Purification of Arachidonate 5-Lipoxygenase from ... - Semantic Scholar

1 downloads 0 Views 5MB Size Report
Purification of Arachidonate 5-Lipoxygenase from Porcine Leukocytes and Its Reactivity with Hydroperoxyeicosatetraenoic Acids*. (Received for publication ...
THEJOURNAL OF BIOLOGICAL CHEMISTRY 0 1986 by The American Society of Biological Chemists, Inc.

Val. 261, No. 17, Issue of June 15,pp. 7982-7988,1986 Printed in U.S. A.

Purification of Arachidonate5-Lipoxygenase from Porcine Leukocytes and Its Reactivity with Hydroperoxyeicosatetraenoic Acids* (Received for publication, October 22, 1985)

Natsuo Ueda, Shuji KanekoS, Tanihiro Yoshimoto, and. Shozo Yamamotog From the Department of Bwchemisty, Tokushimu University School of Medicine, Kurumoto-cho, Tokushimu770, Japan

Arachidonate 5-lipoxygenase was purified to near finity chromatography, we have been successful in the purihomogeneityfromthe 105,000 X g supernatantof fication of 12-lipoxygenasefrom porcine leukocytes (8).This porcine leukocyte homogenate by immunoaffinity paperreports the application of the same method to the chromatography using a monoclonal anti-5-lipoxygen- purification of 5-lipoxygenase from porcine leukocytes and ase antibody. Reaction of the purified enzyme with also describes the reactivity of the purified enzyme with arachidonic acid produced predominantly 5-hydropervarious hydroperoxy-eicosatetraenoicacids with special refo~y-6,8,l1,14-eicosatetraenoic acid with concomitant erence to theLT&l synthase activity. formation ofseveral more polar compounds in smaller amounts. These minor productswere identified as the EXPERIMENTALPROCEDURES degradationproductsofleukotriene A4, namely, 6M~teriuls-[l-’~C]Arachidonic acid (59.6 mCi/mmol) was purtruns-leukotriene B4 (epimeric at C-12) and an epimeric mixture of 5,6-dihydroxy-7,9,11,14-eicosa- chased from Amersham International (Amersham) and arachidonic acid from Nu-Chek-Prep(Elysian).6-truns-LTB4, 12-epi-6-trunstetraenoic acids. These compounds were also produced LTB4, 5,6-diHETEs and 12-0-methyl-5,12-diHETEs prepared from by reaction of the enzyme with B-hydroperoxy-eico- LTA, methyl ester were kindly provided by Dr. S. Terao of Takeda satetraenoic acid. Association ofthe 5-lipoxygenase Research Laboratories, and 5-HETE (racemic mixture) by Dr. M. andleukotriene A synthase activities was demon- Hayashi of Ono Research Institute. ATP, N,O-bis(trimethylsily1)strated by several experiments: heat inactivation of trifluoroacetamide, trimethylchlorosilane, and sodium borohydride enzyme, effect of selective 5-lipoxygenase inhibitors, were obtained from Wako Pure Chemical Industries (Osaka), Coorequirements of calcium ion and ATP, and self-cata- massie Brilliant Blue R-250 and sodium deoxycholate from Nakarai lyzed inactivation of enzyme.Theenzyme was also Chemicals (Kyoto), 3,3’-diaminobenzidine.4 HCl from Dojin (Kuactive with 12- and 15-hydroperoxy-eicosatetraenoic mamoto), high molecular weight standard mixture for SDS-polyacrylamide gel electrophoresis and soybean lipoxidase (type I) from acids producing (5S,12S)- and (5S,lEiS)-dihydroper- Sigma, dithiothreitol from Seikagaku Kogyo (Tokyo), and horse oxy acids, respectively. Maximal velocities of the re- biotinylated anti-mouse IgG and Vectastain ABC kit from Vector actions with these hydroperoxy acids as compared with Laboratories (Burliigame, CA). Cirsiliol was provided by Dr. T. Horie that of arachidonic acid (loo%,0.6 pmo1/3 min/mg of of our university and AA861 byDrs. S. Terao and Y. Maki of Takeda protein) were as follows: 5-hydroperoxy acid, 3.5%, Research Laboratories. Protein A-bearing Cowan I strain of Stuphy12-hydroperoxy acid,22%,and 15-hydroperoxy acid, lococcus uureus (NCTC 8530) was a gift from Dr. W. L. Smith of Michigan State University. The bacterial cells were fixed in 1.5% 30%.

The physiological role of arachidonate 5-lipoxygenase to initiate the biosynthesis of leukotrienes has been well established (1).Attempts have been reported to purify and characterize the enzyme from various animal species: rat basophilic leukemia cells (2-5), guinea pig (6), and human (7) polymorphonuclear leukocytes. The requirement of cofactors such as calcium ion (2) and ATP (6) was also described. According to a strategyto raise monoclonal anti-lipoxygenase antibodies with a partiallypurified enzyme as an antigen and to apply them to purification of lipoxygenases by immunoaf-

* This research was supported by grants-in-aid from the Ministry of Education, Science, and Culture and Ministry of Health and Welfare of Japan,and grants from the JapaneseFoundation of Metabolism and Diseases, Takeda Science Foundation, and Otsuka Pharmaceutical Company, Tokushima Research Institute. The costs of publication of this article were defrayed in partby the payment of page charges. This article must therefore be hereby marked “uduertisement” in accordance with 18U.S.C. Section 1734 solelyto indicate this fact. This work was preliminarily presented at the 6thAnnual Meeting of the Japanese Inflammation Society, Tokyo, July 12, 1985, and a t the 58th Annual Meeting of the Japanese Biochemical Society, Sendai, September 27, 1985. $ On leave from Ono Pharmaceutical Company, Research Institute. f To whom correspondence should be addressed.

formaldehyde and then heat-inactivated a t 80 “C for 5 min (9). Myeloma cell line (SP2/0-Ag14) was provided by Salk Institute (San Diego). Precoated Silica Gel60F254 glass plates for TLC were purchased from Merck (Darmstadt), silicic acid (100 mesh) for column chromatography from Mallinckrodt Chemical Works, Affi-Gel 10 from Bio-Rad, and protein A-Sepharose CL-4B from Pharmacia. C3H/He strain weregiven Preparation of Antibodies-Miceof intraperitoneally a partially purified preparation of 5-lipoxygenase from porcine leukocytes (10). Spleen cells of the immunized mice were fused with myeloma cells (SP-2/0-Ag14) with the aid of polyethylene glycol according to the method of Goding (11).The fusion mixture was subjected to the “HAT selection” (ll), and the hybridoma cells producing anti-5-lipoxygenase antibody were cloned in soft agar by the method of Kennet (12). A clone producing a species

