Effects of active compounds isolated from Angelica shikokiana on lipid metabolism in fat cells

Effects of active compounds isolated from Angelica shikokiana on lipid metabolism in fat cells

Journal of Ethnophurmacology, 25 (1989) 269-280 Elsevier Scientific Publishers Ireland Ltd. 269 EFFECTS OF ACTIVE COMPOUNDS ISOLATED FROM ANGELZCA S...

802KB Sizes 0 Downloads 7 Views

Journal of Ethnophurmacology, 25 (1989) 269-280 Elsevier Scientific Publishers Ireland Ltd.

269

EFFECTS OF ACTIVE COMPOUNDS ISOLATED FROM ANGELZCA SHIKOKIANA ON LIPID METABOLISM IN FAT CELIS

YOSHIYUKI KIMURA and HIROMICHI OKUDA Second Deqrtment of Medical Biochemistry, School of Medicine, Ehime University, Shigenobu-cho, Onsen-gun, Ehime 791-02 (Japan) (Accepted October 13, 1988)

Summary It was found that the EtOAc fraction of the roots of Angelica shikokiana inhibited adrenaline-induced lipolysis in rat fat cells, while having no effect on ACTH-induced lipolysis. On the other hand, the MeOH fraction was found to inhibit ACTH-induced lipolysis but not to inhibit adrenalineinduced lipolysis. The active substances isolated from this root were elucidated to be psoralen (I), bergapten (II), [email protected]‘-angeloyloxy-4’acetoxy-3’,4’-dihydroseselin) (III) and 3’-epoxyangeloyloxy-4’-acetoxy-3’,4’dihydroseselin (IV), respectively. The two furocoumarins, psoralen and bergapten, were found to enhance ACTH- and adrenaline-induced lipolysis in fat cells respectively. In contrast, a coumarin derivative of khellactone type, isopteryxin (III), was found to inhibit the adrenaline-induced lipolysis but with no effect on ACTH-induced lipolysis and insulin-induced lipogenesis from glucose. The other coumarin of khellactone type, 3’(R), 4’(R)3’-epoxyangeloyloxy-4’-acetoxy-3’,4’-dihydroseseline (IV), was found to inhibit the insulin-induced lipogenesis from glucose but it had no effect on adrenaline- and ACTH-induced lipolysis .

Introduction Angelica shikokiana Makino (Umbelliferae), called “yama-ninjin”, is cultivated in the Oita Prefecture of Japan and used as a substitute drug for ginseng roots. It has been reported that two coumarins, anomalin (3’,4’diangeloyloxy-3’,4’dihydroseselin) and isopteryxin (3’-angeloyloxy-4’acetoxy-3’,4’dihydroseselin), have been isolated from the roots of A. shik-

Correspondence to: Y. Kimura. 0378-8741/89/$04.55 @ 1989 EIsevier Scientific Publishers Ireland Ltd. Published and Printed in Ireland

270

okiana (Hata et al., 1967, 1973). In a previous paper (Kimura et al., 1987), we found that a new coumarin isolated from this root inhibited leukotriene (LTBI and LTC,) biosynthesis from human polymorphonuclear leukocytes and the chemical structure was elucidated to be 3’(R), 4’(R)-3’-epoxyangeloyloxy-4’-acetoxy-3’,4’-dihydroseselin. In the present work, we report the structural elucidation of several active substances on lipid metabolism as isolated from the roots of A. shikokiana. Materials and methods

Estimation of adrenaline- and ACTH-induced lipolysis, and insulininduced lipogenesis from [ U-14Cjglucose in rat fat cells Rat fat cells were isolated from the rat epididymal adipose tissue by the procedure of Rodbell (Rodbell, 1964). Hormone-induced lipolysis and insulininduced lipogenesis from [U-‘4C]glucose were determined as described previously (Kimura et al., 1982).

