Proc. Natl. Acad. Sci. USA Vol. 74, No. 1, pp. 144-148, January 1977
Biochemistry
Inhibition of prostaglandin endoperoxide synthetase by thiol analogues of prostaglandin (prostaglandin synthase/oxygenase)
SHIRO OHKI*t, NOBUCHIKA OGINOt, SHOZO YAMAMOTOt, OSAMU HAYAISHIt, HISASHI YAMAMOTOt, HAZIMU MIYAKE§, AND MASAKI HAYASHI§ t Department of Medical Chemistry, Kyoto University Faculty of Medicine, Kyoto, Japan; * Department of Industrial Chemistry, Kyoto University Faculty of Engineering, Kyoto, Japan; and § The Ono Central Research Institute, Shimamoto-cho, Mishima-gun, Osaka, Japan
Contributed by Osamu Hayaishi, October 27, 1976
ABSTRACT A variety of thiol compounds inhibited the enzymatic bis-oxygenation of 8,11,14-eicosatrienoic acid to prostaglandin GI, as examined with a purified preparation of prostaglandin endoperoxide synthetase (prostaglandin synthase; 8,11,14-eicosatrienoate, hydrogen-donor:oxygen oxidoreductase; EC 1.14.99.1) from bovine vesicular gland. The hydroperoxide cleavage of prostaglandin G1 producing prostaglandin H1 was not affected by these thiol compounds. Several prostaglandin analogues with a thiol group (9,11-dihydroxy-15S- or 15R-mercaptoprosta-5,13-dienoic acid, 1-mercapto-9,11,15-trihydroxyprosta-5,13-diene, and 1-mercapto-9-oxo-11,15-dihydroxyprosta-5,13-diene) were most potent inhibitors, showing almost complete inhibition at concentrations on the order of 1 t&M. Other thiol compounds, such as 2,3-dimercaptopropanol, dithiothreitol, and dihydrolipoic acid, were also inhibitory but were much less effective. The inhibition, as examined with 9,11-dihydroxy-15S-mercaptoprosta-5,13-dienoic acid and 2,3dimercaptopropanol, was noncompetitive.
Beef blood hemoglobin (type I), hematin, cysteine, 2-mercaptoethanol, dithiothreitol, and dihydrolipoic acid were purchased from Sigma; tryptophan, ascorbic acid, and 2,3dimercaptopropanol from Nakarai Chemicals (Kyoto); 5,5'dithio-bis-(2-nitrobenzoic acid) from Tokyo Kasei Kogyo (Tokyo); and precoated silica gel plate 60 Fws4 for thin-layer chromatography from E. Merck (Darmstadt). Glutathione was donated by the Yamanouchi Central Research Institute. Synthesis of (5Z,9a,lla,13E,15)-l-Mercapto-9,11,15trihydroxyprosta-5,13-diene (V in Fig. 1). Reduction of I by diisobutylaluminum hydride produced the 1,9-diol II, which was subjected to selective tosylation using 1.2 equivalents of p-toluenesulfonyl chloride in pyridine at 00 for 20 hr (68% yield). The tosylate III was converted to the thioacetate IV with excess sodium thioacetate in dimethyl sulfoxide-dimethylformamide at 250 for 2 hr (83% yield). Hydrolysis of IV by K2CO3 in methanol, followed by removal of tetrahydropyranyl group in acetic acid-tetrahydrofuran-water, gave the 1-mercapto derivative V (in 73% yield from IV): nuclear magnetic resonance spectrum in CDC13 due to 2H of CH2S at S 2.55 ppm. Synthesis of (5Zlla,13E,15S)l-Mercapto-9-oxo-11,15dihydroxyprosta-5,13-diene (VIII). The alcohol III was oxidized by Collins reagent in methylene chloride at 00 for 15 min. The resulting ketone VI was transformed to the keto diol VII in acetic acid-tetrahydrofuran-water (71% yield from III). Treatment of VII with a large excess of sodium hydrosulfide in dimethylformamide at -300 for 1 hr, followed by chromatography on silica gel, produced the keto mercaptan VIII as a crystalline solid: melting point 45-;0; infrared absorption due to C=O at 1735 cm-I; nuclear magnetic resonance spectrum in CDC13 due to 2H of CH2S at 5 2.5 ppm. Synthesis of (5Z,9a,lla,13E,15S)9,1 1-Dihydroxy-15mercaptoprosta-5,13-dienoic acid (XV). The optically active lactone IX prepared by the method of Corey et al. (14), was converted to the 15R-bromide X with 1.1 equivalents of phosphorous tribromide in ether at 00 and then to 15S-thioacetate XI by treatment with sodium thioacetate in dimethyl sulfoxide-dimethylformamide at 250 for 1 hr. The latter compound XI was hydrolyzed to the thiol XII by K2CO3 in methanol at 250 for 1 hr (64% yield from IX). Reaction of XII with ethyl vinyl ether in methylene chloride at 25° for 20 min yielded the sulfide derivative XIII, which was transformed to the corresponding lactol by exposure to diisobutylaluminum hydride in toluene in 97% yield. Treatment of the lactol with the Wittig reagent prepared from 5-triphenylphosphonovaleric acid in dimethyl sulfoxide produced the acid XIV in 88% yield. This compound was hydrolyzed in aqueous acetic acid at 400 for 2 hr to the 15S-mercapto derivative XV (50% yield): nuclear magnetic resonance spectrum in CDC13 due to 1H of C(15)H
The hypothetical mechanism of prostaglandin biosynthesis proposed earlier (1, 2) has recently been substantiated by the isolation of two prostaglandin endoperoxides (3-5). The intermediary nature of these endoperoxides was further supported by the solubilization and purification of two enzyme fractions in our laboratory (6-8). One catalyzes the bis-oxygenation of an unsaturated fatty acid producing prostaglandin G, which has both an endoperoxy and a hydroperoxy group. Heme is required for this reaction. The same enzyme preparation also catalyzes the cleavage of the hydroperoxy group of prostaglandin G, producing prostaglandin H with an endoperoxy and a hydroxy group. This reaction also requires heme and is markedly stimulated by tryptophan (7). The other enzyme isomerizes the endoperoxy group of prostaglandin H to produce prostaglandin E (6, 8). In view of the role of these endoperoxides in prostaglandin biosynthesis, attempts have been made to synthesize analogues of prostaglandin endoperoxides that would have any biological activity and effect on enzyme (9-12). In the course of studies on various analogues of prostaglandin and prostaglandin endoperoxides, we found that some analogues with thiol group(s) as well as other thiol compounds were potent inhibitors of the bis-oxygenation of an unsaturated fatty acid to prostaglandin G. MATERIALS AND METHODS Chemicals. 8,11,14-[1-14C]Eicosatrienoic acid (57.5 mCi/ mmol) was purchased from New England Nuclear and purified by thin-layer chromatography. 8,11,14-Eicosatrienoic acid was prepared by the method of Klenk and Mohrhauer (13), and prostaglandin F2a according to the method of Corey et al. (14). *
On leave from the Ono Central Research Institute. 144
Proc. Natl. Acad. Sci. USA 74 (1977)
Biochemistry: Ohki et al. OH R
Ox
ox
(I) RzCOOMe,XaTHP (11) R:CH20H,X:THP (111) R=CH2OTs,X=THP (IV) R CH2SAc, X= THP (V) RzCH2SH,X=H
0 R OX
OX
(VI) R-CH20Ts,X=THP (VII) RuCH20Ts,XzH (ViII) RzCH2SH,X*H
. s.o (IX) X AcY DO
'-H (X) XzAc, Y/H
(XI) X=Ac, Y=S'SAc (X I) X - H. Y (H
X=,~SH
(XIII) X=CH(OEt)Me, *SCH(OEt)Me
(XVI) XzAc,Y3...H
(XVII) X.H y(,,SH
OH
(XIV) XaCH(OEt)Me,
K>2--2 COH
ox IXY
Y *SCH(OEt)Me /H XuHY (XV) XSH XxH,Ya
(XVIII)
FIG. 1. Structures of prostaglandin analogues and intermediates of their organic syntheses. Abbreviations: Me, methyl; Et, ethyl; THP, tetrahydropyranyl; Ts, p-toluenesulfonyl; Ac, acetyl. at a 3.2-3.6 ppm; homogeneous as examined by thin-layer chromatography. Synthesis of (5Z,9a,lla,13E,15R)9,1 1-Dihydroxy-1Smercaptoprosta-5,13-dienoic acid (XVIII). Although the bromination of the 15S-IX by phosphorous tribromide proceeded regio- and stereospecifically with inversion of configuration at C-15, similar treatment with the 15R-alcohol XVI gave rise to a mixture of 13- and 15-bromide. The mixture was directly converted to a mixture of 13- and 15-thioacetate in 11% and 65% yield, respectively. The 15-thioacetate was readily separated from the 13-isomer by chromatography on silica gel and hydrolyzed with K2CO3 in methanol (76% yield). The 15R-mercaptan XVII thus produced was separated from the contaminating diastereomer at C-15 (15R/15S = 3:2). The same sequence of reactions from XII to XV was carried out for the conversion of XVII to the 15R-mercapto acid XVIII. Thinlayer chromatography on silica gel in a solvent system of chloroform-tetrahydrofuran-acetic acid (10:2:1) gave RF values of 0.66 and 0.59 for XV and XVIII, respectively. The assignment of stereochemistry at C-15 is anticipated on mechanistic consequences as well as chromatographic mobili-
ties.
