Biosci. Biotechnol. Biochem., 70 (10), 2494–2500, 2006
A Novel Labdane-Type Trialdehyde from Myoga (Zingiber mioga Roscoe) That Potently Inhibits Human Platelet Aggregation and Human 5-Lipoxygenase
y Masako ABE,1;3; Yoshio OZAWA,1 Yasushi UDA,2 Yasujiro MORIMITSU,3 Yoshimasa NAKAMURA,4 and Toshihiko OSAWA5
1Department of Health and Nutrition, Takasaki University of Health and Welfare, 37-1 Nakaoorui-machi, Takasaki-shi, Gunma 370-0033, Japan 2Deparment of Bioproductive Sciences, Utsunomiya University, 350 Mine-machi, Utsunomiya-shi, Tochigi 321-8505, Japan 3Laboratory of Food Chemistry, Graduate School of Humanities and Sciences, Ochanomizu University, 2-1-1 Ohtsuka, Bunkyo-ku, Tokyo 112-8610, Japan 4Department of Biofunctional Chemistry, Division of Bioscience, Graduate School of Natural Science and Technology, Okayama University, 1-1-1 Tsushimanaka, Okayama-shi, Okayama 700-8530, Japan 5Laboratory of Food and Biodynamics, Nagoya University Graduate School of Bioagricultural Sciences, Furou-cho, Chikusa-ku, Nagoya-shi, Aichi 464-8601, Japan
Received April 18, 2006; Accepted June 20, 2006; Online Publication, October 7, 2006 [doi:10.1271/bbb.60226]
We screened myoga extracts for inhibitors of human and pungent flavor, the flower buds are used as a spice platelet aggregation and human 5-lipoxygenase. We and pickled vegetable. The characteristic flavor of the identified a novel labdane type of diterpene, together flower buds has been reported to be due to 2-alkyl-3- with three known diterpenes (miogadial and galanals A methoxypyrazine (alkyl = isobutyl, isopropyl) and 8 - and B) from the flower buds of myoga. Spectroscopic 17-epoxy-12E-labdene-15,16-dial (miogadial 2).1,2) Of data indicated the structure of the new compound to these compounds, we have previously found 2 to be be 12(E)-labdene-15,16,(8 )17-trial (miogatrial). Mioga- the dominant contributor to the hot taste of the flower trial and miogadial were potent inhibitors of human buds.2) In addition, the leaves of the myoga plant are platelet aggregation and human 5-lipoxygenase (5- used to wrap and preserve ‘‘manjyu,’’ a bun filled with LOX). The sesquiterpene, polygodial, also exhibited sweetened bean paste. Previous examination of the strong inhibitory activity against human platelet aggre- antimicrobial activities of the constituents of myoga gation and 5-LOX. On the other hand, galanals A and B against several strains of bacteria, yeast, and mold did not have inhibitory activity in either experimental indicated that miogadial had potent activity against system. It thus appears that a 3-formyl-3-butenal yeasts and Gram-positive bacteria.3) structure was essential for the potent inhibition of A number of investigations have recently been human platelet aggregation and human 5-LOX. focused on the health benefits of dietary ingredients; for example, in combatting cancer, atherosclerosis, and Key words: Zingiber mioga Roscoe; platelet aggrega- diabetes. Under pathophysiological conditions, platelet tion; 5-lipoxygenase; labdane-type diter- activation may result in peripheral, cardiovascular, or pene; polygodial cerebrovascular thrombosis, with serious consequen- ces.4) Aspirin has been used for anti-platelet therapy for Myoga (Zingiber mioga Roscoe; Zingiberaceae fam- several decades, but it can promote the formation of ily) is a perennial herb indigenous to eastern Asia and gastric ulcers and induce gastrointestinal bleeding.4) The widely cultivated throughout Japan. It has a stalk that natural