Chemical analysis IR, UV and CD spectra

were

measured on a Shimadzu IR-400 specand JASCO J-500 spectropolarimeter , respectively. ‘H-NMR (200 MHz) and ‘H-NMR (300MHz) spectra were recorded in CDC& and DzO on a JEOL-FX and Varian XL-300 spectrometer. Tetramethylsilane (TMS) was used as internal standard and chemical shifts are reported on the 6 scale (ppm and Hz). Column chromatography was carried out using silica gel 60 (70-230 mesh, ASTM, Merck Co.) as the adsorbent. Melting points were determined on a Yamato MP-21 capillary apparatus and are uncorrected.

trometer , Hitachi 220A spectrometer,

Preparation of primary extract The dried and crushed roots (1 kg) of plants collected in the Oita Prefecture of Japan were successively extracted with EtOAc (11 x 3) and MeOH (5 1x 2) for 3 h under reflux, respectively. The residue was further extracted with Hz0 (5 1) at room temperature for 1 week. The EtOAc, MeOH and Hz0 extracts were concentrated in vacua to give 9.5 g, 51.3 g and 54.5 g residues, respectively.

Psoralen (I) Colorless needles from a mixture of n-hexane and EtOAc, m.p. 162-163” (yield 100mg). The melting point showed no depression on admixture with an authentic sample of psoralen. From the IR and ‘H-NMR spectra data, it was identical with that reported for psoralen.

271

Bergup ten (II)

Colorless needles from a mixture of n-hexane and EtOAc, m.p. 192-193” (yield 120 mg). The melting point should no depression on admixture with an authentic sample of bergapten. From the IR and ‘H-NMR spectra data, it was identical with that reported for bergapten. Isopteryxin

(III)

Colorless viscous oil (yield 2.7 g). It gave one spot on TLC and exhibited a blue-violet fluorescence under a W lamp. IR(CHCL, max)v cm-l: 174O(br. C=O), 1640 (C=C), 1605(aromatic ring). W(EtOH, max)h(loge) nm: 205(4.49), 214(sh.)(4.28), 240(sh.)(3.66), 252(sh.)(3.48), 299(sh.)(3.91), 322(4.12). Mass spectrum: m/z 386(M+). ‘H-NMR(CDCL, 200 MHz)S ppm: 1.44, 148(3Hx2,

s, (C&),C<),

1.88(3H, t, J = 1.5Hz,

CH=C-CH,),

1.97(3H, q, J = 1.5 and 7 .OHz, C=CH-CH,), 2.11(3H, s, OCO-CH& 6.14(1H, q, J= 1.5 and 7.0Hz), 5.42, 6.62(1Hx2, d, J= 5.0Hz, -O-CH-CH-O-), 6.23, 7.64(1H x 2, d, J = 9.5 Hz, -CH=CH-), 6.81, 7.39(1H x 2, J = 8.5 Hz, arom. H). The CD spectra of III are discussed in the Results and Discussion section. From the ‘H-NMR spectra, this compound was identical with that reported for isopteryxin(3’-angeloyloxy-4’-acetoxy-3’,4’-dihydroseselin). Saponification

of IIIa with ethanolic NaOH

To the solution of 200 mg of III in 30 ml of EtOH, 10 ml of 1 N NaOH(EtOH) was added and the mixture was allowed to stand at room temperature for 10 min. The reaction was then stopped by adding 200 ml of ice-cold water, acidified with 20% HzS04, and extracted with Et20 . The Et20 solution was washed with 5% Na&03 and evaporated. The residue upon chromatograph over silica gel (20 g) with n-hexane : EtOAc (3 : 1) gave four products, IIIa, IIIb, 111~ and IIId. IIIb, 111~ and IIId were identified as (+)-cis-ethylkhellactone, (-)- truns-ethylkhellactone and (+)-cis-khellactone, respectively, by comparison with the spectral data of authentic samples. IIIa, colorless, viscous oil. ‘H-NMR(CDCl&? ppm: 1.31(3H, t, J = 7.0 Hz,

-CH,-C&),

1.47, 1.52(38x 2, s, (C&-C<),

1.80(3H, t, J= 1.5Hz,

C=C-C&), 1.85(3H, q, J = 1.5 and 7.0 Hz, C=CH-CH& 4.08(2H, q, J = 7.0 Hz, 0-CH,CH,), 4.63, 5.27(1H x 2, d, J = 2.1 Hz, -O-CHCH-0-), 6.11(1H, q, J=1.5 and7.0Hz), 6.23, 7.63(1Hx2, d, J=9.5Hz, -CEj=CH-), 6.79,7.34(1H x 2, d, J = 8.5 Hz, arom. H). From the ‘H-NMR spectra it was identical with that reported for 3’-angeloyloxy-4’-ethoxy-3’,4’dihydroseselin.