Synthesis of Other Compounds. Procedures for synthesis of methyl (5Z,9a,lla,13E,1SS)-9,11-dimercapto-15-hydroxyprosta-5,13-dienoate, methyl (5Z,9,3,110,13E,15S)-9,11dimercapto-15-hydroxyprosta-5,13-dienoate, methyl (5Z, 9a, 1 1a, 13E,15S)-9, 1 1-epidithio-15-hydroxyprosta-5, 13-di-
145
enoate, and methyl (SZ,9fl,11f,13E,15S)-9,11-epidithio-15hydroxyprosta-5,13-dienoate will be published elsewhere. Enzymes. Prostaglandin endoperoxide synthetase (prostaglandin synthase; 8,11,14-eicosatrienoate, hydrogen-donor: oxygen oxidoreductase; EC 1.14.99.1) was prepared by the method described previously (7). The enzyme solution was concentrated to a protein concentration of 1.5 mg/ml with the aid of a Diaflow membrane (PM-10). Iron-containing superoxide dismutase from Escherichia coli (15) and copper-containing superoxide dismutase from bovine blood (16) were kindly provided by Drs. J. Oka and T. Yoshimoto of this laboratory. Beef liver catalase was purchased from Sigma. Preparation of Prostaglandin G1. The method of preparation described (7) was used on a 10-fold larger scale; i.e., in a 10-ml mixture containing 8,11,14-[1-14C]eicosatrienoic acid (1.1Itmol, 1.25 X 107 cpm) and the purified enzyme (2.5 mg of protein). Enzyme Assays. Conversions from 8,11,14-eicosatrienoic acid to prostaglandin G1 and from prostaglandin G1 to prostaglandin H1 were assayed by the methods using radioactive substrates as described (7). For the determination of oxygen consumption, a galvanic oxygen electrode (type SBOA, the Kyusui Kagaku Institute, Tokyo) was used. The sensitivity of oxygen consumption measurement was increased with the aid of an offset amplifier according to the method of Miyake and Fukuyama (17). The reaction cell was thermostated by circulating water at 24°. The reaction mixture contained Tris-HCl buffer at pH 8.0 (175 ,qmol), 8,11,14-eicosatrienoic acid (250 nmol), hemoglobin or hematin (2 nmol), tryptophan (8.6 Amol), and enzyme in a final volume of 1.75 ml. Reaction was usually started by the addition of enzyme that had been equilibrated at 240. Initial velocity was determined within a time range in which the reaction proceeded in a linear fashion. Miscellaneous Determinations. Protein concentration was determined by the method of Lowry et al. (18), with bovine serum albumin as a standard. Thiol content was determined by titration using 5,5'-dithiobis(2-nitrobenzoic acid) developed by Ellman (19). The concentration of heme was determined spectrophotometrically as reduced pyridine hemochromogen, using an extinction coefficient of 34.4 mM-' cm-' at 557 nm (20). Spectra of heme were recorded with a Union Giken spectrophotometer model SM-401.
RESULTS As shown in Table 1, effects of various thiol compounds were screened by measurement of oxygen consumption upon incubation of the enzyme with 8,11,14-eicosatrienoic acid in the presence of hemoglobin and tryptophan. Prostaglandin H1 rather than G1 was the major product under these conditions (7). Several prostaglandin analogues with thiol group(s) markedly inhibited the oxygen consumption. 9,11-Dihydroxy-15S-mercaptoprosta-5,13-dienoic acid (compound c) and 9,11-dihydroxy-15R-mercaptoprosta-5,13-dienoic acid (compound d) were the most potent inhibitors, giving 50% inhibition at concentrations on the order of 0.1 gM. 1-Mercapto9,11,15-trihydroxyprosta-5,13-diene (compound e) and 1mercapto-9-oxo-11,15-dihydroxyprosta-5,13-diene (compound f) were also effective inhibitors, but with higher concentrations to give 50% inhibition (approximately 1,gM). Analogues with two thiol groups at C-9 and C-11 (compounds a and b) were much less effective irrespective of the configuration of the thiol groups. Their concentrations for 50% inhibition were on the order of 10,uM. It should be noted that compounds with a disulfide bridge between C-9 and C-11 were ineffective either
146
|Oe)
Biochemistry: Ohki et al. Compound
Table 1. Inhibition of oxygen consumption by various thiol compounds
(PM)
i|
C;
MN
()(M
20 |LC
|50 10 20 50
18 31 56
40 1
2
m,
46 g) 40
-
50 h) 40 0.)1~~o4 80 82 90 100 I~~~~~
0 0 0
-
05
0.2
38
2
8
i)
|i 1
2 6
35 72 79
(CI)COH
SiH- SH
0 0 0
0.1
d)
C.8?
