Polyoxygenated Steroids from a Formosan Soft Coral Sinularia facile
Bo-Wei Chen,1 Jui-Hsin Su,1;2 Chang-Feng Dai,3 Yang-Chang Wu,4 and Jyh-Horng Sheu1;21Department of Marine Biotechnology and Resources, National Sun Yat-sen University, Kaohsiung 804, Taiwan 2Asia-Pacific Ocean Research Center, National Sun Yat-sen University, Kaohsiung 804, Taiwan
3Institute of Oceanography, National Taiwan University, Taipei 112, Taiwan
4Graduate Institute of Natural Products, Kaohsiung Medical University, Kaohsiung 807, Taiwan
Received April 3, 2008; E-mail: [email protected]
Five new polyoxygenated steroids 1–5 have been isolated from a Formosan soft coral Sinularia facile. The struc-tures of new metabolites were elucidated on the basis of extensive spectroscopic analysis and cytotoxic activity of 1–5 against the proliferation of a limited panel of cancer cell lines was measured. Metabolite 4 has been shown to exhibit weak cytotoxicity against Hep G2, Hep G3B, MDA-MB-231, and Ca9-22 cancer cell lines.
Previous chemical investigations of the Formosan soft cor-als of the genus Sinularia have afforded several polyoxygenat-ed steroids.1 In a previous study, two cembranes have been isolated from the soft coral Sinularia facile (Durivault).2Our
current chemical investigation on S. facile has also led to the isolation of five new polyoxygenated sterols 1–5 (Chart 1) from its EtOH extract. The structures of the new metabolites were determined on the basis of extensive spectroscopic analy-sis, including 2D NMR (1H–1H COSY, HMQC, HMBC, and NOESY) spectroscopy. Cytotoxicity of metabolites 1–5 against a limited panel of human tumor cell lines including liver (Hep G2 and Hep G3B), breast (MDA-MB-23), and gingival (Ca9-22) carcinoma cells are also reported.
The sliced bodies of the soft coral S. facile were extracted exhaustively with EtOH, and then the concentrated EtOH ex-tract was partitioned between EtOAc and H2O. The combined
EtOAc-soluble fraction was concentrated under reduced pres-sure and the residue was repeatedly chromatographed to yield metabolites 1–5.
Compound 1 was isolated as white power. Its molecular formula, C29H48O4, was established by HR-ESI-MS (m=z
483.3448 [M + Na]þ) and 13C NMR data, implying six
de-grees of unsaturation. IR absorptions were observed at 3422 and 1731 cm1, suggesting the presence of hydroxy and
car-bonyl groups. The structure of this compound was deduced from its 13C NMR and DEPT spectra, which showed that the
compound has 29 carbons, including six methyls, nine sp3
methylenes, one sp2 methine, nine sp3 methines (including
three oxymethines), and two sp2 and two sp3 quaternary
carbons. From 1H and 13C NMR spectra (Tables 1 and 2), 1 was found to possess one acetoxy group [
H 2.05, s; C169.8 (C), 21.8 (CH3)], in addition to one trisubstituted olefin
[
H 5.62, (br d, J ¼ 5:5 Hz),C 137.7 (C), 125.1 (CH)].De-tailed analysis of the 1H–1H COSY and HMBC correlations
(Figure 1) further established the planar structure of 1 as a cholesterol derivative bearing two hydroxy groups at C-1 and C-3, one acetoxy group at C-11, and one 5,6-trisubstituted double bond. In the NOESY spectrum of 1 (Figure 2), the
HO H H H 3 : R = Ac AcO 4 : R = H 2 HO H H H RO 5 HO H H H AcO AcO 1 HO H H H 1 3 5 7 9 10 11 13 14 18 19 17 AcO 20 21 22 25 26 27 24 HO HO HO HO Chart 1.