The abbreviations used are: B-H(P)ETE, (5S)-hydroxy- or hydroperoxy-6-truns-8,11,14-cis-eicosatetraenoicacid; 12-HPETE, (12S)hydroperoxy-5,8-cis-l0-truns-14-cis-eicosatetraenoicacid; IEi-HPETE, (15S)-hydroperoxy-5,8,11-cis-13-trunss-eicosatetraenoic acid; (5S,12S)-diH(P)ETE, (5S,12S)-dihydroxy- or dihydroperoxy6-truns-8-cis-10-truns-14-cis-eicosatetraenoicacid; (5S,15S)diH(P)ETE, (5S,15S)-dihydroxy- or dihydroperoxy-6-truns-8,ll-cis13-truns-eicosatetraenoicacid; LT, leukotriene; LTA4, (5SI-truns5,6-oxido-7,9-truns-11,14-cis-eicosatetraenoic acid; 6-truns-LTB1, (5S,12R)-dihydroxy-6,8,10-truns-14-cis-eicosatetraenoic acid; 12-epi6-truns-LTB4, (5S,12S)-dihydroxy-6,8,10-truns-14-cis-eicosatetraenoic acid; 5,6-diHETE, 5,6-dihydroxy-7,9-truns-ll,l4-cis-eicosatetraenoic acid; 12-0-methyl-5,12-diHETE, (5S)-hydroxy-12-methoxy6,8,10-tr~ns-14-cis-eicosatetraenoic acid; HPLC, high-performance liquid chromatography; TLC, thin layer chromatography; SDS, SOdium dodecyl sulfate.

7982

5-Lipoxygenme and HydroperoxyeicosatetraenoicAcids of antibody referredto as5Lox-6,5Lox-I, or 5Lox-9 was grown in the peritoneal cavity of mouse, and theantibody in the ascites fluid was precipitated with ammonium sulfate and purified by the use of a protein A-Sepharose column (13). The details of the preparation of the monoclonal anti-5-lipoxygenase willbe reported elsewhere. A monoclonal antibody against 12-lipoxygenase of porcine leukocytes (lox-2) was prepared essentially as described above. Preparation and Assay of Enzyme-Porcine leukocytes were collected by the method described for the preparation of 12-lipoxygenase (14). About 65 g (wet weight) of leukocytes were obtained from 10 liters of whole blood, and the cells were suspended in 7 times the volume of 20 mM Tris-HC1 buffer, pH 7.4, containing 0.5 mM dithiothreitol and 0.5 mM EDTA. The cell suspension was subjected to sonic disruption a t 20 kHz for 15 s by the use of a Branson sonifier model 185D. The sonicate was centrifuged at 10,000 X g for 10 min, and the supernatant was further centrifuged at 105,000 X g for 60 min. The high-speed supernatant (350 ml, 7.6 mg of protein/ml) was referred to as thecytosol fraction and could be stored at -70 "C for a t least 1 month without an appreciable loss of 5-lipoxygenase activity. For immunoaffinity chromatography of the enzyme, the anti-5lipoxygenase antibody 5Lor-6 was conjugated to Affi-Gel 10 according to themanufacturer's instruction. The immunoaffinity column (10 X 70 mm) was equilibrated with 20 mM Tris-HC1 buffer at pH 7.4 containing 0.5 mM dithiothreitol and 0.5 mM EDTA. The cytosol fraction (20 ml) was applied to the column at a flow rate of 40 ml/h. The chromatography was performed at 28 "C. After the column was washed with 45 ml of the same buffer at a flow rate of 75 ml/h, the enzyme was eluted a t a flow rate of40 ml/h with 50 mM sodium carbonate buffer at pH 10.0 containing 0.2% sodium deoxycholate, 0.5 mM dithiothreitol, and 1mM EDTA. The eluate was immediately adjusted to pH 8.1 by the addition of one-sixth the volume of 0.5 M Tris-HC1 buffer at pH 7.4. Active fractions were collected and used as thepurified enzyme without further concentration in most of the experiments unless otherwise stated. Since the purified enzyme was unstable, usually the enzyme preparation was used immediately after elution from the column. The standard mixture for the 5-lipoxygenase assay contained 50 mM potassium phosphate buffer at pH 7.4, 2 mM CaC12, 2 mM ATP, 25 p M [l-'4C]arachidonic acid (50,000 cpm, 5 pl of ethanol solution), and enzyme in a final volume of 200 pl. The reaction was started by the addition of [l-'4C]arachidonic acid and continued a t 24 "C for 3 min. Termination of the reaction, extraction and separation of reaction products, and determination of radioactivity were performed as described previously (6). TLC was carried out using solvent system A: ethyl ether/petroleum ether/aceticacid (85:15:0.1). In thereaction without enzyme, about 3.5% of the totalradioactivity appeared in the product regions, and theenzyme activity was corrected for the background value. The intra-assay coefficient of variation of the standard assay was 4.4% (n = 10). For the LTA synthase assay, the enzyme was allowed to react a t 24 "C for 3 min in the 5-lipoxygenase assay mixture in which arachidonic acid was replaced by 2.5 p~ [1-14C]5-HPETE(60,000 cpm, 6 pl of ethanol solution). TLC was performed using solvent system B, the organic phase of a mixture of iso-octane/ethyl acetate/water/acetic acid (5011010020). The sections on the silica gel plate corresponding to 6-trans-LTB4 ( a n epimeric mixture at C-12) and 5,6-diHETE (an epimeric mixture) were scraped for radioactivity determination, and the ratio of the radioactivity of these productsto the total radioactivity was calculated. In thereaction without enzyme, about 1.8%of the total radioactivity appeared in the product regions, and the enzyme activity was corrected for the background value. The intra-assay coefficient of variation in the assay was 6.3% (n = 6 ) . A 12-lipoxygenase-free cytosol fraction of porcine leukocytes was prepared as follows. A 2% S. aureus suspension (1 ml) and a solution of anti-12-lipoxygenase antibody (lox-2, 21 pg in 0.5 ml) were mixed. The mixture was centrifuged at 1,500 X g for 10 min. The precipitate was suspended in 1 ml of the cytosol fraction of porcine leukocytes (7.6 mg of protein) andkept in an ice bath for 20 min. After centrifugation the supernatant solution was used as a 12-lipoxygenase-free preparation. 12-Lipoxygenasewas prepared by ammonium sulfate fractionation (2550%) of the cytosol fraction of porcine leukocytes which passed through the 5-lipoxygenase immunoaffinity column. The activity was assayed at 24 "C as described previously (14). In all these assays the enzyme activity was expressed in terms of nmol/3 min/mg of .protein unless otherwise stated. The protein