products of the Zingiberaceae family, such as extends up to 1 m, with slender leaves reaching 30 cm in ginger, have therefore been studied by many investi- length. The flower buds are produced at ground level gators as a source of new anti-platelet agents.5–13) from underground rhizomes, and they grow to 7–10 cm Although the flower buds and rhizomes of myoga have in length with a pine cone-like shape and are edible been used since ancient times in Japan and China as during the summer or autumn. Due to their distinctive a herbal medicine against cough and rheumatism,14,15)
y To whom correspondence should be addressed. Fax: +81-27-352-1169; E-mail: [email protected] Novel Labdane-Type Trialdehyde from Myoga 2495 there have been only a few studies on its physiological 16 16 CHO CHO 16) 12 12 effects. Iwashita et al. have reported that a myoga 13 13 11 11 extract significantly reduced the glycerol-3-phosphate 20 CHO 20 CHO 1 14 15 1 17 14 15 dehydrogenase activity and triglyceride accumulation 2 9 CHO 2 9 O in 3T3-L1 cells and prevented the increase in body 10 8 17 10 8 3 3 4 5 7 4 5 7 weight and epididymal fat weight of male ICR mice. 6 6 17) 18 19 18 19 In addition, Miyoshi et al. have reported that galanals 1 2 A 3 and B 4 isolated from the flower buds of myoga had potent cytotoxic activity in Jurkat human T 16 16 CHO CHO lymphoma cells due to the induction of apoptosis. Most 12 13 12 18,19) 13 recently, Kim et al. have demonstrated that the 20 11 11 14 20 14 CHO 8 -17-epoxy-12E-labdene-15,16-dial (2) from a myoga 1 15 1 15 2 9 2 9 CH 10 8 R 2 extract strongly suppressed the 12-O-tetradecanoylphor- 17 10 8 17 CHO 3 bol-13-acetate-induced superoxide generation in HL-60 3 4 5 7 4 5 7 6 6 18 19 cells, lipopolysaccharide/interferon- -induced nitric ox- 18 19 ide generation in RAW264.7 cells, and HL-60 cell- 3=OH 5 induced mutagenicity in AS52 cells. Despite this work, 4=OH there have been few studies on the effects of a myoga 16 16 12 CH OH extract on platelet aggregation and 5-lipoxygenase (5- 12 CHO 2 13 11 13 11 LOX) affecting inflammatory responses. 20 20 17 17 CH OH 1 CH OH 1 14 2 We isolated in the current study a novel labdane 14 2 15 2 9 O 15 2 9 O type of diterpene from the flower buds of myoga that 10 8 10 8 4 7 3 4 5 7 3 5 inhibited human platelet aggregation and human 5-LOX. 6 6 We also determined its structure and compared its 18 19 6 18 19 7 inhibitory activity with that of its related analogues including polygodial. 11 15 CHO 1 8 CHO 2 9 Materials and Methods 10 12
3 4 5 7 Materials and reagents. Fresh flower buds of myoga 6 13 14 were purchased from a market in Takasaki City, Japan. 8 2, 3 and 4 were isolated as reported.2) The fresh flower buds of myoga (500 g) were homogenized in ethyl Fig. 1. Chemical Structures of the Tested Compounds. acetate with a blender. The homogenate was allowed to 1, 12(E)-labdene-15,16,(8 )17-trial (miogatrial); 2, 8 (17)-ep- oxy-12(E)-labdene-15, 16-dial (miogadial); 3, galanal A; 4, galanal stand for 3 h at room temperature. After filtration, the B; 5,(E)-labda-8(17),12-diene-15,16-dial; 6, 8 (17)-epoxy-15-hy- filtrate was concentrated under reduced pressure and droxy-12(E)-labden-16-al; 7, 8 (17)-epoxy-12(E)-labdene-15,16-di- then fractionated by flash chromatography in a silica gel ol (reduced miogadial); 8, polygodial FL-100DX column (Fuji Silysia Co.; 2 40 cm) with ethyl acetate–benzene–chloroform (1:4:6) as the solvent. into ice-cooled water. The mixture was then extracted The most active fraction was separated by HPLC using with diethyl ether, and the diethyl ether solution was 4% acetone in n-hexane as the mobile phase at a flow washed, dried and evaporated. The residue was sepa- rate of 4.0 ml/min, to give 2 and 3 and 4.