1304 Bull. Chem. Soc. Jpn. Vol. 81, No. 10, 1304–1307 (2008) 2008 The Chemical Society of Japan
NOE correlations between H-11 and H-8, H3-18, and H3-19;
H3-19 and H-1, H-2
(1.72), and H-4(2.26) as well asbetween H3-18 and H-20 indicated that these protons adapt
a
-orientation.3 This was further supported by comparisonof these NOE correlations with those of the corresponding NOE interactions displayed between the protons of a known compound 24-methylenecholest-5-ene-1
,3,11-triol 11 -acetate.4 The above data fully established the structure ofcompound 1 as cholest-5-ene-1
,3,11-triol 11-acetate (1). The HR-ESI-MS of compound 2 showed the pseudomolec-ular ion at m=z 497.3605 ([M + Na]þ), which indicated themolecular formula C30H50O4. Thus, six degrees of
unsatura-tion were determined for the molecule 2. It was shown that the NMR spectral data of 2 (Tables 1 and 2) is almost identical with those of 1, except that the1H and13C NMR spectra of 2
exhibited an additional methyl group [
H 0.77 (3H, d, J ¼6:7 Hz, H-28),
C15.5 (CH3, C-28)]. The1H–1H COSYcorre-lations between H2-23 and H-24; H-24 and H3-28 and HMBC
correlations from both H3-26 and H3-27 to C-24 and C-25
confirmed that this methyl group should be positioned at C-24. The stereochemistry of compound 2 was established by comparing the very similar NOESY correlations to those of 1. Furthermore, the 24S configuration of 2 was determined by comparison of NMR data with those of yonarasterol B which was isolated from the soft coral Clavularia viridis.5
The proton shift of H3-28,
H0.77, was found to be identicalwith that of yonarasterol B. Also, the carbon shifts of C24–C28 are in excellent agreement with those of yonarasterol B and (24S)-24-methylcholestanol (vs. those of (24R)-24-methylcho-lestanol).6The structure of compound 2 was thus established
as (24S)-24-methylcholest-5-ene-1
,3,11-triol 11-acetate.Table 1. 1H NMR Data for Sterols 1–5
No. 1aÞ 2bÞ 3aÞ 4bÞ 5aÞ
1 3.70 t (3.0)cÞ 3.70 br s 3.70 br s 4.21 br s 3.74 br s 2
: 2.10 m;: 1.72 m : 2.12 m;: 1.72 m : 2.10 m;: 1.72 m : 2.14 m;: 1.72 m : 2.10 m;: 1.72 m 3 3.96 tt (11.5, 5.0) 3.97 m 3.95 tt (11.5, 5.0) 3.98 m 3.96 tt (11.5, 5.0) 4 : 2.40 m;: 2.26 m : 2.40 m;: 2.33 m : 2.40 m;: 2.26 m : 2.38 m;: 2.27 m : 2.38 m;: 2.26 m 6 5.62 br d (5.5) 5.62 br d (5.3) 5.62 br d (5.5) 5.56 br d (5.5) 5.61 br d (5.5) 7 2.00 m; 1.67 m 1.99 m; 1.68 m 2.01 m; 1.67 m 2.02 m; 1.67 m 2.03 m; 1.70 m 8 1.54 m 1.53 m 1.55 m 1.53 m 1.69 m 9 1.87 m 1.87 m 1.88 m 1.65 m 1.96 m 11 5.29 dt (5.5, 11.0) 5.30 dt (5.4, 10.8) 5.29 dt (5.5, 11.0) 4.08 m 5.16 dt (5.0, 11.0) 12 : 1.18 m; : 1.18 m; : 1.18 m; : 1.25 m; : 1.16 m; : 2.38 dd (12.5, 5.0) : 2.39 dd (12.0, 5.1) : 2.38 dd (11.5, 5.0) : 2.34 m : 2.71 dd (12.5, 5.0) 14 1.16 m 1.17 m 1.16 m 1.15 m 1.38 m 15 1.62 m; 1.07 m 1.63 m; 1.08 m 1.62 m; 1.07 m 1.68 m; 1.07 m 1.69 m; 1.07 m 16 1.88 m; 1.31 m 1.89 m; 1.32 m 1.90 m; 1.30 m 1.88 m; 1.32 m 2.01 m; 1.40 m 17 1.16 m 1.17 m 1.16 m 1.15 m 1.34 m 18 0.74 s 0.74 s 0.74 s 0.67 s 4.18 d (12.0); 3.84 d (12.0) 19 1.12 s 1.12 s 1.12 s 1.12 s 1.11 s 20 1.35 m 1.33 m 1.37 m 1.37 m 1.47 m 21 0.90 d (6.5) 0.90 d (6.4) 0.92 d (6.5) 0.95 d (6.4) 1.03 d (6.5) 22 1.30 m; 0.97 m 1.38 m; 0.94 m 1.38 m; 1.02 m 1.37 m; 1.03 m 1.50 m; 1.17 m 23 1.32 m; 1.12 m 1.35 m; 0.96 m 2.00 m; 1.38 m 2.03 m; 1.38 m 2.08 m; 1.88 m 24 1.11 m 1.18 m 5.07 t (7.0) 5.08 t (7.0) 25 1.51 m 1.58 m 2.22 m 26 0.87 d (6.0) 0.85 d (6.8) 1.68 s 1.68 s 1.02 d (6.5) 27 0.86 d (6.0) 0.78 d (6.7) 1.60 s 1.60 s 1.01 d (6.5) 28 0.77 d (6.7) 4.72 s; 4.65 s 11-OAc 2.05 s 2.05 s 2.05 s 2.05 s 18-OAc 2.14 sa) Spectra recorded at 500 MHz in CDCl3. b) 300 MHz in CDCl3. c) J values (in Hz) parentheses.
