15-LOX inactivation by 15S-HpETE. The rabbit 15-LOX (1 mg/ml) was incubated with a 10-fold molar excess of 15S-HpETE in 0.1 M phosphate buffer, pH 7.4. At the times indicated, aliquots were taken and the residual linoleic acid oxygenase activity was assayed in the standard assay system (see Materials and Methods). Each data point represents the mean of duplicate measurements. It should be stressed that the time axis is in log-scale.

group. Under anaerobic conditions (data not shown), a similar extent of enzyme inactivation was reached, excluding a role of secondary oxygenase reactions. Taken together, these data suggest that the epoxy-leukotriene synthase activity, which also proceeds under anaerobic conditions, may be involved in suicidal inactivation. More detailed studies of the inactivation kinetics (Fig. 2) indicated a rapid process. During the first incubation minute almost 75% of the enzyme were inactivated. Suicidal inactivation is accompanied by covalent enzyme modification Since previous studies on the porcine leukocyte 12LOX [31] suggested that its suicidal inactivation is accompanied by covalent enzyme modification, we investigated whether the rabbit 15-LOX may behave similarly. For this purpose, the enzyme was incubated with radiolabeled 15S-HpETE, the protein was precipitated, washed extensively to remove unbound radioactivity, and run on SDS-PAGE under strongly denaturing conditions. From Fig. 3 it can be seen that the LOX protein migrating in the 75 kDa region was radio-labeled (trace A), indicating a covalent modification of the enzyme. A similar labeling was observed when the incubation was carried out under anaerobic conditions (trace B). When the 15-LOX was boiled before incubating with 15SHpETE, a 65% reduction of enzyme labeling was observed in this particular experiment (trace C). In additional experiments, we attempted to rule out the possibility that labeling might be an unspecific process, which is simply due to the fact that the reactive hydroperoxy group attacks amino acid side chains. For this purpose defatted bovine serum albumin (trace D) and human ␥-globulin (not shown) were incubated with 15S-HpETE under identical experimental conditions;

Fig. 3. Radioactive labeling of the rabbit 15-LOX during incubation with [1-14C]-15S-HpETE. The rabbit 15-LOX (5 ␮M) was incubated with a 5-fold molar excess of 15S-HpETE in 0.1 M phosphate buffer, pH 7.4, under aerobic and anaerobic conditions. After 60 min, protein was precipitated, washed, and the pellet was reconstituted in electrophoresis loading buffer (see Materials and Methods). SDS-PAGE was carried out on a 10% polyacrylamide gel and the pure 15-LOX was used as marker. After electrophoresis, the gel was fixed in 12% trichloroacetic acid and radioactivity was quantified as described in Materials and Methods. (A) 15-LOX ⫹ 15S-HpETE aerobic conditions, (B) 15-LOX ⫹ 15S-HpETE anaerobic conditions, (C) heatinactivated 15-LOX (5 min boiling) ⫹ 15S-HpETE aerobic conditions, and (D) bovine serum albumin ⫹ 15S-HpETE.

but, we never detected significant labeling of these proteins. These data suggest that covalent modification of the LOX is a rather specific process, which requires the catalytic activity of the enzyme. Since incubation with [1-14C]-15S-HETE did not lead to enzyme modification and labeling did not depend on the presence of oxygen, the epoxy-leukotriene synthase activity of the enzyme appears to be involved. Moreover, the lower rate of inactivation induced by 13S-HpODE (Fig. 1), which can not be converted to epoxy leukotrienes, suggests the involvement of this enzyme activity. To obtain additional evidence for covalent modification of the 15-LOX matrix-assisted laser desorption, mass spectrometry of the native and inactivated 15-LOX was carried out. From Fig. 4 it can be seen that native enzyme migrated with a m/z of about 76,000 Da, which is in the range of its theoretical mass. In contrast, the inactivated enzyme species migrated at 76,371 Da. These data are compatible with the assumption that an oxygenated arachidonic acid derivative is covalently linked to the enzyme during suicidal inactivation, but the chemical structure of the label could not be determined. In principle, 12S-HpETE is also a substrate for epoxyleukotriene formation, since it contains a reactive peroxy group at C12 and a bisallylic methylen (C7). Thus, hydrogen abstraction from C7 in connection with a homolytic cleavage of the peroxy group may lead to the formation of an 11,12-epoxy-leukotriene. Since the rabbit 15-LOX is capable of catalyzing C7-hydrogen ab-

Lipoxygenase affinity labeling

309

Fig. 5. Radioactive labeling of the rabbit 15-LOX by 12S- and 15SHpETE. The rabbit 15-LOX (5 ␮M) was incubated with a 5-fold molar excess of 15S-HpETE (A) and 12S-HpETE (B) in 0.1 M phosphate buffer, pH 7.4. After 60 min, protein was precipitated, washed, and the pellet was reconstituted in electrophoresis loading buffer (see Materials and Methods). SDS-PAGE was carried out on a 10% polyacrylamide gel and the pure 15-LOX was used as marker. After electrophoresis, the gel was fixed in 12% trichloroacetic acid and radioactivity was quantified as described in Materials and Methods. (C) To exclude artificial protein labeling induced during the workup procedure, the enzyme was first precipitated with acetone and then labeled 15S-HpETE was added. Subsequent workup as described for samples (A) and (B). Fig. 4. Mass spectroscopic analysis of native and 15S-HpETE-inactivated 15-LOX. The rabbit 15-LOX (1 mg/ml) was incubated in the presence or absence of a 10-fold molar excess of 15S-HpETE in 0.1 M phosphate buffer, pH 7.4. After 30 min of incubation, aliquots were analyzed by mass spectroscopy as described in Materials and Methods.