272

3’-Epoxyangeloyloxy-4’-acetoxy-3’,4’-dihydroseselin

(IV)

Colorless, viscous oil. It gave one spot on TLC and exhibited a blue-violet fluorescence under a UV lamp. IR (CHCL, max)vcm-‘: 1730(br.C=0), 1600(aromatic ring). UV @OH, max)h(log E) nm: 207(4.53), 217(sh.)(4.26), 245(3.72), 255(3.65), 301(sh.)(4.06), 324(4.34). Mass spectrum: m/z 402(M+). ‘H-NMR(CDCb)S ppm: 1.39(3H, d, J = 5.6 Hz, -CH-(-O-)CHS), 1.44, l/48(38 x 2, s, (CHa)&(),

1.57(3H, s, -C--J-O-)-C&),

2.14(3H, s, T-OCO-CHB), 3.07(lH, q, J = 5.6 Hz, CH-(-O-)-CHs), 5.40, 6.59(1Hx2, d, J=5.0Hz, -O-C&CF&O-), 6.25, 7.61(1Hx2, d, 6.82, 7.37(1H x 2, d, J = 8.5 Hz, arom. m. Yield J = 9.4 Hz, -CH=CH-), 4.Og. The CD spectra are discussed in the Results and Discussion section. From the IR and ‘H-NMR spectra, it was identical with an authentic sample of 3’(R),4’(R)-3’-epoxyangeloyloxy-4’-acetoxy-3’,4’-dihydroseselin. Saponification

of IV with ethanolic

NaOH

Saponification of IV was performed using the methods described above resulting in four products (IVa, IVb, IVc and IVd). IVb, IVc and IVd were identified as (+)-cis-ethylkhellactone, (-)- Puns-ethylkhellactone and (+)-ciskhellactone, respectively, by comparison with authentic samples. IVa, colorless, viscous oil. ‘H-NMR(CDCb)S ppm: 1.28(3H, d, J = 5.5 Hz, -CH-_(---O--)-C&), 1.30(3H, t, J = 7.0 Hz, -OCH,-CH,), 1.47, 1.51(3Hx2,

s, (C&&-C<),

1.52(38,

s, ‘X&-C--(-0-),

2.99(1H,

q,

J = 5.5 Hz, -CH-(-0-)-CH,), 4.02(28, m, -0--CI&-CHJ, 4.47, 6.26, 7.63(1Hx2, d, J= 5.27(1H x 2, d, J = 2.1 Hz, -0-C&--C&-0-), 9.6 Hz, -CH=CH-), 6.81, 7.34(1H x 2, d, J = 8.5 Hz, arom. H). From the ‘H-NMR. spectra data it was identical with that reported for 3’-epoxyangeloyloxy-4’-ethoxy-3’,4’-dihydroseselin (Kimura et al., 1987). Results and discussion

Isolation of the active substances in isolated fat cells

using hormone-induced

lipid

metabolism

As shown in Fig. 1, the EtOAc primary extract inhibited adrenalineinduced lipolysis at a concentration of 400 pg/ml but had no effect on the ACTH-induced lipolysis. The MeOH extract was found to inhibit the ACTHinduced lipolysis but did not inhibit the adrenaline-induced lipolysis. The EtOAc extract (9.0 g), inhibiting the adrenaline-induced lipolysis in rat fat cells, was then chromatographed on a column of silica gel (200 g) with n-hexane/EtOAc (3: 2) as the eluent. The part eluted with n-hexane/EtOAc (3:2) was divided into five fractions, Fr. 1(700mg), Fr.

273

a

b

OLI’

0 20

I

100 Concentration(pg/ml)

I

400

20

1,. 0 20

l

100 Concentration(pg/ml)

I

400

Fig. 1. Effects of JZtOAc, MeOH and He0 extracts on adrenaline- and ACTH-induced lipolysis in fat cells. Rat fat cells were incubated with the indicated amounts of EtOAc, MeOH or H,O extract in the presence of (a) adrenaline (0.5 pg/ml) or (h) ACTH (0.5 pglml) for 2 h at 37 “C. 0, EtOAc; 0, MeOH; +, HzO.