(E)-labda- rated by TLC to afford a colorless powder, which was 8(17),12-diene-15,16-dial 5 and 8 (17)-epoxy-15-hy- confirmed by MS and NMR to be a diol compound droxy-12(E)-labden-16-al 6 were prepared from the corresponding to 2. The structure of this compound is rhizomes of ginger. Sliced ginger (500 g) was extracted shown in Fig. 1. with n-hexane for 24 h at room temperature. After filtration and evaporation, the residue was subjected to Equipment. HPLC was performed with a Shimadzu HPLC in a Develosil ODS-5 column (8 250 mm) LC-10AS instrument equipped with a Tosoh silica-60 with methanol/H2O/acetic acid (80/20/0.1) as the column (7:8 300 mm) and a UV detector (233 nm). mobile phase at a flow rate of 3 ml/min to give 5 and UV spectra were recorded with a Shimadzu UV-160 6.20) Polygodial 8 was prepared from the leaves of spectrophotometer, optical rotation was measure with a Polygonum hydropiper L. by the method Barnes and Jasco P-1010 polarimeter, and IR spectra were recorded Loder.21) with a Jasco FTIR-420 spectrometer. 1H- (400 MHz) and 13C- (100 MHz) NMR spectra were respectively ob- Preparation of reduced miogadial 7. 2 was reduced as tained with a Jeol EX-400 NMR spectrometer and a 2) reported. To an ice-cold solution of 2 (20 mg) in dry Bruker ARX-400 spectrometer in CDCl3, with tetra- THF (3 ml), NaBH4 (20 mg) was added, and the reaction methylsilane used as the internal standard. MS spectra mixture was stirred at 0 C for 1.5 h, before being poured were recorded with a Jeol JMS 700 mass spectrometer. 2496 M. ABE et al. Extraction and isolation of 1. Fresh inflorescences of inhibitory effect of a test compound on human 5-LOX myoga (500 g) were sliced and homogenized in 1.5-liter was measured by following the production of 5- of ethyl acetate with a blender. After soaking for 3 h at hydroxy-6,8,11,14-eicosatetraenoic acid (5-HETE) from room temperature, the ethyl acetate extract was passed AA. The assay mixture (100 ml total volume) contained through filter paper (No. 2) under reduced pressure. This 0.1 M Tris–HCl (pH 8.0), 10 mM CaCl2,10mM ATP, 10 extraction procedure was conducted twice. The com- mg of sonicated phosphatidylcholine, the test sample in bined ethyl acetate extracts were dried on anhydrous DMSO, and 100 mg of 5-LOX (Cayman Chemical). The Na2SO4 and then concentrated in vacuo at 30 C to give mixture was warmed to 30 C for 5 min, after which an oily residue. The concentrate was successively the reaction was initiated by adding 50 mmol AA. After divided into three fractions (A–C) by silica gel column 10 min at 30 C, the reaction was terminated by adding chromatography in a Kieselgel 60 column (2 20 cm), 0.3 ml of cold ( 20 C) methanol containing 0.2 mmol using 15% acetone in n-hexane as the solvent. All the of 13-hydroxylinoleic acid (Cayman Chemical) as an fractions were concentrated and tested for their inhib- internal standard and 1 mlof1M acetic acid. Following ition of platelet aggregation. Fraction B showed the centrifugation at 11,000 rpm for 10 min, the supernatant strongest ability to inhibit platelet aggregation. This was analyzed by HPLC. fraction was concentrated under reduced pressure at 30 C and then subjected to HPLC, using 4% acetone in Analytical procedure. The level of 5-HETE derived n-hexane