7983

concentration of enzymes was determined by the method of Lowry et al. (15). Electrophoresis of 5-lipoxygenase was performed on a 1-mm thick polyacrylamide gel (7.5%) in the presence of 0.1% SDS (16). Proteins on the gel were stained using Coomassie Brilliant Blue R-250. Electrophoretic transfer of protein bands from the gel to a nitrocellulose membrane (17) was carried out by the use of a Marysol gel electrophoresis apparatus model KS-8440GMT. The blot was stained with the aid of a Vectastain ABC kit and 3,3'-diaminobenzidine as peroxidase substrate either in the presence of monoclonal anti-5-lipoxygenase or in its absence according to themanufacturer's instruction. Preparation of HPETEs-[~-'~C]~-HPETEwas prepared by incubation of the purified 5-lipoxygenase (0.15 mg of protein) with 25 pM [l-'4C]arachidonic acid in a 2-ml reaction mixture with the same composition described for the standardassay. Several tubes were run simultaneously. After 5 min at 24 "C, the reaction mixture was acidified with 0.2 ml of 1 M citric acid and extracted twice each with 8 ml of ethyl ether. The solvent was evaporated, and theresidue was applied to a silicic acid column (0.8 g) pre-equilibrated with petroleum ether/ethyl ether(91).5-HPETE was eluted with a mixture of petroleum ether/ethyl ether (7:3). The solvent was evaporated, and the dried material was repurified by the same procedure. The purity of [1-14C]5-HPETEthus prepared was examined by TLC (more than 92%). For the preparation of [1-14C]12-HPETE, 12-lipoxygenase (0.80 mg of protein) was incubated with 40 p~ [l-14C]arachidonicacid in the standard assay mixture (14), which was scaled up to 2 ml as described for 5-HPETE. Elution of 12-HPETE from silicic acid column was performed with a solvent mixture of petroleum ether/ ethyl ether (82). [1-14C]15-HPETEwas prepared by incubation of soybean lipoxygenase (0.10 mg of protein) with 20 p~ [l-14C]arachidonicacid in the presence of 50 mM Tris-HC1 buffer at pH 9.0 in a total volume of 2 ml. After the reaction at 24 "C for 3 min, the mixture was acidified and extracted as described for 5-HPETE. Elution from silicic acid column was performed with a solvent mixture of petroleum ether/ ethyl ether (752.5). For the preparation of (5S,12S)-diHETE, 12-lipoxygenase (0.80 mg of protein) was incubated with 5-HPETE in the same reaction mixture described for 12-HPETE. When the reaction was complete, sodium borohydride (1 mg) was added, and the ethereal extract was applied to a silicic acid column. The ethyl acetate eluate from the column was used as anauthentic (5S,12S)-diHETE which was identified previously (14). (5S,lSS)-diHETEwas produced by incubation of soybean lipoxygenase (0.10 mg of protein) with 5-HPETE (18) as described for 15-HPETE and isolated as described for (5S,12S)diHETE. High-Performance Liquid Chromatography (HPLC)-For the analyses of various reaction products, reverse-phase HPLC was performed by using a TSK-GEL column (type ODs-12OT, 4.6 X 250 mm) equipped with a Waters 6000A pump and a JASCO spectrophotometer model Uvidec-10011. The solvent system was a mixture of methanol/water/acetic acid (75:25:0.01) at a flow rate of 1.0 ml/min. UV spectra of the products were recorded as a solution in the above solvent system. Gas Chromatography-Mass Spectrometry-For identification of various reaction products by gas chromatography-mass spectrometry, the purified 5-lipoxygenase was incubated with an appropriate substrate in the 2-ml mixture mentioned for 5-HPETE preparation. The ethereal extract was purified by a silicic acid chromatography, and the ethyl acetate eluatewas further purified by reverse-phase HPLC. Each product andthe corresponding authentic compound were treated with diazomethane and then with a (4:l) mixture of N,Obis(trimethylsily1)trifluoroacetamide and trimethylchlorosilane in pyridine. The trimethylsilyl derivative of methyl ester of each product was analyzed with the aid of a Hitachi gas chromatograph-mass spectrometer model M-80A equipped with a 3 X 1000-mm column of 1.5% OV-1 Gas Chrom Q. Ionizing energy was 20 eV.

RESULTS AND DISCUSSION Enzyme Purificationby Immunoaffinity Chromatography-

When the cytosol fraction of porcine leukocytes was applied to a column of Affi-Gel10 to which monoclonal anti-5lipoxygenase (5Lox-6) was linked, a bulk of protein passed through the column (Fig. 1).The 12-lipoxygenasewas found in this fraction, which didnotcontain 5-lipoxygenase as

7984

5-Lipoxygenaseand Hydroperoxyeicosatetraenoic Acids

Elution Volume(m1)

FIG. 1. Purification of 5-lipoxygenase by immunoaffinity chromatography. Chromatography with 10 ml of the cytosol fraction and the assays of 5-lipoxygenase (open circles and dotted line) and 12-lipoxygenase (open triangles and broken line) were performed as described under "ExperimentalProcedures." Protein concentration of each fraction was determined (dots and solid line). Elution buffer a t pH 10.0 containing 0.2% sodium deoxycholate was applied as indicated by arrow.