75 77
;,80 100
0.OOH 79 ,NW~~~~~~t7 0.4 0779 1
niito trtio khtio(tionM)50 Compound tration | (0/ ) p) 1)(M) (PM) (PM)pM tO 46 0.4 27 f0) k)
Compound
33 50 76
10 20
D0Concen-InitonD5
Concen-
Concen- niito D tration Inhibition ID5
a)
Q" e\OC
Proc. Natl. Acad. Sci. USA 74 (1977)
CH29HCH20H SH SH
HM)
OHO SH
0
30
I)
jooH
n)
30
66
50
74
5 10 20
46 76
5 20 50
25 62 76
20
0
NH2CHCOOH 0)
1)0
15
20 50
45
23 |
64
1
H2
20
40 60
H
6
86
1cOH OOI2CH2CHNHi2 40 0 60 MCHCONHCH2COOH0 0 60 K:~~7$^Voo/ 110 0 60 0
A OH 01| 14 I|H2H6HCH2 H ~
12
8 10 10
10
-
-
Rate of oxygen consumption was determined with an oxygen electrode in a reaction mixture described in Materials and Methods. Each thiol compound was present as indicated. Glutathione and cysteine were dissolved in water, and other compounds in acetone. The reaction was started by the addition of 40 pg of purified enzyme. The enzyme activity in the presence of inhibitor is expressed as % of the uninhibited control (0.17 ,mol of oxygen consumed per min). Nonenzymatic oxygen consumption in the presence of each thiol compound was less than 0.002 Mmol/min under these conditions. ID50 refers to the concentration of inhibitor that gives 50% inhibition. Names of compounds are as follows: (a) methyl (5Z,9a,lla,13E,15S)-9,11-dimercapto-15-hydroxyprosta-5,13-dienoate; (b) methyl (5Z,9f,110,13E,15S)-9,11-dimercapto15-hydroxyprosta-5,13-dienoate; (c) (5Z,9a,lla,13E,15S)-9,11-dihydroxy-15-mercaptoprosta-5,13-dienoic acid; (d) (5Z,9a,11a,13E,15R)9,11-dihydroxy-15-mercaptoprosta-5,13-dienoic acid; (e) (5Z,9a,llac,13E,15S)-1-mercapto-9,11,15-trihydroxyprosta-5,13-diene; (f) (5Z,1la,13E,15S)-1-mercapto-9-oxo-11,15-dihydroxyprosta-5,13-diene; (g) methyl (5Z,9a,lla,13E,15S)-9,11-epidithio-15-hydroxyprosta-5,13dienoate; (h) methyl (5Z,9%,11,3,13E,15S)-9,11-epidithio-15-hydroxyprosta-5,13-dienoate; (i) prostaglandin F2a; (j) dithiothreitol; (k) dihydrolipoic acid; (1) 2,3-dimercaptopropanol; (m) 2-mercaptoethanol; (n) glutathione; and (o) cysteine.
in a- or in f-configuration (compounds g and h). Moreover, prostaglandin F~a (compound i) as a control was not inhibitory at all. These observations suggest that the thiol group of those analogues described above is involved in the inhibition of oxygen consumption. Indeed, several thiol compounds that are not prostaglandin analogues were inhibitory. Namely, dithiothreitol (compound j), dihydrolipoic acid (compound k), 2,3-dimercaptopropanol (compound 1) and 2-mercaptoethanol (compound m) showed 50% inhibition at concentrations on the order of 10 ,gM. Inhibitory effect by glutathione (compound n) and cysteine (compound o) was scarcely observed even at 60 ,uM. Ascorbic acid, which is a reducing agent without a thiol group, was inhibitory to some extent. The results described above indicate that many thiol compounds are inhibitors of prostaglandin endoperoxide synthetase and this inhibitory effect is more prominent with analogues of prostaglandin. Since the initial bis-oxygenation of unsaturated fatty acid is the oxygen-consuming step in prostaglandin HI biosynthesis, the above experimental results, demonstrated witl}an oxygen electrode, suggest inhibition of prostaglandin G1 synthesis. In order to confirm this and to examine further the effect of thiol compounds on the other step (conversion from prostaglandin G1 to Hi), we incubated the enzyme either with 8,11,14-[114Cjeicosatrienoic acid in the presence of heme alone (production of prostaglandin G1) or with [1-_4C prostaglandin G1 in the simultaneous presence of heme and tryptophan (conversion of prostaglandin GI to HI). As shown in Fig. 2 (open circles), 9,11-dihydroxy-15S-mercaptoprosta-5,13-dienoic acid
(panel A), 1-mercapto-9,11,15-trihydroxyprosta-5,13-diene (panel B), dihydrolipoic acid (panel C), and 2,3-dimercaptopropanol (panel D) inhibited the formation of prostaglandin GI from 8,11,14-eicosatrienoic acid. In contrast, with any of these compounds there was no marked inhibition in the con-
version of prostaglandin G1 to H1 (closed circles in Fig. 2). In order to investigate the reversibility of inhibition, we
preincubated the enzyme with 9,11-dihydroxy-15S-mercaptoprosta-5,13-dienoic acid at a concentration of 0.52 ,uM (a concentration giving 80-90% inhibition). After preincubation at 240 for 2, 5, and 10 min, a 20-,gl aliquot was removed each time and added to an assay mixture (1.75 ml), thus resulting in 1
100 (B)
(A)
(D)
(C)
50
0
20
40 0
20
40 0
20
40 0
23
40
Inhibitor (PM) FIG. 2. Effect of various thiol compounds on the production of prostaglandin G1 from 8,11,14-eicosatrienoic acid and the conversion of prostaglandin G1 to HI. Each thiol compound was present at concentrations indicated: (A) 9,11-dihydroxy-15S-mercaptoprosta5,13-dienoic acid; (B) 1-mercapto-9-oxo-11,15-dihydroxyprosta5,13-diene; (C) dihydrolipoic acid; and (D) 2,3-dimercaptopropanol. The production of prostaglandin GI from 8,11,14-eicosatrienoic acid (0) with 20 Ag of enzyme and the conversion of prostaglandin GI to HI (-) with 7 gg of enzyme were assayed as described (7). The enzyme activity in the presence of inhibitor is expressed as % of the uninhibited control (2.5 nmol of prostaglandin G1 produced per min and 2.2 nmol prostaglandin Hi produced per min).
Proc. Natl. Acad. Sci. USA 74 (1977)
Biochemistry: Ohki et al. an about 90-fold dilution of the inhibitor. Since there was no inhibition regardless of the preincubation time, the inhibition seemed to be reversible. Similar results were observed with 2,3-dimercaptopropanol. Moreover, when the enzyme was preincubated with either of the above two thiol compounds in the standard reaction mixture without hematin and then the reaction was started by the addition of hematin, the inhibitory effect was not influenced by the preincubation. The results indicate that the enzyme is not inactivated by thiol compounds during the preincubation both in the absence of substrate and in its presence. Rate of oxygen consumption was determined at various concentrations of substrate and thiol compounds, and results were treated by Lineweaver-Burk plot. As shown in Fig. 3A and B, both 9,11-dihydroxy-15S-mercaptoprosta-5,13-dienoic acid and 2,3-dimercaptopropanol exhibited noncompetitive type of inhibition. The Ki value was 0.08 AtM and 4.20 ,M, respectively. Several experiments were carried out in order to examine whether the inhibitory effect of thiol compounds is due to an interaction of these thiol compounds with heme, which is required for the bis-oxygenation of 8,11,14-eicosatrienoic acid to prostaglandin G1. First, when 0.78 tM 9,11-dihydroxy15S-mercaptoprosta-5,13-dienoic acid or 10 qM 2,3-dimercaptopropanol was present in a reaction mixture containing 8,11,14-eicosatrienoic acid, hemoglobin, and tryptophan under standard conditions, there was essentially no change in the visible absorption spectrum of hemoglobin during a 5-min incubation time. The presence of enzyme in this reaction mixture caused only a slow degradation of heme. Second, the heme concentration curve was obtained either in the presence of thiol compound or in its absence. There was no significant difference in the heme concentration to give a half-maximal velocity of reaction. These results suggest that the inhibitory effect of various thiol compounds is not attributable to destruction of heme associated with the oxidation of thiol compounds. Furthermore, it is unlikely that the enzyme is inactivated by some active species of oxygen, such as superoxide anion or hydrogen peroxide, produced by a possible heme-catalyzed autooxidation of thiol compound, since the addition of superoxide dismutase or catalase did not reverse the inhibition by 1-mercapto9,11,15-trihydroxyprosta-5,13-diene or 2,3-dimercaptopropanol. DISCUSSION Effects of various thiol compounds have been investigated in various aspects of prostaglandin biosynthesis. -It was earlier found that glutathione was required for the over-all synthesis of prostaglandin E by a microsomal preparation (1, 21-23). This requirement of glutathione was later assigned to the step of isomerization of prostaglandin H to E (3, 6, 8). It was also reported that a complex of thiol and copper accelerated the formation of prostaglandin F2, rather than E2 (24), but this was attributed to a nonenzymatic degradation of prostaglandin endoperoxide (25). Another interesting finding was the inhibition of prostaglandin synthesis as demonstrated with various systems; i.e., partial inhibition by several thiol compounds in prostaglandin E2 synthesis catalyzed by bovine vesicular gland microsome (22), inhibition by 2,3-dimercaptopropanol in the co-oxygenation of various compounds coupled with prostaglandin biosynthesis (26), and inhibition by 2-mercaptoethanol in the formation of thromboxane B2 via prostaglandin endoperoxide by brain tissue (27). As described in this paper, a variety of thiol compounds were shown to inhibit the initial bis-oxygenatior of an unsaturated
3
147
A
E 2 E 0
ii
0.1
0.2
8J1 14-Eicosotrienoic acid (PMF'
0.1 8,1 114-Ecosatrienoec acid
0.2 F' (WM
FIG. 3. Inhibition of oxygen consumption by thiel compounds as presented by Lineweaver-Burk plot. Initial rate of oxygen consumption was determined in the standard reaction mixture described in Materials and Methods. Reaction was started by the addition of 30 Ag of enzyme. (A) 9,11-Dihydroxy-15S-mercaptoprosta-5,13-dienoic acid was present at 0 nM (0), 86 nM (X), and 114 nM (A). (B) 2,3-Dimercaptopropanol was present at O AM (0), 5.7 MM (X), and 11.4 AM (A).