as the mobile phase at a flow rate of 4 ml/min. from AA was measured by HPLC with a Shimadzu 2 and 1 were respectively eluted at retention times of LC-10AT equipped with an autosampler and a Capcell- 20.5 and 22.5 min. To purify compound 1, the HPLC pak C18 column (4:6 150 mm; Shiseido), using procedure was conducted three times, resulting in a methanol/water/acetic acid (80:20:0.01) as the solvent colorless oily compound (about 20 mg). The ethanol and detecting at 233 nm. The control was performed solution of 1 did not have any strong taste. in triplicate. The IC50 value was determined as the 25 UVmax (EtOH) nm (") 233 (18000); ½ D þ33 concentration of the compound that gave 50% inhibition þ (c1.6, CH3COCH3). HR-MS m=z (M ): calculated (for of the production of 5-HETE compared with the C20H30O3), 318.2195; found, 318.2197. EI-MS m=z (%): negative control (no added test compound). 318 (38, Mþ), 304 (50), 261 (56), 137 (55), 123 (100), 109 (78), 95 (80), 69 (92). IRmax (CHCl3 on NaCl) Results and Discussion cm 1: 3027, 2927, 2854, 1716, 1683, 1642, 1461, 1444, 1389, 1368, 1290, 1232, 1059. Structure of 1 The results of 13C-NMR and 1H-NMR of 1 are shown Preparation of human platelets.22) Venous blood was in Table 1. Based on the EI-MS and HR-MS measure- obtained from healthy donors by withdrawing into plastic syringes containing 1/10 final volume of 3.8% 1 13 sodium citrate as an anticoagulant. Platelet-rich plasma Table 1. H- and C-NMR Data for 1 was prepared by centrifuging the citrate-treated blood at Position H C 120 g for 10 min. Platelet-poor plasma was obtained 1.15 (1H, m) 1 39.88 by centrifuging the residue at 1100 g for 15 min. All 1.75 (1H, m) procedures were performed at room temperature. 2 1.40 (2H, m) 19.82 1.22 (1H, m) 3 43.16 Platelet aggregation measurement and IC50 determi- 1.42 (1H, m) nation.22) Platelet aggregation was measured by tur- 4 34.25 5 1.92 (1H, dd, J ¼ 4, 13) 55.03 bidimetry with a dual-channel aggregometer (PA-20 6 1.60 (2H, m) 20.11 Aggregation Analyzer; Kowa). Platelet-poor plasma 1.40 (1H, m) 7 27.65 was used as a control for 100% transmittance. In all 2.35 (1H, m) experiments, 270 ml of platelet-rich plasma (250,000 8 2.50 (1H, m) 49.23 platelets/ml) and 1 ml of the test compound in methanol 9 1.05 (1H, ddd, J ¼ 4, 12, 12) 56.87 10 39.88 were incubated at 37 C for 1 min while stirring. 2.65 (1H, ddd, J ¼ 4, 6, 15.2) 11 27.14 Aggregation was induced by adding 20 mlofanADP 2.85 (1H, ddd, J ¼ 8:8, 12, 15.2) solution (final concentration of 10 mM)or20mlof 12 7.10 (1H, dd, J ¼ 6:0, 8.8) 158.72 arachidonic acid (AA) solution (final concentration of 13 138.35 3.48 (1H, d, J ¼ 17) 0.5 mM). The IC value was determined as the 14 40.29 50 3.54 (1H, d, J ¼ 17) aggregation compared to that of the negative control 15 9.62 (1H, s) 198.84 (no added test compound). These experiments were 16 9.50 (1H, s) 194.80 carried out under experimental and ethical guidance for 17 9.96 (1H, s) 205.37 biohazard safety. 18 0.88 (3H, s) 34.25 19 0.82 (3H, s) 22.31 20 0.80 (3H, s) 16.16 Assay of human 5-LOX inhibition.23,24) Briefly, the Novel Labdane-Type Trialdehyde from Myoga 2497
H 16 CHO 13 11 12 H CH H 20 3 CHO 16 12 13 C O 1 9 14 17 H H H CHO 15 20 15 2 H CH 11 H 10 H 3 H 14 C 8 19 H O H H CH3 8 C H 17 O 3 4 7 2 10 5 1 9 H H H H 6 4 6 H 3 5 H 7 H3C H H3C CH3 18 H H 18 19 H
Fig. 2. HMBC and NOESY Correlations of 1. Arrows denote the key HMBC correlations. Bold lines indicate 1H–1H correlations. Dashed lines show the selected NOESY correlations.