Lane A

B

C

D

E

F

FIG. 2. Thin layer chromatograms of the reaction products examined after removal of 12-lipoxygenase by the use of anti- from arachidonic acid and 5-HPETE. The purified enzyme (5 pg of protein) was incubated with 2.5 p~ [l-'4C]arachidonic acid under 12-lipoxygenase antibody. Then, thecolumn was washed with the standard conditions (A-C). The enzyme (19 pg of protein) was sodium carbonate buffer at pH 10.0 containing 0.2% sodium also incubated with 2.5 p~ [1-14C]5-HPETE (D-F). Ethereal extracdeoxycholate. The 5-lipoxygenase activity was detected in the tion and TLC using solvent B were carried out as described under eluate, but the 12-lipoxygenase activity was not detectable. "Experimental Procedures." The silica gel plate was exposed to anxElution was performed a t 28 "C rather than in thecold room ray film for autoradiography. Sodium borohydride (about 0.1 mg) was at 4 "C since proteins came out in sharperpeaks. The column added after termination of reaction ( B and E). The enzyme was kept a t 60 "C for 5 min and then incubated with substrate (C and F). could be used for at least 1year without an appreciable loss Bands 0,2,4, and 5 indicate migration of authentic arachidonic acid, of its binding capacity. 5-HETE, 6-truns-LTB4 (epimeric a t C-12), and 5,6-diHETE (epiThis enzyme preparation was allowed to react with [ l-"C] meric mixture), respectively, as visualized by iodine vapor or UV arachidonic acid, and theproducts were analyzed by silica gel absorption.

TLC aspresented in Fig. 2 (lane A ) . In addition to themajor product in band 1, several minor productswere found in bands 2-5. When sodium borohydride was added at the end of reaction (lane B ) , the major product (band 1 in lane A ) was converted to a more polar compound(band 2 in lane B ) ,which cochromatographed with authentic 5-HETE. Identity of the reduced product with 5-HETE was confirmed by gas chromatography-mass spectrometry as described below. Thus, the immediate reaction product (band 1 in lane A ) was identified as 5-HPETE. The enzyme reaction as followed by the formation of 5HPETE did not proceed in a linear fashion. As soon as the reaction started, it slowed down gradually and ceased after about 5 min even though theenzyme was still saturated with the substrate. Thus, a precise determination of the initial velocity was difficult. In routine assays the enzyme activity was determined by a 3-min reaction rather than a prolonged incubation, and expressed in termsof pmol of product/3 min/ mg of protein. The specific activity of the purified enzyme determined by 3-min reaction was about 0.6 pmo1/3 min/mg of protein and that determined by 1-min reaction was about 0.3 pmol/min/mg of protein. The characteristic time course of suicide-type was also described for other preparationsof 5lipoxygenase (4-7). The final specific activity of the previously purified enzymes was not described in view of such apeculiar time course. Ref. 5 reported a value of 350 & 100 ng of 5HETE (5-HPETE)/20min/pg of protein a t 37 "C,which was equivalent to about1.1pmo1/20 min/mg of protein. When the maximal reaction rate at the initial phase is read from the time course presented in Fig. 6 of Ref. 5, the enzyme activity per minute is about 0.3 pmol/min/mg of protein. Considering the difference in the incubation temperature (37 and 24 "C), this reported value of the purified enzyme of rat basophilic

leukemia cells may be lower than thefinal specific activity of our enzyme preparation. The other paper (3) also reporting the purification of the enzyme from rat basophilic leukemia cells did not provide us with sufficientparameters which were necessary for the calculation of the specific enzyme activity in termsof pmol/min/mg of protein. As examined by SDS-polyacrylamide gel electrophoresis (Fig. 3), the enzyme after the first immunochromatography showed a protein band with molecular weight of about 72,000 and several other bands (lane B ) . Rechromatography of the enzyme under the same conditions brought about a preparation of near homogeneity exhibiting a major protein band with a molecular weight of 72,000 and several band in trace amounts (lane A ) . The presence of a high activity of 12lipoxygenase did not allow the determination of the 5-lipoxygenase activity in thecytosol fraction. Therefore,the cytosol fraction was treated with anti-12-lipoxygenase antibody to remove 12-lipoxygenase. By the first immunoaffinity chromatography of the 12-lipoxygenase-free cytosol fraction, the enzyme activity was recovered in 50% yield and the total protein in0.49% yield. This stepincreased the specific enzyme activity from 6.4 nmol/3 min/mg of protein (the 12-lipoxygenase-free cytosol fraction) to 0.66 pmo1/3 min/mg of protein. At the second chromatography the yield of the enzyme activity was 27% and thatof protein was 28%, and thespecific enzyme activity remained unchanged. Thus, the immunoaffinity chromatography was an efficient method to purify the enzyme protein, but rechromatography did not bring about further increase of the specific enzyme activity due to the instability of the purified enzyme as described below. The enzyme purified by the firstimmunochromatography was used in most of the present work, and the importantfindings were