fatty acid to prostaglandin G1. The second step (conversion of prostaglandin GI to HI) was not significantly affected. It should be noted that several prostaglandin analogues with thiol group were much more potent inhibitors. The inhibition by these analogues was attributable to the function of the thiol group on the basis of the following observations. Various other thiol compounds unrelated to prostaglandin in structure were also inhibitory. The inhibition, as examined with methyl 9a,11adimercapto-15S-hydroxyprosta-5,13-dienoate, was reversed by the presence of an excess amount of 5,5'-dithiobis(2-nitrobenzoic acid).1 Analogues with a disulfide between 9- and 1 1-positions were without effect. Prostaglandin F2, showed no inhibition. Such an inhibition by the thiol group was fortified at least in part by a prostaglandin-like structure of the molecule, although these analogues were noncompetitive inhibitors. In the experiment using a homogenate of sheep vesicular gland it was reported that the addition of both glutathione peroxidase and glutathione inhibited the synthesis of prostaglandin E (28). The inhibition was attributed to the functioning glutathione peroxidase system with glutathione as a hydrogen donor and presumably with prostaglandin G as a hydrogen acceptor. It seems to be difficult to explain the thiol-dependent inhibition of our enzyme system by such a mechanism. The highly purified enzyme preparation was shown to be free of the glutathione peroxidase activity (7). Moreover, as described in Fig. 2, the addition of various thiol compounds to the enzyme and prostaglandin G1 did not stimulate the production of prostaglandin HI, which should be produced by the contaminating peroxidase, if any, with prostaglandin G1 as hydrogen acceptor, suggesting that the thiol-dependent peroxidase was not operative. On the othpr hand, the inhibitory effect of the thiol group might be due to the interaction of the thiol group with copper or iron which might be contained in the enzyme as an essential prosthetic group. The presence of the metal prosthetic group and the precise mechanism of inhibition remain to be elucidated. In view of the anti-inflammatory effect of penicillamine, which has a thiol group (29) and inhibits prostaglandin biosynthesis (30), a possible anti-inflammatory effect of these thiol analogues of prostaglandin should be investigated. ¶ S. Ohki, unpublished experiment.
148
Biochemistry: Ohki et al.
This work has been supported in part by a Grant-in-Aid for Scientific Research from the Ministry of Education, Science and Culture of Japan and by grants from the Japanese Foundation on Metabolism and Diseases, the Fujiwara Memorial Foundation, and the Kanae Foundation of the New Medicine Research Association. 1. Nugteren, D. H., Beerthuis, R. K. & van Dorp, D. A. (1966) Rec. Trav. Chim. Pays-Bas Beig. 85,405-419. 2. Hamberg, M. & Samuelsson, B. (1967) J. Biol. Chem. 242, 5336-5343. 3. Nugteren, D. H. & Hazelhof, E. (1973) Biochim. Biophys. Acta 326, 448-461. 4. Hamberg, M. & Samuelsson, B. (1973) Proc. Nati. Acad. Sci. USA
70,899-903. 5. Hamberg, M., Svensson, J., Wakabayashi, T. & Samuelsson, B. (1974) Proc. Natl. Acad. Sci. USA 71, 345-349. 6. Miyamoto, T., Yamamoto, S. & Hayaishi, 0. (1974) Proc. Natl. Acad. Sci. USA 71, 3645-3648. 7. Miyamoto, T., Ogino, N., Yamamoto, S. & Hayaishi, 0. (1976) J. Biol. Chem. 251, 2629-2636. 8. Ogino, N., Miyamoto, T., Yamamoto, S. & Hayaishi, 0. (1977) J. Biol. Chem., in press. 9. Wlodawer, P., Samuelsson, B., Albonico, S. M. & Corey, E. J. (1971) J. Am. Chem. Soc. 93,2815-2816. 10. Corey, E. J., Nicolaou, K. C., Machida, Y., Malmsten, C. L. & Samuelsson, B. (1975) Proc. NatI. Acad. Sci. USA 72, 33553358. 11. Bundy, G. L. (1975) Tetrahedron Lett. 24, 1957-1960. 12. Malmsten, C. (1976) Life Sci. 18, 169-176. 13. Klenk, E. & Mohrhauer, H. (1960) Z. Physiol. Chem. 320, 218-232.
Proc. Natl. Acad. Sci. USA 74 (1977) 14. Corey, E. J., Weinshenker, N. M., Schaaf, T. K. & Huber, W. (1969) J. Am. Chem. Soc. 91,5675-5677. 15. Yost, F. J., Jr. & Fridovich, I. (1973) J. Biol. Chem. 248, 49054908. 16. McCord, J. M. & Fridovich, I. (1969) J. Biol. Chem. 244, 6049-6055. 17. Miyake, Y. & Fukuyama, M. (1976) J. Biochem. (Tokyo) 79, 613-620. 18. Lowry, 0. H., Rosebrough, N. J., Farr, A. L. & Randall, R. J. (1951) J. Biol. Chem. 193, 265-275. 19. Ellman, G. L. (1959) Arch. Biochem. Biophys. 82,70-77. 20. Paul, K.-G., Theorell, H. & Akeson, A. (1953) Acta Chem. Scand. 7, 1284-1287. 21. Samuelsson, B. (1967) Prog. Biochem. Pharmacol. 3,59-70. 22. Takeguchi, C., Kohno, E. & Sih, C. J. (1971) Biochemistry 10, 2372-2376. 23. Yoshimoto, A., Ito, H. & Tomita, K. (1970) J. Biochem. (Tokyo) 68,487-499. 24. Lee, R. E. & Lands, W. M. (1972) Biochim. Biophys. Acta 260, 203-211. 25. Chan, J. A., Nagasawa, M., Takeguchi, C. & Sih, C. J. (1975) Biochemistry 14, 2987-2991. 26. Marnett, L. J., Wlodawer, P. & Samuelsson, B. (1975) J. Biol. Chem. 250, 8510-8517. 27. Wolfe, L. S., Rostworoski, K. & Marison, J. (1976) Biochem. Biophys. Res. Commun. 70,907-913. 28. Lands, W., Lee, R. & Smith, W. (1971) Ann. N.Y. Acad. Sci. 180, 107-122. 29. Shen, T. Y. (1974) in Medicinal Chemistry, eds. Scherren, R. A. & Whitehouse, M. W. (Academic Press, New York), Vol. 13, pp. 203-204. 30. Maddox, I. S. (1973) Biochim. Biophys. Acta 306,74-81.