straction [35], it was straightforward to assume that 12S-HpETE may also constitute an activity label for the 15-LOX protein. From Fig. 5 it can be seen that 12SHpETE (trace B) induced a similar protein labeling as 15S-HpETE (trace A). To exclude that LOX labeling may be due to methodological artifacts, a control experiment was carried out in which the 15-LOX was precipitated with acetone before the addition of radio-labeled 15S-HpETE. In this case (trace C), we did not observe any protein labeling. To obtain further evidence for the specificity of the labeling process, we quantified the labeling stoichiometry under variable experimental conditions (variation of the 15S-HpETE/enzyme ratio between 1:1 and 12:1, variation of the incubation period between 6 and 60 min). In all experiments (n ⫽ 10), we calculated a molar labeling ratio of 1.3 ⫾ 0.2. These data indicate that little more than 1 inactivator molecule was bound per mole enzyme and this ratio did neither depend on the concentration of 15S-HpETE nor on the duration of the incubation period.

From Fig. 3 it can be seen that boiling of the enzyme impaired protein labeling. When the molar labeling ratio of the control incubations (1.3 ⫾ 0.2 mole inactivator/ mole enzyme, n ⫽ 10) was set at 100%, an 80.9 ⫾ 6.8% (n ⫽ 6) reduction of enzyme labeling was achieved when boiled enzyme was used. This labeling ratio was surprisingly high if one considers that the catalytic activity of the enzyme was completely abolished. One possibility to explain this unexpected finding was the fact that boiling the enzyme releases the catalytically active nonheme iron, which may act as catalyst for radical-mediated decomposition of 15S-HpETE. If this scenario were true, one would expect a reduction of labeling intensity if the incubation is carried out in the presence of iron-chelating agents. To test this hypothesis, we compared the extent of protein labeling of the native LOX with that of the boiled enzyme in the absence and presence of EDTA (5 mM final concentration). From Fig. 6 it can be seen that in this particular experiment heat denaturation in the absence of EDTA reduced enzyme labeling by about 65%. However, when EDTA was present labeling intensity was even lower (16%). In addition, we measured a decrease in absorbance at 235 nm when the heat-denatured enzyme was incubated with 15S-HpODE, indicating a decomposition of conjugated dienes.

310

R. WIESNER et al.

Fig. 6. Effect of heat denaturation of 15-LOX on protein labeling. The rabbit 15-LOX (5 ␮M) was incubated with a 5-fold molar excess of 15S-HpETE in 0.1 M phosphate buffer, pH 7.4. After 60 min, protein was precipitated, washed, and the pellet was reconstituted in electrophoresis loading buffer (see Materials and Methods). SDS-PAGE was carried out on a 10% polyacrylamide gel and the pure 15-LOX was used as marker. After electrophoresis, the gel was fixed in 12% trichloroacetic acid and radioactivity was quantified as described in Materials and Methods. (A) 15-LOX ⫹ 15S-HpETE (control incubation, labeling was set at 100% for quantification), (B) heat-inactivated 15-LOX (5 min boiling) ⫹ 15S-HpETE, and (C) heat-inactivated 15-LOX (5 min boiling in the presence of 5 mM EDTA) ⫹ 15SHpETE.

The results of the above-described experiments indicated two aspects of parallelism of 15-LOX inactivation and enzyme labeling linking the two processes to each other: (i) 15S-HpETE inactivates and labels the 15-LOX, but 15-HETE fails to do both; and, (ii) oxygen is neither required for inactivation nor for enzyme labeling. To show further parallelism, we studied the time course of protein labeling. From Fig. 7 it can be seen that, with enzyme inactivation (Fig. 2), protein labeling was rather rapid; within the first 6 min of incubation maximal labeling of the enzyme was reached. These data suggest that covalent protein modification may contribute to suicidal inactivation of the enzyme. Active site peptides are labeled during 15-LOX/ 15-HpETE interaction Quantification of the labeling stoichiometry indicated a 1.3:1 molar labeling ratio and such a ratio can hardly be reconciled with random protein modification. Instead, the requirement of the enzymatic activity for both, enzyme inactivation and protein labeling, suggested the involvement of the catalytic activity (LTA4-synthase activity) and, thus, preferential labeling of active site peptides was likely. To obtain experimental evidence for this hypothesis, the labeled 15-LOX was digested with Lys-C endoproteinase and the cleavage fragments were separated by SDS-PAGE and checked for radioactivity.