5-Lipoxygenaseand Hydroperoxyeicosatetraenoic Acids

7985

products in small amounts (bunds 2-5). Borohydride reduction did not change the positions of bunds 3-5 (lane B ) . Subsequently, the reaction products were reduced and analyzed by reverse-phase HPLC (Fig. 4). As monitored by absorption a t 270 nm and then at 240 nm, a doublet peak (Iu -205K and b) appeared with a retention time of 11.9 min and 12.8 min, respectively, followed by two small peaks (IIIu and b, 24.1 and 26.2 min). Apeak detected at 240 nm (IV)was eluted at a retention time of 47.6 min. The radioactive materials in these peaks were subjected to TLC. Materials from peaks Ia and Ib appeared in the position corresponding to band 4 of the experiment of Fig. 2, and those from peaks IIIa and b migrated to the height of band 5. Peak IV corresponded to bund 2. On the other hand, bands 4, 5, and 2 on the TLC plate (Fig. 2, lane A ) were scraped, and the ethylacetate extract from each band was analyzed by HPLC. The extract from bands 4,5, and 2 gave the sameelution profile as described above for peaks Ia andb, IIIa andb, and IV. Compounds from HPLC peaks Iaand bexhibited UV spectra withan absorption maximum a t 269 nm andshoulders 45K at 258 and 280 nm. These spectra were identical with those of authentic 6-truns-LTB4 and its 12-epimer prepared from LT&. HPLC of authentic 6-truns-LTB4 and its 12-epimer gave peaks at the position of Ia and b.Ref. 19 described separation of 6-truns-LTB4andits 12-epimer by reverseA CD phase HPLC on which the former compound was eluted FIG. 3. SDS-polyacrylamide gel electrophoresisof 5-lipoxygenase. The 5-lipoxygenase preparations ( A ) after the second im- earlier than theother. The mass spectra of compounds Ia and munochromatography (15 pg) and ( B ) after the first immunochro- b (trimethylsilyl ether derivative of the methyl ester) showed matography (45 pg) were concentrated to 50 pl by lyophilization. The ion peaks at m/e 494 (M), 479 (M - 15), 404 (M - go), 383 immunochromatography (150 (M - 111, .CHz-CH=CH-(CH,)4-CH3),293 pass-through fraction(C) after the first (M - (111 pg) and (D)a molecular weight standard mixture containing myosin go)), 267, 229, 217, 203, 191, 169, and 129 (base peak), which from rabbit muscle (molecular weight, 205,000), @-galactosidasefrom were described for 6-truns-LTB4 and 12-epi-6-truns-LTB4 Escherichia coli (116,000), phosphorylase b from rabbit muscle (97,400),bovine albumin (66,000), and egg albumin (45,000). Electro- (19). Amixture of compounds IIIa and IIIb exhibiteda UV phoresis was carried out as described under “Experimental Procedures,” and protein bands were stained with Coomassie Brilliant Blue spectrum with a maximum at 272 nm and shoulders at 264 R-250. and 283 nm. The spectrum was identical withthat of authentic 5,6-diHETE (an epimeric mixture), which gave peaks IIIa confirmed with the preparation obtained by the second im- and b upon HPLC. Gas chromatography-mass spectrometry munochromatography. of a mixture of compounds IIIa andb (trimethylsilyl ether of The final preparation of the enzyme showed a band corre- methyl ester derivative) gave ion peaks a t m/e 479 (M - 15), sponding to a molecular weight of 72,000 f 3,000 (n = 7) 463 (M - 31), 404 (M - go), 291 (loss of -CH(OSiMe3)upon SDS-polyacrylamide gel electrophoresiswith marker (CH2),-COOCH3), 203 (base peak, Me3SiO+=CH-(CH2)3proteins. This value was close to themolecular weight of the COOCH3),and 171. An identical mass spectrum was recorded purified 5-lipoxygenase of rat basophilic leukemia cells re- with authentic compounds. These UV and mass spectra were ported by oneresearch group (73,000) (5)but somewhat previously described for 5,6-diHETE (19). The stereochemical different from the value described by the othergroup (90,000 structures of compounds IIIa andb were not determined. as a monomeric enzyme) (3). The protein band was transCompound IV showed an absorption maximum at 235 nm. ferred from the polyacrylamide gel to a nitrocellulose membrane, and theblot was stained by the avidin-biotin complex method described under “Experimental Procedures.” As compared with the control run without anti-5-lipoxygenase antibody, a colored band was observed selectively at theposition corresponding to thepurified enzyme with amolecular weight of 72,000 in the presence of any of the threemonoclonal anti5-lipoxygenase antibodies ( 5 h x - 6 , 5 h x - l ,and 5 h x - 9 ) . ” The enzyme purified by the first immunoaffinity chroma0 IO 20 30 4 0 50 tography was most active around pH 7.5-8.0. When stored a t Retention Time (mid 24 “C, the enzyme was inactivated rapidly. Half of the enzyme FIG. 4. High-performance liquid chromatography of rethe activity was lost for about 1 h, and the enzyme was almost action products from arachidonic acid. The purified enzyme (0.20 inactive after 3 h. The enzyme inactivation was slower at mgof protein) was incubated with 25 p~ [l-14C]arachidonicacid 2 “C, and about50% of the enzyme activity was retained after (500,000 cpm) for 5 min in the 2-ml standard assay mixture. After 24 h. No appreciable inactivation was observed by storage of the reaction sodium borohydride was added to the mixture. The ethereal extract was purified by silicic acid chromatography and the enzyme a t -70 “C for 1week. analyzed by reverse-phase HPLC as described under “Experimental Identification of Reaction Products-As shown in lane A of Procedures.” Elution of authentic compounds was indicated by numFig. 2, the production of 5-HPETE (bund I) from [l-“C] bers; 6-truns-LTBd (Zu),12-epi-6-truns-LTB4 (Zb), 5S,12S-diHETE arachidonic acid was accompanied by several more polar (ZZ),5,6-diHETE (ZZZuand b) and 5-HETE(ZV).

- r8

+

1

7986

5-Lipoxygenme and HydroperoxyeicosatetraenoicAcids

The mass spectrum of the trimethylsilyl ether derivative of the methyl ester had peaks a t m/e 406 (M), 391 (M - 15), 375 (M - 31), 316 (M - W), 305 (M - 101, loss of -CHz(CHZ)z-COOCH3),255 (M - 151, loss of .CH,-(CH=CHCHz)z-(CH2)3-CH3), 216, 215, and 203 (Me,-SiO+=CH(CHz)3-COOCH3). The UV and mass spectra were identical with those of authentic 5-HETE and consistent with those previously described (20). The compound corresponding to band 3 (Fig. 2, lane A ) remains unidentified. Migration of band 3 was not affected by borohydride reduction, ruling out a possibility that the compound in band 3 was a hydroperoxy product. A UV absorbing peak corresponding to (5S,12S)-diHETE was hardly detectable upon HPLC (position indicated by I1 in Fig. 4). This observation was in sharp contrast to theproduction of (5S,12S)-diHPETE in a larger amount by potato 5-lipoxygenase (21). The production of 6-trans-LTB1 and its12-epimer and 5,6diHETEs suggested an intermediate formation of LT&, and an attempt was made to trap LTA, as 12-0-methyl-5,12diHETE by the method of Borgeat and Samuelsson (22). After incubation of the enzyme with arachidonic acid, an excess amount of methanol was added to thereaction mixture. Upon reverse-phase HPLC a doublet peak cochromatographing with that of authentic12-0-methyl-5,12-diHETE(an epimeric mixture) was detected by monitoring absorption at 270 nm. The compounds as a mixture exhibited an absorption spectrum with a maximum at 269 nm and shoulders a t 260 and 279 nm. Essentially the same spectrum was described previously (22). The trimethylsilyl ether derivative of the methyl estergave a mass spectrum with ion peaks at m/e 436 (M), 421 (M - 15), 404 (M - 32), 325 (M - 111,loss of -CHz-CH = CH-(CH,),-CHz), 293 (M - (32 l l l ) ) , 235 (M - (111 + go)), 159, and 133. These ions were described previously for 12-0-methyl-5,12-diHETE (22). Reactivity with 5-HPETE-As described above, the reaction of the 5-lipoxygenasewith arachidonic acid brought about an accumulation of 5-HPETE and several degradation products of LT&. Therefore, the purified enzyme was incubated with [1-14C]5-HPETE, and the products were analyzed by TLC (Fig. 2, lanes D-F). Incubation with a heat-denatured enzyme (lane F) caused a nonenzymatic conversion of part of the added 5-HPETE (band I) to 5-HETE (band 2) and an unidentified compound (band 3). In contrast, 6-trans-LTBd (a 12-epimeric mixture) and 5,6-diHETE (an epimeric mixture) were observed in bands 4 and 5 only in an enzymatic reaction (lane D). The product profile was unaffected by the addition of sodium borohydride ( l a n e E).The results indicated that thepurified preparation of 5-lipoxygenase also catalyzed the conversion of 5-HPETE to LTA4 (LTA synthase reaction). Since the association of the 5-lipoxygenase and LTA synthase activitieswas reported previously with a 5-lipoxygenase of potato tuber (21), these two enzyme activities were examined with our purified enzyme from porcine leukocytes under various experimental conditions. The LTA synthase reaction proceeded in a timecourse of suicide-type reaction described above for the 5-lipoxygenase reaction. When the enzyme was kept at various temperatures for 5 min, the 5-lipoxygenase and LTA synthase activities were lost almost in parallel as the temperature was raised (Fig. 5). We earlier reported two types of compounds, a benzoquinone derivative (AA861) (23) and a flavone derivative (cirsiliol) (24) asinhibitors of 5lipoxygenase. These compounds a t p~ concentrations inhibited 5-lipoxygenase while inhibition of 12-lipoxygenase required much higher concentrations. Effect of these com-