Biochemistry
Inhibition of prostaglandin endoperoxide synthetase by thiol analogues of prostaglandin (prostaglandin synthase/oxygenase)
SHIRO OHKI*t, NOBUCHIKA OGINOt, SHOZO YAMAMOTOt, OSAMU HAYAISHIt, HISASHI YAMAMOTOt, HAZIMU MIYAKE§, AND MASAKI HAYASHI§ t Department of Medical Chemistry, Kyoto University Faculty of Medicine, Kyoto, Japan; * Department of Industrial Chemistry, Kyoto University Faculty of Engineering, Kyoto, Japan; and § The Ono Central Research Institute, Shimamoto-cho, Mishima-gun, Osaka, Japan
Contributed by Osamu Hayaishi, October 27, 1976
ABSTRACT A variety of thiol compounds inhibited the enzymatic bis-oxygenation of 8,11,14-eicosatrienoic acid to prostaglandin GI, as examined with a purified preparation of prostaglandin endoperoxide synthetase (prostaglandin synthase; 8,11,14-eicosatrienoate, hydrogen-donor:oxygen oxidoreductase; EC 1.14.99.1) from bovine vesicular gland. The hydroperoxide cleavage of prostaglandin G1 producing prostaglandin H1 was not affected by these thiol compounds. Several prostaglandin analogues with a thiol group (9,11-dihydroxy-15S- or 15R-mercaptoprosta-5,13-dienoic acid, 1-mercapto-9,11,15-trihydroxyprosta-5,13-diene, and 1-mercapto-9-oxo-11,15-dihydroxyprosta-5,13-diene) were most potent inhibitors, showing almost complete inhibition at concentrations on the order of 1 t&M. Other thiol compounds, such as 2,3-dimercaptopropanol, dithiothreitol, and dihydrolipoic acid, were also inhibitory but were much less effective. The inhibition, as examined with 9,11-dihydroxy-15S-mercaptoprosta-5,13-dienoic acid and 2,3dimercaptopropanol, was noncompetitive.
Beef blood hemoglobin (type I), hematin, cysteine, 2-mercaptoethanol, dithiothreitol, and dihydrolipoic acid were purchased from Sigma; tryptophan, ascorbic acid, and 2,3dimercaptopropanol from Nakarai Chemicals (Kyoto); 5,5'dithio-bis-(2-nitrobenzoic acid) from Tokyo Kasei Kogyo (Tokyo); and precoated silica gel plate 60 Fws4 for thin-layer chromatography from E. Merck (Darmstadt). Glutathione was donated by the Yamanouchi Central Research Institute. Synthesis of (5Z,9a,lla,13E,15)-l-Mercapto-9,11,15trihydroxyprosta-5,13-diene (V in Fig. 1). Reduction of I by diisobutylaluminum hydride produced the 1,9-diol II, which was subjected to selective tosylation using 1.2 equivalents of p-toluenesulfonyl chloride in pyridine at 00 for 20 hr (68% yield). The tosylate III was converted to the thioacetate IV with excess sodium thioacetate in dimethyl sulfoxide-dimethylformamide at 250 for 2 hr (83% yield). Hydrolysis of IV by K2CO3 in methanol, followed by removal of tetrahydropyranyl group in acetic acid-tetrahydrofuran-water, gave the 1-mercapto derivative V (in 73% yield from IV): nuclear magnetic resonance spectrum in CDC13 due to 2H of CH2S at S 2.55 ppm. Synthesis of (5Zlla,13E,15S)l-Mercapto-9-oxo-11,15dihydroxyprosta-5,13-diene (VIII). The alcohol III was oxidized by Collins reagent in methylene chloride at 00 for 15 min. The resulting ketone VI was transformed to the keto diol VII in acetic acid-tetrahydrofuran-water (71% yield from III). Treatment of VII with a large excess of sodium hydrosulfide in dimethylformamide at -300 for 1 hr, followed by chromatography on silica gel, produced the keto mercaptan VIII as a crystalline solid: melting point 45-;0; infrared absorption due to C=O at 1735 cm-I; nuclear magnetic resonance spectrum in CDC13 due to 2H of CH2S at 5 2.5 ppm. Synthesis of (5Z,9a,lla,13E,15S)9,1 1-Dihydroxy-15mercaptoprosta-5,13-dienoic acid (XV). The optically active lactone IX prepared by the method of Corey et al. (14), was converted to the 15R-bromide X with 1.1 equivalents of phosphorous tribromide in ether at 00 and then to 15S-thioacetate XI by treatment with sodium thioacetate in dimethyl sulfoxide-dimethylformamide at 250 for 1 hr. The latter compound XI was hydrolyzed to the thiol XII by K2CO3 in methanol at 250 for 1 hr (64% yield from IX). Reaction of XII with ethyl vinyl ether in methylene chloride at 25° for 20 min yielded the sulfide derivative XIII, which was transformed to the corresponding lactol by exposure to diisobutylaluminum hydride in toluene in 97% yield. Treatment of the lactol with the Wittig reagent prepared from 5-triphenylphosphonovaleric acid in dimethyl sulfoxide produced the acid XIV in 88% yield. This compound was hydrolyzed in aqueous acetic acid at 400 for 2 hr to the 15S-mercapto derivative XV (50% yield): nuclear magnetic resonance spectrum in CDC13 due to 1H of C(15)H
The hypothetical mechanism of prostaglandin biosynthesis proposed earlier (1, 2) has recently been substantiated by the isolation of two prostaglandin endoperoxides (3-5). The intermediary nature of these endoperoxides was further supported by the solubilization and purification of two enzyme fractions in our laboratory (6-8). One catalyzes the bis-oxygenation of an unsaturated fatty acid producing prostaglandin G, which has both an endoperoxy and a hydroperoxy group. Heme is required for this reaction. The same enzyme preparation also catalyzes the cleavage of the hydroperoxy group of prostaglandin G, producing prostaglandin H with an endoperoxy and a hydroxy group. This reaction also requires heme and is markedly stimulated by tryptophan (7). The other enzyme isomerizes the endoperoxy group of prostaglandin H to produce prostaglandin E (6, 8). In view of the role of these endoperoxides in prostaglandin biosynthesis, attempts have been made to synthesize analogues of prostaglandin endoperoxides that would have any biological activity and effect on enzyme (9-12). In the course of studies on various analogues of prostaglandin and prostaglandin endoperoxides, we found that some analogues with thiol group(s) as well as other thiol compounds were potent inhibitors of the bis-oxygenation of an unsaturated fatty acid to prostaglandin G. MATERIALS AND METHODS Chemicals. 8,11,14-[1-14C]Eicosatrienoic acid (57.5 mCi/ mmol) was purchased from New England Nuclear and purified by thin-layer chromatography. 8,11,14-Eicosatrienoic acid was prepared by the method of Klenk and Mohrhauer (13), and prostaglandin F2a according to the method of Corey et al. (14). *
On leave from the Ono Central Research Institute. 144
Proc. Natl. Acad. Sci. USA 74 (1977)
Biochemistry: Ohki et al. OH R
Ox
ox
(I) RzCOOMe,XaTHP (11) R:CH20H,X:THP (111) R=CH2OTs,X=THP (IV) R CH2SAc, X= THP (V) RzCH2SH,X=H
0 R OX
OX
(VI) R-CH20Ts,X=THP (VII) RuCH20Ts,XzH (ViII) RzCH2SH,X*H
. s.o (IX) X AcY DO
'-H (X) XzAc, Y/H
(XI) X=Ac, Y=S'SAc (X I) X - H. Y (H
X=,~SH
(XIII) X=CH(OEt)Me, *SCH(OEt)Me
(XVI) XzAc,Y3...H
(XVII) X.H y(,,SH
OH
(XIV) XaCH(OEt)Me,
K>2--2 COH
ox IXY
Y *SCH(OEt)Me /H XuHY (XV) XSH XxH,Ya
(XVIII)
FIG. 1. Structures of prostaglandin analogues and intermediates of their organic syntheses. Abbreviations: Me, methyl; Et, ethyl; THP, tetrahydropyranyl; Ts, p-toluenesulfonyl; Ac, acetyl. at a 3.2-3.6 ppm; homogeneous as examined by thin-layer chromatography. Synthesis of (5Z,9a,lla,13E,15R)9,1 1-Dihydroxy-1Smercaptoprosta-5,13-dienoic acid (XVIII). Although the bromination of the 15S-IX by phosphorous tribromide proceeded regio- and stereospecifically with inversion of configuration at C-15, similar treatment with the 15R-alcohol XVI gave rise to a mixture of 13- and 15-bromide. The mixture was directly converted to a mixture of 13- and 15-thioacetate in 11% and 65% yield, respectively. The 15-thioacetate was readily separated from the 13-isomer by chromatography on silica gel and hydrolyzed with K2CO3 in methanol (76% yield). The 15R-mercaptan XVII thus produced was separated from the contaminating diastereomer at C-15 (15R/15S = 3:2). The same sequence of reactions from XII to XV was carried out for the conversion of XVII to the 15R-mercapto acid XVIII. Thinlayer chromatography on silica gel in a solvent system of chloroform-tetrahydrofuran-acetic acid (10:2:1) gave RF values of 0.66 and 0.59 for XV and XVIII, respectively. The assignment of stereochemistry at C-15 is anticipated on mechanistic consequences as well as chromatographic mobili-
ties.