Fig. 7. Time course of 15S-HpETE-induced protein labeling. The rabbit 15-LOX (5 ␮M) was incubated with a 5-fold molar excess of 15SHpETE in 0.1 M phosphate buffer, pH 7.4. At the time points indicated, aliquots were taken and the protein was precipitated. Further workup procedure, electrophoresis, and data evaluation are described in Materials and Methods.

To optimize the cleavage and separation procedure, we carried out seven experiments along this line and two representative electropherograms are shown in Fig. 8. Typically, we observed two major radioactive cleavage fragments: a fragment migrating with an apparent molecular weight of 16 kDa and one at about 6 kDa. The relative labeling intensity of the two major fragments varied in different experiment (traces B and C), but in either case the two bands were dominant. In addition, several less intensely labeled bands were observed but they never became prominent under any of the experimental conditions. The theoretical Lys-C cleavage pattern, which is summarized in Table 2, indicated that no 16 kDa fragment is expected if digestion was complete. However, if one considers the possibility of partial cleavage, the most probable candidate to explain the 16 kDa fragment is P26 ⫹ P27. For the 6 kDa cleavage fragments three major candidates (P19, P29, P30) had to be considered. For direct identification, the cleavage peptides were blotted to an Immobilin membrane and their N-terminal amino acid sequences were determined by Edman degradation. The data obtained are summarized in Table 3. For the 16 kDa cleavage peptides, we obtained two different N-terminal sequences. The major sequence, which accounted for about 70%, was identical with the N-terminus of the 15-LOX. Considering the molecular weight of about 16 kDa, one has to conclude that this digestion fragment comprises the theoretical cleavage peptides P1-P10/P11. These peptides form the N-terminal ␤-barrel domain of the enzyme and the condensed structure of this domain may be discussed as possible reason for its relative resistance toward Lys C-endopro-

Lipoxygenase affinity labeling

311

Table 2. Theoretical LysC-Endoproteinase Cleavage Fragments of the Rabbit 15-LOX

Fragment P1 P2 P3 P4 P5 P6 P7 P8 P9

Fig. 8. Lys-C endoproteinase cleavage of 15-LOX labeled with [1-14C]15S-HpETE. The rabbit 15-LOX (5 ␮M) was incubated with a 5-fold molar excess of [1-14C]-15S-HpETE in 0.1 M phosphate buffer, pH 7.4. After 60 min, protein was precipitated, washed, and the pellet was reconstituted in 4 M guanidinium hydrochloride. Protein derivatization, Lys-C endoproteinase cleavage, and tricine SDS-PAGE were carried out as described in Materials and Methods. After electrophoresis, the gel was fixed in 12% trichloroacetic acid and radioactivity was quantified using a BAS200 imaging analyzer. Two independent experiments indicating the variability of labeling intensity of the 6 kDa and 16 kDa bands are shown. (A) Coomassie staining, first experiment; (B) radioactivity imaging, first experiment; (C) radioactivity imaging, second experiment; and, (D) Coomassie staining, second experiment.

teinase cleavage. The second sequence, which accounted for about 30% of the peptides found in the 16 kDa region, indicated P26 (Table 3). However, its molecular weight is only about 5 kDa and this result did not fit the 16 kDa molecular weight range. However, if one considers incomplete cleavage the molecular weight of P26 ⫹ 27 is about 16 kDa. P26 ⫹ 27 contains important active site structural elements, such as F353 and I418, which have been identified as sequence determinants for the positional specificity [33,36]. Together with I593 [33] they form the bottom of the substrate-binding pocket and were predicted to physically contact LOX substrates. In addition, P26 ⫹ 27 contains H361 and H366 that have been identified as ligands for the catalytic nonheme iron [37]. The localization of P26 ⫹ 27 within the active site of the enzyme is shown in Fig. 9A. In the 6 kDa molecular weight region, three major cleavage peptides (P19, P29, P30) were expected (Table 2). In fact, N-terminal amino acid analysis revealed that all of them could be detected. P19 does not contain any active site structural elements and, in the 3D structure, it was located mainly at the surface of the protein. In contrast, P30 contains I593, which has been identified as sequence determinant of the positional specificity [33]. P29 contains iron ligands H541 and H545 [37] and, thus, the peptide is involved in structuring the active site (Fig. 9). Comigration of the three peptides in the labeled region does not mean that all of them are radioactively

P10 P11 P12 P13 P14 P15 P16 P17 P18 P19 P20 P21 P22 P23 P24 P25 P26 P27 P28 P29 P30 P31 P32 P33

Amino acids

MW (Da)

1–18 19–20 21–44 45–49 50–54 55–66 67 68–71 72–91 92–126

1818 260 2721 681 546 1465 145 544 2254 3862

127–137 135–145 146–156 157–170 171 172–188 189–198 199–213 214–266 267–269 270–285 286–304

1510 1067 1100 1690 146 1847 1155 1858 6173 389 1738 2152

305–310 311–322 323–344 345–387 388–484 485–512 513–569 570–625 626–625 636–643 644–662