+

Temperature (TI

FIG. 5. Heat inactivation of 5-lipoxygenase and LTA synthase. The purified enzyme was kept a t various temperatures for 5 min. An aliquot (3 and 19 pg of protein) was removed for the standard assay of 5-lipoxygenase with 2.5 PM [l-'4C]arachidonic acid (closed circles) and LTA synthase with 2.5 p~ [ 1-14C]5-HPETE(opencircles). The activity of untreated enzyme was expressed as 100%; 5-lipoxygenase, 0.32 nrnol/3 min and LTA synthase, 0.068 nmol,& min.

"0' 0.1 I 1-0 00.01 0.1 I IO AA861 (pM) Cirsiliol &MI

FIG. 6. Inhibition of LTA synthase activity by selective inhibitors of 5-lipoxygenase. The purified enzyme (8 and 19 pg of protein) was preincubated for 3 min in an ice bath with AA861 (left) or cirsiliol (right). Both compounds were dissolved in 4 pl of methanol. In the control run the enzyme was preincubated with methanol (4 pl). Then, the enzyme was assayed either for the 5-lipoxygenase activity with 2.5 pM [l-14C]arachidonicacid (closed circles)or for the LTA synthase activity with 2.5 pM [1-14C]5-HPETE (open circles). The enzyme activity in the control run was expressed as 100%; 5lipoxygenase,0.27 nmol/3 min, and LTA synthase, 0.080 nmol/3 min.

pounds was tested both by the 5-lipoxygenase assay and by the LTA synthase assay. As shown in Fig. 6, both AA861 and cirsiliol inhibited the LTA synthase reaction as well as the5lipoxygenase reaction. Several papers have reported a requirement of calcium ion for 5-lipoxygenase from various animal species (2-7). Calcium ion in the order of millimolar was necessary for 5-lipoxygenase from porcine leukocytes. The LTA synthase activity of the enzyme also required calcium ion. Barium and manganese ions showed about 60% of the effect of calcium ion on both the 5-lipoxygenase and LTA synthase reactions. We earlier reported that ATP and other nucleotides stimulated the calcium-dependent 5-lipoxygenase activity (4, 6). As illustrated in Fig. 7, the LTA synthase activity was also stimulated by the addition of ATP. For technical reasons the assays of both enzyme activities in the experiments described above (Figs. 5-7) were performed a t a subsaturation level of substrate (2.5 PM). Effect of the concentration of arachidonic acid and 5-HPETE on the 5lipoxygenase and LTA synthase activities, respectively, was examined (Fig. 8). Reaction of 25 p~ arachidonic acid or 5HPETE with a heat-denatured enzyme (60 "C, 5 min) proceeded to a negligible extent. As compared at 40 p ~ the , 5lipoxygenase activity was about 25-fold higher than the LTA synthase activity. Such a big difference of the relative activities of the twoenzyme reactions was consistent with an appreciable accumulation of 5-HPETE when the enzyme was incubated with arachidonic acid (Fig. 2, lane A). It should be noted that the rateof LTA, production was higher when the enzyme was incubated with arachidonic acid (crosses and

5-Lipoxygenaseand Hydroperoxyeicosatetraenoic Acids

7987

Fig. 9 indicated that thepurified enzyme was also active with both12-HPETEand15-HPETE. When the enzyme was incubated with [ 1-14C]12-HPETE, therewas a major product (band I in lane A ) , which was indistinguishable from the c .compound produced from 5-HPETE by 12-lipoxygenase of .-5 porcine leukocytes (lune B ) . Thelatter compound was (5S,12S)-diHPETE as preliminarily reported (8). The compound which was produced from 12-HPETE by 5-lipoxygenase and then reduced with sodium borohydride, cochromatographed with the enzymatically prepared (5S,12S)-diHETE on reverse-phase HPLC. The borohydride-reduced product 0 ' showed an absorption maximum a t 269 nm with shoulders at 0 " 0.01 0.1 I 10 260 and 279 nm. The spectrum was previously described (14, ATP (mM) 26). These observationstogetherwith the following mass FIG.7. Stimulatory effect of ATP on the 5-lipoxygenase spectrum of the borohydride-reduced product (trimethylsilyl and LTA synthase activities. The purified enzyme (5 and 10 pg of etherand methyl ester) supported the identity of the 5protein) was assayed for the 5-lipoxygenase activity with 2.5 p~ [ l "Clarachidonic acid (closed circles) and the LTA synthase activity lipoxygenase product from 12-HPETE with (5S,12S)-diwith 2.5 p~ [1-I4C]5-HPETE (open circles)in the presence of various HPETE: at m/e 479 (M - 15), 463 (M - 31), 404 (M - go), (M concentrations of ATP. The enzyme activity determined inthe pres- 383 (M - 111, loss of .CH,-CH=CH-(CH,)4-CH,), 293 ence of 10 mM ATP was expressed as 100%; 5-lipoxygenase, 0.30 - (111 go)), 267,229,217,203 (Me,SiO+=CH-(CH,),nmol/3 min, and LTA synthase, 0.052 nmol/3 min. COOCH,), 171, and 129 (base peak), which were described for (5S,12S)-diHETE (14,26). Inaddition to (5S,12S)-diHPETE thus identified, minor products with a little higher polarity were also observed in lanes A and B, which were presumably derived from (5S,12S)-diHPETEby nonenzymatic reduction. There was no conversion of 12-HPETE with heat-denatured 5-lipoxygenase (60 "C, 5 min). The enzyme was also active with 15-HPETE as shown in lune C in Fig. 9. Heat-denatured enzyme (60 "C, 5 min) was totally inactive. The reaction product (band II in lane C) comigrated with (5S,15S)-diHPETE prepared from 5HPETE by soybean lipoxygenase (band II in lane D).ReversephaseHPLC showed the same retention time for the 5lipoxygenase product from 15-HPETE and the15-lipoxygen0 IO 20 30 40 5 0 ase product from 5-HPETE, bothof which were reduced with Substrate (JJM) sodium borohydride to more polar compounds with the same FIG.8. Effect of substrate concentration of the enzyme re- polarity. Identification of both of the reduced compounds as actions with various HPETEs. [l-'4C]Arachidonic acid (closed (5S,15S)-diHETE was confirmed by their spectra with abcircles), 5-HPETE (open circles),12-HPETE (closed triangles),or 15100 -