: HMBC : 1H-1H COSY HO OH O 1 3 6 8 9 11 17 18 19 21 24 25 26 27 4 7 10 13 15 16 20 O
Figure 1. Selective1H–1H COSY and HMBC correlations of 1.
20 18 19 8 14 17 HO H AcO H H H H HO H H 1 H H H H H H H 1 3 11 9
Figure 2. Selective NOESY correlations of 1. B.-W. Chen et al. Bull. Chem. Soc. Jpn. Vol. 81, No. 10 (2008) 1305
Compound 3 has the molecular formula C29H46O4, as
deter-mined by HR-ESI-MS and NMR spectral data. Both the 1H
and13C NMR signals of 3 were found to be very closely relat-ed to those of compound 1, suggesting a very similar steroid skeleton. The only difference observed is that the side chain methyls of the isopropyl group (
0.87, d, J ¼ 6:0 Hz, H3-26;
0.86, d, J ¼ 6:0 Hz, H3-27) of 1 were replaced by twovinyl methyls (
1.60 and 1.68, each 3H, s) of 3. The above observation and the signal of an additional olefinic proton ( 5.07, t, J ¼ 7:0 Hz) showed the presence of a 24,25-trisubsti-tuted double bond in 3. These results revealed that the structure of 3 should be established as cholesta-5,24-diene-1,3,11 -triol 11-acetate. A structurally similar metabolite 4, was fur-ther isolated as a white solid. It’s molecular formula, C27H44-O3 was established by HR-ESI-MS. The1H and13C NMR of
4was very similar to that of 3 except that an acetyl group in 3was lost and the chemical shift of H-11 of 4 was shifted to upper field by 1.21 ppm, in comparison with that of 3. It was thus suggested that 4 is the 11-deacetyl derivative of 3.
The molecular formula of metabolite 5 was assigned as C32H50O6 from the HR-ESI-MS and NMR data (Tables 1
and 2). The 1H and 13C NMR spectral data of A–D rings in
5were nearly identical with those of 1 except for the replace-ment of a methyl substitution at C-13 in 1 by an acetoxymethyl group [
H 4.18 (d, J ¼ 12:0 Hz), 3.84 (d, J ¼ 12:0 Hz); C62.2 (CH2)] in 5. This was further confirmed by the HMBC
correlations from both H2-18 and the acetoxyl methyl to the
ester carbonyl carbon appeared at 171.4 (C). Furthermore, the structure of side chain (C-20 to C-28) was fully established by the1H–1H COSY correlations from H-20 to H
3-21; H2-22
to H2-23; H-25 to H3-26 and H3-27, and HMBC correlations
from H3-21 to C-17, C-20, C-22; H2-23 to C-24; H3-26 and
H3-27 to C-24, C-25; and H2-28 to C-23, C-25 and by
compar-ing the NMR data to those of known compounds.7Thus, the
structure of steroid 5 was established as 24-methylenechol-est-5-ene-1