657 1420 2377 4994 10879 3205 6320 6327 1216 1016 2191

Secondary structural elements Sheet 1, strand 1-1 Sheet Sheet Sheet Sheet

2, 1, 1, 2,

strand strand strand strand

2-1 and 2-2 1-2 1-2 2

Sheet 1, strand 1-3 Sheet 1, strand 1-4; sheet 2, strand 2-4, helix 1 Helix 2 Sheet 3, strand 3-1 Sheet 3, strand 3-2 Helices 3 and 4 Helix 5 Helix 5 Helix 6 Helices 7–9 Sheet 4, strand 4-2; helix 10 Sheet 4, strand 4-3; sheet 5, strand 5-1; sheet 5, strand 5-2 Sheet 4, strand 4-3 Sheet 4, strand 4-4 Sheet 4, strand 4-4; helix 11 Helices 11–14 Helices 14–21 Helices 22 and 23 Helices 24–28 Helices 29–31 Helix 31 Helix 31

Fragments were determined from the cDNA sequence of the rabbit 15-LOX using the ProMass program.

labeled. For further identification, we carried out analytical isoelectric focusing. From Fig. 10 it can be seen that three major bands of radioactive peptides (a– c) were identified. N-terminal amino acid sequencing indicated cleavage peptide P16 as major constituent of Fig. 10 band (a). Its theoretical PI is 4.5 and, in our experiments, it migrated in the predicted region. In SDS-PAGE it escaped detection because of its low molecular weight. In the 3D structure, this cleavage peptide is the major constituent of helix 5, which is exposed at the protein surface (Fig. 9B). It is located close to the entrance into the substrate-binding pocket and contains W181, which has recently been identified as sequence determinant for membrane binding activity of the enzyme [38]. N-terminal amino acid sequencing of Fig. 10 band (b), which migrated with a PI of 5.2, indicated the presence of cleavage peptide P30. This peptide, which contains I593 (sequence determinant for positional specificity),

R. WIESNER et al.

312

Table 3. N-terminal Amino Acid Analysis of Radioactively Labeled Lys-C Cleavage Peptides Fragment size (kDa) 16

6

Amino acid sequence

Share (%)

Location in the protein

G V Y R V (C) V S T G A

70

XXVRSSDFQV

30

DATLETVMATL

40

LARRVRDSWQE

40

XGFPT

20

Peptide 1-10/11 (N-terminus) G1-R137 Peptide 26-27 (active site) C345-Y484 Peptide 30 (active site) D500-E625 Peptide 19 (protein surface) L214-E266 Peptide 29 (active site) Q513-T569

Contains these structural active-site elements Mainly ␤-barrel domain I418, M419, F353 (determinants of positional specificity), H361, H366 (iron ligands) I593 (determinant of positional specificity) None H541, H545 (iron ligands)

The 15-LOX was inactivated with a 5-fold molar excess of 15S-HpETE. The protein was precipitated with ice-cold acetone, and the sample was worked up as described in Materials and Methods.

has a theoretical PI of 5.3 and migrated in SDS-PAGE in the strongly labeled 6 kDa region. As indicated in Fig. 9A, it is localized at the active site of the LOX. Figure 10 band (c), which migrated in isoelectric focusing with a PI of 6.9, is a rather diffuse peptide fraction and we were unable to identify matching N-terminal peptide sequences. Considering the theoretical PI of the Lys-C cleavage fragments, one may conclude that P27 with a PI of 7.0 migrates in this region. All other complete cleavage fragments should either trail P27 (P32, PI ⫽ 7.3; P11, PI ⫽ 7.5; P3, PI ⫽ 7.5) in isoelectric focusing or lead it (P15, PI ⫽ 6.1; P7, PI ⫽ 6.1). Thus, the cleavage peptide P27, which contains the sequence determinants of positional specificity I418 and M419, is likely to be covalently modified. Summarizing these data, one can conclude that the active site peptide P30 is covalently modified as indi-

cated by SDS-PAGE and isoelectric focusing. In addition, SDS page provided evidence for radiolabeling of the active site peptides P26 ⫹ 27 and P29. Moreover, isoelectric focusing indicated covalent modification of the surface peptide P16. The fact that at least two active site peptides (P26 ⫹ 27 and P30) are covalently modified suggests that the inactivating intermediate has the freedom to react with more than one partner and, thus, this interaction may not be absolutely specific. However, the close to equimolar stoichiometry and the saturation kinetics of enzyme labeling suggest that covalent modification of the enzyme constitutes one mechanism of suicidal enzyme inactivation. In addition to active site constituents, surface peptides may also be modified to a certain extent and this may contribute to the observed labeling stoichiometry of higher than 1:1.

Fig. 9. Structural aspects of rabbit 15-LOX. The X-ray coordinates of the rabbit 15-LOX were used to construct this figure. Since the data set is incomplete, the lacking features (amino acids C210, G211, E601, P602 and the residues 177–187) were modeled in. (A) Structure of the active site. The peptides detected in the 6 kDa (P29, P30) and 16 kDa region (P26⫹27) of the SDS-PAGE (Fig. 8) are shown as helices. Functionally important active site amino acids [iron ligands (H366, H361, H541, H545) and sequence determinants of the positional specificity (F353, I418, M419, I593)] are indicated as CPK model. (B) Overall enzyme structure indicating the surface-exposed position of helix 5 (Lys-C cleavage peptide 16). R403, which is supposed to be involved in substrate binding (may interact with carboxylic group of the fatty acids), is shown in yellow.