Synthesis of Other Compounds. Procedures for synthesis of methyl (5Z,9a,lla,13E,1SS)-9,11-dimercapto-15-hydroxyprosta-5,13-dienoate, methyl (5Z,9,3,110,13E,15S)-9,11dimercapto-15-hydroxyprosta-5,13-dienoate, methyl (5Z, 9a, 1 1a, 13E,15S)-9, 1 1-epidithio-15-hydroxyprosta-5, 13-di-
145
enoate, and methyl (SZ,9fl,11f,13E,15S)-9,11-epidithio-15hydroxyprosta-5,13-dienoate will be published elsewhere. Enzymes. Prostaglandin endoperoxide synthetase (prostaglandin synthase; 8,11,14-eicosatrienoate, hydrogen-donor: oxygen oxidoreductase; EC 1.14.99.1) was prepared by the method described previously (7). The enzyme solution was concentrated to a protein concentration of 1.5 mg/ml with the aid of a Diaflow membrane (PM-10). Iron-containing superoxide dismutase from Escherichia coli (15) and copper-containing superoxide dismutase from bovine blood (16) were kindly provided by Drs. J. Oka and T. Yoshimoto of this laboratory. Beef liver catalase was purchased from Sigma. Preparation of Prostaglandin G1. The method of preparation described (7) was used on a 10-fold larger scale; i.e., in a 10-ml mixture containing 8,11,14-[1-14C]eicosatrienoic acid (1.1Itmol, 1.25 X 107 cpm) and the purified enzyme (2.5 mg of protein). Enzyme Assays. Conversions from 8,11,14-eicosatrienoic acid to prostaglandin G1 and from prostaglandin G1 to prostaglandin H1 were assayed by the methods using radioactive substrates as described (7). For the determination of oxygen consumption, a galvanic oxygen electrode (type SBOA, the Kyusui Kagaku Institute, Tokyo) was used. The sensitivity of oxygen consumption measurement was increased with the aid of an offset amplifier according to the method of Miyake and Fukuyama (17). The reaction cell was thermostated by circulating water at 24°. The reaction mixture contained Tris-HCl buffer at pH 8.0 (175 ,qmol), 8,11,14-eicosatrienoic acid (250 nmol), hemoglobin or hematin (2 nmol), tryptophan (8.6 Amol), and enzyme in a final volume of 1.75 ml. Reaction was usually started by the addition of enzyme that had been equilibrated at 240. Initial velocity was determined within a time range in which the reaction proceeded in a linear fashion. Miscellaneous Determinations. Protein concentration was determined by the method of Lowry et al. (18), with bovine serum albumin as a standard. Thiol content was determined by titration using 5,5'-dithiobis(2-nitrobenzoic acid) developed by Ellman (19). The concentration of heme was determined spectrophotometrically as reduced pyridine hemochromogen, using an extinction coefficient of 34.4 mM-' cm-' at 557 nm (20). Spectra of heme were recorded with a Union Giken spectrophotometer model SM-401.
RESULTS As shown in Table 1, effects of various thiol compounds were screened by measurement of oxygen consumption upon incubation of the enzyme with 8,11,14-eicosatrienoic acid in the presence of hemoglobin and tryptophan. Prostaglandin H1 rather than G1 was the major product under these conditions (7). Several prostaglandin analogues with thiol group(s) markedly inhibited the oxygen consumption. 9,11-Dihydroxy-15S-mercaptoprosta-5,13-dienoic acid (compound c) and 9,11-dihydroxy-15R-mercaptoprosta-5,13-dienoic acid (compound d) were the most potent inhibitors, giving 50% inhibition at concentrations on the order of 0.1 gM. 1-Mercapto9,11,15-trihydroxyprosta-5,13-diene (compound e) and 1mercapto-9-oxo-11,15-dihydroxyprosta-5,13-diene (compound f) were also effective inhibitors, but with higher concentrations to give 50% inhibition (approximately 1,gM). Analogues with two thiol groups at C-9 and C-11 (compounds a and b) were much less effective irrespective of the configuration of the thiol groups. Their concentrations for 50% inhibition were on the order of 10,uM. It should be noted that compounds with a disulfide bridge between C-9 and C-11 were ineffective either
146
|Oe)
Biochemistry: Ohki et al. Compound
Table 1. Inhibition of oxygen consumption by various thiol compounds
(PM)
i|
C;
MN
()(M
20 |LC
|50 10 20 50
18 31 56
40 1
2
m,
46 g) 40
-
50 h) 40 0.)1~~o4 80 82 90 100 I~~~~~
0 0 0
-
05
0.2
38
2
8
i)
|i 1
2 6
35 72 79
(CI)COH
SiH- SH
0 0 0
0.1
d)
C.8?