,3,11,18-tetraol 11,18-diacetate.The cytotoxicity of compounds 1–5 against the proliferation of a limited panel of cancer cell lines, including human liver (Hep G2 and Hep G3B), breast (MDA-MB-23) and gingival (Ca9-22) carcinoma cells, was evaluated. The results showed that compound 4, the more potent one of compounds 1–5, ex-hibited cytotoxicity towards Hep G2, Hep G3B, MDA-MB-23, and Ca9-22 cancer cell lines with IC50’s 12.8, 12.0, 9.6, and
10.8mg mL1, respectively. Metabolites 1 and 5 also were
Table 2. 13C NMR Data for Sterols 1–5
No. 1aÞ 2bÞ 3aÞ 4bÞ 5aÞ
1 74.4 (CH)cÞ 74.5 (CH) 74.4 (CH) 75.0 (CH) 74.4 (CH) 2 37.9 (CH2) 38.0 (CH2) 37.9 (CH2) 38.3 (CH2) 38.0 (CH2) 3 66.2 (CH) 66.3 (CH) 66.2 (CH) 66.5 (CH) 66.2 (CH) 4 42.0 (CH2) 42.7 (CH2) 42.0 (CH2) 42.2 (CH2) 42.1 (CH2) 5 137.7 (C) 137.7 (C) 137.7 (C) 138.7 (C) 137.8 (C) 6 125.1 (CH) 125.1 (CH) 125.1 (CH) 124.8 (CH) 125.1 (CH) 7 32.1 (CH2) 32.2 (CH2) 32.1 (CH2) 32.5 (CH2) 32.2 (CH2) 8 31.9 (CH) 32.0 (CH) 31.9 (CH) 31.9 (CH) 32.1 (CH) 9 45.3 (CH) 45.3 (CH) 45.3 (CH) 45.3 (CH) 45.2 (CH) 10 43.0 (C) 43.1 (C) 43.0 (C) 43.1 (C) 43.0 (C) 11 72.3 (CH) 72.3 (CH) 72.3 (CH) 68.3 (CH) 71.7 (CH) 12 45.6 (CH2) 45.6 (CH2) 45.5 (CH2) 50.8 (CH2) 41.1 (CH2) 13 42.6 (C) 42.9 (C) 42.7 (C) 42.7 (C) 45.8 (C) 14 55.4 (CH) 55.4 (CH) 55.3 (CH) 55.6 (CH) 54.9 (CH) 15 24.2 (CH2) 24.3 (CH2) 24.2 (CH2) 24.3 (CH2) 24.0 (CH2) 16 28.3 (CH2) 28.3 (CH2) 28.3 (CH2) 28.4 (CH2) 28.0 (CH2) 17 55.8 (CH) 55.8 (CH) 55.8 (CH) 55.8 (CH) 56.0 (CH) 18 12.4 (CH3) 12.5 (CH3) 12.4 (CH3) 12.6 (CH3) 62.2 (CH2) 19 19.1 (CH3) 19.2 (CH3) 19.1 (CH3) 19.3 (CH3) 19.1 (CH3) 20 35.5 (CH) 36.0 (CH) 35.3 (CH) 35.5 (CH) 35.5 (CH) 21 18.7 (CH3) 18.9 (CH3) 18.6 (CH3) 18.7 (CH3) 19.0 (CH3) 22 36.0 (CH2) 33.6 (CH2) 35.9 (CH2) 36.0 (CH2) 34.4 (CH2) 23 23.7 (CH2) 30.6 (CH2) 24.6 (CH2) 24.7 (CH2) 30.6 (CH2) 24 39.4 (CH2) 39.1 (CH) 125.0 (CH) 125.1 (CH) 156.5 (C) 25 28.0 (CH) 31.6 (CH) 131.1 (C) 131.1 (C) 33.8 (CH) 26 22.5 (CH3) 20.5 (CH3) 25.7 (CH3) 25.8 (CH3) 22.0 (CH3) 27 22.8 (CH3) 17.7 (CH3) 17.6 (CH3) 17.7 (CH3) 21.8 (CH3) 28 15.5 (CH3) 106.1 (CH2) 11-OAc 21.8 (CH3) 21.9 (CH3) 21.8 (CH3) 21.8 (CH3) 169.8 (C) 169.9 (C) 169.8 (C) 169.5 (C) 18-OAc 21.0 (CH3) 171.4 (C) a) Spectra recorded at 125 MHz in CDCl3. b) 75 MHz in CDCl3. c) Attached protons were deduced by DEPT
experiments.
found to show weak cytotoxicity toward some of the above four cancer cells (Table 3).