ments, 1 had the molecular formula of C20H30O3 NOE correlation was observed between the 20-methyl (MW ¼ 318) which indicates six degrees of unsatura- proton and 17-aldehyde proton, and between 5-H and tion. The UV spectrum [(EtOH) nm(") 233 (18000)] of 1 both 8-H and 9-H. The orientations of the 8-aldehyde indicates the presence of an , -unsaturated aldehyde group and 9-H were confirmed to be -equatorial and group. The IR spectrum of 1 shows absorption bands -axial, respectively.25–27) Thus, by examining the mo- at 1684 and 1716 cm 1, which could be respectively lecular model generated from the foregoing data, the attributed to an , -unsaturated aldehyde and an relative structure of 1 was determined for the first time aliphatic aldehyde group. The 1H-NMR spectrum of 1 to be 12(E)-labdene-15,16,(8 )17-trial, which we have shows signals originating from three methyl groups ( named miogatrial 1. Unsaturated diterpenes with three 0.80 [3H, s], 0.82 [3H, s], 0.88 [3H, s]), an olefinic aldehyde groups are rare in natural products.29) One hydrogen ( 7.10, 1H, dd, J ¼ 6:0, 8.8), and three example, halimedatrial, has been isolated from the aldehyde groups ( 9.50 [1H, s], 9.62 [1H, s], 9.96 [1H, tropical marine algae, Halimeda, as an effective feeding s]). On the basis of the 13C-NMR and DEPT spectral deterrent to herbivorous fishes.30) It was therefore data, 1 consisted of seven methylene (1, 2, 3, 6, 7, 11, assumed that 1 would possess some biological activities, 14-H2), three methyl (18, 19, 20-H3), three methine (5, even though it did not have the hot taste characteristic of 8, 9-H), three carbonyl (15, 16, 17-H), two alkene (12-H, myoga. 13-C), and two quaternary carbons (4, 10-C). To satisfy the degree of unsaturation, 1 should have two rings in Inhibition of human platelet aggregation the structure. As shown in Table 2, we examined the ability of the As shown in Table 1, there are some similarities in diterpene compounds isolated from myoga to inhibit the the NMR data between 1 and 2.2) The 1H–1H COSY and AA- and ADP-induced aggregation of human platelets. HMQC experiments on 1 show the presence of four Aspirin and indomethacin were used as positive controls partial structures (shown as bold lines in Fig. 2; 1-C– because they are known inhibitors of platelet aggrega- 3-C, 7-C–17-C, 11-C–12-C, and 14-C–15-C). In the tion.31) We found that 1 and 2 inhibited the ADP- HMBC experiment, long-range correlation was observed induced platelet activation with IC50 values of 3.5 and between the following protons and carbons of 1: 5-H and 3.0 mM, respectively. The IC50 values of 3 and 4 were 1-C, 7-C, 10-C; 9-H and 7-C; 12-H and 16-C; 16-H and 268 and 287 mM, respectively. Furthermore, 5 inhibited 13-C; 17-H and 8-C; 18-H and 3-C; and 20-H and 1-C. human platelet aggregation with an IC50 value of 3.2 mM. In addition, the EI-MS spectrum of 1 showed the We also found for the first time that polygodial 8,a characteristic fragment ion peak (m=z 137) of a labdane pungent component of Polygonum hydropiper L., 25) type of diterpene. These data confirm that the planar strongly inhibited human platelet aggregation (IC50 ¼ structure of 1 was labd-12-ene-15,16,17-trial. 4:9 mM). In contrast, 6 and 7 did not inhibit platelet In the NOESY experiment for 1, NOE correlation was aggregation. In respect of the AA-induced platelet observed between the 12-olefin proton and the 16- aggregation, 1 and 2 had IC50 values of 11.9 and 10.7 aldehyde proton. The data indicate that the double bond mM, respectively, being approximately 25% as potent as (12-C to 13-C) had an E geometry.26–28) Furthermore, indomethacin. Interestingly, 1 and 2 were significantly 2498 M. ABE et al. Table 2. Inhibition of Human Platelet Aggregation Table 3. Inhibition of Human 5-LOX
Inducer Compound IC50 (mM) AA ADP Compound miogatrial (1) 7.5 IC50 (mM)IC50 (mM) miogadial (2) 4.0 galanal A (3) >100 miogatrial (1) 11.9 3.5 galanal B (4) >100 miogadial (2) 10.7 3.0 E-labda-8(17),12-diene-15,16-dial (5) 18.9 galanal A (3) >500 268 8 ,(17)-epoxy-15-hydroxy-12(E)-labden-16-al (6) >100 galanal B (4) >500 287 reduced miogadial (7)— E-labda-8(17),12-diene-15,16-dial (5)— 3.2 polygodial (8) 8.6 8 ,(17)-epoxy-15-hydroxy-12(E)- — >500 AA861 (positive control) 0.3 labden-16-al (6) NDGA (positive control) 16.5 reduced miogadial (7)— >500 polygodial (8)—4.9 AA861, 2-(12-hydroxydodeca-5,10-diynyl)-3,5,6-trimethyl-p-bezoquinone aspirin (positive control) 41.2 — NDGA, nordihydroguaiaretic acid indomethacin (positive control) 2.1 — Not determined