2(800 mg), Fr. 3(6.Og), Fr. 4(200mg) and Fr. 5(300mg). As shown in Fig. 2, Fr. 2 enhanced the ACTH-induced lipolysis at a concentration of 400 pg/ml, but did not inhibit adrenaline-induced lipolysis. Furthermore, Fr. 3 inhibited both adrenaline- and ACTH-induced lipolysis at concentrations of 400 pg/ml and 100 pg/ml (Fig. 2). Fractions 1, 4 and 5 had no effect on adrenalineinduced lipolysis. Two active compounds (I and II) enhancing ACTH-induced lipolysis were obtained from Fr. 2(400mg) by silica gel column chromatography with n-hexane/EtOAc (2: 1) as the eluent. As shown in Fig. 3, compound I enhanced the ACTH-induced lipolysis at concentrations of 150 pg/ml and 75 pglml; however, it had no effect on the adrenaline-induced lipolysis and the insulin-induced lipogenesis from glucose. Compound II enhanced the adrenaline-induced lipolysis at concentrations of 150 pg/ml and 75 pg/ml; however, it had no effect on the ACTH-induced lipolysis and insulin-induced lipogenesis from glucose (Fig. 3). Two other components (III and IV) were isolated from the Fr. 3(5.5 g) by silica gel column chromatography with n-hexane/EtOAc (2: 1). As shown in Fig. 3, compound III inhibited adrenaline-induced lipolysis at concentrations of 150 pglml and 75 pg/ml; however, it had no effect on ACTH-induced lipolysis and insulin-induced lipogenesis from glucose. On the other hand, compound IV inhibited lipo-

274 a

Wml Fr. 1

m/ml

Wml

Wml

Fr. 3

Fr. 4

Fr.5

b

0 ACTH (O.Spg/ml)

20 lob 400 iwml Fr.1

20 160 400 Wml Fr. 2

20 l&y0 Fr. 3

20 l&o Fr. 4

20 160 400 uglml Fr.5

Fig. 2. Effects of Fractions 1, 2, 3, 4 and 5 on (a) adrenaline- and (b) ACTH-induced lipolysis in fat cells.

275

37.5 ;5 150 Wmi II

b

37&l 150 )rglml I

37.5 7's 150 it+1

37.57: 150 )rglml

II

III

Fig. 3a, b.

37.:9+mEo

IV

276

m +-0

;

alo

x. In ._ IJI

w C

:

.-a -I

0 None

37.5 75 150 Mm1

37.5 75 150 &ml

rr

I

III

d

None Ins. (loop IU 1

I

Ii

375 i5 150 w/ml

37.5 ;5 150 Wml

III

IV

Fig, 3. IBeets of compounds I, II, III and IV on (a) adrenaline- and (bf ACTH-induced Iipolysis and on (c) lipogenesis and (d) insulin-induced lipogenesis from [LJ-%]glucose in fat cells. Significantly different from the respective values of adrenaline, ACTH or insulin alone; *p
genesis and insulin-induced lipogenesis at concentrations of 150 pg/ml and 75 pglml; however, it had no effect on adrenaline- and ACTH-induced lipolysis. Chemical structures of active substances I, II, III and IV Compounds I and II were identified as psoralen and bergapten, respectively, by comparison with authentic samples. Compounds III and IV, colorless, viscous oils, exhibited blue-violet fluorescence under a UV lamp. The infrared (IR) spectra of III and IV showed the presence of carbonyl and aromatic ring moities. The ultraviolet (UV) spectra of III and IV exhibited similarities to those of 7-hydroxycoumarin derivatives (Smith et al., 1957; Willette and Soine, 1962). The proton nuclear magnetic resonance (‘H-NMR) spectrum (CDCla) of III shows two pairs of doublet with intensities corresponding to one proton each; one pair appearing at 6.23 and 7.64 ppm (J = 9.5 Hz) can be assigned to protons at C-3 and C-4 positions of the coumarin ring and the other at 6.81 and 7.39 ppm (J = 8.5 Hz) for C-5 and C-6 positions of the coumarin ring, respectively. Further signals at 5.42 and 6.62 ppm (each lH, d, J = 5.0 Hz) can be attributed to protons at C-3’ and C-4’ positions, respectively. In the ‘H-NMR spectrum of III, signals corresponding to angeloyloxy groups were observed at 6.14ppm (lH, q, J= 1.5 and 7.0Hz), 1.97(3H, q, J= 1.5 and 7.0 Hz) and 1.88(3H, t, J = 1.5 Hz) as well as signals due to an acetoxy group at 2.11(3H, s) and gem-dimethyl group at 1.44 and 1.48(each 3H, s). These findings suggest that the structure of III could be either 3’-acetoxy-4’angeloyloxy-3’,4’-dihydroseselin (pteryxin) or 3’-angeloyloxy-4’-acetoxy-3’,4’dihydroseselin (isopteryxin). Saponification of III with 1N ethanolic NaOH at room temperature afforded four products: IIIa, IIIb, 111~and IIId. Among them, IIIb, 111~and IIId were identical with (+)-cis-ethylkhellactone, (-)trans-ethylkhellactone and (+)-cis-khellactone by comparison of spectral data with those of authentic samples, respectively. Furthermore, IIIa was noted to be identical with truns-3’-angeloyloxy-4’-ethoxy-3’,4’-dihydroseselin after comparison of its spectral data with those of an authentic sample (Hata et al., 1973). Therefore, III was concluded to be 3’-angeloyloxy-4’acetoxy-3’,4’-dihydroseselin. The ‘H-NMR spectrum (CDCL) of IV also showed the signals of two pairs of doublet attributed to the 3- and 4-positions of coumarin ring at 6.25 and 7.61ppm (each lH, d, J = 9.4 Hz) and the 5- and 6-positions of coumarin ring at 6.82 and 7.37 ppm (each lH, d, J = 8.5 Hz), respectively, as well as the signals due to the 3’- and 4’-positions at 5.40 and 6.59ppm (each lH, d, J = 5.0 Hz), gem-dimethyl group at 1.44 and 1.48 ppm (each 3H, s) and an acetoxyl group at 2.14 (3H, s). Further the signals were observed at 3.07 (lH, q, J = 5.6 Hz), 1.39(3H, d, J = 5.6 Hz) and 1.57(3H, s). These data of IV are very similar to those of epoxypteryxin (3’-acetoxy-4’-epoxyangeloyloxy3’,4’-dihydroseselin) isolated from the roots of Luserpitium archungelica

278

Wulf. (Bohlmann and Thefeld, 1970). Therefore, the structure of IV seemed to be either 3’-epoxyangeloyloxy-4’-acetoxy-3’,4’-dihydroseselin or 3’acetoxy-4’-epoxyangeloyloxy-3’,4’-dihydroseselin. Saponification of IV also afforded four products. Among them, three products were identical with (+)-cis-ethylkhellactone, (-)-truns-ethylkhellactone and (+)-cis-khellactone by direct comparison with authentic samples. The structure of IVa, afforded by the saponification of IV, was also identified as 3’-epoxyangeloyloxy-4’ethoxy-3’,4’-dihydroseselin by comparison with an authentic sample. ThereIV was identified as 3’-epoxyangeloyloxy-4’-acetoxy-3’,4’-dihyfore, droseselin. The relative configuration of the C-3’ and C-4’ positions of III and IV were determined to be cis by observation of the coupling constant between 3’-H and 4’-H (J= 5.0Hz) in the ‘H-NMR spectrum. Moreover, the absolute configuration of III and IV were concluded to be 3’(R) and 4’(R) from the evidence that (+)-cis-ethylkhellactone and (-)- trcms-ethylkhellactone were obtained by saponification of III and IV and that the circular dichroism (CD) spectrum of III and IV showed multi-Cotton effects very similar to that of anomalin (3’(R) 4’(R)-3’,4’-diangeloyloxy-3’,4’-dihydroseselin) (Hata et al., 1974) (Fig. 4). Therefore, the structures of III and IV were established to be 3’(R),4’(R)-3’-angeloyloxy-4’-acetoxy-3’,4’-dihydroseselin and 3’(R),4’(R)-3’epoxyangeloyloxy-4’-acetoxy-3’,4’-dihydroseselii-i, respectively. In summary, the roots of Angelica skikokiana Makino have been used in

d&lo9a-

223nm r

76-

i’

5-

ii*

4-

;

3_

j

; \ 325nm

350nm

-4 -

245nm

-5Fig. 4. CD spectrum

of III and IV.