75 77
;,80 100
0.OOH 79 ,NW~~~~~~t7 0.4 0779 1
niito trtio khtio(tionM)50 Compound tration | (0/ ) p) 1)(M) (PM) (PM)pM tO 46 0.4 27 f0) k)
Compound
33 50 76
10 20
D0Concen-InitonD5
Concen-
Concen- niito D tration Inhibition ID5
a)
Q" e\OC
Proc. Natl. Acad. Sci. USA 74 (1977)
CH29HCH20H SH SH
HM)
OHO SH
0
30
I)
jooH
n)
30
66
50
74
5 10 20
46 76
5 20 50
25 62 76
20
0
NH2CHCOOH 0)
1)0
15
20 50
45
23 |
64
1
H2
20
40 60
H
6
86
1cOH OOI2CH2CHNHi2 40 0 60 MCHCONHCH2COOH0 0 60 K:~~7$^Voo/ 110 0 60 0
A OH 01| 14 I|H2H6HCH2 H ~
12
8 10 10
10
-
-
Rate of oxygen consumption was determined with an oxygen electrode in a reaction mixture described in Materials and Methods. Each thiol compound was present as indicated. Glutathione and cysteine were dissolved in water, and other compounds in acetone. The reaction was started by the addition of 40 pg of purified enzyme. The enzyme activity in the presence of inhibitor is expressed as % of the uninhibited control (0.17 ,mol of oxygen consumed per min). Nonenzymatic oxygen consumption in the presence of each thiol compound was less than 0.002 Mmol/min under these conditions. ID50 refers to the concentration of inhibitor that gives 50% inhibition. Names of compounds are as follows: (a) methyl (5Z,9a,lla,13E,15S)-9,11-dimercapto-15-hydroxyprosta-5,13-dienoate; (b) methyl (5Z,9f,110,13E,15S)-9,11-dimercapto15-hydroxyprosta-5,13-dienoate; (c) (5Z,9a,lla,13E,15S)-9,11-dihydroxy-15-mercaptoprosta-5,13-dienoic acid; (d) (5Z,9a,11a,13E,15R)9,11-dihydroxy-15-mercaptoprosta-5,13-dienoic acid; (e) (5Z,9a,llac,13E,15S)-1-mercapto-9,11,15-trihydroxyprosta-5,13-diene; (f) (5Z,1la,13E,15S)-1-mercapto-9-oxo-11,15-dihydroxyprosta-5,13-diene; (g) methyl (5Z,9a,lla,13E,15S)-9,11-epidithio-15-hydroxyprosta-5,13dienoate; (h) methyl (5Z,9%,11,3,13E,15S)-9,11-epidithio-15-hydroxyprosta-5,13-dienoate; (i) prostaglandin F2a; (j) dithiothreitol; (k) dihydrolipoic acid; (1) 2,3-dimercaptopropanol; (m) 2-mercaptoethanol; (n) glutathione; and (o) cysteine.
in a- or in f-configuration (compounds g and h). Moreover, prostaglandin F~a (compound i) as a control was not inhibitory at all. These observations suggest that the thiol group of those analogues described above is involved in the inhibition of oxygen consumption. Indeed, several thiol compounds that are not prostaglandin analogues were inhibitory. Namely, dithiothreitol (compound j), dihydrolipoic acid (compound k), 2,3-dimercaptopropanol (compound 1) and 2-mercaptoethanol (compound m) showed 50% inhibition at concentrations on the order of 10 ,gM. Inhibitory effect by glutathione (compound n) and cysteine (compound o) was scarcely observed even at 60 ,uM. Ascorbic acid, which is a reducing agent without a thiol group, was inhibitory to some extent. The results described above indicate that many thiol compounds are inhibitors of prostaglandin endoperoxide synthetase and this inhibitory effect is more prominent with analogues of prostaglandin. Since the initial bis-oxygenation of unsaturated fatty acid is the oxygen-consuming step in prostaglandin HI biosynthesis, the above experimental results, demonstrated witl}an oxygen electrode, suggest inhibition of prostaglandin G1 synthesis. In order to confirm this and to examine further the effect of thiol compounds on the other step (conversion from prostaglandin G1 to Hi), we incubated the enzyme either with 8,11,14-[114Cjeicosatrienoic acid in the presence of heme alone (production of prostaglandin G1) or with [1-_4C prostaglandin G1 in the simultaneous presence of heme and tryptophan (conversion of prostaglandin GI to HI). As shown in Fig. 2 (open circles), 9,11-dihydroxy-15S-mercaptoprosta-5,13-dienoic acid
(panel A), 1-mercapto-9,11,15-trihydroxyprosta-5,13-diene (panel B), dihydrolipoic acid (panel C), and 2,3-dimercaptopropanol (panel D) inhibited the formation of prostaglandin GI from 8,11,14-eicosatrienoic acid. In contrast, with any of these compounds there was no marked inhibition in the con-
version of prostaglandin G1 to H1 (closed circles in Fig. 2). In order to investigate the reversibility of inhibition, we
preincubated the enzyme with 9,11-dihydroxy-15S-mercaptoprosta-5,13-dienoic acid at a concentration of 0.52 ,uM (a concentration giving 80-90% inhibition). After preincubation at 240 for 2, 5, and 10 min, a 20-,gl aliquot was removed each time and added to an assay mixture (1.75 ml), thus resulting in 1
100 (B)
(A)
(D)
(C)
50
0
20
40 0
20
40 0
20
40 0
23
40
Inhibitor (PM) FIG. 2. Effect of various thiol compounds on the production of prostaglandin G1 from 8,11,14-eicosatrienoic acid and the conversion of prostaglandin G1 to HI. Each thiol compound was present at concentrations indicated: (A) 9,11-dihydroxy-15S-mercaptoprosta5,13-dienoic acid; (B) 1-mercapto-9-oxo-11,15-dihydroxyprosta5,13-diene; (C) dihydrolipoic acid; and (D) 2,3-dimercaptopropanol. The production of prostaglandin GI from 8,11,14-eicosatrienoic acid (0) with 20 Ag of enzyme and the conversion of prostaglandin GI to HI (-) with 7 gg of enzyme were assayed as described (7). The enzyme activity in the presence of inhibitor is expressed as % of the uninhibited control (2.5 nmol of prostaglandin G1 produced per min and 2.2 nmol prostaglandin Hi produced per min).
Proc. Natl. Acad. Sci. USA 74 (1977)
Biochemistry: Ohki et al. an about 90-fold dilution of the inhibitor. Since there was no inhibition regardless of the preincubation time, the inhibition seemed to be reversible. Similar results were observed with 2,3-dimercaptopropanol. Moreover, when the enzyme was preincubated with either of the above two thiol compounds in the standard reaction mixture without hematin and then the reaction was started by the addition of hematin, the inhibitory effect was not influenced by the preincubation. The results indicate that the enzyme is not inactivated by thiol compounds during the preincubation both in the absence of substrate and in its presence. Rate of oxygen consumption was determined at various concentrations of substrate and thiol compounds, and results were treated by Lineweaver-Burk plot. As shown in Fig. 3A and B, both 9,11-dihydroxy-15S-mercaptoprosta-5,13-dienoic acid and 2,3-dimercaptopropanol exhibited noncompetitive type of inhibition. The Ki value was 0.08 AtM and 4.20 ,M, respectively. Several experiments were carried out in order to examine whether the inhibitory effect of thiol compounds is due to an interaction of these thiol compounds with heme, which is required for the bis-oxygenation of 8,11,14-eicosatrienoic acid to prostaglandin G1. First, when 0.78 tM 9,11-dihydroxy15S-mercaptoprosta-5,13-dienoic acid or 10 qM 2,3-dimercaptopropanol was present in a reaction mixture containing 8,11,14-eicosatrienoic acid, hemoglobin, and tryptophan under standard conditions, there was essentially no change in the visible absorption spectrum of hemoglobin during a 5-min incubation time. The presence of enzyme in this reaction mixture caused only a slow degradation of heme. Second, the heme concentration curve was obtained either in the presence of thiol compound or in its absence. There was no significant difference in the heme concentration to give a half-maximal velocity of reaction. These results suggest that the inhibitory effect of various thiol compounds is not attributable to destruction of heme associated with the oxidation of thiol compounds. Furthermore, it is unlikely that the enzyme is inactivated by some active species of oxygen, such as superoxide anion or hydrogen peroxide, produced by a possible heme-catalyzed autooxidation of thiol compound, since the addition of superoxide dismutase or catalase did not reverse the inhibition by 1-mercapto9,11,15-trihydroxyprosta-5,13-diene or 2,3-dimercaptopropanol. DISCUSSION Effects of various thiol compounds have been investigated in various aspects of prostaglandin biosynthesis. -It was earlier found that glutathione was required for the over-all synthesis of prostaglandin E by a microsomal preparation (1, 21-23). This requirement of glutathione was later assigned to the step of isomerization of prostaglandin H to E (3, 6, 8). It was also reported that a complex of thiol and copper accelerated the formation of prostaglandin F2, rather than E2 (24), but this was attributed to a nonenzymatic degradation of prostaglandin endoperoxide (25). Another interesting finding was the inhibition of prostaglandin synthesis as demonstrated with various systems; i.e., partial inhibition by several thiol compounds in prostaglandin E2 synthesis catalyzed by bovine vesicular gland microsome (22), inhibition by 2,3-dimercaptopropanol in the co-oxygenation of various compounds coupled with prostaglandin biosynthesis (26), and inhibition by 2-mercaptoethanol in the formation of thromboxane B2 via prostaglandin endoperoxide by brain tissue (27). As described in this paper, a variety of thiol compounds were shown to inhibit the initial bis-oxygenatior of an unsaturated
3
147
A
E 2 E 0
ii
0.1
0.2
8J1 14-Eicosotrienoic acid (PMF'
0.1 8,1 114-Ecosatrienoec acid
0.2 F' (WM
FIG. 3. Inhibition of oxygen consumption by thiel compounds as presented by Lineweaver-Burk plot. Initial rate of oxygen consumption was determined in the standard reaction mixture described in Materials and Methods. Reaction was started by the addition of 30 Ag of enzyme. (A) 9,11-Dihydroxy-15S-mercaptoprosta-5,13-dienoic acid was present at 0 nM (0), 86 nM (X), and 114 nM (A). (B) 2,3-Dimercaptopropanol was present at O AM (0), 5.7 MM (X), and 11.4 AM (A).