Lipoxygenase affinity labeling

313

Scheme 2. Principle mechanisms of suicidal inactivation of LOXs.

Fig. 10. Isoelectric focusing of the radioactive Lys-C cleavage peptides. The rabbit 15-LOX (5 ␮M) was incubated with a 5-fold molar excess of [1-14C]-15S-HpETE in 0.1 M phosphate buffer, pH 7.4. After 60 min, protein was precipitated, washed, and the pellet was reconstituted in 4 M guanidinium hydrochloride solution. Protein derivatization and Lys-C endoproteinase cleavage were carried out as described in Materials and Methods. An aliquot of the cleavage sample was applied to isoelectric focusing using a Biorad Ready gel with a pH range of 3–10. After focusing, the gel was fixed in 12% trichloroacetic acid, stained with 0.2% Serva-blue, dried, and exposed to a radiographic film for radioactivity imaging (see Materials and Methods).

DISCUSSION

Suicidal inactivation is a common property of enzymes involves in eicosanoid biosynthesis. Most mammalian LOXs [21,31], cyclooxygenase isoforms [39 – 42], and other enzymes of eicosanoid synthesis [43– 45] are rapidly inactivated when reacting with polyenoic fatty acids. For the leukotriene A4 hydrolase, an active site peptide was covalently modified during inactivation [45]. Surprisingly, some LOX isoforms, such as the platelet-type 12-LOX of various mammalian species [27,31], are remarkably resistant to suicidal inactivation, whereas others [21–26,28 –31] are quite susceptible. The molecular mechanism of suicidal inactivation has been a matter of discussion for several years, but for the time being there is no common mechanistic scenario for this process. In fact, summarizing the literature data and considering the results presented in this study, one has to conclude that there is not just one but several mechanisms of suicidal inactivation. In principle, two inactivating mechanisms may occur (Scheme 2): 1) Syncatalytic inactivation. In this case, the enzyme is inactivated by radical intermediates formed during the LOX reaction. At hypobaric oxygen concentrations or anaerobiosis, oxygenation of the carbon-centered fatty acid radical (allylic or pentadienylic) formed via initial hydrogen abstraction may become rate limiting. Because of their extremely low half-life, fatty acid radicals are unlikely to escape the active site and, thus, may attack

amino acid residues of the substrate-binding pocket, which may lead to irreversible enzyme inactivation. Similarly, the oxygen-centered hydroperoxy radical, which is also an intermediate of the oxygenase reaction, may lead to syncatalytic enzyme inactivation. However, because of its longer half-life it may even escape the active site to modify surface peptides or surrounding xenobiotics [16,17]. 2) Postcatalytic inactivation. In this case, the enzyme reacts with the hydroperoxy fatty acid formed during the oxygenase reaction and there are three different principle routes: (i) chemical modification of amino acid residues by the peroxide without the involvement of enzymatic activity [46]; (ii) formation of inactivating intermediates (e.g., alkoxy radical) during the hydroperoxidase reaction [3]; and, (iii) formation of reactive intermediates during epoxy-leukotriene formation. It should be stressed that hydroperoxy fatty acids containing bisallylic methylenes (e.g., 15-HpETE and 5-HpETE) can be metabolized by LOXs via the hydroperoxidase and epoxy-leukotriene synthase pathway. Thus, the enzymes can be inactivated by either mechanism. In contrast, fatty acid hydroperoxides that lack bisallylic methylenes (13S-HpODE) can only be converted via the hydroperoxidase pathway so that the epoxy-leukotriene inactivation mechanism is not possible. In summary, one may conclude that there is no uniform mechanism of suicidal inactivation of LOXs. In contrast, it appears to be a complex process in which several mechanisms may overlay. However, under defined experimental conditions certain mechanisms may be dominant while others may not occur. In fact, under our experimental conditions (no oxygen limitations) syncatalytic does not play a major role. When arachidonic oxygenation is carried out in the presence of peroxide-reducing enzyme (phospholipid hydroperoxide glutathione peroxidase), suicidal inactivation is slowed down (not shown). In addition, we showed that covalent labeling is strongly reduced when heatinactivated enzyme was used. These data suggest that the