279

q&x0wg& 1 :

R’=H

11 : R’=OCHs

RZ 111:R*= -OCO~C=C~W H+.Z’

‘H AH3

IV:R*=

-oco-c-c H3C’

‘0’

‘H

Fig. 5. Structures of I, II, III and IV.

treatment of adult diseases such as hyperlipemia and diabetes, inflammatory and allergic diseases as a substitutional drug for ginseng roots in the Kyushu region of Japan. In a previous paper (Kimura et al., 1987) 3’(R),4’(R)-3’-epoxyangeloyloxy-4’-acetoxy-3’,li’dihydroseselin isolated from the roots was found to be an inhibitory substance on leukotriene (LTB, and LT&) formation from human polymorphonuclear leukocytes. In the present study, four coumarin derivatives, as active substances on hormone-induced lipid metabolism in isolated fat cells, were isolated from the roots of A. shikokiunu. The structures of active substances were elucidated to be two furocoumarins, psoralen and bergapten, and two derivatives, khellactone 3’(R),4’(R)-3’-angeloyloxy-4’-acetoxy-3’,4’-dihydroseselin (isopteryxin) and 3’(R),4’(R)-3’-epoxyangeloyloxy-4’-acetoxy-3’,4’dihydroseselin, respectively (Fig. 5). Bergapten and psoralen enhanced adrenaline- and ACTH-induced lipolysis in fat cells, respectively, but had no effect on the insulin-induced lipogenesis from glucose. In contrast, 3’(R),4’(R)-3’-angeloyloxy-4’-acetoxy-3’,4’-dihydroseselin was found to inhibit adrenaline-induced lipolysis, but had no effect on ACTHinduced lipolysis and insulin-induced lipogenesis. On the other hand, of the two khellactone derivatives, 3’(R), 4’(R)-3’-epoxyangeloyloxy-4’-acetoxy-3’,4’dihydroseselin inhibited both lipogenesis and insulin-induced lipogenesis from glucose in fat cells, but had no effect on either adrenaline- or ACTHinduced lipolysis. Further work is needed to clarify the relationships between the structures and physiological actions of such coumarin derivatives on hormone-induced lipid metabolism in fat cells. Acknowledgements

The authors are grateful to Prof. Dr. M. Koxawa and Dr. K. Baba, Osaka University of Pharmaceutical Sciences, for taking the ‘H-NMR and CD spectra, and for helpful discussions on the chemical structures.

280

References Bohlmann,

F. and Thefeld,

W. (1970)

Neue dihydroseselin-derivate

aus Laserpitum

archan-

gelica Wulf. Tetrahedron Letters 41, 3577-3580. Hata,

K.,

Kozawa,

M. and Ikeshiro,

Y. (1967)

New coumarins

Angelica anomala Lall. and Angelica cartilaginomarginata Yakugaku Zasshi 87, 1118-1124. Hata, K., Kozawa,

Ligusticum 251.

isolated

(Makino)

from

the roots

of

Nakai (Umbelliferae).

M., Baba, K. and Mitsui, M. (1973) Chemical components of the roots of Fernald and Angelica shikokiana Makino. Yakugaku Zasshi 93, 248-

hultenii

Hata, K., Korawa,

M., Baba, K., Yen, K.Y. and Yang, L.L. (1974) Coumarins from the roots of Chemical and Pharmaceutical Bulletin 22, 957-961. Kimura, Y., Ohminami, H., Arichi, H., Okuda, H., Baba, K., Kozawa, M. and Arichi, S. (1982) Effects of various coumarins from roots of Angelica duhurica on actions of adrenaline, ACTH and insulin in fat cells. Plantu Medica 45, 183-187. Kimura, Y., Okuda, H., Baba, K., Kozawa, M. and Arichi, S. (1987) Effects of an active substance isolated from the roots of Angelica shikokiana on leukotriene and monohydroxyeicosatetraenoic acid biosyntheses in human polymorphonuclear leukocytes. Planta Medica

Angelica morii Hayata.

53, 521-525. Rodbell, M. (1964) Metabolism of isolated fat cells. I: ism and lipolysis. Journal of Biological Chemistry Smith, E., Hosansky, N., Bywater, W.G. and van samidin, dihydrosamidin and visnadin. Journal

Effects of hormones

on glucose

metabol-

239, 375380. Tamelen,

E.E. (1957)

Constitution

of

Chemical Society

79,

Willette, R.E. and Soine, T.O. (1962) Isolation, purification, and structure determination pteryxin and suksdorfin. Journal of Pharmaceutical Sciences 51, 149-156.

of

of the American

3534-3540.