fatty acid to prostaglandin G1. The second step (conversion of prostaglandin GI to HI) was not significantly affected. It should be noted that several prostaglandin analogues with thiol group were much more potent inhibitors. The inhibition by these analogues was attributable to the function of the thiol group on the basis of the following observations. Various other thiol compounds unrelated to prostaglandin in structure were also inhibitory. The inhibition, as examined with methyl 9a,11adimercapto-15S-hydroxyprosta-5,13-dienoate, was reversed by the presence of an excess amount of 5,5'-dithiobis(2-nitrobenzoic acid).1 Analogues with a disulfide between 9- and 1 1-positions were without effect. Prostaglandin F2, showed no inhibition. Such an inhibition by the thiol group was fortified at least in part by a prostaglandin-like structure of the molecule, although these analogues were noncompetitive inhibitors. In the experiment using a homogenate of sheep vesicular gland it was reported that the addition of both glutathione peroxidase and glutathione inhibited the synthesis of prostaglandin E (28). The inhibition was attributed to the functioning glutathione peroxidase system with glutathione as a hydrogen donor and presumably with prostaglandin G as a hydrogen acceptor. It seems to be difficult to explain the thiol-dependent inhibition of our enzyme system by such a mechanism. The highly purified enzyme preparation was shown to be free of the glutathione peroxidase activity (7). Moreover, as described in Fig. 2, the addition of various thiol compounds to the enzyme and prostaglandin G1 did not stimulate the production of prostaglandin HI, which should be produced by the contaminating peroxidase, if any, with prostaglandin G1 as hydrogen acceptor, suggesting that the thiol-dependent peroxidase was not operative. On the othpr hand, the inhibitory effect of the thiol group might be due to the interaction of the thiol group with copper or iron which might be contained in the enzyme as an essential prosthetic group. The presence of the metal prosthetic group and the precise mechanism of inhibition remain to be elucidated. In view of the anti-inflammatory effect of penicillamine, which has a thiol group (29) and inhibits prostaglandin biosynthesis (30), a possible anti-inflammatory effect of these thiol analogues of prostaglandin should be investigated. ¶ S. Ohki, unpublished experiment.
148
Biochemistry: Ohki et al.
This work has been supported in part by a Grant-in-Aid for Scientific Research from the Ministry of Education, Science and Culture of Japan and by grants from the Japanese Foundation on Metabolism and Diseases, the Fujiwara Memorial Foundation, and the Kanae Foundation of the New Medicine Research Association. 1. Nugteren, D. H., Beerthuis, R. K. & van Dorp, D. A. (1966) Rec. Trav. Chim. Pays-Bas Beig. 85,405-419. 2. Hamberg, M. & Samuelsson, B. (1967) J. Biol. Chem. 242, 5336-5343. 3. Nugteren, D. H. & Hazelhof, E. (1973) Biochim. Biophys. Acta 326, 448-461. 4. Hamberg, M. & Samuelsson, B. (1973) Proc. Nati. Acad. Sci. USA
70,899-903. 5. Hamberg, M., Svensson, J., Wakabayashi, T. & Samuelsson, B. (1974) Proc. Natl. Acad. Sci. USA 71, 345-349. 6. Miyamoto, T., Yamamoto, S. & Hayaishi, 0. (1974) Proc. Natl. Acad. Sci. USA 71, 3645-3648. 7. Miyamoto, T., Ogino, N., Yamamoto, S. & Hayaishi, 0. (1976) J. Biol. Chem. 251, 2629-2636. 8. Ogino, N., Miyamoto, T., Yamamoto, S. & Hayaishi, 0. (1977) J. Biol. Chem., in press. 9. Wlodawer, P., Samuelsson, B., Albonico, S. M. & Corey, E. J. (1971) J. Am. Chem. Soc. 93,2815-2816. 10. Corey, E. J., Nicolaou, K. C., Machida, Y., Malmsten, C. L. & Samuelsson, B. (1975) Proc. NatI. Acad. Sci. USA 72, 33553358. 11. Bundy, G. L. (1975) Tetrahedron Lett. 24, 1957-1960. 12. Malmsten, C. (1976) Life Sci. 18, 169-176. 13. Klenk, E. & Mohrhauer, H. (1960) Z. Physiol. Chem. 320, 218-232.
Proc. Natl. Acad. Sci. USA 74 (1977) 14. Corey, E. J., Weinshenker, N. M., Schaaf, T. K. & Huber, W. (1969) J. Am. Chem. Soc. 91,5675-5677. 15. Yost, F. J., Jr. & Fridovich, I. (1973) J. Biol. Chem. 248, 49054908. 16. McCord, J. M. & Fridovich, I. (1969) J. Biol. Chem. 244, 6049-6055. 17. Miyake, Y. & Fukuyama, M. (1976) J. Biochem. (Tokyo) 79, 613-620. 18. Lowry, 0. H., Rosebrough, N. J., Farr, A. L. & Randall, R. J. (1951) J. Biol. Chem. 193, 265-275. 19. Ellman, G. L. (1959) Arch. Biochem. Biophys. 82,70-77. 20. Paul, K.-G., Theorell, H. & Akeson, A. (1953) Acta Chem. Scand. 7, 1284-1287. 21. Samuelsson, B. (1967) Prog. Biochem. Pharmacol. 3,59-70. 22. Takeguchi, C., Kohno, E. & Sih, C. J. (1971) Biochemistry 10, 2372-2376. 23. Yoshimoto, A., Ito, H. & Tomita, K. (1970) J. Biochem. (Tokyo) 68,487-499. 24. Lee, R. E. & Lands, W. M. (1972) Biochim. Biophys. Acta 260, 203-211. 25. Chan, J. A., Nagasawa, M., Takeguchi, C. & Sih, C. J. (1975) Biochemistry 14, 2987-2991. 26. Marnett, L. J., Wlodawer, P. & Samuelsson, B. (1975) J. Biol. Chem. 250, 8510-8517. 27. Wolfe, L. S., Rostworoski, K. & Marison, J. (1976) Biochem. Biophys. Res. Commun. 70,907-913. 28. Lands, W., Lee, R. & Smith, W. (1971) Ann. N.Y. Acad. Sci. 180, 107-122. 29. Shen, T. Y. (1974) in Medicinal Chemistry, eds. Scherren, R. A. & Whitehouse, M. W. (Academic Press, New York), Vol. 13, pp. 203-204. 30. Maddox, I. S. (1973) Biochim. Biophys. Acta 306,74-81.