314

R. WIESNER et al.

chemical modification inactivation mechanism may not be of major importance. Thus, suicidal inactivation of the 15-LOX during arachidonic acid oxygenation proceeds via the hydroperoxidase inactivation mechanisms and/or epoxy-leukotriene synthase. It is impossible to calculate the relative contribution of the two pathways, but comparison of the inactivation rates observed during linoleic acid and arachidonic acid oxygenation suggests a significant contribution of the leukotriene synthase pathway. Another possibility to evaluate the relative shares of either inactivation mechanisms is to compare radioactive labeling of the enzyme protein by 15S-HpETE and 13S-HpODE. In this study, we have not tested radioactive labeling of the rabbit 15-LOX by 13S-HpODE, but corresponding experiments had been carried out for the porcine ortolog. For this LOX isoform, a considerable reduction of enzyme labeling with 13-HpODE [31] and these data are in line with our kinetic measurements (Fig. 1). The experiments with heat-denatured enzyme indicated a requirement for enzymatic activity for protein labeling. Moreover, calculations of the labeling stoichiometry indicated a close to, but not a perfect, 1:1 molar ratio. These data suggest the enzyme labeling may not be a random process, but may be restricted to specific sites, most probably to active site peptides. In fact, SDS-PAGE of the labeled Lys-C cleavage fragments revealed five different peptides (P1 ⫺ 10, P19, P26 ⫹ 27, P29, and P30) and three of them (P26 ⫹ 27, P29, and P30) contribute to the active site structure. Isoelectric focusing confirmed covalent modification of P30, but with this method we did not obtain evidence for labeling of the other peptides. In addition, another covalently modified cleavage peptide (P16) was found in isoelectric focusing and this peptide is located at the protein surface. It is quite plausible that covalent modification of active site peptides is accompanied by irreversible inactivation. In contrast, modification of surface peptides does not necessarily lead to enzyme inactivation. In fact, sitedirected mutagenesis of a P16 amino acid (W181E) is well tolerated by the enzyme [38]. On the other hand, linkage of a long fatty chain to a peptide, which is in close spatial neighborhood to the entry of the substrate-binding pocket, may impact the enzymatic activity. In additional experiments (not shown), we isolated the labeled peptide P30 and attempted to identify the target amino acid. Unfortunately, we faced two major methodological problems: (i) electrophoresis did not yield sufficient amounts of labeled peptide and numerous attempts failed to isolate modified P30 by RP-HPLC; and, (ii) when labeled cleavage peptides were isolated by SDS-PAGE and subsequently subjected to amino acid analysis, we lost the label almost completely. From these data, we concluded that acidic and alkaline hydrolysis might cleave the covalent

linkage between the amino acid and the radioactive probe. Thus, identification of the modified amino acid was not possible with this particular method. A similar loss of radioactive labeling was also observed when the precipitated enzyme was treated for 2 h with 1 M NaOH. Our mechanistic studies on 15-LOX/15S-HpETE interaction suggested that hydroperoxy fatty acids, which serve as substrate for epoxy-leukotriene formation, might constitute suitable probes for active site labeling. For mammalian LOXs, 3D data are only available for the rabbit 15-LOX and, thus, structural information on the active site of other LOX isoforms is limited. Unfortunately, the amino acid homology to other LOX isoforms is rather low so that structural models may not be very informative. Thus, reliable active site probes may be of particular interest to characterize the active site of other LOX isoforms. It should, for instance, be possible to affinity label the substrate-binding pocket of the human 5-LOX using radioactive 5S-HpETE as substrate. It is well known that this enzyme converts 5S-HpETE to 5,6-epoxy LTA4 [20] and this reaction is very similar to 14,15-LTA4 formation from 15-HpETE by the 15-LOX. Since site-directed mutagenesis did already provide useful information on the active site structure of this isoenzyme [47,48], our strategy may help to further characterize this functional environment. Acknowledgements — Financial support for this study was provided by Deutsche Forschungsgemeinschaft (Ku 961/7-1/2). Parts of this work were carried out during a sabbatical of H. K. at Tokushima University. The excellent technical assistance of Mrs. M. Anton is acknowledged. This article is dedicated to Prof. S. M. Rapoport on occasion of his 90th birthday.

REFERENCES [1] Brash, A. R. Lipoxygenases: occurrence, functions, catalysis, and acquisition of substrate. J. Biol. Chem. 274:23679 –23682; 1999. [2] Grechkin, A. Recent developments in biochemistry of the plant lipoxygenase pathway. Prog. Lipid Res. 37:317–352; 1998. [3] de Groot, J. J.; Veldink, G. A.; Vliegenthart, J. F.; Boldingh, J.; Wever, R.; van Gelder, B. F. Demonstration by EPR spectroscopy of the functional role of iron in soybean lipoxygenase-1. Biochim. Biophys. Acta 377:71–79; 1975. [4] Rickert, K. W.; Klinman, J. P. Nature of hydrogen transfer in soybean lipoxygenase 1: separation of primary and secondary isotope effects. Biochemistry 38:12218 –12228; 1999. [5] Funk, C. D. Prostaglandins and leukotrienes: advances in eikosanoid biology. Science 294:1871–1875; 2001. [6] Drazen, J. M.; Isreal, E.; O’Byrne, P. M. Treatment of asthma with drugs modifying the leukotriene pathway. N. Engl. J. Med. 340:197–206; 1999. [7] Rapoport, S. M.; Schewe, T.; Thiele, B. J. Maturational breakdown of mitochondria and other organelles in reticulocytes. In: Harris, J. R., ed. Blood cell biochemistry. New York: Plenum Press; 1990:151–194. [8] van Leyen, K.; Duvoisin, R. M.; Engelhardt, H.; Wiedmann, M. A function for lipoxygenase in programmed organelle degradation. Nature 395:392–395; 1998. [9] Maccarrone, M.; Melino, G.; Finazzi-Agro, A. Lipoxygenases