2,

I

+

HPETE (open squares) a t various concentrations was incubated for 3 min with 2.5-5,24,17, and 16 pg ofthe purified enzyme, respectively, under the standard conditions for 5-lipoxygenase assay. Solvent B was used for TLC. Production of LT& (actually detected with its degradation products) starting from arachidonic acid was also plotted (crosses and broken line).

broken line in Fig. 8) rather than 5-HPETE, indicatinga better coupling of the two enzyme reactions. Several lines of enzymological evidence support theconcept that both the 5-lipoxygenase activity and the LTA synthase activity of the enzyme purified from porcine leukocytes can be attributed to the same single enzyme as proposed and discussed for the potatoenzyme (21). Since the LTA synthase activity was much lower than the 5-lipoxygenase activity as described in Fig. 8, the production of 5-HPETE by the 5lipoxygenase reaction was not efficiently coupled to the LTA synthase reaction.Aquestion may be raisedwhether the enzyme was modified during the purification procedures resulting in an uncoupling of the two enzyme activities or a certain coupling factor was removed from the enzyme preparation. However, the overproduction of 5-HPETE (mostly detected as 5-HETE)from arachidonic acid was also described with the whole cells (25) and thecytosol (7) of human leukocytes. Reactivity with 12-HPETE and 15-HPETE-The finding that 5-lipoxygenase of porcine leukocytes was also active with 5-HPETE as substrate,prompted us to examine the activity with other HPETEs. The TLC autoradiogram presented in

I2-HPETE"" 5-HPETY'

Lane

A

B

C

D

FIG.9. Thin layer chromatograms of the 5-lipoxygenase products from 12-HPETE and 15-HPETE. The purified enzyme (12 pg of protein) was incubated with 10 PM [l-'*C]12-HPETE (lane A ) , and 16 pgof enzyme with 10 ~ L M[1-'4C]15-HPETE (lune C) in thestandard 5-lipoxygenase assay system. 12-Lipoxygenase from porcine leukocytes (40 pg of protein) was incubated with 2.5 p~ 11"C15-HPETE (lane B ) , and soybean lipoxygenase (5 pg of protein) with 5 p~ [1-14C]5-HPETE (lane D).Reaction was carried out for 3 min in a 0.2- and 0.1-ml reaction mixture described respectively for preparation of 12-HPETE and 15-HPETE under "Experimental Procedures." TLC was carried out in solvent B, and the silica gel plate was exposed to anx-ray film for autoradiography. In lane D authentic [1-14C]5-HPETEwas placed on the silica gel plate together with the reaction product. Migration of authentic 5-, 12- and 15-HPETEwas indicated.

7988

5-Lipoxygenaseand Hydroperoxyeicosatetraenoic Acids

sorption maxima at 244 nm (18) and mass spectra (trimethylsilyl ether and methyl ester) (18)as follows: at m/e 494 (M), 479 (M - 15), 463 (M - 31), 404 (M - go), 394, 393 (M 101, IOSS of -(CH2)3COOCH3), 333(M - (90 + 71)), 314 (M - 2 x go), 225 (+(CH=CH-)2CH(OSiMe3)-(CH2)4-CH3), 203 (Me3SiO'=CH-(CH2)3-COOCH3) and 173 (base peak, Me3SiO'=CH-(CH&-CH3). In addition to (5S,15S)-diHPETE asa major product, therewere minor products a little below band 11, which were presumably produced by nonenzymatic reduction of hydroperoxy groups. Moreover, trace amounts of unidentified products were observed midway between band I1 and theorigin. Relative activities of the 5-lipoxygenation with arachidonic acid, 12-HPETE, and 15-HPETE are shown in Fig. 8. Both 12-HPETE and 15-HPETEwere oxygenated by the purified 5-lipoxygenase from porcine leukocytes with 22 and 30% maximal velocity of the arachidonate 5-lipoxygenation, respectively. Although the enzyme protein was highly purified by immunoaffinity chromatography, instability of the purified enzyme hindered a reasonable increase in the specific enzyme activity. The immunoaffinity chromatography may have dissociated a certain factorwhich is required to keep the enzyme activity, for example, phosphatidylcholine as reported by Goetze et al. ( 5 ) . Results of our investigations with such a purified enzyme demonstrated reactivity of 5-lipoxygenase with various HPETEs. The reaction with 5-HPETE showed the LTA synthase activity, and association of the LTA synthase activity with the 5-lipoxygenase activity was supported by several experiments. It should be carefully considered whether or not theLTA synthase activityassociated with the purified 5-lipoxygenase can account for most of the activity in the cytosol fraction. Although a precise determination of the LTA synthase activity in thecytosol fraction was difficult due to the presence of a certain peroxidase activity which decomposed 5-HPETE added as such or generated from arachidonic acid and theincubation of arachidonic acid with the 12-lipoxygenase-freecytosol fraction accumulated 5-HPETE (part of which was reduced to 5-HETE), LTAl as judged by its degradation products was produced only in asmall amount as in thereaction withthe purified 5-lipoxygenase. The passthrough fraction at the immunochromatography step had a hardly detectable activity of LTA synthase after removal of 12-lipoxygenaseby the use of its antibody. These observations may rule out the presence of a separate LTA synthase in the cytosol fraction. The enzyme activity of 12-lipoxygenase to synthesize 14,15-epoxycompound of L T G type was predicted with porcine leukocytes (27). Involvement of 12-lipoxygenase in the LTA synthase reaction may be excluded because the 12-lipoxygenase activity was not detectable in the purified preparation of 5-lipoxygenase and the 12-lipoxygenase purified from porcine leukocytes did not produce LTA, (5,6epoxide) from 5-HPETE? A possible contribution of a certain contaminating hemoprotein to theLTA synthase activity (28) cannot be rigorously ruled out. However, a heat inactivation ofthe LTA synthase activityand a characteristicrequirement of calcium ion and ATP distinguished the activity of our enzyme from such a casual activity of contaminating hemoprotein. The enzyme was also active with 12-HPETE and15HPETE. The versatile reactivity with various HPETEs is also the case of other lipoxygenases, i.e. rabbit reticulocyte 15-lipoxygenase (29), porcine leukocyte 12-lipoxygenase (8, 27), and soybean lipoxygenase (30). Experiments which allow a precise chemical insight intothe reaction mechanism of the C. Yokoyama et al., unpublished observation.