Experimental
General Experimental Procedures. Melting points were de-termined using a Fisher–Johns melting point apparatus. IR spectra were recorded on a Jasco FT/IR-4100 infrared spectrophotometer. Optical rotations were measured on a Jasco P-1020 polarimeter. NMR spectra were recorded on a Bruker AMX-300 FT-NMR at 300 MHz for1H and 75 MHz for13C or on a Varian Unity INOVA
500 FT-NMR at 500 MHz for1H and 125 MHz for13C, in CDCl 3.
LRMS and HRMS were obtained by ESI on a Bruker APEX II mass spectrometer. Silica gel 60 (Merck, 230–400 mesh) was used for column chromatography. Precoated Silica gel plates (Merck Kieselgel 60 F254 0.2 mm) were used for analytical TLC.
High-performance liquid chromatography (HPLC) was performed on a Shimadzu LC-10ATVP apparatus equipped with a Shimadzu
SPD-10AVPUV detector. The columns used in HPLC separation
are YMC-Pack Pro C18 (reverse-phase column, 250 10 mm, 5mm) and Varian Dynamax, Si-60 (normal-phase column, 250 21:4 mm, 100 A˚ , 5 mm).
Animal Material. The soft coral S. facile was collected by hand using SCUBA off the coast of Pingtung County, located in southern Taiwan, in July 2001, at depths of 2–5 m and stored in a freezer until extraction. A voucher sample (20010719-1) was deposited at the Department of Marine Biotechnology and Resour-ces, National Sun Yat-sen University.
Extraction and Isolation. The frozen bodies of S. facile (1.05 kg, wet wt) were sliced and exhaustively extracted with EtOH (1 L 4). The combined organic layer was filtered and con-centrated by a rotorary evaporator, and the residue of the resulting aqueous suspension was partitioned between EtOAc and H2O. The
EtOAc extract was dried with anhydrous Na2SO4. After removal
of solvent in vacuo, the residue (15 g) was subjected to column chromatography on Si gel and eluted with EtOAc in n-hexane (0–100% of EtOAc, gradient) to yield 26 fractions. Fraction 23, eluted with EtOAc–MeOH (3:1), was rechromatographed over a Sephadex LH-20 column, using acetone as the mobile phase to afford five subfractions (F1–F5). Subfraction F2 was separated by reverse-phase HPLC (CH3CN–H2O, 6:1 to 9:1) to afford
com-pounds 1 (3.0 mg), 2 (5.0 mg), 3 (2.2 mg), 4 (4.1 mg), and 5 (2.3 mg), respectively.
Cholest-5-ene-1
,3
,11
-triol 11-Acetate (1). White pow-der; mp 128–130C; ½
25
D ¼ 23 (c 0.3, CHCl3); IR (neat)
max3422, 1731 cm1;1H NMR (CDCl3, 500 MHz) and13C NMR(CDCl3, 125 MHz), see Tables 1 and 2; ESIMS m=z 483
(M + Na)þ; HRESIMS m=z 483.3448 (calcd for C
29H48O4Na,
483.3450).
24(S)-24-Methylcholest-5-ene-1
,3
,11
-triol 11-Acetate (2). White powder; mp 130–133C; ½
25
D ¼ 58 (c 0.5, CHCl3);
IR (neat)
max 3437, 1733 cm1; 1H NMR (CDCl3, 300 MHz)and 13C NMR (CDCl
3, 75 MHz), see Tables 1 and 2; ESIMS
m=z 497 (M + Na)þ; HRESIMS m=z 497.3605 (calcd for
C30H50O4Na, 497.3607).
Cholesta-5,24-diene-1
,3
,11
-triol 11-Acetate (3). White powder; mp 147–150C; ½
25
D ¼ 20 (c 0.2, CHCl3); IR (neat)
max3422, 1728 cm1;1H NMR (CDCl3, 500 MHz) and13C NMR(CDCl3, 125 MHz), see Tables 1 and 2; ESIMS m=z 481
(M + Na)þ; HRESIMS m=z 481.3292 (calcd for C
29H46O4Na,
481.3294).