Lipoxygenase affinity labeling

[10] [11] [12]

[13] [14] [15]

[16] [17]

[18] [19]

[20] [21]

[22] [23] [24] [25] [26]

[27]

[28] [29] [30]

and their involvement in programmed cell death. Cell Death Differ. 8:776 –784; 2001. Chisolm, G. M.; Steinberg, D. The oxidative modification hypothesis of atherogenesis: an overview. Free Radic. Biol. Med. 28: 1815–1826; 2000. Cathcart, M. K.; Folcik, V. A. Lipoxygenases and atherosclerosis: protection versus pathogenesis. Free Radic. Biol. Med. 28:1726 – 1734; 2000. Maas, R. L.; Brash, A. R. Evidence for a lipoxygenase mechanism in the biosynthesis of epoxide and dihydroxy leukotrienes from 15(S)-hydroperoxyicosatetraenoic acid by human platelets and porcine leukocytes. Proc. Natl. Acad. Sci. USA 80:2884 –2888; 1983. K¨uhn, H.; Wiesner, R.; Alder, L.; Fitzsimmons, B. J.; Rokach, J.; Brash, A. R. Formation of lipoxin B by the pure reticulocyte lipoxygenase. Eur. J. Biochem. 169:593– 601; 1987. Joseph, P.; Srinivasan, S. N.; Kulkarni, A. P. Purification and partial characterization of lipoxygenase with dual catalytic activities from human term placenta. Biochem. J. 293:83–91; 1993. Nunez-Delicado, E.; Sojo, M. M.; Sanchez-Ferrer, A.; GarciaCarmona, F. Hydroperoxidase activity of lipoxygenase in the presence of cyclodextrins. Arch. Biochem. Biophys. 367:274 – 280; 1999. Kulkarni, A. P. Lipoxygenase: a versatile biocatalyst for biotransformation of endobiotics and xenobiotics. Cell. Mol. Life Sci. 58:1805–1825; 2001. Nunez-Delicado, E.; Sanchez-Ferrer, A.; Garcia-Carmona, F. A. A kinetic study of the one-electron oxidation of Trolox C by the hydroperoxidase activity of lipoxygenase. Biochim. Biophys. Acta 1335:127–134; 1997. Shimizu, T.; Radmark, O.; Samuelsson, B. Enzyme with dual lipoxygenase activities catalyzes leukotriene A4 synthesis from arachidonic acid. Proc. Natl. Acad. Sci. USA 81:689 – 693; 1984. Bryant, R. W.; Schewe, T.; Rapoport, S. M.; Bailey, J. M. Leukotriene formation by a purified reticulocyte lipoxygenase enzyme. Conversion of arachidonic acid and 15-hydroperoxyeicosatetraenoic acid to 14, 15-leukotriene A4. J. Biol. Chem. 260:3548 –3555; 1985. Shimizu, T. Enzymes functional in the synthesis of leukotrienes and related compounds. Int. J. Biochem. 20:661– 666; 1998. H¨artel, B.; Ludwig, P.; Schewe, T.; Rapoport, S. M. Self-inactivation by 13-hydroperoxylinoleic acid and lipohydroperoxidase activity of the reticulocyte lipoxygenase. Eur. J. Biochem. 126: 353–357; 1982. Rouzer, C. A.; Kargman, S. Translocation of 5-lipoxygenase to the membrane in human leukocytes challenged with ionophore A23187. J. Biol. Chem. 263:10980 –10988; 1988. de Carolis, E.; Denis, D.; Riendeau, D. Oxidative inactivation of human 5-lipoxygenase in phosphatidylcholine vesicles. Eur. J. Biochem. 235:416 – 423; 1996. Kim, M. R.; Kim, S. H.; Sok, D. E. Inactivation of potato lipoxygenase by hydroperoxy acids as suicide substrates. Biochem. Biophys. Res. Commun. 164:1384 –1390; 1989. Furukawa, M.; Yoshimoto, T.; Ochi, K.; Yamamoto, S. Studies on arachidonate 5-lipoxygenase of rat basophilic leukemia cells. Biochim. Biophys. Acta 795:458 – 465; 1984. Yoshimoto, T.; Miyamoto, Y.; Ochi, K.; Yamamoto, S. Arachidonate 12-lipoxygenase of porcine leukocyte with activity for 5-hydroxyeicosatetraenoic acid. Biochim. Biophys. Acta 713:638 – 646; 1982. Takahashi, Y.; Ueda, N.; Yamamoto, S. Two immunologically and catalytically distinct arachidonate 12-lipoxygenases of bovine platelets and leukocytes. Arch. Biochem. Biophys. 266:613– 621; 1988. Yamamoto, S.; Suzuki, H.; Ueda, N. Arachidonate 12-lipoxygenases. Prog. Lipid Res. 36:23– 41; 1997. Rapoport, S.; Hartel, B.; Hausdorf, G. Methionine sulfoxide formation: the cause of self-inactivation of reticulocyte. Eur. J. Biochem. 139:573–576; 1984. Gan, Q. F.; Witkop, G. L.; Sloane, D. L.; Straub, K. M.; Sigal, E. Identification of a specific methionine in mammalian 15-lipoxy-