enzymes, for example, the stereoselective hydrogen abstraction from the substrates,are required to understand the apparently versatile function of these lipoxygenases by a simplified principle. Acknoulledgements-The authors are grateful to Drs. S. Terao and K. Takeda for their expertassistance to record mass spectra of various 5-lipoxygenaseproducts. REFERENCES 1. SamueIsson, B. (1983) Science 220, 568-575 2. Jakschik, B. A., and Lee, L. H. (1980) Nature 287, 51-52 3. Parker, C.W., and Aykent, S. (1982) Biochem.Biophys.Res. Commun. 109,1011-1016 4. Furukawa, M., Yoshimoto, T., Ochi, K.,and Yamamoto, S. (1984) Biochim. Biophys.Acta 795,458-465 5. Goetze, A. M., Fayer, L., Bouska, J., Bornemeier, D., and Carter, G. W. (1985) Prostaglandins 2 9 , 689-701 6. Ochi,K., Yoshimoto, T., Yamamoto, S., Taniguchi, K., and Miyamoto, T. (1983) J. Biol. Chem. 258,5754-5758 7. Soberman, R. J., Harper, T. W., Betteridge, D., Lewis, R. A., and Austen, K. F. (1985) J. Bwl. Chem. 2 6 0 , 4508-4515

8. Yokoyama, C., Shinjo, F., Yoshimoto, T., Yamamoto, S., Izumi, S., Komatsu, N., and Watanabe, K. (1985) in Advances in Prostaglandin, Thromboxane,and Leukotriene Research (Hayaishi, 0. and Yamamoto, s., eds) Vol. 15, pp. 193-195, Raven Press, New York 9. Kessler, S. W. (1976) J. Immunol. 1 1 7 , 1482-1490 10. Yoshimoto, T., Kaneko, S., Magata, K., Shinjo, F., Mizuno, K., Ueda, N., Ehara, H., Yokoyama, C., and Yamamoto, S. (1985) in Advances in Prostaglandin, Thromboxane and Leukotriene Research (Hayaishi, 0.and Yamamoto, S., eds) Vol. 15, pp. 173-175, Raven Press, New York 11. Goding, J. W. (1980) J. Immunol. Methods 39,285-308 12. Kennett, R. H. (1980) in Monoclonal Antibodies (Kennett, R. H., McKearn, T. J., and Bechtol, K. B., eds), pp. 372-373, Plenum Publishing Corp., New York 13. Oi,V. T., and Herzenberg, L. A. (1980) in Selected Methods in Cellular Immunology (Mishell, B. B., and Shiigi, S. M., eds), pp. 351-372, Freeman Publications, San Francisco 14. Yoshimoto, T., Miyamoto, Y., Ochi, K., and Yamamoto, S. (1982) Bwchim. Biophys. Acta 713,638-646 15. Lowry, 0.H., Rosebrough, N. J., Farr, A. L., and Randall, R. J. (1951) J. Biol. Chem. 193,265-275 16. Laemmli, U. K. (1970) Nature 227,680-685 17. Towbin. H.. Staehelin, T.. and Gordon. J. (1979) Proc. Natl. Acad. Sci. U. S.' A. 76,4350-4354 18. Mass, R. L., Turk, J., Oates, J. A., and Brash, A.R. (1982) J. Biol. Chem. 257,7056-7067 19. Borgeat, P., and Samuelsson, B. (1979) J. Bwl. Chem. 254,78657869 20. Borgeat, P., Hamberg, M., and Samuelsson, B. (1976) J. Biol. Chem. 251,7816-7820 21. Shimizu. T.. Ridmark, O., and Samuelsson, B. (1984) Proc. Natl. Acad. Sci.. U. S. A. 81,689-693 22. Boreeat. P.. and Samuelsson., B. .(1979). Proc. Natl. Acad. Sci. U. S.-A. 76,'3213-3217 23. Yoshimoto, T., Yokoyama, C., Ochi, K., Yamamoto, S., Maki, Y., Ashida, Y., Terao, S., and Shiraishi, M. (1982) Biochim. Biophys. Acta 713,470-473 24. Yoshimoto, T., Furukawa, M., Yamamoto, S., Horie, T., Watanabe-Kohno, S. (1983) Biochem. Biophys. Res. Commun. 116, 612-618 25. Sun, F.F., and McGuire, J. C. (1984) Biochim. Biophys. Acta 794,56-64 26. Borgeat, P., Picard, S., Vallerand, P., and Sirois, P. (1981) Prostaglandins Leukotrienes Med.6,557-570 27. Maas, R. L., and Brash, A. R. (1983) Proc. NatE. Acad. Sci. U. S. A. 80,2884-2888 28. RHdmark, O., Shimizu, T., Fitzpatrick, F., and Samuelsson, B. (1984) Biochim. Bwphys. Acta 792,324-329 29. Bryant, R.W., Schewe, T., Rapoport, S. M., and Bailey, J. M. (1985) J.Biol. Chem. 260,3548-3555 30. van Os, C. P. A., Pijke-Schilder, G. P. M., van Halbeek, H., Verhagen, J., and Vliegenthart, J. F. G. (1981) Biochim. Biophys. Acta 663, 177-193