Cholesta-5,24-diene-1
,3
-11
-triol (4). White powder; mp 120–123C; ½
25
D ¼ 60 (c 0.4, CHCl3); IR (neat)
max 3370cm1; 1H NMR (CDCl
3, 300 MHz) and 13C NMR (CDCl3, 75
MHz), see Tables 1 and 2; ESIMS m=z 439 (M + Na)þ;
HRESIMS m=z 439.3185 (calcd for C27H44O3Na, 439.3188).
24-Methylenecholest-5-ene-1
,3
,11
,18-tetraol 11,18-Di-acetate (5). White powder; mp 137–140C; ½
25
D ¼ 42 (c
0.2, CHCl3); IR (neat)
max 3411, 1738 cm1;1H NMR (CDCl3,500 MHz) and13C NMR (CDCl
3, 125 MHz), see Tables 1 and 2;
ESIMS m=z 553 (M + Na)þ; HRESIMS m=z 553.3502 (calcd
for C32H50O6Na, 553.3505).
Cytotoxicity Testing. Cell lines were purchased from the American Type Culture Collection (ATCC). Cytotoxicity assays of the test compounds 1–5 were performed using the MTT [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide] col-orimetric method.8,9
Financial support was provided by the Ministry of Educa-tion (No. 96C031702) and the NaEduca-tional Science Council of Taiwan (NSC 95-2113-M-110-011-MY3) awarded to J.-H. Sheu.
References
1 a) A. F. Ahmed, C.-F. Dai, Y.-H. Kuo, J.-H. Sheu, Steroids 2003, 68, 377. b) A. F. Ahmed, Y.-T. Hsieh, Z.-H. Wen, Y.-C. Wu, J.-H. Sheu, J. Nat. Prod. 2006, 69, 1275. c) J.-H. Su, Y.-J. Tseng, H.-H. Huang, A. F. Ahmed, C.-K. Lu, Y.-C. Wu, J.-H. Sheu, J. Nat. Prod. 2006, 69, 850. d) A. F. Ahmed, S.-H. Tai, Y.-C. Wu, J.-H. Sheu, Steroids 2007, 72, 368. e) J.-H. Sheu, K.-C. Chang, C.-Y. Duh, J. Nat. Prod. 2000, 63, 149.
2 B. F. Bowden, J. C. Coll, S. J. Mitchell, R. Kazlauskas, Aust. J. Chem. 1981, 34, 1551.
3 W. Zhang, Y.-W. Guo, E. Mollo, A. Fontana, G. Cimino, J. Nat. Prod. 2004, 67, 2083.
4 B. M. Jagodzinska, J. S. Trimmer, W. Fenical, C. Djerassi, J. Org. Chem. 1985, 50, 1435.
5 M. Iwashima, K. Nara, K. Iguchi, Steroids 2000, 65, 130. 6 N. Koizumi, Y. Fujimoto, T. Takeshita, N. Ikekawa, Chem. Pharm. Bull. 1979, 27, 38.
7 B. M. Jagodzinska, J. S. Trimmer, W. Fenical, C. Djerassi, J. Org. Chem. 1985, 50, 2988.
8 M. C. Alley, D. A. Scudiero, A. Monks, M. L. Hursey, M. J. Czerwinski, D. L. Fine, B. J. Abbott, J. G. Mayo, R. H. Shoemaker, M. R. Boyd, Cancer Res. 1988, 48, 589.
9 D. A. Scudiero, R. H. Shoemaker, K. D. Paull, A. Monks, S. Tierney, T. H. Nofziger, M. J. Currens, D. Seniff, M. R. Boyd, Cancer Res. 1988, 48, 4827.
Table 3. Cytotoxicity Data of Compounds 1–5
Cell lines IC50/mg mL1
Compound
Hep G2 Hep G3B MDA-MB-231 Ca9-22
1 —aÞ 16.7 17.3 18.6 2 — — — — 3 — — — — 4 12.8 12.0 9.6 10.8 5 17.9 16.4 — 19.5 Doxorubicin 1.6 0.2 0.2 0.1 a) IC50> 20mg mL1.