[31]

[32] [33]

[34]

[35]

[36]

[37]

[38]

[39] [40] [41]

[42] [43]

[44] [45]

[46] [47]

[48]

315

genase which is oxygenated by the enzyme product 13-HPODE: dissociation of sulfoxide formation from self-inactivation. Biochemistry 34:7069 –7079; 1995. Kishimoto, K.; Nakamura, M.; Suzuki, H.; Yoshimoto, T.; Yamamoto, S.; Takao, T.; Shimonishi, Y.; Tanabe, T. Suicide inactivation of porcine leukocyte 12-lipoxygenase associated with its incorporation of 15-hydroperoxy-5,8,11,13-eicosatetraenoic acid derivative. Biochim. Biophys. Acta 1300:56 – 62; 1996. Kim, M. R.; Sok, D. E. Irreversible inhibition of soybean lipoxygenase-1 by hydroperoxy acids as substrates. Arch. Biochem. Biophys. 288:270 –275; 1991. Borngr¨aber, S.; Browner, M.; Gillmor, S.; Gerth, C.; Anton, M.; Fletterick, R.; K¨uhn, H. Shape and specificity in mammalian 15-lipoxygenase active site. The functional interplay of sequence determinants for the reaction specificity. J. Biol. Chem. 274: 37345–37350; 1999. Sch¨agger, H.; von Jagow, G. Tricine-sodium dodecyl sulfatepolyacrylamide gel electrophoresis for the separation of proteins in the range from 1 to 100 kDa. Anal. Biochem. 166:368 –379; 1987. Schwarz, K.; Borngr¨aber, S.; Anton, M.; K¨uhn, H. Probing the substrate alignment at the active site of 15-lipoxygenases by targeted substrate modification and site-directed mutagenesis. Evidence for an inverse substrate orientation. Biochemistry 37:15327–15335; 1998. Borngr¨aber, S.; Kuban, R. J.; Anton, M.; K¨uhn; H. Phenylalanine 353 is a primary determinant for the positional specificity of mammalian 15-lipoxygenases. J. Mol. Biol. 264:1145–1153; 1996. Gillmor, S. A.; Villasenor, A.; Fletterick, R.; Sigal, E.; Browner, M. F. The structure of mammalian 15-lipoxygenase reveals similarity to the lipases and the determinants of substrate specificity. Nature Struct. Biol. 4:1003–1009; 1997. Walther, M.; Anton, M.; Wiedmann, M.; Fletterick, R.; K¨uhn, H. The N-terminal domain of the reticulocyte-type 15-lipoxygenase is not essential for enzymatic activity but contains determinants for membrane binding. J. Biol. Chem. 277:27360 –27366; 2002. Miyamoto, T.; Ogino, N.; Yamamoto, S.; Hayaishi, O. Purification of prostaglandin endoperoxide synthetase from bovine vesicular gland microsomes. J. Biol. Chem. 251:2629 –2636; 1976. Lands, W. E. Biological consequences of fatty acid oxygenase reaction mechanisms. Prostaglandins Leukot. Med. 13:35– 46; 1984. Tong, H.; Richards, E.; Wood, C. E. Prostaglandin endoperoxide synthase-2 abundance is increased in brain tissues of late-gestation fetal sheep in response to cerebral hypoperfusion. J. Soc. Gynecol. Investig. 6:127–135; 1999. Smith, W. L.; DeWitt, D. L.; Garavito, R. M. Cyclooxygenases: structural, cellular, and molecular biology. Ann. Rev. Biochem. 69:145–182; 2000. Wade, M. L.; Voelkel, N. F.; Fitzpatrick, F. A. “Suicide” inactivation of prostaglandin I2 synthase: characterization of mechanism-based inactivation with isolated enzyme and endothelial cells. Arch. Biochem. Biophys. 321:453– 458; 1995. Jones, D. A.; Fitzpatrick, F. A. Thromboxane A2 synthase. Modification during “suicide” inactivation. J. Biol. Chem. 266:23510 – 23514; 1991. Mueller, M. J.; Wetterholm, A.; Blomster, M.; Jornvall, H.; Samuelson, B.; Haeggstrom, J. Z. Leukotriene A4 hydrolase: mapping of a henicosapeptide involved in mechanism-based inactivation. Proc. Natl. Acad. Sci. USA 92:8383– 8387; 1995. Rapoport, S. M.; H¨artel, B.; Schewe, T.; K¨uhn, H. Hydroperoxy fatty acids inactivate the reticulocyte lipoxygenase independently of a hydroperoxidase activity. FEBS Lett. 202:201–206; 1986. Nguyen, T.; Falgueyret, J. P.; Abramovitz, M.; Riendeau, D. Evaluation of the role of conserved His and Met residues among lipoxygenases by site-directed mutagenesis of recombinant human 5-lipoxygenase. J. Biol. Chem. 266:22057–22062; 1991. Schwarz, K.; Walther, M.; Anton, M.; Gerth, C.; Feussner, I.; K¨uhn, H. Structural basis of lipoxygenase specificity. J. Biol. Chem. 276:773–779; 2001.